- Email: [email protected]
A neural pathway for the shift response in the cat
Brain Research, 337 (1985) 51-58 Elsevier
51
BRE 10800
A Neural Pathway for the Shift Response in the Cat D. I. HAMASAKI and GREGORY W. MAGUIRE
William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, Department of Ophthalmology, University of Miami, Miami, FL 33101 (U.S.A.) (Accepted September llth, 1984)
Key words: shift response - - cat retinal ganglion cell - - neural pathway - - Mcllwain effect
The shift response (Mcllwain effect) was elicited by moving a large grating situated greater than 15° from the conventional receptive field center (RFC). We examined the change in the amplitude of the shift response induced by placing a steady target on the RFC or the RF surround. We found that appropriate stimulation of the RFC or the RF surround will increase and inappropriate stimulation will decrease the amplitude of the shift response in a graded manner. The amplitude of the shift response was not correlated with the maintained activity but was correlated with the transient peak firing rate which is evoked by flashing the enhancing stimulus. A shift stimulus which elicits a strong shift response can be blocked by inappropriate stimulation. The results suggest that the shift signal is modulating a tonic signal which is present in the RF. Because the shift response is a transient excitation, we suggest that the shift response results from a disinhibitory process. A possible neural pathway for the shift response is presented.
INTRODUCTION The shift response ( p e r i p h e r y effect, Mcllwain effect) is a transient excitation of the retinal ganglion cells commonly elicited by a r a p i d displacement of a large target situated far away (greater than 15 °) from the conventional R F (receptive field). If the same target is flashed, a transient excitation occurs at light on and light off 10. The neural p a t h w a y involved in the transmission of the shift signals laterally in the retina has not been established although most investigators agree that the amacrine cells r a t h e r than the horizontal cells are the signal carriers. Of the different types of amacrine cells, I k e d a and Wright14 and D e r r i n g t o n et al. 6 suggested that the transient amacrine cells form part of the p a t h w a y because of the similarity in the properties of the shift response to those of the intracellularly d e t e r m i n e d p r o p e r t i e s of the transient amacrine cells. Kriiger 19 has p r o p o s e d that the laterally conducting n e t w o r k is fed by the bipolar cells which, in turn, feeds the ganglion cells. H e suggested that the ganglion cells receive input from O N - c e n t e r and O F F - c e n t e r bipolar cells via amacrine cells. In
this arrangement, the sustained portion of the bipolar cells' response would cancel to leave only the transient ON and O F F signals. The purpose of this study was to examine in detail the effect of a steady stimulus placed either on the center or surround of the R F (receptive field) on the amplitude of the shift response. The shift response was elicited by a large grating situated 15 ° from the R F C of the retinal ganglion cells of cats. W e compared the a m p l i t u d e of the shift response elicited with and without an enhancing stimulus, and we shall show that the amplitude of the shift response is directly p r o p o r t i o n a l to the strength of the enhancing stimulus. The p r o p e r t i e s of the shift response suggest that it results from a disinhibitory process, and the O N - O F F transient amacrine cells form a necessary element in the pathway. A possible pathway for the shift response is presented. MATERIALS AND METHODS
Animals The data were collected from 6 n o r m a l adult cats.
Correspondence: D. I. Hamasaki, William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, Dept. of Ophthalmology, University of Miami, Miami, FL 33101, U.S.A. 0006-8993/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
52 The animals were initially anesthetized with ketamine (20 mg/kg) and a catheter was placed in the femoral vein. A n e s t h e s i a was m a i n t a i n e d during surgery by intravenous u r e t h a n e (250 mg/kg), and during the experiments by including u r e t h a n e (15 mg/kg/h) in the infusion. In addition, the animals b r e a t h e d 75% N 2 0 and 25% O 2 during the 36 h of the experiment. Flaxedil (17 mg/kg/h) was infused continuously to reduce eye movements. The femoral arterial pressure was continuously m o n i t o r e d . Rectal t e m p e r a ture was m a i n t a i n e d at 3 8 - 3 9 °C, and the end tidal CO 2 at 4.0%. The pupils were dilated with topical atropine and the nictitating m e m b r a n e retracted with neosynephrine. The corneas were p r o t e c t e d from drying by contact lenses which had 4 m m d i a m e t e r artificial pupils. Lenses were used to m a k e the retina conjugate with a translucent screen situated 114 cm in front of the cat. The stimuli were p r e s e n t e d on this screen. The fundus of each eye was p r o j e c t e d back onto the screen 19, and the position of the optic disc and the area centralis (blood vessel free zone) were m a r k e d on the screen. A l l of the R F were t a b u l a t e d with reference to the area centralis.
50 ° vertically and 77 ° horizontally at the cat's e y O 0. Square waves from a waveform g e n e r a t o r were used to move a m i r r o r attached to a g a l v a n o m e t e r 1/2 cycle to elicit the shift response. The luminance of the bright bars was 61.0 cd/m z and that of the dark bars 15.3 cd/m z. The b a c k g r o u n d was diffusely lit from the front and had a luminance of 15.3 cd/m 2.
Procedure A f t e r a unit was isolated, the b o r d e r s of the R F C were plotted and m a r k e d on the screen. The eye of origin was d e t e r m i n e d , and the cell was classified as an X- or Y-cell by the contrast reversal stimulus n. A 30 ° o p a q u e mask was then c e n t e r e d on the R F which blocked the grating over this region. A 2 ° hole was
Recordings Insulex-coated tungsten m i c r o e l e c t r o d e s were used to isolate single optic tract fibers. The nerve impulses triggered artificial spikes which were fed to a Texas I n s t r u m e n t $960A m i n i c o m p u t e r p r o g r a m m e d to give average response histograms (25 sums, 10 ms bins).
Stimulus The light for the stimulus was o b t a i n e d from 15 V 10 A quartz iodide lamps. Two i n d e p e n d e n t channels were used. The first channel was used to classify the cells into X- and Y-cells by a contrast reversal stimulus 11, and to stimulate the R F center. The unatt e n u a t e d luminance of this b e a m was 61.0 cd/m 2, and neutral density filters were used to reduce the stimulus intensity in 0.5 log unit steps. A shutter, controlled by a pulse g e n e r a t o r , was used to deliver 1 s duration flashes at 3.2 s intervals. The second channel was used to elicit the shift response. A grating (0.14 cycles/deg) was p r o j e c t e d to cover the entire translucent screen which s u b t e n d e d
l ! Fig. 1. Average response histograms recorded from an ONcenter X-cell. All of the responses were elicited by the full intensity grating stimulus. The response recorded without the enhancing spot is shown as the upper histogram, and the responses elicited with different intensity enhancing spot are shown below. The numbers represent the value of the neutral density filters used to attenuate the full intensity spot (0 = 61.0 cd/m2). Calibration, 40 spikes/s and 60 ms.
53 cut in the center of this mask so that the center of the RF could be stimulated. A small disc was placed in the optical path to occlude the grating from the hole. The mask was taped to the screen on the side opposite the cat. After the mask had been positioned, the responses elicited by shifting the grating were recorded. It is known that the amplitude of the shift response can be enhanced by placing a steady target of appropriate contrast on the RFC, i.e., a bright spot for ON-center cells or a dark spot for OFF-center cells. If the target used enhances the shift response, we have termed this as appropriate stimulation and if it depresses the shift response, as inappropriate stimulation of the RF. RESULTS
Enhancement of the shift response by an appropriate stimulus It has been reported that placing an appropriate stimulus (bright spot for ON-center cells) on the R F C will increase the amplitude of the shift response6, 7. During the course of earlier studies 10, it was noted that the degree of enhancement was related to the intensity of the enhancing stimulus. Inasmuch as this
relationship has important bearing on the neural circuitry for the shift response, we examined this relationship in more detail. Examples of the effect of the intensity of the enhancing stimulus on the shift response are shown for an ON-center X-cell in Fig. 1, and for an ON-center Y-cell in Fig. 2. The shift response elicited by the grating without an enhancing stimulus is shown as the histogram without a number. In both X- and Y-cells, the shift response consists of a transient excitation which occurs with each shift of the grating. After the control responses were recorded, a small spot of light was placed on the RFC. The intensity of the spot was increased in 0.5 log unit steps by neutral density filters, and the numbers next to the histograms represent the value of the filter used to attenuate the full intensity enhancing stimulus (0 -- 61.0 cd/m2). The same grating stimulus was used to elicit all of the responses. When the full intensity spot of light was reduced by 1.5 log units, the shift response of the X-cell was similar to that without an enhancing stimulus. A n increase in the intensity by 0.5 log units resulted in a slight increase in the shift response, and with further increases in the intensity of the enhancing stimulus, there were further increases in the amplitude of the
Fig. 2. Average response histograms recorded from an ON-center Y-cell. All of the responses were elicited by the full intensity grating. The response recorded without an enhancing spot is shown as the upper left histogram, and the responses recorded with the different intensity enhancing spot are shown below. The numbers represent the value of the neutral density filter used to attenuate the full intensity spot. Calibration, 40 spikes/s and 60 ms.
54 shift response. The largest shift response was ob-
O F F - c e n t e r cells. The control response elicited from
tained with the full intensity enhancing spot.
an O F F - c e n t e r Y-cell without the enhancing a n n u l u s
Although the amplitude of the shift response was
is shown in the upper histogram in Fig. 3. With the
significantly larger for the Y-cell, the pattern of change in the amp!i!ude of the shift response with increasing intensity of the enhancing stimulus was similar to that seen in the X-cell. For O F F - c e n t e r cells, either a dark spot in the R F C or a bright annulus in the surround of the R F led to an e n h a n c e m e n t of the shift response. Because it was easier to alter the intensity of a bright annulus, we used an annulus to study the effect of the intensity of the enhancing stimulus on the shift response of
1.5
O Fig. 3. Average response histograms recorded from an OFFcenter Y-cell. All of the responses were elicited by the same grating stimulus. The response recorded without the enhancing annulus is shown at the top, and the responses elicited with different intensities of the steady enhancing annulus are shown below. The numbers represent the value of the neutral density filters used to attenuate the full intensity annulus (0 = 61.0 cd/m2). Calibration, 40 spikes/s and 60 ms.
Fig. 4. Average response histograms recorded from an ONcenter Y-cell. All of the responses were elicited by the same grating stimulus. The response recorded without the annulus is shown at the top, and the responses elicited with increasing intensities of the steady annulus are shown below. The numbers represent the value of the neutral density filters used to attenuate the full intensity annulus (0 = 61 cd/mZ). Calibration, 40 spikes/s and 60 ms.
55 steady full intensity annulus (0), there was a strong enhancement of the shift response elicited by the same grating stimulus. As the intensity of the annulus is reduced, there was a reduction in the shift response. However, with the annulus reduced by 1.0 and 1.5 log units, the shift response was smaller than that elicited with no annulus present. This reduction in the shift response probably arises from the stimulation of the center mechanism by the dim annulus as there is strong evidence that the borders of the center and surround mechanisms are coextensive in Ycells]2,13. The results of these experiments demonstrate not only the influence of the enhancing stimulus but also show that similar effects can be obtained by placing an appropriate stimulus on the center or the surround of the RF.
Effect of an inappropriate stimulus on the shift response The question then arose whether an inappropriate stimulus would reduce or block the shift response, and whether it would do this in a graded manner. The shift responses recorded from an ON-center Y-cell are shown in Fig. 4 with the response recorded without any enhancing stimulus shown in the upper histogram. With increasing intensity of the annulus, there is a progressive decrease in the amplitude of the shift response. With the full intensity annulus, the same grating which gave a strong shift response elicits only a weak shift response. Similar findings were made for X-cells. Relationship between the enhancing stimulus and the amplitude o f the shift response To examine the relationship between the enhancing stimulus and the shift response, we plotted the amplitude of the shift response (see below) as a func-
140
60
X-cells
a~ Q.
,,I 40
?oo
/ .e" / ,;
2o
. _A/
e-.:.:..'. E <
o
I 1.5
I 1.0
,,i "'*"
.,"/""..o ":"'~ ' n .¢,"
.," .;
, ~-'" /
/"
e~
zx
~,~:...._,m" .,,.-" ~
/
Y-cells
60
.."
,.! A-" .." E : ~ 4 J~ • "" "
L.~ .d
.d
....
I 0.5
I 0
I 1.5
I 1.0
0.5
I 0
- L o g intensity of enhancing stimulus
Fig. 5. Effect of the intensity of the enhancing stimulus on the amplitude of the shift response of 5 X- and 5 Y-cells. The intensity of the steady stimulus is shown on the abscissa and the amplitude of the shift response on the ordinate. All of the responses were elicited by the same grating stimulus. Four of the X-cells and one of the Y-cells were ON-center cells.
lion of the intensity of the enhancing stimulus for 5 Xcells (left) and for 5 Y-cells (right) in Fig. 5. Four of the X-cells and one of the Y-cells were ON-center cells. For both types of cells, there was an increase in the amplitude of the shift response with increasing intensities of the enhancing stimulus.
Coefficient o f correlation between the enhancing stimulus and the amplitude of the shift response We then asked two questions: how strong is the relationship between the shift response and the enhancing stimulus; and what aspect of the enhancing stimulus is related to the increase in the amplitude of the shift response? Initially, we considered the increased level of maintained activity induced by the enhancing stimulus. A coefficient of correlation was determined between the amplitude of the shift response and the lev-
TABLE I
Coefficient of correlation between the amplitude of the shift response and the maintained activity, the transient peak firing rate, the sustained firing rate and the inhibitory activity
ONX OFF X ON Y OFF Y * P < 0.01.
No. of units
No. of points
Maintained activity
Peak firing
Sustained firing
Inhibitory activity
20 8 7 10
81 31 35 36
0.02 0.23 0.11 0.18
0.41" 0.58* 0.50* 0.64*
0.48* 0.34 0.29 0.30
0.38* 0.60* 0.51" 0.66*
56 el of maintained activity. The r-values, obtained for the 4 different types of units, were found to be low and not significant. (Table I). To determine which aspect of the enhancing stimulus was related to the enhancing effect, we recorded the shift response with an enhancing stimulus, and then recorded the center response elicited by flashing the enhancing stimulus with a duration of i s. For the shift response, we measured the firing rate during a 50 ms period when it was largest and subtracted the maintained firing level. From the response elicited by flashing the enhancing stimulus, we measured the peak firing rate (maximum-maintained firing rate), and the sustained firing rate (mean firing during the last 50 ms of the 1 s stimulus). A coefficient of correlation was calculated between each of these values and the amplitude of the shift response (Table I). For the sustained firing level, the only significant correlation was found for the ON X-cells (P < 0.01). For the transient peak firing rate, on the other hand, the coefficient of correlation was significant (P < 0.01) for all types of units. The r-value was highest for the OFF Y-cells and lowest for the ON X-cells. For these calculations, all of the units from the different cats were combined. When the coefficient of correlation was determined for individual units, the values were considerably higher. For example, for 3 ON Y-cells, r = 0.78 (n = 7; 0.05 > P > 0.01), 0.77 (n = 8; P < 0.01) and 0.99 (n = 7; P < 0.01). Similar calculations were made for two ON X-cells, and r = 0.51 (n = 9; n.s.) and 0.93 (n = 6; P < 0.01). These findings show that the coefficient of correlations may be stronger than that found when all of the units were combined. When the RFC is stimulated, the transient peak firing is followed by a rapid decrease of the firing rate to the sustained firing level. This decrease is probably due to the activation of an inhibitory process 22-24. To obtain a measure of this inhibitory activity, we calculated the difference between the transient peak and the sustained firing level. A coefficient of correlation was then determined between this value and the amplitude of the shift response. The results showed that the r-values were high, and comparable to that obtained for the transient peak (Table I). These findings demonstrate that the enhancing effect induced was correlated with the signal evoked by the stimulus when it was first turned on, and this tonic
A C+)
(÷)
.......... -
_Q__ ÷
(÷)
b m
Fig. 6. Suggested neural pathway for the shift response of an OFF-center GC. The inner plexiform layer (IPL) is divided into sublaminas a and b. ON-OFF AC, ON-OFF amacrine cell; hCB, hyperpolarizing cone bipolar cell; dCB, depolarizing cone bipolar cell; GC, ganglion cell.
signal is present for the duration of the enhancing stimulus. DISCUSSION The main observations of this study are: (1) appropriate stimulation of the RFC or the RF surround will increase and inappropriate stimulation will decrease the amplitude of the shift response; (2) the shift response is graded and is strongly correlated with the capacity of the enhancing stimulus to activate the RF; (3) a shift stimulus which elicits a strong shift response can be blocked by strong inappropriate stimulation; and (4) there is no correlation between the ongoing activity (maintained activity) and the amplitude of the shift response. These results then suggest that the shift signal is modulating a tonic signal which is present in the RF. When this signal is strong then the shift response will be large, and when the signal is weak (as during inappropriate stimulation) the shift response will also be weak. The tonic signal is not the on-going activity as this was found to be poorly correlated with the ampli-
57 tude of the shift response. Because of the strong correlation between the amplitude of the shift response and the maximum firing rate, and also with the inhibitory activity following the maximum firing rate, it is more likely that some aspect of this part of the signal is present in the RF. How then does the shift stimulus elicit a transient excitation? The O N - O F F response characteristic of the amacrine cells in lower vertebrates3,15, 25 makes them a likely candidate to form one element of the neural pathway that carries the shift signal from the peripheral regions to the RFC 6,14. Intracellular recordings from the O N - O F F amacrine cells in the cat have now been accomplished, and they have been found to be monostratified in sublamina a with widespread dendrites 17. They have been identified as the A19 amacrine cell 16. A19 amacrine cells are coupled to other A19 amacrine cells by gap junctions, and make conventional synapses only on other amacrine cells. They receive input from other amacrine cells, and also from only the hyperpolarizing bipolar cells. Uptake studies of G A B A 8 and muscimo121 indicate that the A19 amacrine cells use G A B A as their neurotransmitter. Because the output of this O N - O F F amacrine cell in the cat is not onto the ganglion cells, the excitatory shift response must arise from its activity at an earlier stage. Because G A B A is, in all likelihood, an inhibitory transmitter, we suggest that the excitatory shift response arises from a disinhibitory process. To accomplish this, we hypothesize that the A19 amacrine cell inhibits a sustained type of amacrine cell which is present in the RFC. A possible neural pathway for the shift response is diagrammed in Fig. 6. The grating stimulates a large number of photoreceptors in the periphery which leads to hyperpolarization of some bipolar cells (hCB) and depolarization of other bipolar cells (dCB). Because the dendrites of the A19 amacrine cell make synapses only in sublamina a, the input to the A19 cell from the dCB must arise from other amacrine cells. In this arrangement, there would be relative depolarization of the A19 amacrine cell at light on (dCB) and off (hCB) when the grating is flashed or is shifted. The shift signal is carried laterally to the RFC by A19 amacrine cells where they make contact with the sustained amacrine cells. These amacrine cells re-
ceive input from the bipolar cells which serve the center mechanism. The output of these amacrine cells is inhibitory, and is either fed back to the bipolar cell or fed forward to the ganglion cell (not shown). Because these amacrine cells receive input from the bipolar cells, the degree of inhibitory activity will be directly related to the activity of the bipolar cell. This pathway can then account for the relationship between the excitatory shift response and the strength of the tonic signal present in the RF. If the activation of the RF is strong, then the activity or inhibitory tonic signal of the sustained amacrine cell will also be strong. The inhibition of this amacrine cell by the shift signal, i.e., a disinhibition, will lead to a strong shift response. In the same manner, inappropriate stimulation of the RF will lead to a weak inhibitory signal by the sustained amacrine cell, and the inhibition of this weak signal by the shift signal can only give a weak shift response. The interposition of the transient amacrine cell as one of the element in the pathway not only accounts for the O N - O F F behavior of the shift response, but also the unusual intensive property of the shift response. Previous studies have shown that the intensity range over which the shift response is graded is very limited so that a small change in contrast of the target can change a threshold response to a saturated response1. 7. Intracellular recordings from the transient amacrine cells of lower vertebrates have shown that these cells respond to a light flash with transient depolarization at light on and light off3,5,15,2s. In the catfish, these are the only retinal cells which always give frequency doubling responses 3. Chan and Naka 3 also reported that the power level of the response was independent of the power level of the input signal over a wide range of stimulus intensities as had been reported earlier by Kaneko and Hashimoto 15. Additional support for the inclusion of the O N OFF amacrine cell in this neural pathway comes from the study of Frishman and Linsenmeier 9 who showed that infusion of picrotoxin, a G A B A antagonist, reduced the amplitude of the shift response in Y-cells of the cat. This pathway, however, does not account for the differences in the shift response of the X- and Y-cells. The Y-cells have stronger shift responses with shorter latencies. In addition, Y-type ganglion cells receive input mainly from amacrine cells while X-type
58 ganglion cells receive input mainly from bipolar cells. The modifications of the proposed pathway to account for these differences in X- and Y-cells must await further experiments.
ACKNOWLEDGEMENTS
REFERENCES
13 Hickey, T. L., Winters, R. W. and Pollack, J. G., Centersurround interaction in two types of on-center retinal ganglion cells in the cat retina, Vision Res., 13 (1973) 1511-1526. 14 Ikeda, H. and Wright, M. J., Functional organization of the periphery effect in retinal ganglion cells, Vision Res., 12 (1972) 1857-1879. 15 Kaneko, A. and Hashimoto, H., Electrophysiological study of single neurons in the inner nuclear layer of the carp retina, Vision Res., 9 (1969) 37-55. 16 Kolb, H. R., Nelson, R. and Mariani, A., Amacrine cells, bipolar cells and ganglion cells of the cat retina. A Golgi study, Vision Res., 21 (1982) 1081-1114. 17 Kolb, H. and Nelson, R., An O N - O F F amacrine cell type in the cat retina, Soc. Neurosci Abstr., 9 (1983) 806. 18 Kriiger, J., The shift-effect enhances X- and suppresses Ytype response characteristics of cat retinal ganglion cells, Brain Research, 201 (1980) 71-84. 19 KriJger, J., The difference between X- and Y-type responses in ganglion cells of the cat's retina, Vision Res., 21 (1981) 1685-1687. 20 Pettigrew, J. D., Cooper, M. L. and Blasdel, G. G., Improved use of tapetal reflection for eye-position monitoring, Invest. Ophthalmol. Vis. Sci., 18 (1979) 490. 21 Pourcho, R. G., A Golgi and autoradiographic study of (3H) muscimol-labeled cells in the cat retina, Soc. Neurosci. Abstr., 8 (1982) 47. 22 Rodieck, R. W., The Vertebrate Retina, W. H. Freeman, San Francisco, 1973. 23 Saito, H. A. and Fukada, Y., Gain control mechanisms within the receptive field center of cat's retinal ganglion cells, Vision Res., 15 (1975) 1407-1410, 24 Stein, A., Millikin, W. and Stevens, J., The spatiotemporal building blocks of X-, Y- and W-ganglion cell receptive fields of the cat's retina, Exp. Brain Res., 49 (1983) 341-352. 25 Werblin, F. S. and Dowling, J. E., Organization of the retina of the mudpuppy, Necturus rnaculosus. II. Intracellular recording, J. Neurophysiol., 32 (1969) 339-355.
1 Barlow, H. B., Derrington, A. M., Harris, L. R. and Lenhie, P., The effects of remote retinal stimulation on the responses of cat retinal ganglion cells, J. Physiol. (Lond.), 269 (1977) 177-194. 2 Belgum, J. H., Dvorak, D. R. and McReynolds, J. S., Light-evoked sustained inhibition in mudpuppy retinal ganglion cells, Vision Res., 22 (1982) 257-260. 3 Chan, R. Y. and Naka, K., The amacrine cell, Vision Res., 16 (1976) 1119-1129. 4 Cleland, B. G., Levick, W. R. and Sanderson, K. J., Properties of sustained and transient ganglion cells in the cat retina, J. Physiol. (Lond.), 228 (1973) 649-680. 5 Dacheux, R. F. and Miller, R. F., An intracellular electrophysiological study of the ontogeny of functional synapses in the rabbit retina. II. Amacrine cells, J. comp. Neurol., 198 (1976) 327-334. 6 Derrington, A. M., Lennie, P. and Wright, M. J., The mechanism of peripherally evoked responses in retinal ganglion cells, J. Physiol. (Lond.), 289 (1981) 299-310. 7 Fischer, B., Kriiger, J. and Droll, W., Quantitative aspects of the shift-effect in cat retinal ganglion cells, Brain Research, 83 (1979) 391-403. 8 Freed, M. A., Nakamura, Y. and Sterling, P., Four types of amacrine in the cat retina that accumulate GABA, J. cornp. Neurol., 219 (1983) 295-304. 9 Frishman, L. J. and Linsenmeier, R. A., Effects of picrotoxin and strychnine on nonlinear responses of Y-type cat retinal ganglion cells, J. Physiol. (Lond.), 324 (1982) 347-363. 10 Hamasaki, D. I. and Hanada, I., A comparison of the shift response of X- and Y-cells in the cat's retina, Exp. Brain Res., 50 (1983) 117-124. 11 Hamasaki, D. I. and Sutija, V. G., Classification of cat retinal ganglion cells into X- and Y-cells with a contrast reversal stimulus, Exp. Brain Res., 35 (1979) 25-36. 12 Hammond, P., Receptive field mechanisms of sustained and transient retinal ganglion cells in cat, Exp. Brain Res., 23 (1975) 113-128.
We wish to thank Mr. O. Novarro for his excellent help with the experiments. Supported in part by a Public Health Service Grant EY-00376 from the National Eye Institute, National Institutes of Health, Bethesda, MD. G.W.M. was supported by an Ophthalmology Specialty Training Grant No. EYO 7021-06.
51
BRE 10800
A Neural Pathway for the Shift Response in the Cat D. I. HAMASAKI and GREGORY W. MAGUIRE
William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, Department of Ophthalmology, University of Miami, Miami, FL 33101 (U.S.A.) (Accepted September llth, 1984)
Key words: shift response - - cat retinal ganglion cell - - neural pathway - - Mcllwain effect
The shift response (Mcllwain effect) was elicited by moving a large grating situated greater than 15° from the conventional receptive field center (RFC). We examined the change in the amplitude of the shift response induced by placing a steady target on the RFC or the RF surround. We found that appropriate stimulation of the RFC or the RF surround will increase and inappropriate stimulation will decrease the amplitude of the shift response in a graded manner. The amplitude of the shift response was not correlated with the maintained activity but was correlated with the transient peak firing rate which is evoked by flashing the enhancing stimulus. A shift stimulus which elicits a strong shift response can be blocked by inappropriate stimulation. The results suggest that the shift signal is modulating a tonic signal which is present in the RF. Because the shift response is a transient excitation, we suggest that the shift response results from a disinhibitory process. A possible neural pathway for the shift response is presented.
INTRODUCTION The shift response ( p e r i p h e r y effect, Mcllwain effect) is a transient excitation of the retinal ganglion cells commonly elicited by a r a p i d displacement of a large target situated far away (greater than 15 °) from the conventional R F (receptive field). If the same target is flashed, a transient excitation occurs at light on and light off 10. The neural p a t h w a y involved in the transmission of the shift signals laterally in the retina has not been established although most investigators agree that the amacrine cells r a t h e r than the horizontal cells are the signal carriers. Of the different types of amacrine cells, I k e d a and Wright14 and D e r r i n g t o n et al. 6 suggested that the transient amacrine cells form part of the p a t h w a y because of the similarity in the properties of the shift response to those of the intracellularly d e t e r m i n e d p r o p e r t i e s of the transient amacrine cells. Kriiger 19 has p r o p o s e d that the laterally conducting n e t w o r k is fed by the bipolar cells which, in turn, feeds the ganglion cells. H e suggested that the ganglion cells receive input from O N - c e n t e r and O F F - c e n t e r bipolar cells via amacrine cells. In
this arrangement, the sustained portion of the bipolar cells' response would cancel to leave only the transient ON and O F F signals. The purpose of this study was to examine in detail the effect of a steady stimulus placed either on the center or surround of the R F (receptive field) on the amplitude of the shift response. The shift response was elicited by a large grating situated 15 ° from the R F C of the retinal ganglion cells of cats. W e compared the a m p l i t u d e of the shift response elicited with and without an enhancing stimulus, and we shall show that the amplitude of the shift response is directly p r o p o r t i o n a l to the strength of the enhancing stimulus. The p r o p e r t i e s of the shift response suggest that it results from a disinhibitory process, and the O N - O F F transient amacrine cells form a necessary element in the pathway. A possible pathway for the shift response is presented. MATERIALS AND METHODS
Animals The data were collected from 6 n o r m a l adult cats.
Correspondence: D. I. Hamasaki, William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, Dept. of Ophthalmology, University of Miami, Miami, FL 33101, U.S.A. 0006-8993/85/$03.30 © 1985 Elsevier Science Publishers B.V. (Biomedical Division)
52 The animals were initially anesthetized with ketamine (20 mg/kg) and a catheter was placed in the femoral vein. A n e s t h e s i a was m a i n t a i n e d during surgery by intravenous u r e t h a n e (250 mg/kg), and during the experiments by including u r e t h a n e (15 mg/kg/h) in the infusion. In addition, the animals b r e a t h e d 75% N 2 0 and 25% O 2 during the 36 h of the experiment. Flaxedil (17 mg/kg/h) was infused continuously to reduce eye movements. The femoral arterial pressure was continuously m o n i t o r e d . Rectal t e m p e r a ture was m a i n t a i n e d at 3 8 - 3 9 °C, and the end tidal CO 2 at 4.0%. The pupils were dilated with topical atropine and the nictitating m e m b r a n e retracted with neosynephrine. The corneas were p r o t e c t e d from drying by contact lenses which had 4 m m d i a m e t e r artificial pupils. Lenses were used to m a k e the retina conjugate with a translucent screen situated 114 cm in front of the cat. The stimuli were p r e s e n t e d on this screen. The fundus of each eye was p r o j e c t e d back onto the screen 19, and the position of the optic disc and the area centralis (blood vessel free zone) were m a r k e d on the screen. A l l of the R F were t a b u l a t e d with reference to the area centralis.
50 ° vertically and 77 ° horizontally at the cat's e y O 0. Square waves from a waveform g e n e r a t o r were used to move a m i r r o r attached to a g a l v a n o m e t e r 1/2 cycle to elicit the shift response. The luminance of the bright bars was 61.0 cd/m z and that of the dark bars 15.3 cd/m z. The b a c k g r o u n d was diffusely lit from the front and had a luminance of 15.3 cd/m 2.
Procedure A f t e r a unit was isolated, the b o r d e r s of the R F C were plotted and m a r k e d on the screen. The eye of origin was d e t e r m i n e d , and the cell was classified as an X- or Y-cell by the contrast reversal stimulus n. A 30 ° o p a q u e mask was then c e n t e r e d on the R F which blocked the grating over this region. A 2 ° hole was
Recordings Insulex-coated tungsten m i c r o e l e c t r o d e s were used to isolate single optic tract fibers. The nerve impulses triggered artificial spikes which were fed to a Texas I n s t r u m e n t $960A m i n i c o m p u t e r p r o g r a m m e d to give average response histograms (25 sums, 10 ms bins).
Stimulus The light for the stimulus was o b t a i n e d from 15 V 10 A quartz iodide lamps. Two i n d e p e n d e n t channels were used. The first channel was used to classify the cells into X- and Y-cells by a contrast reversal stimulus 11, and to stimulate the R F center. The unatt e n u a t e d luminance of this b e a m was 61.0 cd/m 2, and neutral density filters were used to reduce the stimulus intensity in 0.5 log unit steps. A shutter, controlled by a pulse g e n e r a t o r , was used to deliver 1 s duration flashes at 3.2 s intervals. The second channel was used to elicit the shift response. A grating (0.14 cycles/deg) was p r o j e c t e d to cover the entire translucent screen which s u b t e n d e d
l ! Fig. 1. Average response histograms recorded from an ONcenter X-cell. All of the responses were elicited by the full intensity grating stimulus. The response recorded without the enhancing spot is shown as the upper histogram, and the responses elicited with different intensity enhancing spot are shown below. The numbers represent the value of the neutral density filters used to attenuate the full intensity spot (0 = 61.0 cd/m2). Calibration, 40 spikes/s and 60 ms.
53 cut in the center of this mask so that the center of the RF could be stimulated. A small disc was placed in the optical path to occlude the grating from the hole. The mask was taped to the screen on the side opposite the cat. After the mask had been positioned, the responses elicited by shifting the grating were recorded. It is known that the amplitude of the shift response can be enhanced by placing a steady target of appropriate contrast on the RFC, i.e., a bright spot for ON-center cells or a dark spot for OFF-center cells. If the target used enhances the shift response, we have termed this as appropriate stimulation and if it depresses the shift response, as inappropriate stimulation of the RF. RESULTS
Enhancement of the shift response by an appropriate stimulus It has been reported that placing an appropriate stimulus (bright spot for ON-center cells) on the R F C will increase the amplitude of the shift response6, 7. During the course of earlier studies 10, it was noted that the degree of enhancement was related to the intensity of the enhancing stimulus. Inasmuch as this
relationship has important bearing on the neural circuitry for the shift response, we examined this relationship in more detail. Examples of the effect of the intensity of the enhancing stimulus on the shift response are shown for an ON-center X-cell in Fig. 1, and for an ON-center Y-cell in Fig. 2. The shift response elicited by the grating without an enhancing stimulus is shown as the histogram without a number. In both X- and Y-cells, the shift response consists of a transient excitation which occurs with each shift of the grating. After the control responses were recorded, a small spot of light was placed on the RFC. The intensity of the spot was increased in 0.5 log unit steps by neutral density filters, and the numbers next to the histograms represent the value of the filter used to attenuate the full intensity enhancing stimulus (0 -- 61.0 cd/m2). The same grating stimulus was used to elicit all of the responses. When the full intensity spot of light was reduced by 1.5 log units, the shift response of the X-cell was similar to that without an enhancing stimulus. A n increase in the intensity by 0.5 log units resulted in a slight increase in the shift response, and with further increases in the intensity of the enhancing stimulus, there were further increases in the amplitude of the
Fig. 2. Average response histograms recorded from an ON-center Y-cell. All of the responses were elicited by the full intensity grating. The response recorded without an enhancing spot is shown as the upper left histogram, and the responses recorded with the different intensity enhancing spot are shown below. The numbers represent the value of the neutral density filter used to attenuate the full intensity spot. Calibration, 40 spikes/s and 60 ms.
54 shift response. The largest shift response was ob-
O F F - c e n t e r cells. The control response elicited from
tained with the full intensity enhancing spot.
an O F F - c e n t e r Y-cell without the enhancing a n n u l u s
Although the amplitude of the shift response was
is shown in the upper histogram in Fig. 3. With the
significantly larger for the Y-cell, the pattern of change in the amp!i!ude of the shift response with increasing intensity of the enhancing stimulus was similar to that seen in the X-cell. For O F F - c e n t e r cells, either a dark spot in the R F C or a bright annulus in the surround of the R F led to an e n h a n c e m e n t of the shift response. Because it was easier to alter the intensity of a bright annulus, we used an annulus to study the effect of the intensity of the enhancing stimulus on the shift response of
1.5
O Fig. 3. Average response histograms recorded from an OFFcenter Y-cell. All of the responses were elicited by the same grating stimulus. The response recorded without the enhancing annulus is shown at the top, and the responses elicited with different intensities of the steady enhancing annulus are shown below. The numbers represent the value of the neutral density filters used to attenuate the full intensity annulus (0 = 61.0 cd/m2). Calibration, 40 spikes/s and 60 ms.
Fig. 4. Average response histograms recorded from an ONcenter Y-cell. All of the responses were elicited by the same grating stimulus. The response recorded without the annulus is shown at the top, and the responses elicited with increasing intensities of the steady annulus are shown below. The numbers represent the value of the neutral density filters used to attenuate the full intensity annulus (0 = 61 cd/mZ). Calibration, 40 spikes/s and 60 ms.
55 steady full intensity annulus (0), there was a strong enhancement of the shift response elicited by the same grating stimulus. As the intensity of the annulus is reduced, there was a reduction in the shift response. However, with the annulus reduced by 1.0 and 1.5 log units, the shift response was smaller than that elicited with no annulus present. This reduction in the shift response probably arises from the stimulation of the center mechanism by the dim annulus as there is strong evidence that the borders of the center and surround mechanisms are coextensive in Ycells]2,13. The results of these experiments demonstrate not only the influence of the enhancing stimulus but also show that similar effects can be obtained by placing an appropriate stimulus on the center or the surround of the RF.
Effect of an inappropriate stimulus on the shift response The question then arose whether an inappropriate stimulus would reduce or block the shift response, and whether it would do this in a graded manner. The shift responses recorded from an ON-center Y-cell are shown in Fig. 4 with the response recorded without any enhancing stimulus shown in the upper histogram. With increasing intensity of the annulus, there is a progressive decrease in the amplitude of the shift response. With the full intensity annulus, the same grating which gave a strong shift response elicits only a weak shift response. Similar findings were made for X-cells. Relationship between the enhancing stimulus and the amplitude o f the shift response To examine the relationship between the enhancing stimulus and the shift response, we plotted the amplitude of the shift response (see below) as a func-
140
60
X-cells
a~ Q.
,,I 40
?oo
/ .e" / ,;
2o
. _A/
e-.:.:..'. E <
o
I 1.5
I 1.0
,,i "'*"
.,"/""..o ":"'~ ' n .¢,"
.," .;
, ~-'" /
/"
e~
zx
~,~:...._,m" .,,.-" ~
/
Y-cells
60
.."
,.! A-" .." E : ~ 4 J~ • "" "
L.~ .d
.d
....
I 0.5
I 0
I 1.5
I 1.0
0.5
I 0
- L o g intensity of enhancing stimulus
Fig. 5. Effect of the intensity of the enhancing stimulus on the amplitude of the shift response of 5 X- and 5 Y-cells. The intensity of the steady stimulus is shown on the abscissa and the amplitude of the shift response on the ordinate. All of the responses were elicited by the same grating stimulus. Four of the X-cells and one of the Y-cells were ON-center cells.
lion of the intensity of the enhancing stimulus for 5 Xcells (left) and for 5 Y-cells (right) in Fig. 5. Four of the X-cells and one of the Y-cells were ON-center cells. For both types of cells, there was an increase in the amplitude of the shift response with increasing intensities of the enhancing stimulus.
Coefficient o f correlation between the enhancing stimulus and the amplitude of the shift response We then asked two questions: how strong is the relationship between the shift response and the enhancing stimulus; and what aspect of the enhancing stimulus is related to the increase in the amplitude of the shift response? Initially, we considered the increased level of maintained activity induced by the enhancing stimulus. A coefficient of correlation was determined between the amplitude of the shift response and the lev-
TABLE I
Coefficient of correlation between the amplitude of the shift response and the maintained activity, the transient peak firing rate, the sustained firing rate and the inhibitory activity
ONX OFF X ON Y OFF Y * P < 0.01.
No. of units
No. of points
Maintained activity
Peak firing
Sustained firing
Inhibitory activity
20 8 7 10
81 31 35 36
0.02 0.23 0.11 0.18
0.41" 0.58* 0.50* 0.64*
0.48* 0.34 0.29 0.30
0.38* 0.60* 0.51" 0.66*
56 el of maintained activity. The r-values, obtained for the 4 different types of units, were found to be low and not significant. (Table I). To determine which aspect of the enhancing stimulus was related to the enhancing effect, we recorded the shift response with an enhancing stimulus, and then recorded the center response elicited by flashing the enhancing stimulus with a duration of i s. For the shift response, we measured the firing rate during a 50 ms period when it was largest and subtracted the maintained firing level. From the response elicited by flashing the enhancing stimulus, we measured the peak firing rate (maximum-maintained firing rate), and the sustained firing rate (mean firing during the last 50 ms of the 1 s stimulus). A coefficient of correlation was calculated between each of these values and the amplitude of the shift response (Table I). For the sustained firing level, the only significant correlation was found for the ON X-cells (P < 0.01). For the transient peak firing rate, on the other hand, the coefficient of correlation was significant (P < 0.01) for all types of units. The r-value was highest for the OFF Y-cells and lowest for the ON X-cells. For these calculations, all of the units from the different cats were combined. When the coefficient of correlation was determined for individual units, the values were considerably higher. For example, for 3 ON Y-cells, r = 0.78 (n = 7; 0.05 > P > 0.01), 0.77 (n = 8; P < 0.01) and 0.99 (n = 7; P < 0.01). Similar calculations were made for two ON X-cells, and r = 0.51 (n = 9; n.s.) and 0.93 (n = 6; P < 0.01). These findings show that the coefficient of correlations may be stronger than that found when all of the units were combined. When the RFC is stimulated, the transient peak firing is followed by a rapid decrease of the firing rate to the sustained firing level. This decrease is probably due to the activation of an inhibitory process 22-24. To obtain a measure of this inhibitory activity, we calculated the difference between the transient peak and the sustained firing level. A coefficient of correlation was then determined between this value and the amplitude of the shift response. The results showed that the r-values were high, and comparable to that obtained for the transient peak (Table I). These findings demonstrate that the enhancing effect induced was correlated with the signal evoked by the stimulus when it was first turned on, and this tonic
A C+)
(÷)
.......... -
_Q__ ÷
(÷)
b m
Fig. 6. Suggested neural pathway for the shift response of an OFF-center GC. The inner plexiform layer (IPL) is divided into sublaminas a and b. ON-OFF AC, ON-OFF amacrine cell; hCB, hyperpolarizing cone bipolar cell; dCB, depolarizing cone bipolar cell; GC, ganglion cell.
signal is present for the duration of the enhancing stimulus. DISCUSSION The main observations of this study are: (1) appropriate stimulation of the RFC or the RF surround will increase and inappropriate stimulation will decrease the amplitude of the shift response; (2) the shift response is graded and is strongly correlated with the capacity of the enhancing stimulus to activate the RF; (3) a shift stimulus which elicits a strong shift response can be blocked by strong inappropriate stimulation; and (4) there is no correlation between the ongoing activity (maintained activity) and the amplitude of the shift response. These results then suggest that the shift signal is modulating a tonic signal which is present in the RF. When this signal is strong then the shift response will be large, and when the signal is weak (as during inappropriate stimulation) the shift response will also be weak. The tonic signal is not the on-going activity as this was found to be poorly correlated with the ampli-
57 tude of the shift response. Because of the strong correlation between the amplitude of the shift response and the maximum firing rate, and also with the inhibitory activity following the maximum firing rate, it is more likely that some aspect of this part of the signal is present in the RF. How then does the shift stimulus elicit a transient excitation? The O N - O F F response characteristic of the amacrine cells in lower vertebrates3,15, 25 makes them a likely candidate to form one element of the neural pathway that carries the shift signal from the peripheral regions to the RFC 6,14. Intracellular recordings from the O N - O F F amacrine cells in the cat have now been accomplished, and they have been found to be monostratified in sublamina a with widespread dendrites 17. They have been identified as the A19 amacrine cell 16. A19 amacrine cells are coupled to other A19 amacrine cells by gap junctions, and make conventional synapses only on other amacrine cells. They receive input from other amacrine cells, and also from only the hyperpolarizing bipolar cells. Uptake studies of G A B A 8 and muscimo121 indicate that the A19 amacrine cells use G A B A as their neurotransmitter. Because the output of this O N - O F F amacrine cell in the cat is not onto the ganglion cells, the excitatory shift response must arise from its activity at an earlier stage. Because G A B A is, in all likelihood, an inhibitory transmitter, we suggest that the excitatory shift response arises from a disinhibitory process. To accomplish this, we hypothesize that the A19 amacrine cell inhibits a sustained type of amacrine cell which is present in the RFC. A possible neural pathway for the shift response is diagrammed in Fig. 6. The grating stimulates a large number of photoreceptors in the periphery which leads to hyperpolarization of some bipolar cells (hCB) and depolarization of other bipolar cells (dCB). Because the dendrites of the A19 amacrine cell make synapses only in sublamina a, the input to the A19 cell from the dCB must arise from other amacrine cells. In this arrangement, there would be relative depolarization of the A19 amacrine cell at light on (dCB) and off (hCB) when the grating is flashed or is shifted. The shift signal is carried laterally to the RFC by A19 amacrine cells where they make contact with the sustained amacrine cells. These amacrine cells re-
ceive input from the bipolar cells which serve the center mechanism. The output of these amacrine cells is inhibitory, and is either fed back to the bipolar cell or fed forward to the ganglion cell (not shown). Because these amacrine cells receive input from the bipolar cells, the degree of inhibitory activity will be directly related to the activity of the bipolar cell. This pathway can then account for the relationship between the excitatory shift response and the strength of the tonic signal present in the RF. If the activation of the RF is strong, then the activity or inhibitory tonic signal of the sustained amacrine cell will also be strong. The inhibition of this amacrine cell by the shift signal, i.e., a disinhibition, will lead to a strong shift response. In the same manner, inappropriate stimulation of the RF will lead to a weak inhibitory signal by the sustained amacrine cell, and the inhibition of this weak signal by the shift signal can only give a weak shift response. The interposition of the transient amacrine cell as one of the element in the pathway not only accounts for the O N - O F F behavior of the shift response, but also the unusual intensive property of the shift response. Previous studies have shown that the intensity range over which the shift response is graded is very limited so that a small change in contrast of the target can change a threshold response to a saturated response1. 7. Intracellular recordings from the transient amacrine cells of lower vertebrates have shown that these cells respond to a light flash with transient depolarization at light on and light off3,5,15,2s. In the catfish, these are the only retinal cells which always give frequency doubling responses 3. Chan and Naka 3 also reported that the power level of the response was independent of the power level of the input signal over a wide range of stimulus intensities as had been reported earlier by Kaneko and Hashimoto 15. Additional support for the inclusion of the O N OFF amacrine cell in this neural pathway comes from the study of Frishman and Linsenmeier 9 who showed that infusion of picrotoxin, a G A B A antagonist, reduced the amplitude of the shift response in Y-cells of the cat. This pathway, however, does not account for the differences in the shift response of the X- and Y-cells. The Y-cells have stronger shift responses with shorter latencies. In addition, Y-type ganglion cells receive input mainly from amacrine cells while X-type
58 ganglion cells receive input mainly from bipolar cells. The modifications of the proposed pathway to account for these differences in X- and Y-cells must await further experiments.
ACKNOWLEDGEMENTS
REFERENCES
13 Hickey, T. L., Winters, R. W. and Pollack, J. G., Centersurround interaction in two types of on-center retinal ganglion cells in the cat retina, Vision Res., 13 (1973) 1511-1526. 14 Ikeda, H. and Wright, M. J., Functional organization of the periphery effect in retinal ganglion cells, Vision Res., 12 (1972) 1857-1879. 15 Kaneko, A. and Hashimoto, H., Electrophysiological study of single neurons in the inner nuclear layer of the carp retina, Vision Res., 9 (1969) 37-55. 16 Kolb, H. R., Nelson, R. and Mariani, A., Amacrine cells, bipolar cells and ganglion cells of the cat retina. A Golgi study, Vision Res., 21 (1982) 1081-1114. 17 Kolb, H. and Nelson, R., An O N - O F F amacrine cell type in the cat retina, Soc. Neurosci Abstr., 9 (1983) 806. 18 Kriiger, J., The shift-effect enhances X- and suppresses Ytype response characteristics of cat retinal ganglion cells, Brain Research, 201 (1980) 71-84. 19 KriJger, J., The difference between X- and Y-type responses in ganglion cells of the cat's retina, Vision Res., 21 (1981) 1685-1687. 20 Pettigrew, J. D., Cooper, M. L. and Blasdel, G. G., Improved use of tapetal reflection for eye-position monitoring, Invest. Ophthalmol. Vis. Sci., 18 (1979) 490. 21 Pourcho, R. G., A Golgi and autoradiographic study of (3H) muscimol-labeled cells in the cat retina, Soc. Neurosci. Abstr., 8 (1982) 47. 22 Rodieck, R. W., The Vertebrate Retina, W. H. Freeman, San Francisco, 1973. 23 Saito, H. A. and Fukada, Y., Gain control mechanisms within the receptive field center of cat's retinal ganglion cells, Vision Res., 15 (1975) 1407-1410, 24 Stein, A., Millikin, W. and Stevens, J., The spatiotemporal building blocks of X-, Y- and W-ganglion cell receptive fields of the cat's retina, Exp. Brain Res., 49 (1983) 341-352. 25 Werblin, F. S. and Dowling, J. E., Organization of the retina of the mudpuppy, Necturus rnaculosus. II. Intracellular recording, J. Neurophysiol., 32 (1969) 339-355.
1 Barlow, H. B., Derrington, A. M., Harris, L. R. and Lenhie, P., The effects of remote retinal stimulation on the responses of cat retinal ganglion cells, J. Physiol. (Lond.), 269 (1977) 177-194. 2 Belgum, J. H., Dvorak, D. R. and McReynolds, J. S., Light-evoked sustained inhibition in mudpuppy retinal ganglion cells, Vision Res., 22 (1982) 257-260. 3 Chan, R. Y. and Naka, K., The amacrine cell, Vision Res., 16 (1976) 1119-1129. 4 Cleland, B. G., Levick, W. R. and Sanderson, K. J., Properties of sustained and transient ganglion cells in the cat retina, J. Physiol. (Lond.), 228 (1973) 649-680. 5 Dacheux, R. F. and Miller, R. F., An intracellular electrophysiological study of the ontogeny of functional synapses in the rabbit retina. II. Amacrine cells, J. comp. Neurol., 198 (1976) 327-334. 6 Derrington, A. M., Lennie, P. and Wright, M. J., The mechanism of peripherally evoked responses in retinal ganglion cells, J. Physiol. (Lond.), 289 (1981) 299-310. 7 Fischer, B., Kriiger, J. and Droll, W., Quantitative aspects of the shift-effect in cat retinal ganglion cells, Brain Research, 83 (1979) 391-403. 8 Freed, M. A., Nakamura, Y. and Sterling, P., Four types of amacrine in the cat retina that accumulate GABA, J. cornp. Neurol., 219 (1983) 295-304. 9 Frishman, L. J. and Linsenmeier, R. A., Effects of picrotoxin and strychnine on nonlinear responses of Y-type cat retinal ganglion cells, J. Physiol. (Lond.), 324 (1982) 347-363. 10 Hamasaki, D. I. and Hanada, I., A comparison of the shift response of X- and Y-cells in the cat's retina, Exp. Brain Res., 50 (1983) 117-124. 11 Hamasaki, D. I. and Sutija, V. G., Classification of cat retinal ganglion cells into X- and Y-cells with a contrast reversal stimulus, Exp. Brain Res., 35 (1979) 25-36. 12 Hammond, P., Receptive field mechanisms of sustained and transient retinal ganglion cells in cat, Exp. Brain Res., 23 (1975) 113-128.
We wish to thank Mr. O. Novarro for his excellent help with the experiments. Supported in part by a Public Health Service Grant EY-00376 from the National Eye Institute, National Institutes of Health, Bethesda, MD. G.W.M. was supported by an Ophthalmology Specialty Training Grant No. EYO 7021-06.