Spectral sensitivity ofL-TypeS-Potentials in a teleost retina

Vision Res. Supplemenf No. 3, pp. 77-86.

SPECTRAL

Pergamon Press1971. Printed inGreatBritain

SENSITIVITY OF L-TYPE S-POTENTIALS IN A TELEOST RETINA M. LAUFER, E. E. MILLAN and H. VANEGAS

Departamento de Neurobiologia, Instituto Venezolano de Investigaciones Cientificas, Caracas, Venezuela IMPORTANT developments have taken place in recent years concerning retinal research, as evidenced in this symposium. A lot has been learned about retinal structure and the functional characteristics of its elements. However, the role played by horizontal cells and Spotentials in the elaboration of the retinal output remains unclear. Furthermore, the information available on S-potentials is highly contradictory. Hyperpolarizing, or L-type Spotentials, have been found with one or with several spectral sensitivities, or response curves, in different species of teleosts (see LAUFER and MILLAN, 1970), and even in one and the same species (MOTOKAWA, OIKAWA and TASAKI, 1957; WITKOVSKY,1967). Much of the data available consist of spectral response curves, ill suited to be compared with photopigment absorption data, which in turn show great variability (LIEBMANand ENTINE, 1964; MARKS, 1965; SVAIZTICHIN,NEGISHI and FATEHCHAND,1965). Spectral characteristics of Lpotentials corresponding to rod photopigments have been reported (MITARAI, SVAETICHIN,VALLECALLE, FATEHCHAND,VILLEGA~and LAUFEX,1961)anddenied(WITKOvsKY, 1967). Purkinjeshifts have been found to occur (MITARAI et al., 1961) and have been related to changes in the resting level of the cells responsible for its generation (SVAETICHIN,1967). In this report, we pose several questions which arise from the data referred to above, and the problem of spectral sensitivity of L-potentials, that is, some of the functional characteristics of the transmission from photoreceptor to horizontal cell are re-examined experimentally. Recently, LAUFER and MILLAN (1970) reported, in light adapted retinas of the teleost Eugerres plumieri, the existence of three different spectral sensitivity curves of L-potentials with maxima at about 476,568 and 606 nm. They correspond very well with the absorption characteristics of cone photopigments in the same species which were found, microspectrophotometrically, to have absorption maxima at about 471, 568 and 604 nm. These results indicate the existence of separate transmission lines for each variety of cone into three functionally isolated horizontal cells. In order to agree with the data, these transmission paths must be independent; no cross talk can take place between receptors with different photopigments. This in turn satisfies the requirement of independent adaptation mechanisms for each cone type to account for human color adaptation (RUSHTON, 1965). Since bipolar cells receive information from the cones at the same synaptic complex as horizontal cells, a similar segregation of cone inputs should exist in them. Spectral characteristics of bipolar cell activity are not known at present. Obviously, they might show mixing of different cone outputs, but in this case the mixing must take place at the bipolar cells, not at the level of the receptors. It was also found (LAUFERand MILLAN, 1970) that, in the scotopic state, slow hyperpolarizing S-potentials with very high sensitivity could be recorded and had a spectral sensitivity curve corresponding well with the absorption of a photopigment with maximum 77

M. LAUFER. E. E. MILLAN AND H. VANEGAS

78

about 500 nm. When the retina was stimulated with very low light intensities, that was the only type of response found, but, upon stimulation with light intensities l-2 log units higher, cells were also found which exhibited faster responses with sensitivity maxima at about 606 nm. Later, the other sensitivity curves, with maxima at 476 and 568, were also found in dark adapted retinas. These results indicate that some of the horizontal cells receive information from cones whatever the state of adaptation of the retina; others do receive a rod input in dark adapted state, but what happens to them as the retina adapts to light? Could we confirm the existence of a Purkinje shift in S-potentials (MIRATAIet nl., 1961) recorded from one and the same cell? How is this phenomenon characterized as to spectral sensitivities? Experiments were carried out in isolated retinas of E. plumieri. Test stimuli and backgrounds consisted of light circles 2-3 mm dia. focused on the retina; neutral density filters were used to modify the intensity in 0.1 log unit steps and narrow band interference filters (Balzers) were used to vary wavelength. Recordings were made with micropipettes placed at the center of the light stimulus and conventional instrumentation. For details, see LAUFER and MILLAN(1970). S-Potentials with maximal sensitivity at 500 nm when the dark adapted retina is explored with dim test lights show a shift in their spectral sensitivity upon application of a strong background light superimposed on the same retinal area, and of the same size as the test flashes. Figure 1 shows two cases where such cells with maximal sensitivity at 500 nm are hyperpolarized by a 502 nm background light, becoming unresponsive while it is maintained. Upon cessation of the background illumination, the retina is less sensitive, the S-potential show a faster time course and its maximal sensitivity is now at 560 nm (A) or 600 nm (B). This shift in spectral sensitivity is progressive, and in cases in which stability of the

A Ib-

_

_

_

__

_

-

_

_

-

-

-

-

_

_

-

_____

----Me---

x500560,66

B

background

502, 1.0 --I

x 560 560, 3.6

u --

_ - _ _ _ --- - _--_--

L

A 5oot-560,‘S.O

background

502, IO--I

x 600 600, 3.6

5 set FIG. 1. Continuous oscillographic recordings of scotopic S-potentials showing the effect of background light upon resting levels, and the subsequent responses to light ftasbes. ‘Test flashes and backgrounds, signalled on the lower tracing, were superimposed on the same retinal area. In this, and all other figures, the numbers preceeded by Aindicate the maximum, in nm, of the spectral sensitivity curve obtained for the cell; under them are spe~i&d the Wavelength in nm and the relative intensity in log units of attenuation of the stimrdi II@. Background wavelength and relative intensity are also indicated. Voltage calibration: 20 mV between horizontal lines.

Spectral Sensitivity of L-Type S-Potentials

in a Teleost Retina

79

“-,

z W

100 .I 80

..

60

..

B

..

20

”

/I

I I

/’

400

500

WAVELENGTH

v ‘\,

d, ov \ ‘?

A

o-

A

I’ /’ .

0: /’

\\ \\ \ A ‘\

o,,’ ;I’

1,

:

u-l 40

,,%,, ,;-;;g,, I’ I’

\ ‘\

600 (nm)

FIG. 2. (A) Measurements of spectral sensitivity of S-potentials from one cell in a dark adapted retina (a), after a 502 nm background maintained for 30 set (A), and after a second, stronger background (V). Dashed lines represent absorption curves of photopigments with maxima at 500, 565 and 604 nm. (B) Measurements of spectral sensitivity of S-potentials from one cell in a dark adapted retina ( l ), during a 502 nm background (A), then, after 5 min of darkness (O), and during a second 502 nm background (V). Dashed lines as in (A). All measurements were normalized and sensitivity expressed in percent of the maximum for each determination.

records was good, and a cell could be held for +lO min, it was found to be reversible. Figure 2 shows sensitivity measurements from two dark adapted retinas before and after adaptive sensitivity shifts occurred. In A, the original determinations (0) fit the rhodopsin absorption curve quite well. After a 30 set background of 502 nm and 15 log units higher intensity than the test lights, the measurements indicated by upwards pointed triangles were made; no shift of the maximum has taken place, sensitivity has decreased by O-6 log units, and the curve is broader to the right. After a second 30-set exposure to the background, this time of 15 log units higher intensity than the first one, the measurements indicated by the downwards pointing triangle were obtained. They fit rather well the curve corresponding to the absorption of a pigment with maximum at 565 nm, and the sensitivity at the maximum was l-6 log units lower than that of the original determination at 500 nm. In Fig. 2B results of a similar experiment are shown, but in this case, after the original determination in a dark adapted retina (a), a 502 nm background light 2.5 log units higher than the test was maintained for 2 min, during which, but after the initial total depression of the S-potential, a second determination of spectral sensitivity was made (A). The curve and its maximum has shifted to the right and the sensitivity decreased by 3 log units. The background illumination was then turned off and the retina maintained in the dark for 5 min, after which the maximum was again found to be at 500 nm (0) and the sensitivity returned to the original values. A second exposure to the same background illumination produced the same results

80

M. LAUFER, E. E. MILLAN AND H. VANEGAS

as the first one (V). The measurements made during the 502 nm backgrounds show maxima corresponding with the 604 nm photopigment absorption curve. Spectral sensitivity shifts can be produced when the test stimuli and adapting background are not superimposed but fall on different receptor populations. Figure 3 shows results of two experiments in which, after recording a slow S-potential with maximum sensitivity

c x500 500,&O

-

-

bockgd*

h.500f

502, I.2

500,40

x600 600,OO 5 set

Bx

--

16 500 500

bockgd 502, 1.2 h

v *

X500f 500,4,2

600

600,OO FIO. 3. (A) and (B). S-Potentialsrecorded from two cellsin dark adaptedretinasbefore,during and immediatelyafter application of a “background” lit on a di&reut retinal area to that on whichthe test flashesfelLSpeciEcationsof stimuliand “background”as in Fig. 1. Under the “background” specXcationsare giventhe valuescorrespondingto stimulideliveredduringits presentation.

at 500 nm, a circle of “background” light the same size as the test lights was projected in a neighbouring retinal area, in such a way that there would be no overlap. Scattered light into the test area, if any, was far less than that which is necessary to produce the spectral sensitivity shifts described above. Under these conditions and only during application of the “background” light, the spectral sensitivity of the S-potentials was found to have a maximum around 600 nm, and the sensitivity was largely reduced. Upon withdrawal of the “background”, the sensitivity recovered and the spectral maximum returned to 500 nm within seconds. Thus, the mechanism responsible for the changes in spectral sensitivity of the S-potentials is not an exclusively photochemical one, since the shift can be produced through the effect of light falling on receptors other than those being tested. There must exist some neural factor able to switch, in neighbouring areas, from a rod input to a cone input into horizontal cells and possibly also, into bipolars. An element morphologically well suited to do this is the horizontal cell, which is strategically distributed in tangential layers and has lateral connections and processes. Our experiment shows the existence of nonphotochemical mechanisms involved in the spectral sensitivity shifts related to adap tation, but clarifies neither the mechanism nor the cells involved.

Spectral Sensitivity of L-Type S-Potentials

Cells that generate

S-potentials

81

in a Teleost Retina

have been claimed to control

retinal excitability

(SVAETICHIN, 1967; BYZOV,1967; MAKSIMOVA, 1969) and have been suggested to be involved in the process of adaptation (MITARAIet al., 1961; LIPETZ,1963; SVAETICHIN, 1967). If so,

one ought to find a parallelism between adaptive changes, such as sensitivity and spectral sensitivity, and their electrical behaviour, such as dark resting potential and S-potentials. Changes in sensitivity during dark adaptation are not dependent upon the dark resting potentials of S-units, as can be seen after the cessation of the backgrounds in Fig. 1, and as was shown by NAKA and RUSHTON (1968), though no spectral sensitivity study was made in their experiments. Comparison of the dark resting potential in the scotopic condition and after spectral sensitivity shifts have occurred shows no correlation either. Figures 1A and B show higher dark resting levels after background than before, the differences being negligible in A and marked in B. However, Fig. 4 shows, after similar shifts in spectral sensitivity, either no difference, or depolarization of the cell.

A

i-500 560,

6.6

-backgd

h 560

502,0.6

560,36

5 see v

_\ _-

x 500 500,60

-backgd 502, I-O

x600 4 96, I.2

FIG.4. (A) and (B). S-Potentials recorded from two cells in dark adapted retinas before, during and after a background light superimposed upon the same retinal area as the test flashes. Specifications of stimuli and background as in Fig. 1.

Another approach to the same question is to compare the dark resting levels of cells which have different spectral sensitivities in either adaptive condition of the retina, or to compare these levels before and after spectral sensitivity shifts have taken place in those cases where the stability of recording enables one to exclude movements of the electrode, either within a cell or between different cells. In the histogram in Fig. 5 no significant differences between the average dark resting levels are observed among cells with different spectral sensitivities, both in light adapted and dark adapted retinas. No differences of dark resting level exists between the ce1I.swith maximal sensitivity at 500 nm and those at 470, 560 and 606 nm.

82

M.LAUFER,E.E.MILLANAND

H.VANEGAS

5 0 15 IO 5 0 IO DARK

20 RESTI

30

40 NG

50

60

mV

POTENTIAL

FIG. 5. Frequency distributions of dark resting potentials of cells with different spectral sensitivities, indicated at the left, in dark adapted (clear bars) and light adapted (hM&ed bars) retinas. Open circles indicate the averages for light adapted retinas; crosses indicate those for dark adapted retinas.

In order to examine whether the mechanisms involved in generating an S-potential out of a light stimulus are similar or not in dark and light adapted conditions, the intensityresponse relationships were determined for the different types of spectral sensitivities in both adaptive conditions. Larger variations were obtained within a given type of S-potential than between the different types, suggesting that the mechanisms operating in each case have similar properties or rather, that their resulting effect is similar. Figure 6 shows results obtained in cases in which responses with maximal sensitivity at 500, 560 and 606 mu were examined in dark adapted retinas, and one case with maximal sensitivity at 560 nm in a light adapted retina, always using stimuli of the wavelength to which they were maximally sensitive. All groups of data points fit reasonably well a common double asymptotic curve. Such intensity-response relationships for any one cell studied are independent of the wavelength of the light stimulus used. Figure 7 shows records obtained under increasing stimulus intensities on a cell with maximal sensitivity at 560 nm in a dark adapted retina, using stimuli of 500 nm in B and 600 nm in C. Figure 8 shows that, if the absoissa is shifted to compensate for the difference in sensitivity to the two wavelengths, both groups of data fit the same curve reasonabiy well. Such a result, representative of several cases studied, is consistent with either of two interpretations: (a) the cell under study receives its input from only one receptor type, the differences observed being di&rences in sen&vity but not in the transducing and generating mechanisms; or (b) all receptors f&ng into the cell under study feed into it through identical mechanisms, thus it is possible to find only one common intensity-response curve. The experiment does not discriminate between the two possibilities but the first one is consistent with our previous results.

Spectral Sensitivity of GType &Potentials

x mm

100

G

in a Teleost Retina

O..*“-“-

”

80

-

2

o

500

rsm

(D.A.)

,,

560

nm

I&Ad

x

560

nm

(L.A.1

.

606

nm

(D.A.)

,? It 1) ,a0

W

,’

UJ

60

I

-

&&O**’ 0

_&o-p6:O

LO

4:8

3:6

2:4

LIGHT

G

:

I NT&

1oTY

PIG. 6. Intensity-response relationship for S-potentials with maximal spectral sensitivity as indicated at the left, in retinas previously dark adapted (D.A.) or light adapted &A.). Complete sets of data points have been shifted laterally to fit the hand drawn curve on a relative intensity scale.

B _

A

-0 -

L-Y--

_--L---

r

20

r

f

-_

_

mV

--------4, -----------

4:8

5%

4:8

3%

5 set

D.A.

_---iL,,

A

c

-

/

560

B 600 c

500

0.0

1:2

C

x560

Stirn.

2:4

/ _------------.4:s

3:6

2:4

172

o-o

FIG.7. S-Potentials recorded from a cell with maximal spectral sensitivity at 560 nm in a dark adapted retina, stimulated with increasing intensities of 600nm in (R) and 5OOnm in (C). Dark resting potential and response to stimulus of 560 nm are shown in (A). The d.c. level was shifted upwards for the recordings shown in (B) and (C).

83

M. LAUFER, E. E. MILLANAMDH. VANEGAS

84

LO

6% G

3:6

476 LIGHT

I

2:4 1:2 N T E N S

0 I .i

Y

FIG. 8. Intensity-response relationship for S-potentials from an ex@ment

such aa that illustrated in Fig. 7. Data pointa wponding to the 500 m stimuli (X ) were shifted laterally to fit the hand drawn curve.

The results presented here open new questions concerning The horizontal

cells from which S-potentials

with cones containing

the same photopigment,

spectral sensitivity corresponds on a cone pigment absorption tor connections

in order to account

to the absorption

of spectral sensitivity from that dependent

S-potentials

for S-potentials

of only one photopigment.

on a rod pigment absorption

whose

But the shift

to one dependent

implies that such cells at least must have a dual input. Recep-

of rod- and cone-horizontal

cells appear to be distinctly segregated in the

teleost retina (STELL, 1967; PARTHE, 1971), which would

indicate that the rod horizontal

can only receive information from cones through a different pathway synapse. Additional morphological information is needed. Simultaneous

and adaptation.

are recorded (KANEKO, 1970) must connect only

pigment concentration

than the receptor

studies are needed to clarify whether the spectral

sensitivity shift and its recovery are dependent

upon bleaching

ments, or whether it is the result of a neural mechanism.

and recovery of visual pig-

Neural adaptation

not only plays

a role in sensitivity changes (see DOWLING, 1967) but also in spectral sensitivity shifts, since spectral sensitivity in a retinal area can be a.&cted by light falling on a neighbouring area. It is not very likely that horizontal cells are the elements responsible for adaptive changes in view of the fact that their potentials do not correlate with changes in sensitivity or spectral sensitivity, and that the different S-potentials appear to be generated through similar mechanisms in either adaptive states, as suggested by identical intensity-response characteristics. The mechanisms for neural adaptation still need to be discovered.

Ac/cnowk&enzenu-The authors acknowledge the valuable cooperation of Dr. Grns~u~ %AETKWNfor lending his laboratory facilitiq to Mr. AWRY+QrtgwDs ANAL CAzoilLn and ToMILSROA for their technical assistance, and to Miss MARQARITA hk2a and Mrs. &rr+MAtiolHBWJRAfor their sacretarial help.

Spectral Sensitivity of L-Type S-Potentials

in a Teleost Retina

85

REFERENCES BYZOV,A. L. (1967). Horizontal cells of the retina as regulators of synaptic transmission. Neurosci. Transl. 3, 268-216. DOWLINO,J. E. (1967). The site of visual adaptation. Science, N. Y. 155,273-279. KANEKO, A. (1970). Physiological and morphological identitication of horizontal, bipolar and amacrine cells in goldfish retina. J. Physiol., Land. 207,623-633. LAUFER,M. and MILLAN, E. (1970). Spectral analysis of L-type S-potentials and their relation to photopigment absorption in a fish (Eugerresplumieri) retina. Vision Res. 10,237-251. LIEBMAN,P, A. and ENTINE,G. (1964). Sensitive low-light level microspcctrophotometer: detection of photosensitive pigments of retinal cones. J. opt. Sot. Am. 54, 1451-1459. LIPETZ, L. E. (1963). Glial control of neuronal activity. IEEE Transact. molec. Electron. 7,144-l% MARKS, W. B. (1965). Visual pigments of single goldfish cones. J. Physiol., Lond. 178, 14-32. MAKSIMOVA,E. M. (1969). Effect of intracellular polarization of horizontal cells on ganglion cell activity in the fish retina. Neurosci. Transl. 11,114-120. MITARAI, G., SVAE~IC~~N,G., VALLECALLE,E., FATEHCHAND,R., VILLEGAS,J. and LAUFER, M. (1961). Glia-neuron interactions and adaptational mechanisms of the retina. In The Visual System: Neurophysiology andPsychophysics pp. 463-481 (edited by R. JUNG and H. KORNHUBER),Springer-Verlag, Berlin. MOTOKAWA,K., OIKAWA,T. andTAs~~1, K. (1957). Receptor potential of vertebrate retina. J. Neurophysiol. 20,186199. NAKA, K. I. and RUSHTON,W. A. H. (1968). S-potential and dark adaptation in fish. J. Physiol., Land. 194 259-269. PARTHE,V. (1971). Horizontal, bipolar and oligopolar cells in the teleost retina. Vision Res., this symposium, RUSHTON,W. A. H. (1965). The Ferrier Lecture: Visual adaptation. Proc. R. Sot. B. 162,20-46. STELL,W. K. (1967). The structure and relationships of horizontal cells and photoreceptor-bipolar synaptic complexes in goldfish retina. Am. J. Anat. 121,401-424. SVAETICHIN,G. (1967). Celulas horizontales y amacrinas de la retina: propiedades y mecanismos de control sobre las bipolares y ganglionares. Acta Cient. Venezolana (Supl.) 3,254276. SVAETICHIN, G., NEGISHI,K. and FATEHCHAND,R. (1965). Cellular mechanisms of the Young-Hering visual system. In Ciba Foundation Symposium on Colour Vision, pp. 178-203 (edited by A. V. S. DE REUCK and J. KNIGHT). Churchill, London. WITKOVSKY,P. (1967). A comparison of ganglion cell and S-potential response properties in carp retina. J. Neurophysiol. 30,546560.

Abstract-A review is presented of studies on spectral sensitivity of hyperpolarizing (L-type) S-potentials, which shows the existence, in Eugerresplumieri, of three spectral sensitivity curves in light adaptation, corresponding to the three cone photopigments; in dark adaptation, a fourth spectral sensitivity was found with maximum corresponding to rhodopsin absorption. It is now shown that cells with the latter spectral sensitivity shift their maximum upon exposure to a background light, that this spectral sensitivity shift is progressive and reversible, and that it can be produced whether the adapting light falls on the same, or on a different receptor population as that tested, indicating that it is not an exclusively photochemical event, but some neural mechanism must mediate such a shift. The involvement of S-potential producing cells in these adaptive changes was explored by studying dark resting potentials and intensity-response relationships. Dark resting potentials before and after spectral sensitivity shifts, both in cells of different spectral sensitivities and in either adaptive state, do not differ significantly, indicating that adaptation is indpcndent of S-unit resting potential. Intensity-response relationships are similar for the different type of S-potentials in either adaptive state, suggesting identical generating mechanisms for all of them.

Resumen-Se presenta una revisi6n de estudios de sensibilidad espectral de potenciales S hiperpolarizantes (tipoL) que demuestran en Ellgerresplttmieri tres curvas de sensibilidad espectral en adaptation a la luz, correspondientes a 10s tres fotopigmentos de cones; en adaptaci6n a la obscuridad existe un cuarto tipo de sensibilidad espectral con maxim0 correspondiente a rodopsina. Se demuestra que celulas de esta uhima modalidad cambian su sensibilidad espectral ante un fondo luminoso, que ese cambio es progresivo y reversible, y que puede producirse cuando el fondo luminoso cae sobre la misma u otra poblacion de receptores diferente a la explorada. Esto indica que el cambio no es un evento exclusivamente fotoquimico; debe ser

M. LAUFER,E. E. MILLAN AND H. VANECAS

mediado por un mecanismo neural. La participacih de cklulas generadoras de potenciales S en tales cambios adaptativos investigose. estudiando potenciales de reposo en oscuridad y las relaciones intensidad-respuesta. Los potenciales de reposo en oscuridad antes y despuks de carnbios de sensibilidad espectral, asi coma en cklulas con diferente sensibilidad espectral en ambos estados adaptativos no difieren significativamente, indicando que la adaptacih es independiente de1 potential de reposo de las unidades S. Las relaciones intensidad-respuesta son similares para 10s diferentes tipos de potential S en ambos estados adaptativos, sugiriendo idhticos mecanismos generadores para todos ellos.