Title: Origin of the ocular light-modulated standing potential in cat
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00098103/00001
 Material Information
Title: Origin of the ocular light-modulated standing potential in cat
Alternate Title: Light-modulated standing potential in cat
Physical Description: vi, 115 leaves : ill., graphs ; 28 cm.
Language: English
Creator: Lieberman, Harris Ritchie
Copyright Date: 1977
Subject: Cats -- Physiology   ( lcsh )
Physiological optics   ( lcsh )
Eye   ( lcsh )
Psychology thesis Ph. D   ( lcsh )
Dissertations, Academic -- Psychology -- UF   ( lcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Statement of Responsibility: by Harris R. Lieberman.
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 108-114.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098103
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000063221
oclc - 04203762
notis - AAG8419


This item has the following downloads:

originofocularli00lieb ( PDF )

Full Text








I would like to thank the Chairman of my committee,

Dr. William W. Dawson, for his guidance. He generously

shared his time providing intellectual stimulation and

advice and assistance that were invaluable to me. I would

also like to thank the other members of my committee,

Dr. Calvin K. Adams, Dr. Jay M. Enoch, Dr. Merle E. Meyer,

and Dr. Horst O. Schwassmann, for their thoughtful criticism

and constructive advice. I am also very grateful to

Dr. Constance Fitzgerald who photocoagulated the discs of

the subjects used in the vascular lesion experiments. I

would also like to thank Minnie Hawthorne who did the

histology and Carl Litzkow who constructed the high

impedence DC amplifier. I wish to thank the Center for

Neurobiological Sciences for financial support.

Finally, my Ellice provided invaluable help actively

assisting throughout.



ACKNOWLEDGEMENTS................................ ii

ABSTRACT..... ............. .............. .............._ v

CHAPTER 1: INTRODUCTION..................... ........., 1

Significance.......................... ..........,. 8

CHAPTER 2: LITERATURE REVIEW........................... 10

CHAPTER 3: GENERAL METHODOLOGY........................ 18

Subjects..................................... 18

Electrode Fabrication............................. 18

Generation and Presentation of Stimuli............ 20

Signal Amplification and Processing............... 22

Preparation and Maintenance of Animal............. 22

THE LMSP OF THE CAT.................................. 26

Rationale and Methodology......................... 26

Results...................................... 27

Discussion................................... 38

FLICKERING STIMULI .................. .................. 41

Rationale.................................... 41

Methodology.................................. 43

Results...................................... 45

Discussion................................... 55


CHAPTER 6: EXPERIMENT 3: ASPARTATE.................... 57

Rationale..................................... 57

Mdethodology.................................. 58

Results....................................... 59

Discussion.................................... 69


Rationale..................................... 72

Methodology................................... 73

Results....................................... 75

Discussion.................................... 91

CHAPTER 8: ANALYSIS AND CONCLUSIONS.................... 94

Conclusions................................... 104

APPENDIX A Subjects Used in Each Experiment........... 105

APPENDIX B Rll31C Glow Tube Spectral Emission......... 106

APPENDIX C Formula for Paralytic Agent................ 107

REFERENCES...................................... 108

BIOGRAPHICAL SKETCH.................................... 115


Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy



Harris R. Lieberman

August, 1977

Chairman: William W. Dawson
Major Department: Psychology

In many vertebrates a direct current (DC) potential or

standing potential (SP) exists across the eye with the

cornea being positive relative to the posterior pole. After

the onset of a light preceded by a period of dark adaptation

a slow damped oscillation of the potential is observed.

The initial peak of this oscillation occurs at about 6

minutes after light onset. At light cessation a similar

damped oscillation of opposite polarity and lower amplitude

results. These modulations in the standing potential in-

duced by light onset and cessation are termed the light

modulated standing potential (LMSP). To directly record

the LMSP in animals a highly DC stable recording system was

developed. It was employed to localize the origin of the

ocular LMSP in the cat. The precise origin of the potential

is controversial although the PE and receptors are generally

considered the source. Three experiments were performed to

determine if other retinal structures participate. In one

experiment the LMSPs elicited by the onset or cessation of

a steady or flickering light of equal time averaged irradiance

were compared. The neural elements in the retina are known

to be responsive to transient changes in irradiance. There-

fore large differences in the LMSPs should be elicited by

the paired flickering and steady stimuli if these neural

elements or an associated process participate in generation

of the LMSP. No differences were found. In two other ex-

periments the inner retina (all retinal elements except the

receptors) was lesioned either chemically with sodium aspar-

tate or by occlusion of the retinal vasculature at the optic

disc. After either type of lesion the LMSP could no longer

be elicited or was reduced significantly. The b-wave of the

electroretinogram (ERG) which originates in the inner retina

was also abolished or similarly reduced while the a-wave

arising mainly in the receptors was preserved. It was con-

cluded an inner retinal component is essential for gener-

ation of the LMSP. Since the LMSP did not respond dif-

ferentially to flickering vs. steady illumination, as

expected if it was of neural origin, it was hypothesized

that the Muller cells (retinal glia) are responsible for its




Since the mid-nineteenth century it has been known that

there is a potential difference between the anterior and

posterior pole of the eye (Skoog, 1975). In the vertebrate

eye the cornea is positive with respect to the funds (Skoog,

1975). This standing potential (SP) across the eye varies

with changes in retinal illumination. Specifically, the SP

voltage increases when illuminance is increased (Fig. la)

and decreases with dark adaptation (Fig. lb). In the human

these changes resemble damped oscillations and have a period

of approximately 25 minutes (Arden and Kelsey, 1962a,b;

Skoog, 1975; Taumer, et al., 1975).

Until recently, the human SP as influenced by changes

in retinal illumination had been measured only by the in-

direct electrooculogram technique because of the technical

difficulties associated with direct current [:DC) recording.

For this indirect method electrodes are placed at the inner

and outer canthus of each eye and, because the eye is a

dipole, horizontal eye movements of a standard amplitude

indirectly monitor standing potential. As the eye rhythm-

ically moves left and right a potential resembling a tri-

angular waveform is generated. The peak to trough amplitude

Figure 1. Tne light-on (A) and light-off (B)
light mo~dulated standing ]potenltials (:U4SPs)
elicited by the onset or cessation of the
steady unattenuated light source. Light-on
and off are indicated by arrowis. Note the long
time course of the changes. In this and all
other illustrations of electrophysiological
signals an upward deflection is a cornea-positive
response .

10 MIH

of this potential is an indirect measure of the SP across

the eye.

A number of problems, however, are associated with this

technique. The subject must be cooperative and able to make

the necessary eye movements. Therefore it is not readily

applicable to children and animals. Even fully cooperative

subjects fatigue rapidly and any variance in the extent of

eye movements may be an artifact.

To avoid confusion the following terminology will be


SP the D.C. voltage across the eye

LMSP the change in SP induced by an increment

or decrement in retinal illumination

EOG indirect measure of the light modulated SP

as described above

Recently, however, it has been demonstrated that high

quality direct recordings of the light modulated SP in the

human are possible (Nilsson and Skoog, 1975; Skoog, 1975).

Directly recording the SP is difficult because of its DC

nature. Direct current amplifiers must be substituted for

conventional capacitively coupled amplifiers and specially

constructed non-polarizing and highly stable electrodes must

be employed.

Both direct and indirect methods of recording SP have

been attempted in non-human subjects by a few investigators.

For indirect recordings mechanical devices have been em-

played to physically move the eyes of animals (Foulds and

Ikeda, 1966; Heck and Papst, 1957).

Both direct and indirect recordings have shown that

like the human SP, the SPs of many vertebrates respond to

changes in retinal illumination and these changes resemble

damped oscillations (Kikwada, 1968).

Due perhaps to the severe technical problems there has

been little basic research on the LMSP and SP in spite of

the wide acceptance of the EOG as a clinically useful diag-

nostic test. We have been successful, however, in develop-

ing a satisfactory system for directly recording the stand-

ing potential. It has been employed here to localize the

LMSP in the cat.

The precise origins of the SP and the LMSP are contro-

versial. Generally it is believed they originate from a

pigment epithelium (PE) and receptor interaction (Arden and

Kelsey, 1962b; Arden, et al., 1962; Noell, 1954; and Taumer,

et al., 1974). It was Noell (1954) using rabbits, who

documented the importance of the PE in generating the SP

(but not the LMSP). His research was largely based on his

accidental discovery of SP responsiveness to a chemical,

sodium azide. When small amounts of this drug are intro-

duced into the circulation, a transient increase in SP

voltage results. Noell found that when the retina was

severely lesioned with iodoacetic acid the azide response of

the SP was preserved although somewhat reduced in amplitude.

However in animals with an intact retina but a PE lesioned

by another chemical (iodate), the SP responsiveness to azide

was not present. Noell concluded that the PE was an es-

sential element in the generation of the SP across the eye.

The importance of the retinal receptors in the gener-

ation of the SP is evident from the human action spectrum of

the light modulated EOG. It closely resembles the C.I.E.

scotopic sensitivity curve (Arden and Kelsey, 1962b). Also

in the rabbit the LMSP is totally abolished after retinal

detachment (Arden and Kelsey, 1962b; Foulds and Ikeda,

1966) .

Gouras and Carr (1965) and Gouras (1969) while con-

ceding the importance of the PE and photoreceptor in the

generation of the SP have stated that the LMSP is influenced

by retinal structures other than the receptors. This con-

clusion was based on studies of monkeys with occluded

central retinal arteries and humans with similar lesions.

Succeeding this occlusion the inner retina degenerates but

the photoreceptors and PE remain intact (Brown, 1968; Gouras

and Carr, 1965). In monkeys Gouras and Carr (1965)

report that one to five weeks following surgery the LMSP

response to "light on" was absent or severely attenuated.

In humans with similar lesions it has been reported that the

light onset rise of the EOG is abolished (Nagaya, 1964) but

the light-off response initially disappears but eventually

returns. Gouras (1969) concluded that the light rise and

dark trough of the SP arise at different loci and that the

light rise requires the presence of a functioning inner

retina. Arden and Kelsey (1962a,b) also suggest that the

light peak and dark trough of the EOG have fundamentally

different properties. They based this conclusion on ap-

parent differences between the two EOG components. Other

investigators, notably Taumer et al. (1974) believe that the

light rise and dark trough of the LMSP represent the same


Since we have developed a method of recording LMSP in

animals, it was employed to begin to localize the LMSP. The

overall approach taken was to determine if certain retinal

layers or processes were necessary for the generation of the

LMSP. Since the retina transmits information serially for

the most part, it seemed advisable to begin at the earliest

stages of this process and move inward.

Four experiments make up the body of this dissertation.

The first characterizes the LMSP of the cat. Specifically a

quantitative description of the time course of the light-on

and light-off LMSP is presented. In addition amplitude by

intensity functions were generated for both the light-on and

off LMSP. This function could then be compared with ampli-

tude by intensity functions at certain stages in the retinal

visual process. A second experiment compared LMSPs elicited

by steady and flickering lights of equal time averaged ir-

radiance. This experiment was designed to determine the

contribution of retinal neural activity to the LMSP as com-

pared to photopigment related processes. The neural retina

is known to be highly sensitive to time varying stimuli

(as reviewed by Brown, 1965) while the photopigment in the

receptors and related metabolic processes are not sensitive

to such stimuli (Campbell and Rushton, 1955).

The third and fourth experiments determined if the

LMSP is generated in the receptors and PE alone or if neural

and/or glial inner retinal components are involved. To

functionally separate the inner and outer retina two tech-

niques, one chemical and one surgical, were applied. The

chemical approach used sodium aspartate which by activity on

the horizontal and bipolar cells prevents conduction of

impulses beyond the receptors (Cervetto and MacNichol, 1972;

Murakami et al., 1972; Murakami et al., 1975). The other

technique consists of surgically occluding the blood supply

to the inner retina by lesioning the retinal vessels as they

enter the eye at the optic disc. The receptors and PE are

supported by the choroidal circulation so they are not af-

fected. This is a standard technique (Brown, 1968; Maffei

and Poppele, 1968). The effectiveness of both types of

inner retinal lesions were confirmed by recording the

electroretinogram (ERG). The a-wave is a measure of outer

retina viability while the b-wave has an inner retinal site

of origin (Brown, 1968).

In all experiments the primary dependent variable was

the LMSP. It was examined as a function of the various

experimental conditions described individually below. In

some experiments ERGs were also recorded.


Research on various field potentials in the eye has a

long history. The major components of the electroretinogram

have been localized and this has led to a number of basic

and applied contributions. For example, because the a-wave

has been shown to reflect the activity of the rods and cones

various researchers and clinicians have used it to study the

receptors. For example, Boynton and his associates (Baron

and Boynton, 1974; Boynton and Whitten, 1970, 1972) have made

important basic contributions by using the a-wave to study

receptor information coding in primates. Their findings

concur with intracellular recordings of Normann and Werblin

(1974) in the mudpuppy.

It is not currently feasible to record from mammalian

receptors and even in other animals it is extremely dif-

ficult. Field potentials are presently the only available

technique for directly studying the electrophysiology of


It is generally believed that the LMSP and SP are a

measure of receptor and PE metabolic activity (Arden and

Kelsey, 1962b; Arden, et al., 1962; Noell, 1954; Taumer, et

al., 1974). This viewpoint has therefore channeled basic re-

search and clinical application regarding the SP and LMSP

in a particular direction. As discussed above this may not

be a correct formulation since Gouras and Carr (1965) have

apparently demonstrated an inner retinal contribution to the

light modulated SP. Recent basic research of Skoog (1975),

Nilsson and Skoog (1975) and Taumer, et al., (1974), for

example, are based on the assumption the LMSP is a measure

of PE and receptor activity. Similarly the clinical com-

munity has considered the EOG as useful only for diagnosing

PE related lesions (Arden et al., 1962).

A more definite localization of the origin of the SP

should be an important step in successfully utilizing it for

basic and applied purposes. In the LMSP is a measure of

inner retinal metabolic activity a powerful new tool may be

available for studying the metabolic concomitants of retinal

activity in normal and pathological states. The potential

importance of such readily recordable, non-invasive measures

of retinal metabolic activity -should not be underestimated.

Very little is known about how retinal neural activity is

related to retina and PE metabolism. It is hoped that

localization of the LMSP will provide a new advance in

studying such relationships.



In this section a thorough review of the classical and

modern literature on the SP and LMSP is presented.

The existence of a standing potential across the eye

was first documented by Du Bois Reymond in 1849. Using a

freshwater fish (the tench), he found that a potential dif-

ference existed between the front and the back of the eye.

During the late 19th and early 20th century other investi-

gators (Dewar and M'Kendrick, 1876; Kuhne and Steiner, 1881)

extended this observation to a variety of species and it

became apparent that this potential was cornea positive for

vertebrates and cornea negative for invertebrates (Arden and

Kelsey, 1962a; Marg, 1951; cited in Miles, 1940). This

polarity difference has been attributed to the reversed

position of the receptor cell layers in the vertebrates and

invertebrates (Arden and Kelsey, 1962a). It was also

demonstrated that this SP was modulated by light (cited in

Arden and Kelsey, 1962a; Himstedt and Nagel, 1902; Kuhne and

Steiner, 1881). The early investigators also generated data

which suggested that the SP originated in the retina and/or

PE. In 1903 DeHaas found that the potential changed dra-

matically at the ora serrata. Additionally it was demon-


strated that large potential differences exist across the

isolated retina (cited in Arden and Kelsey, 1962a; DeHaas,

1903; Dewar and M'Kendrick, 1876; Kuhne and Steiner,

1881). Dewar (1877) was the first to record the human SP

(Miles, 1939). All of these early investigators were re-

cording the SP directly and because of the difficulties

associated with such DC recordings and discovery of the more

easily recorded E:RG, a decline in interest in the SP occur-

red (Ten Doesschate and Ten Doesschate, 1956).

Interest in the SP was reawakened in the 1930's in con-

nection with attempts to monitor human eye movements (Jacob-

sen, 1930; Meyers, 1929). Electrodes were placed at the

inner and outer canthus of each eye and when horizontal eye

movements were made voltage changes resulted. These inves-

tigators believed that these electrical changes were due to

activity of the extraocular muscles. However, Mowrer, et

al. (1936) disproved this hypothesis. They demonstrated

that passive movements of the eyes of anesthetized cats led

to similar potentials and injection of chromic acid into the

eye abolished these potentials. After Mowrer et al. (1936)

established the intraocular origin of the SP, interest in it


Miles (1939) was one of the first to apply this new

technique for other than eye movement research. He demon-

strated that the EOG in humans is modulated by changes in

retinal illuminance confirming the classical work on the SP.

Additionally he determined that a variety of other parame-


ters influenced the EOG, including pressure on the eye and

changes in blood pressure.

In the middle and late fifties a number of papers

appeared which investigated the effects of various visual

parameters on the EOG and attempted to develop it as a

clinical test of visual function (Aserinsky, 1955; Francois,

et al., 1955, 1956; Kris, 1958; Ten Doesschate and Ten

Doesschate, 1956). This activity culminated with important

papers by Arden and his associates which established the EOG

as a useful clinical test (Arden and Kelsey, 1962a,b; Arden,

et al., 1962).

Aserinsky (1955) evaluated the effect of illumination

and sleep on the amplitude of the EOG. He found that it was

significantly elevated by light and depressed by dark.

Additionally he found that a long period of dark adaptation

and sleep did not further reduce the amplitude of the SP

beyond the trough reached during a period of dark adapt-

ation. Instead the EOG on awakening was considerably

greater in amplitude than the pre-sleep dark adapted vol-

tage. This may reflect diurnal variations in EOG which have

been reported but not quantified by Kris (1958).

Ten Doesschate and Ten Doesschate (1956), confirmed the

findings of Aserinsky (1955) with respect to the light peak

and dark trough of the human EOG. Additionally they stated

that the dark trough was followed by a rise in EOG poten-

tial. Kris (1958) and Kolder (1959) confirmed these obser-

vations and concluded that both light on and off responses

resembled damped oscillations. Kris noted that the first

peak of the light adapted EOG occurred at the 10th minute

plus or minus 2 minutes after the change in illumination,

the first trough at 25th minute plus or minus 4 minutes and

the second peak at the 36th minute plus or minus 4 minutes.

She also found that the dark response reached an initial

trough at 8-10 minutes after light off and, in agreement

with Ten Doesschate and Ten Doesschate (1956), observed that

it returned to a steady state level below the preceding

light adapted baseline value.

The first studies on clinical application of the EOG

appeared in 1955 and 1956. Francois et al. (1955, 1956a,b)

studied the effect of light on the EOG. In general their

findings are in agreement with the other studies mentioned

above. They also found abnormal EOG Ss with various retinal

ocular pathologies. The EOG sensitivity to light onset or

cessation was abnormal in cases of retinal detachment,

central retinal artery occlusion and glaucoma but normal in

cases of papillitis, retrobulbar neuritis, and lesions of

the visual cortex. Additionally they noted a dissociation

of the SP and the light modulated SP. For example, in cen-

tral retinal artery occlusion the SP was intact but could

not be modulated by light.

These papers, published in the mid and late 1950's laid

the groundwork for the pivotal works of Arden and associates

on the EOG and its clinical application (Arden and Barrada,

1962; Arden and Kelsey, 1962a,b; Arden, et al., 1962).


Arden's group published the first detailed parameteric study

of the EOG. The parameters explored included light inten-

sity, length of period of illumination, the relationship

between the light peak and dark trough and the wave length

of the light stimulus employed. In addition, they proposed

a standardized clinical EOG testing procedure which has been

widely accepted.

Arden and Kelsey (1962a), based on a variety of experi-

mental results, suggest that both the light peak and dark

trough components of the EOG are generated in one structure.

However, they also conclude the dark trough and light rise

differ fundamentally in their properties.

Support of a common structural site for the origin of

the SP comes from their discovery that the two opposite

waves interact rather than sum algebraically in the manner

that the a- and b-waves of the ERG, generated at different

loci, summate. They also found what they considered to be

two fundamental differences between the light peak and dark

trough. The dark trough had an all-or-nothing nature with

respect to the brightness of the preceding light period.

Its amplitude was independent of the duration to the pre-

ceding illum~ination. The light rise, however, was graded in

amplitude depending on the brightness and duration of the

evoking stimulus. They concluded that light peak and dark

trough are the result of processes acting upon a single

generator of the LM1SP.

In a second paper, Arden and Kelsey (1962b) concen-

trated on exploring various aspects of the light peak.


Preliminary to determining the action spectrum of the light

rise they did an amplitude by intensity function over a 7

log unit range on the light peak. They found the amplitude

of the light rise to vary linearly with the logarithm of the

light intensity from 20 to 10,000 trolands (td). Above that

range the light rise was saturated. They then went on to

measure the EOG action spectrum and found it was nearly

identical to the C.I.E. scotopic visibility curve (the

average dark adapted spectral sensitivity of the human

visual system). They concluded that the first step in the

generation of the SP light rise was the rhodopsin in the

rods of the retina.

Arden and Kelsey (1962b) also investigated in some

detail the influence of the length of the dark period on the

succeeding light rise. With increasingly long periods of

dark adaptation (up to 20 minutes) the light rise increased

in amplitude. When these data were compared to estimates of

the percent of rhodopsin regenerated in the human following

a total bleach (Rushton, 1961) the correspondence was fairly


Because Arden and Kelsey (1962b) hypothesized that the

light rise was, in a very general sense, due to an increase

in metabolic activity, they inferred that it would be very

sensitive to changes in blood supply to the eye. Therefore,

another portion of the Arden and Kelsey (1962b) paper is

devoted to measuring the effects of decreased blood supply

to the eye created by pressure exerted on the globe. The


resulting ischemia caused a fall in SP. In addition, when

pressure was applied at various times during the light-dark

cycle, the magnitude of the fall but not the final resting

value varied. Regardless of whether pressure was applied to

the eye during a light peak or a dark trough the SP fell to

the same level which was about equal to the dark trough

value. They concluded that the light sensitive and pressure

sensitive portions of the SP were similar.

Arden and Kelsey (1962b) and Arden et al. (1962) also

presented data from various patients with ocular lesions to

further examine the effect of metabolic changes on the SP

and to localize the site of its generation. The light rise

was severely reduced in cases of central retinal vein

occlusion which they thought indicated the importance of

vascular factors. It was also observed, in early cases of

trauma induced retinal detachment where light perception and

small ERGs were still present, that the EOG was absent. In

cases of choroiditis where the ERG, remained normal the light

rise of the EOG was abnormally small. It was concluded that

the light rise of the SP requires a functional PE and PE-

retina interface and the primary event eliciting the light

rise occurred in the pigment epithelium.

Since the appearance of the papers by Arden and his co-

workers in 1962, there have only been a few non-clinical

studies on the light modulated SP in spite of the wide

acceptance of the EOG as a clinically useful procedure.

Some of these papers have contradicted data and interpre-


station of the studies by Arden and associates in 1962. For

example, in 1965, Gouras and Carr demonstrated the light

rise was abolished by CRA occlusion in the rhesus monkey.

In a later paper Gouras (1969) suggested, based mainly on

clinical data, that the light rise and dark trough arise in

different structures. Taumer et al. (1974) argue against

Arden and Kelsey's (1962b) conclusion that the light rise

and dark trough differ fundamentally. They found that the

two light modulated components of the human EOG are similar,

both being damped oscillations with a common cycle time of

about 25 minutes.

Foulds and Ikeda (1966) examined the EOG of the rabbit

following experimentally induced retinal detachment. They

noted a decrease in SP amplitude and the complete absence of

the light peak if the detachment was complete. This is in

agreement with what Arden and Kelsey (1962b) found on humans

with retinal detachments.

Some of the most recent basic research on the SP demon-

strated that it could be recorded directly from the human by

using a highly stable calomel half-cell electrode system

(Skoog, 1975). In agreement with other investigators,

notably Taumer et al. (1974), he found the light on and off

components of the SP to be similar in time course, to be

damped oscillations with a period of about 30 minutes, and

to be of opposite polarity.




The thirteen subjects in all experiments on the LMSP

were adult cats of either sex weighing 2.5-4.0 kilograms.

The cats were housed in individual cages and were maintained

on a cycle consisting of 12 hours of light and 12 hours of

dark. Appendix A lists the various experiments for which

each cat was used.

Electrode Fabrication

The single greatest obstacle to high quality DC record-

ing was construction of high stability electrodes with no DC

drift. DC drift is a slow constant change in potential seen

across electrodes when their stability is measured in physio-

logical saline. This type of drift is not normally a

problem for electrophysiological recordings because the

potentials usually measured last no more than a few seconds.

However for the measurement of the LMSP, which has a time

course many minutes long, such instability is unacceptable.

Therefore high DC stability Ag-AgCl electrodes were suc-

cessfully fabricated. The method used in their construct-

ion, as described below, was derived in part from Geddes


(1972). Ten mil platinum wires about 4 cm long with a small

loop at one end were first cleaned by sonication in Freon

12. In order to avoid the deposition of organic materials

on the electrodes they were only handled with jeweler's

forceps once this initial cleaning was completed. Following

the cleaning the loops on the end of the wires were immersed

in a mixture of pure AgO and a small amount of distilled

water. This mixture adhered to the loop and formed an

irregularly shaped cylinder which was approximately 4 mm

long and approximately 1.5 mm in diameter. The electrode

was then placed in a 5000C oven for 1 hour. This converted

the AgO to pure silver. The next process was to coat the

platinum but not the silver cylinder with an insulator,

Robert's epoxy. Finally the silver was chlorided by using a

constant current DC power supply. To accomplish this the

electrode was placed in a bath of physiological saline and

connected to the positive terminal of the power supply. A

platinum plate also immersed in the saline was connected to

the negative pole of the power supply. A current of 10

milliamps at 10 volts was passed across the electrode for 5

minutes. The electrodes were then stored in a bath of

physiological saline and grounded together until used.

Prior to an experiment they were checked for DC stability

and noise level in a physiological saline bath. DC drift of

the complete recording system and electrode measured across

a saline bridge was less than 65 uV/hr. These same elec-


trodes were also used to record the ERGs. Their low noise

characteristics proved excellent for this purpose.

Generation and Presentation of Visual Stimuli

A M~axwellian view stimulator with a Sylvania R1131C

glow tube (Buchmann-Olsen and Rosenfalck, 1957) was the

source that provided the steady or flashing visual stimulus

used to modulate the SP. This stimulus subtended 27.20 of

visual angle. The total irradiance of the stimulus when

unattenuated was 5.4 mW/steradian as measured by an Epply

thermopile. The spectral emission of the Rll31C glow tube

at 30 mA, the current at which it was always run, is pre-

sented in Appendix B. When necessary the stimulus was

attenuated with Kodak #96 written neutral density filters.

The unattenuated irradiance will be referred to throughout

as the 0.0 intensity level and all other irradiance values

are calculated with reference to this value.

Since a human dark adaptation experiment was performed

the stimulus was also calibrated in photometric units. The

retinal illuminance of the stimulator was 57,000 photopic tds

when unattenuated. This was calculated by the method

described by Westheimer (1966). A Salford Electrical

Industries photometer was used for the necessary luminance

measurements. The glow tube was driven by a Roush Corpor-

ation timer and flash generator. Current was continuously

monitored. A Beckman Universal timer and counter Model

5510-11 monitored flicker rate.


In all cases at least twenty for dark adaptation and

twenty-five minutes for light adaptation were necessary

between stimulus alterations, i.e. turning the light on or

off. Such long intervals of adaptation were mandatory

because of the oscillation of the LMSP. Taumer et al.

(1974) have shown that the oscillatory rises and troughs of

the LMSP will interact with a subsequently elicited change

in illumination unless a sufficient time period is allowed

for the SP to reach an equilibrium. The time period for

light adaptation was longer than for dark adaptation because

light adaptation produced greater amplitude oscillations

which required a longer time period to reach an equilibrium.

As a rule an LMSP eliciting stimulus was not presented

unless a stable baseline, defined as less than 0.2 mV of DC

drift over a 4 minute period was obtained.

ERGs were generated by placing a Grass PS-2 xenon

photostimulator approximately 25 cm from the eye. The flash

had a 20 v second duration and was operated at the W-16

(brightest) setting. An interflash interval (IFI) of 10

seconds was generated by the Roush timer and monitored on

the Beckman timer. In all cases ERGs were recorded when the

S was dark adapted for at least 10 minutes, the opposite eye

was always patched to prevent the recording of spurious


Signal Amplification and Processing

A locally designed and constructed high input impedance

DC differential amplifier provided the first stage of am-

plification. Its gain was set at 10 and the bandpass was 0-

14 Khz at the 3 db point. The input impedence was 1012 ohms.

The signal was further amplified by a Redcor Corp 371

differential DC amplifier. Depending on whether SPs or ERGs

were being recorded its gain was set respectively at 10 or

20 and the upper bandpass was 0.1 or 1.0 khz at the 3 db


The SP and ERG signals were processed on a Nicolet Nic-

80 minicomputer, displayed on a monitor oscilloscope and

digitally stored on disc for subsequent analysis. SP data

were also recorded on an Ominiscribe strip recorder. ERGs

were summed sixteen times using the Nic-80.

Preparation and Maintenance of Animal

All Ss were deprived of food for at least twelve hours

prior to the experiment. Anesthesia was initiated by an

intramuscular (IM) injection of 8-10 mg/kg of Ketamine HC1

(Bristol). Also at this time the S's pupil was dilated and

accommodation paralysed with a few drops of 1% cyclopento-

late HCI instilled in the conjunctival sac. Occasionally it

was also necessary to use a similar amount of viscous 10%

phenylephrine HCI to retract the nicitating membrane. The

S was also given an IM injection of 0.1 mg/kg of atropine

sulfate to suppress salivation. When the S was sufficiently

anesthetized (five to ten minutes after the initial inject-

ion) a catheter was inserted in the femoral vein and the cat

was intubated with an appropriately sized trachea tube. The

outside of the tube was coated with 2% lidocaine jelly to

prevent discomfort after the anesthetic wore off. Other

necessary minor invasive procedures were all performed while

the cat was still anesthetized. A small (3-4 mm.) incision

was made above the supraorbital ridge of each eye for

insertion of the indifferent and ground electrodes. These

areas were infiltrated with 2% lidocaine HC1 to prevent any

subsequent discomfort. On occasions when an intraocular

injection was planned a small incision was made at the

external canthus of the eye to be injected. This area was

also infiltrated with lidocaine. When these procedures were

completed the animal was placed in a Baltimore stereotaxic

holder and points of pressure, notably the region of the

external auditory meatus, were covered with 2% lidocaine

jelly. Then an intravenous (IV) infusion of a tubocurarine-

gallamine mixture was initiated (See Appendix C for for-

mula). After an initial dose of 1 cc/kg a slow infusion of

the paralytic agent delivered via a persitaltic pump (Extra-

corporeal Inc. Model 907) was begun. During the first 30

minutes a dose equivalent to the initial dose of the para-

lytic agent was infused. After that time the rate of

delivery was reduced by a factor of 1/2 by dilution of the

mixture. Repeat doses of atropine (0.07 mg/kg) were admin-

istered IM every four or five hours.


Artificial respiration using a Harvard apparatus small

animal respirator was begun when the initial dose of tubo-

curarine-gallamine had caused a marked decrease in voluntary

respiration. Percent expired CO2 was continuously monitored

by a Beckman LB-1 medical gas analyser and was maintained at

4.3 + 0.2 by varying the respiratory rate and volume. To

prevent atelectasis the S was hyperventilated occasionally.

While the S's physiological condition was stabilizing,

the electrodes were placed on the cat. The active (non-

inverting) electrode was placed in one end of an agar-saline

filled bridge in a 3 mm inner diameter (ID) polyethylene

tube. The other end of the agar-saline bridge was connected

to a specially constructed tran-sparent contact lens. The

lens was moistened with a 2.5% hydroxypropy methylcellulose

solution. The indifferent and ground electrodes (Ag-AgC1)

were in 4 mm ID polyethylene tubes filled with agar saline

bridges. They were also placed in the incisions made above

the S's orbital ridge~s. The indifferent electrode (in-

verting input) was located above the eye to be recorded

from. To insure a low impedance, and a high stability

interface, the indifferent and ground electrode tubes had

agar saline soaked cotton wicks inserted in the ends which

were in contact with the cat. Body temperature was moni-

tored with an electronic thermometer via a rectal probe and

was maintained at 38 + 20C with a fluid filled heating pad.

Once the electrodes were in place and the animal was

stable, the Mlaxwellian view stimulator was aligned. The


focal point of the stimulator was located at the center of

the entrance pupil of the S's eye approximately aligned on

the visual axis.



Rationale and Methodology

The objective of this experiment was to characterize

the directly recorded LMSP of the cat. After a thorough

literature search no data on the LMSP of the cat were found.

In fact no quantitative data on directly recorded LMSP in

any species were found although the characteristics of the

human EOG have been adequately described (Arden and Kelsey,

1962a,b; Kris, 1958; Taumer, et al., 1974).

In pilot experiments the light-on and light-off LMSPs

of the cat were observed to be damped oscillations. To

quantify the time course of these oscillations and to permit

comparison with human LMSP oscillations the latencies of the

various peaks and troughs of the cat were measured. In

addition to quantifying the latencies of the peaks and

troughs of the light-on and light-off LMSP, amplitude by

intensity functions were generated over a 3-5 log unit range

of irradiance for the first peak of the on LMSP and the

first trough of the off LMSP.

The LMSP eliciting stimulus used in this experiment was

always the onset or cessation of a steady light. The

stimulus intervals were as described in Chapter 3.


The waveforms of both the on and off LMSP of the cat

resembled damped oscillations although they had opposite

initial polarities (Figure 2). The time course of both

types of responses were extremely long compared to other

electrophysiological responses.

Table I summarizes the mean latencies, standard devia-

tions, ranges, and sample sizes of the various peaks and

troughs of the on and off LMSP. In all cases a peak refers

to a cornea-positive inversion point and a trough refers to

a cornea-negative inversion. Responses were pooled for 8

cats. The steady stimulus which elicited these light-on or

light-off L.MSPs ranged from 5.4-.0054 mW~/steradian in ir-

radiance. The responses elicited by the various intensities

were pooled since no consistent differences in latency as a

function of stimulus intensity were observed. There were

some consistent intersubject differences in latency however.

Sample size decreases with each successive peak or trough

because in many cases, especially when less intense stimuli

were presented, peaks and troughs were not observed. Typi-

cally the light-on response of the SP (Figure 2a) was a

cornea-positive signal which reached a first mean peak at

6.4 minutes after light onset (Table 1). Following this

initial peak the potential dropped for a few minutes until

the first trough was reached at a mean of 17.4 minutes.

Generally the rising phase of the first peak had a steeper

slope than the subsequent falling phase. After this first

Figure 2. The light-on (A) and light-off (B)
light modulated standing potentials (IMSPs)
elicited by the onset or cessation of the steady
unattenuated light source. Light-on and off are
indicated by arrows. Note th~e long time course
of the changes. In this and all other illustrations
of electrop~hysiological signals an upward deflection
is a cornea-positive response.

10 MIN

mean latencyr (rmum) 6. 44 17.45 26.39 33.31 42.30

standard deviation .71 2. 43 3.99 4. 24 6.84

range (min) 2.8 8.8 10.7 10.3 11. 99

sample size 24 13 9 5 3

B- Light-Off

Trough 1 Peak 1 Trough 2 Peak 2 Trough 3

IIean latency (min) 8.09 16.60 23.11 30. 97 40.70

standard deviation 1.76 .88 1.6 1.79

range (min) 6.6 2.5 4.5 3.4

sample size 18 9 6 3 1

Table 1. &ban latencies, standard deviations, ranges and sample sizes
for the various peaks and troughs of the light-on LNSP (A) and light-
off IMSP (B) .

A- Light-On

Peak 2

Peak 3

Peak 1

Trough 1

Trough 2

trough was reached the SP rose to a second peak at a mean of

26.4 minutes after light onset. This peak was always of

smaller amplitude than the first peak. This second peak was

followed by another trough at a mean of 33.3 minutes after

light onset. By this time in most recordings it became

difficult to distinguish further oscillations. When present

a third peak occurred at a mean of 42.3 minutes after light

onset (Figure 2a) and appeared to be smaller in amplitude

than the second peak. The troughs of the light-on LMSP

rarely fell to a value which was as low as the SP value

preceding the light presentation (Figure 2a).

The light-off response was of opposite polarity than

the light-on response. It began with a trough and then rose

to a peak. The light-off response (Figure 2b), like the

light-on response, consisted of oscillations of decreasing

amplitude though their time course and waveform differed

somewhat from the light-on LMSP. The light-off response

reached a first trough at a mean of 8.1 minutes after light

onset (Table 1). This was followed by a first peak at a

mean of 16.6 minutes and then a second trough at a mean of

23.1 minutes. The initial off response rose to a first peak

more gradually than the corresponding fall of the first peak

of the light-on response making it more difficult to score

the off response for amplitude and latency. It was also

more difficult to observe the succeeding oscillations of the

light-off LMSP because the total off response was of a

smaller amplitude than the light-on response elicited by the


same irradiance. Nevertheless in one case when the signal

was cessation of the brightest light and the recordings were

of exceptional stability a third oscillation could be

plainly observed at 41 minutes after onset (Table 1). Gen-

erally no portion of an off response reached a voltage which

was as great as the light adapted SP amplitude which immed-

iately preceded cessation of the light.

In addition to the major components of the light-on and

light-off responses, certain early minor components of the

light-on and off LMSP are also consistently present. These

components can be observed in Figure 2a but are better seen

when displayed on an expanded time scale as in Figure 3.

The early minor components of the light-on LMSP (Figure 3a)

consist of an early cornea-positive component (Pl) followed

by a slower, cornea-negative component (Nl). The Nl com-

ponent is followed by the first major cornea-positive com-

ponent of the light-on LMSP. The early minor components of

the off response tend to be difficult to observe. They are

present in Figure 2b though they are more easily observed on

an expanded time scale (Figure 3b). They consist of an

early cornea-negative component (Nl) followed by a slower

cornea-positive wave (Pl) which in turn is followed by the

first major component of the off response.

A third minor component in the light-on SP is a small

irregularity in the first peak at 1.6 minutes after light

onset. This component is indicated by the curved arrows in

Figure 4a and was elicited by the highest levels of illumin-


0 5 Y


Figure 3. The early minor components of the LMSP.
Waveform A is the response to light onset. The two~ early
components are identified as positive one (Pl) and negative
one (NI). Wavefonn B is the response to light cessation.
The nomenclature is thle same as waveform A. Note the time
scale is more expanded than the previous figure. Both the
on and~ off responses were elicited by the onset or cessation
of a steady light (as indicated by the straight arrows) with
a log relative irradiance of T.3.

Figure 4. Illustration of another minor component seen in the
IMSP. Figure 4A is the light-on response and Figure 4B is the
light-off response to steady light with a log relative irradiance
of 1.3. The records are the same as Figure 2: however, the time
scale has been lengthened. The curved arrows indicate the location
of these additional early components. This third minor component
is seen in both on and off response. The straight arrows indicate
light onset and cessation.


i -E


ation presented. It was observed consistently in most

light-on LMSPs. A similar component can occasionally be

seen in light-off LMSPs. It is illustrated in Figure 4b as

indicated by the curved arrows.

To further characterize the LMSP of the cat, amplitude

by intensity functions for the first major peak of the

light-on LMSP of 4 cats are presented in Figure Sa. Over

this range the light-on response varies in an essentially

linear fashion with the logarithm of stimulus intensity.

Figure 5b presents amplitude by intensity functions for

the first major trough of the light-off response of the LMSP

for the same eyes as Figure Sa. The off response functions

are of a lower amplitude and not as orderly as the light-on

LMSP data. In one case, cat 46, the response appears

saturated at higher intensities. In most cases, however, a

clear off response could not be recorded below 3.3 log

units. This is partly attributable to longer time course of

the first oscillation of the off LMSP which made it dif-

ficult to be certain a response had occurred when scoring

data. The overall lower amplitude of the off response also

contributed to the problem of measuring the off response.

It does appear, however, that the off LMSP responds in a

graded fashion to decrements in irradiance. In the 2 cases,

cats 22 and 216, where clear cut graded changes occur, the

decrement in amplitude appears to be linear with respect to

the logarithm of stimulus intensity.

Figure 5. Amplitude by intensity function for the first major peak
of the light-on (Figure 5A) and first major trough light-off
(Figure 5B) INSP for 4 Ss. Amplitude in mV is on the Y- axis and it
is plotted as a function to log relative irradiance on the Xi- axis.
Each function represents data froma a single S. Each point is the
am~plitude of a single response. The stimulus was always the onset
or cessation of a steady light.

4.3 3.3 2.3 1.3



As illustrated by Figure 2 the light-on LMSP and light-

off LMSP of the cat are of opposite initial polarities and

are damped oscillations with similar time courses (Table 1).

The first major component of the light-on response is cornea-

positive and the first major component of the light-off

response is cornea-negative. Similar findings with respect

to waveform have been reported in all vertebrate species

studies (Gouras, 1969; Kikawada, 1968; Skoog, 1975). The

initial major peak of the light-on LMSP of the cat occurs at

6-7 minutes after light onset. The first major trough of

the light-off response occurs at 8 minutes after light

cessation. In the human Kris (1958) reported the first

light-on peak occurs at 10 plus or minus 2 minutes and the

first light-off trough occurs at 8-10 minutes when measured

indirectly by the EOG. Skoog (1975) recording the LMSP of

the human directly, shows records where the first light peak

occurs about 6-10 minutes after light onset and the first

dark trough occurs 8-12 minutes after light cessation. The

amplitude of the off response in the cat was lower than the

on response elicited by the same stimulus. The same holds

true for the directly recorded human LMSP and EOG (Arden and

Kelsey, 1962a; Skoog, 1975; Taumer, et al., 1974). The

human on and off LMSP takes 36 plus or minus 4 minutes to

reach the second light-on peak or light-off trough (Kris,

1958). This is longer than the cat LMSP which required 10-

13 minutes less to reach the equivalent inversion. It is

important to note that in the cat, like the human (Skoog,

1975; Taumer et al., 1974), the on and off responses are

similar in waveform and time course but are opposite in

polarity. Gouras and Carr (1965) found oscillations with a

similar time course in the rhesus monkey eye. Kikawada

(1968) studied the first peak and trough of the light-on and

off LMSP of a variety of species from all major vertebrate

classes. He found considerable variability within and

across the major classes of vertebrates (3-13 minutes for

the first light peak and 7-20 minutes for the first dark

trough). The similar period of the on and off LMSP of the

cat and other species indicates that pigment bleaching or a

directly related metabolic process is probably not the basis

for the on and off LMSP. Pigment bleaching is a very rapid

process while pigment regeneration requires many minutes of

time. Therefore if the on and off LMSP are of similar

origin it is unlikely they arise from pigment regeneration

or a directly related process.

The two early minor components of the cat light-on LMSP

(Figure 3) are also present in the human EOG and are similar

in waveform and time course (Taumer et al., 1974). The

third component which occurs riding on the major initial

component of the on and off LMSP about 1.6 minutes after

light onset also has been observed in the human LMSP (Skoog,


It is well established that the amplitude of the first

major light-on peak of the human EOG varies linearly with

the logarithm of stimulus intensity (Arden and Kelsey,

1962b; Taumer, et al., 1974). This also is the case for the

cat light-on LMSP as shown in Figure Sa. The log-linear re-

lationship between stimulus intensity and light-on LMSP

amplitude is another indication that the photopigment in the

receptors or a directly related process is not the source of

the light-on LMSP. As noted above the extent of pigment

bleached by a stimulus is linearly related to the stimulus

intensity not the logarithm of the stimulus intensity

(Dowling and Ripps, 1970). If the light-on LMSP is directly

related to visual pigment kinetics then it also should be

directly related in a linear fashion to stimulus intensity.

There is some controversy about the graded nature of

the light-off response of the LMSP. Arden and Kelsey

(1962a) found it to be an all or none response, always

falling to the same initial trough regardless of the stimulus

intensity that evoked it. They also found a great deal of

response variability in the light-off EOG amplitude unre-

lated to stimulus variations. Taumer et al. (1974) found

graded changes in light-off EOG linearly related to the

logarithm of stimulus intensity and dispute the findings of

Arden and Kelsey (1962a). The results of this study on the

cat LMSP tend to support the findings of Taumer et al.

(1974) since graded changes in the amplitude of the off

response were recorded (Figure 5). There was a great deal

of variability however, as Arden and Kelsey (1962a) found

for the light-off EOG.




This experiment was designed to determine if the LMSP

elicited by steady as opposed to flickering stimuli of equal

time average irradiance (quanta) are equivalent. The neural

components of the retina are highly sensitive to time

varying stimuli while other retinal processes such as visual

pigment related metabolism and glial cell activity are not

as sensitive (Campbell and Rushton, 1955; Miller and Dowling,

1970). Therefore, by studying LMSP responsiveness to

flickering vs equivalent steady stimuli it is possible to

indirectly determine the importance of various elements in

generating the LMSP.

For example the first component of the visual process

is the photochemical response of the rods and cones. One

part of this process is pigment regeneration which is fairly

well understood. Since the PE is known to participate in

pigment regeneration, a comparatively slow process (Dowling

and Gibbons, 1961), it is possible the LMISP, which has a

long time course reflects related metabolic processes (Arden

and Kelsey, 1962b; Taumer et al., 1974). One purpose of

this experiment will be to compare the LMSP elicited by


stimuli which bleach equal overall amounts of pigment but

effect the rest of the neural components of the retina dif-

ferentially. It is apparent from intracellular recording in

many lower vertebrates that neural cells in the retina

respond differentially to steady and flickering light

(Rodieck, 1973; Werblin and Dowling, 1969). The stimuli

therefore were a steady light and flickering light of twice

the luminance and half the duration (50% duty cycle). It

has been demonstrated that the fraction of pigment bleached

is proportional to the product of the intensity of a light

and its duration. This relationship is termed the Bunsen-

Roscoe Law. Campbell and Rushton (1955) demonstrated its

holds for human rod pigment for durations ranging from 0.3

to 48 seconds. They did not investigate shorter durations.

It is known that at microsecond flash durations and/or high

intensities the Bunsen-Roscoe Law does not hold (Hagins,

1957; Pugh, 1975). For this reason very short durations or

high intensities were not employed in this experiment.

A Variety of light intensities and rates of flicker

were employed in order to be certain that any flicker vs.

steady differences or similarities were of a general nature

and not unique to particular stimulus situations. Addition-

ally the equivalence of the bleach produced by the flickering

and steady light was confirmed by presenting each type of

stimulus to a human observer and measuring the subsequent

dark adaptation function.


Cat Experiments

The object of this experiment was to investigate the

effects of square wave flicker versus an equivalent steady

retinal irradiance. The flicker and steady stimuli were

equivalent with respect to averaged irradiance and therefore

at the durations employed bleached equivalent amounts of

pigment (Rushton and Campbell, 1955). The flickering stim-

ulus had a square wave configuration with a 50% duty cycle.

Since it was on one half as long as the equivalent steady

state stimulus, its irradiance was increased by 0.3 log

units. This increased the irradiance by a factor of two.

The equivalence of the two types of stimuli with respect to

extent of photopigment bleached was empirically confirmed by

a human psychophysical dark adaptation experiment which is

described separately.

In addition to varying flicker rates from 0.5 to 30 Hz,

certain flicker rates were tested over a wide range of

retinal illuminations (3-4 log units). Therefore, there were

two basic protocols for the flicker versus steady experi-

ments. In one type of experiment a particular rate of

flicker was tested over a 3 or 4 log unit range of irradi-

ances. In this type of experiment, a flickering stimulus of

a particular irradiance was always preceded or succeeded by

a steady stimulus of the equivalent time averaged irradiance.

The other type of flicker versus steady experiment was

designed to explore a large range of flicker rates. In


these experiments different rates of flicker were presented

at one irradiance. The equivalent steady irradiance con-

dition was also presented periodically throughout the experi-

ment for comparison purposes and to monitor any change in

overall responsiveness. The interstimulus intervals of both

types of experiments followed the guidelines described in

Chapter 3.

Human Dark Adaptation Experiment

The object of this experiment was to test the equiv-

alence of the bleach produced by the steady and flickering

stimuli. The S in the human psychophysical dark adaptation

study was a 25 year old female who had normal vision. She

had little previous psychophysical experience. The observer

was first light adapted with the same Maxwellian view

stimulator used to generate the LMSP in the cat experiments.

Then immediately after light adaptation dark adaptation

functions were measured on a dark adaptometer.

For the preliminary light adaptation the observer was

instructed in and then practiced proper alignment of the

Maxwellian view stimulator. She was instructed to fixate

the center of the field with her right eye. Stable head

position was maintained by a head and chin rest. The left

eye was patched. The S was light adapted by either a steady

light attenuated by 0.3 log units or an unattenuated light

flickering at 2 cps. After 5 minutes of light adaptation

the S moved to the dark adaptometer. The adaptometer was

constructed locally and was based on a design described by


Hecht and Shlear (1936). The observer fixated a red light

emitting diode. The test stimulus was a 120 diameter disc

of unfiltered tungsten light. It was centered 140 nasal to

the point of fixation on the horizontal meridian of the

right eye. Since the adapting stimulus subtended 27.20, the

dark adaptation test stimulus fell fully within the adapted

area. The maximum luminance of the test stimulus was set at

1.65 log foot lamberts as measured with a Salford Electrical

Industries photometer. It could be attenuated in 0.1 log

unit steps with Kodak #96 written neutral density filters

and had a duration of approximately 200 msec. For 35

minutes dark adaptation thresholds were determined with this

test stimulus. The ascending method of limits was employed

and approximately 3 measurements were made each minute.


For two cats amplitude by intensity functions were

generated at 2 and 8 Hz respectively. For three cats flicker

rates ranging from 0.5-32 Hz were tested individually. In

both types of experiment responses to the equivalent steady

light conditions were also recorded. Figures 6a and b

present the on and off LMSP to steady and flickering light

(2 Hz) of equivalent time averaged irradiance (equal total

quantal flux). The waveforms are similar especially the

first major component of the on and off response. Second

peaks and troughs are present in the flicker on and off

response and their time course is similar to the oscilla-

tions elicited by steady stimuli.

Figure 6. The light-on (Figure 6A and B)
and ligjht-off (Figure 6C and D) INSP
elicited by the onset and cessation of a
steady light (A and C) or a flickering
light (B anld D) of a corresponding time
averaged ir-radiance.


10 MI


Figure 7 presents the data from the experiment where on

and off LMSP amplitude by intensity functions were generated

over a 3 log unit range for a stimulus flickering at 8 Hz

and a steady light of equivalent time averaged irradiance.

In Figure 7a the light-on data demonstrate that over this 3

log unit range of stimuli there were no consistent differ-

ences between equivalent steady and flickering stimuli. The

light-off steady vs flicker data recorded during the same

experiment are presented in Figure 7b. There are no con-

sistent differences across the adaptation levels for the off

response to flicker vs steady stimulation.

Figure Sa graphically presents a comparison of the

amplitudes of the first peaks of the light-on LMSP elicited

by the equivalent steady and flicker light conditions. The

data from five Ss were pooled. At each flicker rate the

response has been compared to the steady stimulus of an

equivalent time averaged irradiance recorded from that S.

The difference in mV between the two was calculated by sub-

tracting the appropriate steady response amplitude from the

corresponding flicker amplitude. For each square wave

flicker rate tested and presented on the abscissa the steady

vs flicker difference in amplitude has been calculated and

is presented on the ordinate. The number of observations

that each point is based upon is presented above each point.

The error bars represent + 1 standard deviation. To deter-

mine if there was a significant difference overall between

the steady and flicker conditions, the light-on data were

Figure 7. Amplitude by intensity function for the first
major peak of the light-on (Figure 7A) and first major trough
of the light-off (Figure 7B) IISP for a single S. The functions
represent: a steady illumination ( ) and a square wave flickering
stimulus of 8 Hz of equivalent time averaged irradiance (o) The
S was cat 216.

1.3 0.3

3.3 2.3






o.s C

3.3 2.3 1.3 0.1


Figure 8. The differences in amplitudes between the first
major peak of the light-on LM~SP (Figure 8A) elicited by a
steady light and the equivalent flickering light stimulus.
To arrive at a value in mV plotted on the X axis the steady
amplitude in my was subtracted from the equivalent flicker
amplitude. Flicker rate is plotted on the y-axis and each
point is a nlan with the sarp~le size indicated directly
above the point. The bars represent + 1 standard deviation.
Figure 8B presents the difference in amplitude between the
off LM1SP elicited by equivalent steady and flickering stimuli.

n=1 n=3 n=12 n=3 N=6 n=1 n=i

n=2 N=7 n=1 N=4 n=1 N=1

0.5 I 2 4 8 16 32


pooled for all flicker rates and a two tailed student's t

test was performed. There was no a significant difference

between the steady and flicker conditions (t = 0.76, df =

41, p > 0.25).

Figure 8b presents the same data for the off steady vs

off flicker conditions. Although the variation in amplitude

is greater as noted previously for the off response of the

LMSP there was no consistent effect of a steady vs a flick-

ering stimulus on the amplitude of the off LMSP from 1-32 Hz

(t = 0.58, df = 25, p > 0.25).

The flicker versus steady experiments are to a certain

extent based on a corollary of the Ferry-Porter law e.g.

that a flickering light with the same equivalent time aver-

aged irradiance as a steady light will bleach an equivalent

amount of photopigment (Campbell and Rushton, 1955). This

assumption was tested for the specific stimuli presented in

this experiment in a dark adaptation experiment on a human

observer. Figure 9 represents the two dark adaptation

curves generated for the human observer preadapted to either

a steady or flickering stimulus of equivalent time averaged

luminance (38,500 photopic trolands). Threshold luminance

in log foot lamberts is plotted as a function of time after

cessation of the preadaptation stimulus. There was no major

difference between the two functions,

Figure 9. Dark adaptation curves following preadaptation to either
a steady (solid circles) or flickering stimulus (open circles) of
equal time averaged lunminance. Threshold lumninance in log foot
lamberts is plotted as a function of time in minutes after the
preadapting light cessation.




0 m



I 080

(Hf1 Sol) 3 3N VN I~ill (I'IGoH S3d1H


The results of these experiments demonstrate that there

are no major significant differences, over the range of

flicker rates tested, in both the light-on and light-off

LMSP elicited by steady and flickering stimuli equated for

time averaged irradiance. This was true for a wide range of

flicker rates and over a range of intensities. The equiv-

alence of these two stimulus conditions was confirmed in the

human dark adaptation experiment. The similarity of the on

and off LMSP to stimuli which produce extremely dissimilar

responses of retinal neural elements indicates that the LMSP

is not a measure of retinal neural activity. There are con-

siderable electrophysiological and psychophysical data on

the cat and many other vertebrate species which demonstrates

that the retina is highly sensitive to the flicker rates em-

ployed in this study as reviewed by Brown (1965). Dodt and

Enroth (1954) report that the cat ERG and retinal ganglion

cell responses follow square wave flickering stimuli up to

50 Hzt. In a behavioral study Kappauf (1936, cited in Brown,

1965) reported cats could distinguish stimuli flickering at

up to 50 Hrz. Wh~en the large differences in the responses of

the neural elements of the retina to flickering as opposed

to steady light are compared to the similarity of the LMSP

elicited by these stimuli, it seems likely these components

do not make a significant contribution to either the light-on

or light-off LMSP.


The equivalence of LMSPs produced by stimuli which

produce an equivalent bleach suggests the LMSP may in

some way be related to this phenomenon. However some results

of experiment 1 (ie. the amplitude by intensity functions for

the LMSP) argue against a direct relationship between bleach-

ing and the LMSP. Also the differences in time course for

bleaching and regeneration when compared to the similarity

seen in the light-on and light-off LMSP time course is, as

noted in experiment 1, further evidence that neither pigment

nor pigment related processes are the source of the LMSP.




Both experiments 3 and 4 were designed to determine if

the LMSP is generated in the receptors and PE alone or if

neural and glial retinal components are involved. In order

to make this distinction two techniques, one chemical and

one surgical, have been applied. The chemical approach

(this experiment) is based on a number of recent papers

which have shown that sodium aspartate isolates the receptors

from the rest of the retina by acting on the horizontal and

bipolar cells. This has been determined in fish and amphib-

ians (Cervetto and MlacNichol, 1972; Murakami et al., 1975;

Mlurakami et al., 1972) via intracellular recordings. When

sodium aspartate is applied to the retina in minimal con-

centrations the horizontal cells depolarize and their re-

sponse to light stimuli is reduced (Cervetto and MacNichol,

1972). Bipolar cells also are sensitive to aspartate. The

nature of the response varies with the type of bipolar cell

but photic sensitivity is completely suppressed (Murakami

et al., 1975). To account for these effects it has been

postulated that aspartate or a similar substance is the

transmitter released by the photoreceptors (Murakami et al.,


1975). Whatever the mechanism, aspartate appears to isolate

the receptors from the rest of the retina. These obser-

vations have been extended to the mammalian retina by Honda

and Dawson (1975) and H~anitzsch (1973) in the rabbit and

Baron and Boynton (1974) in the monkey. Their studies have

shown that the a-wave of the ERG is not effected by aspartate

but that the b-wave is extinguished. The a-wave of the ERG

has been shown to be a measure of receptor activity while the

b-wave has an inner retinal origin (Brown, 1968; Miller and

Dowling, 1970).

Therefore, sodium aspartate was injected in a single,

or in one case multiple, dose into the vitreous of the cat.

Prior to and after the injections) the LM1SP of that eye was

examined. To confirm that aspartate was effecting the inner

retina but not the receptors simultaneous ERG recording be-

fore and after the aspartate injection were made. It has

been noted that aspartate has other ocular effects in the

monkey, notably constriction of the retinal vasculature and

formation of cataracts (Baron, 1974). All Ss were examined

for such effects.

In these acute experiments the S was set up and stabil-

ized as described above. Preliminary to the injections) of

aspartate the LMSP and dark adapted ERGs were recorded.

After satisfactory records were obtained an intravitreous

injection of sodium aspartate was given via a tuberculin


syringe with a 27 guage needle. Dosages of sodium aspartate

ranged from 3.27 mg. to 33.0. For most experiments an

intermediate dosage of 16.4 mg. in 0.04 cc of sterile water

was employed. The sodium aspartate was prepared from

aspartic acid (Sigma). To achieve the desired concentration

at a pH of 7.3 the aspartic acid was added to sterile water

for injection and the pH was increased by addition of con-

centrated sodium hydroxide.

Before and after injection of the sodium aspartate the

dark adapted ERG was monitored. Since aspartate causes a

decrement in the b-wave of the mammalian ERG (Hanitzsch,

1973; H~onda and Dawson, 1975) this provided the necessary

indication of aspartate effectiveness. At various times

following the injection the light on and off components of

the LMSP were recorded. Interstimulus intervals for both

LMSPs and ERGs were as described in Chapter 3.


The dosage of sodium aspartate employed in this experi-

ment was arrived at in a pilot experiment (data not pre-

sented). An initial dosage of 4.1 mgs, the same dose as

Baron (1974) administered per hour in the macaque monkey,

was initially injected into the vitreous. Following this

injection the b-wave of the ERG was monitored since its de-

crease in amplitude is indicative of aspartate effective-

ness. When no significant effect on the b-wave was ob-

served, increasing dosages of aspartate were administered at


1 hour intervals until a significant effect on the b-wave

(defined as a reduction in b-wave amplitude of at least 50%)

was noted. This effect occurred after a total dosage of

16.4 mg. Therefore a single injection of 16.4 mg of sodium

aspartate was administered in one eye of 3 cats. In all

cases this single injection of 16.4 mg of sodium aspartate

had a profound effect on the ERG and LMSP. In a fourth cat,

to induce a larger effect, a larger dose of 49.4 mg was

administered in 2 intravitreous injections separated by 1

hour in time. This larger dose had an even greater effect

on both dependent variables. In all 3 cases where 16.4 mg

of aspartate was injected, the b-wave of the ERG was reduced

in amplitude by at least 70% while the a-wave remained

largely intact. This is illustrated in Figure 10 which

presents the change in ERG due to a single injection of 16.4

mg of sodium aspartate in the left eye of cat 421. Figure

10a is the ERG recorded from the same eye immediately

preceding the injection of aspartate. A normal a- and b-

wave are present and are so labeled. Records b through f

illustrate a large decline in b-wave due to the aspartate


Figure 11 presents some LMSP traces recorded from cat

421 during the same experiment presented in Figure 10.

Figure Ila is the light-on LMSP recorded in response to the

onset of the unattenuated light. It was recorded preceding

the aspartate injection. The response appears normal.

Figure llb is the response to the same level of irradiance

Figure 10. The effect of sodium aspartate (16.4 mg) on
th~e electroretinogiram (ERG) Record A is the ERG recorded
prior to the aspartate injection. Rourd B through F
demonstrate the change in the ERG as a function of time
after the injection indicated in minutes with each record.
Each record is the sum of 16 signals.





+56 0V




10 MI

Fiue1.Peet helgto n f NS eoddbfr n
afte apr tt.RcrsAadBaetepeadpsaatteo

repne resecivl~y. eodCadDaete r n otsatt

after jection areset idcth ed wig tho eachf ~ record. dbfr n

201 minutes after aspartate injection. The SP light-on

response is severely attenuated. The ERG b-wave was also

severely attenuated by that time after the injection as

illustrated in Figure 10.

Figure 11c presents the light-off LMSP recorded prior

to aspartate injection. It too is large and appears normal

with respect to waveform. Figure 11d presents the off

response 255 minutes after aspartate injection. A cornea-

negative off response is no longer present. The b-wave is

also severely reduced by this time as demonstrated by Figure


A summary graph (Figure 12) presents both the ERG and

LMSP data from four aspartate experiments. Since the rele-

vant parameter with respect to ERG is the decline in b-wave

relative to a-wave amplitude, the b- to a-wave ratio was

calculated for the pre and postaspartate ERGs. The data were

then normalized by using the preaspartate b-wave to a-wave

ratio as 100%. This permitted more direct comparison with

the decline in the on and off LMSP which were also normal-

ized. This was accomplished by employing the preaspartate

on and off responses as the baseline 100% level. Figure 12

therefore represents the effect with time of an injection of

sodium aspartate on the b- to a-wave ratio of the ERG, the

light-on LMSP, and light-off LMSP, all normalized to percent

of preaspartate amplitude.

The data in Figure 12 are from animals which, with the

exception of S 321, received a single ocular injection of

Figure 12. Thle effect of sodium aspartate on the b- to a-wave ratio of the
EPG, light-on LMS~P and light-off U minutes after an injection. For comparison purposes each dependent variable
has been nounalized with the preaspartate value equaling 100%,. The dashed
line (s) in each figure indicate an injection of aspartate. Note that S 321
had a second injection of aspartate. Each symbol represents on of the
dependent variables; ERG, light-on UBSP, light-off UISP,.

100 -r.0 4 9 o 4


oo o

C 20-O

1111illi IIA [ I

100 -*.0 ao 44 "~' 321

80~ -- a

60C I OO




8 0 160 240 320 80 160 240 320

T IME (mi a)

16.4 mgs of aspartate in 0.5 ml of solution at time 0. Some

of the data in Figure 12 from cat 421 are presented pictor-

ially as Figures 10 and 11. Cat 321 had 2 injections of

sodium aspartate. The first was 16.4 mg at time 0; the

second given 62 minutes later was 32.8 mg in 0.1 cc of


The ERG data from cat 419 best illustrate the orderly

and gradual decline in b-wave amplitude relative to a-wave

amplitude following a single injection of sodium aspartate.

In this cat by 120 minutes after the injection the b- to a-

wave ratio has declined from the preaspartate value over

50%. The large gaps in time for ERG records are unavoidable

since it is not possible to simultaneously record ERGs and

LMSP without the stimuli which elicit each interacting and

contaminating the results of both. The ERGs of the other 3

cats follow the same pattern with the time course and the

ultimate reduction in b- to a-wave ratio varying somewhat.

The greatest decrement in ERG occurred for cat 321 who

received the second injection of aspartate.

In all cats (Figure 12) the decrease in both the light-

on and off LMSP consistently occurred in tandem with the

decrease in b- to a-wave ratio. In cat 44 where a light-on

and off response were collected in the initial 80 minutes

after the injection the decrement is moderate (40-60%) as

would be expected from the reduction in b- to a-wave ratio

usually seen at these times after a single aspartate in-

jection and the ERGs recorded from this cat. In cat 419 the


decrease in both LMSP and b- to a-wave ratio was moderate

and of approximately the same value, both reaching about 30%

at 200 min (Figure 12). In cat 421 and 44 (Figure 12) the

b-wave reduction after 120 minutes is greater than cat 419

and the LMSP values are also somewhat lower (0-25%6). These

intersubject variations in the time course of the effect are

probably related to the precise place of the injection in

the vitreous and the associated diffusion gradient. In cat

321 where a second injection of aspartate was administered

to induce a larger decrement in b-wave than seen in the

other experiments the ERG b-wave and LMSP were both essen-

tially abolished by 120 minutes after the initial injection

(60 minutes after the second injection). It can be con-

cluded that sodium aspartate severely reduces the b-wave of

the ERG relative to the a-wave and that the LMSP is sim-

ilarly reduced.

As a control experiment a cat was injected with 0.5 ml

of sterile water in the vitreous. The pH was adjusted to

7.3 to equal the pH of the aspartate solution. The ERG and

LMSP were recorded prior to and at various times after the

injection. No effects of the control injection were obsery-

ed on either the ERG Or LMSP, up to 150 minutes after the


At the conclusion of all experiments the cat was exam-

ined ophthalmoscopically for retinal vascular constriction

and cataracts which have been reported to occur after

aspartate injection (Baron, 1974). In cat 321, which re-

ceived a double injection of aspartate, a cataract was

present. Due to the reduced visibility of the funds of

this cat it was impossible to determine if the retinal

vasculature had constricted. Neither cataracts nor vaso-

constriction were noted in the other cats.


The effects of sodium aspartate on the ERG and LM1SP

were unequivocal. Following an injection of aspartate into

the vitreous the b-wave began to decline steadily. A simul-

taneous decrement in on and off LMSPs also occurred. Most

importantly when the b-wave was reduced to about 20% or less

of its preinjection amplitude relative to the a-wave, the

light-on and off LMSPs were similarly reduced or absent.

The near abolishment of the b-wave and retention of the a-

wave is a characteristic effect of sodium aspartate on the

retina of vertebrates and invertebrates (Baron and Boynton,

1974; Cervetto and MacNichol, 1972; Hanitzsch, 1973; Honda

and Dawson, 1975; Murakami, 1972; Murakami, et al., 1975).

Based on intracellular recording in invertebrates it has

been concluded that sodium aspartate isolates the receptors

from the rest of the retina by saturating the horizontal and

bipolar cells (Cervetto and MacNichol, 1972; Murakami 1972;

Murakami, et al., 1975). The precise mechanism of this

effect is not known although it has been postulated that

aspartate or a similar substance is the transmitter released

by the photoreceptors. In mammals, where intracellular


recording techniques have not yet been perfected, the effect

of aspartate on the ERG is the same as seen in inverte-

brates. Furthermore, a large body of literature exists

which demonstrates that in mammals the a-wave is a receptor

based phenomenon while the locus of the b-wave is the inner

retina (Brown, 1968; Granit, 1933). It is clear, therefore,

that when an ERG is recorded with a severely reduced or

absent b-wave but a remaining a-wave, the inner and outer

retina have effectively been separated. The presence of an

a-wave in such a preparation indicates that the receptors

are still functional and that the preparation has not yet

generally deteriorated. Additionally the control experi-

ments, with an injection of an equivalent volume of a

control solution instead of sodium aspartate, had no effect

on the ERG or LMSP indicating that the aspartate effects

were not artifactual. If intraocular pressure changes or

other injection induced trauma induced the alterations in

the dependent variables then similar effects would result

following the control injection. It is concluded, there-

fore, that some postreceptor components of the retina either

neural or glial must be functioning normally for the gener-

ation of LMSPs. This conclusion is further reinforced by

the results of the retinal vascular occlusion experiments.

In the only other in vivo experiments on the effect of

sodium aspartate, Baron (1974) and Baron and Boynton (1974)

reported that in monkeys certain other ocular effects of the

chemical were observed. Specifically by 30 minutes after


the start of infusion of sodium aspartate into the vitreous

vasoconstriction of the retinal vasculature was noted. In

addition within 3 hours they noted the formation of a

posterior cataract in the eye being infused. No such

effects were observed in these experiments with the ex-

ception of cat 321 which received a second injection of

sodium aspartate. It is difficult to account for this

species difference since the dosage here was larger than

Baron's (1974). However, the effect of aspartate on the ERG

is constant across all vertebrates examined (frog, rabbit,

monkey, and cat) whether the preparation is in vivo or in

vitro (Baron and Boynton, 1974; Cervetto and MlacNichol,

1972; Hanitzsch, 1973; Honda and Dawson, 1975; Murakami, et

al., 1972; Murakami, et al., 1975). It is therefore unlikely

that the in vivo effect is due to retinal vasoconstriction

although this cannot be ruled out since such vasoconstriction

would at least partially lesion the inner retina while

leaving the receptors intact.




To surgically lesion the inner retina, the retinal

blood supply was interrupted thereby destroying the inner

retina while leaving the receptors intact. The receptors

are preserved because they are nourished by the choroidal

circulation (Brown, 1968). The effectiveness of such a

lesion has been confirmed histologically in the macaque

monkey by Brown (1968) and squirrel monkey by Kroll (1968).

Using electron microscopic techniques Kroll (1968) found

that by 16 hours after occlusion all neural and glial cells

vitread to the outer nuclear layer were severely autolyzed.

The receptors were spared. This technique therefore has

been employed to isolate receptor driven electrical po-

tentials in the cat and other species (Brown, 1968). When

the retinal blood supply is occluded in cat and monkey, the

b-wave of the ERG is eliminated very rapidly but the a-wave

remains intact (Brown, 1968; Maffei and Poppele, 1968).

After 20 days Brown (1968) using light microscopy, found the

inner layers to have degenerated with the exception of only

a few cells while the photoreceptors appeared intact.

This technique was therefore applied in the cat to

create the same type of lesion. At various times after the

lesion (ranging from 1-55 days) the LMSP and ERG were exam-

ined. The extent of damage was verified histologically at

various times after the lesion.

Experiment 4, like experiment 3, therefore determined

using a completely different technique if the inner retina

was participating in the generation of the LMSP.


A variety of techniques were tested before a successful

lesion of the inner retina could be generated. The retinal

blood supply of the cat is unlike that of monkey or human in

that there is no single central retinal artery in the optic

nerve where it enters the eye (Wong and Macri, 1964).

Rather a number of arteries (3-4) enter the eye between the

optic nerve and adjacent sclera. Because of the complex

nature of the retinal vasculature first attempts to sur-

gically tie off or cut the vessels were unsuccessful. At

this point an attempt was made to electrocoagulate the

retinal vessels using a specially designed probe inserted

into the eye and placed above the major retinal vessels as

they left the disc. This invariably resulted in severe

bleeding into the vitreous and the procedure was discon-

tinued. Finally, an attempt was made to lesion the retinal

vasculature with a xenon arc photocoagulator. In this

procedure brief flashes of high intensity light were di-

rected at the optic disc successfully sealing the retinal

blood supply of one eye of the cat.

Three cats were the Ss for this experiment. The lesions

were all made in a single experimental session. At various

times after the lesion was made, ranging from one to fifty-

five days, the LMSP was measured in response to the onset

and cessation of the unattenuated stimulus. The general

methodology followed is described above in Chapter 3. For

control purposes the LMSP of the other eye was also re-

corded. In addition dark adapted ERGs were recorded from

both eyes to confirm the extent and nature of the lesion.

After Each Ss had been tested in this manner on at

least one occasion following the photocoagulation of the

optic disc, it was sacrificed. The 3 subjects were sacri-

ficed at 12, 21 and 55 days after the initial lesion was

made. The eyes were immediately removed, a slit was made at

the ora serrata and they were placed in Karnovsky's fixative

(Karnovsky, 1965). After at least 24 hours of fixation at

4oC the anterior chambers were removed and portions of the

posterior pole were embedded in celloidin. Various regions

of peripheral and central retina of each eye were sampled

for all eyes. Tangential sections were cut at 20 9, stained,

examined, and photographed. The stain employed was cresyl

violet. Some tissue was also embedded in plastic, sectioned

at 2 p, and stained with toluidine blue.


Immediately after the lesion the disc region appeared

very pale. It was apparent that the retinal vasculature was

lesioned since blood was no longer flowing in the large

vessels. In one of the cats, 6H44, a small amount of

bleeding into the vitreous occurred. After an hour it

appeared to be about 3 disc diameters wide. Subsequently it

was largely reabsorbed. At various times after the lesion

the ERG and LMSP of both eyes of each cat were recorded.

Cat 6H44 and 2B9 were each recorded from on two separate


Figure 13 presents the ERGs recorded from cat 7B9 one

day after the lesion. Figure 13a recorded from the control

eye is a normal ERG. Figure 13b is the ERG recorded from

the lesioned eye. The b-wave is severely depressed but a

small component is still present. The a-wave is intact.

The on and off LMSP recorded from this eye during the same

session were also reduced as illustrated by Figure 14.

Figure 14a is the on LMSP from the control eye and Figure

14b is the light-on LMSP recorded from the lesioned eye.

The amplitude of the response of the normal eye is 2.1 mV

while the amplitude of the lesioned eye's response is 1.6

mV. The off LMSP recorded from the control and experimental

eyes are presented in Figure 14c and d respectively. The

control eye amplitude is 0.6 mV and the lesioned eye ampli-

tude is 0.35 mV.

Figure 13. ERG recorded fromn the control eye
(Figure 13A) and from the lesionedl eye (Figure 13B)
of cat 7B9 one day after the optic disc was photo-
coagulated .


10 MIN

Figure 14. A, B, is the light-on INSP recorded respectively
fromn the control eye and experimental eyes of cat 7B9 one day
after the optic diisc was photomoagulated. Traces C and D
present the light-off response of the control and exp~erimental
eyes .


The LMSP and ERG data collected from all the Ss in the

vascular lesion experiments are summarized in Figure 15.

Figure 15a,b,c represent respectively the b- to a-wave

ratio, the light-on LMSP and the light-off LMSP recorded in

the control and experimental eyes as a function of time in

days secceeding the lesion. Since like the aspartate

experiments, the relevant parameter for the ERG is the

decline in b-wave relative to the a-wave amplitude, the b-

to a-wave ratio was calculated for the control and experi-

mental eyes. The data were then normalized by using the

mean of the control eye value as 100%. This permits more

direct comparison with the on and off LMSP data which were

also normalized in the same way i.e. the mean amplitude

recorded from the control eye equaling 100%. On day 1, cat

7B9, a small b-wave was present and the on and off LMSP of

the lesioned eye had declined to about 50 or 60% of the

control eye amplitude. These data were displayed in part in

Figures 13 and 14. On day 6, cat 6H44 was the S and his b-

wave and both the on and off LMSP were absent (Figure 15).

Twelve days after the lesion cat ICH was run. The b-wave

was absent, the off LMSP was not present but a small on re-

sponse, 20% or less than the control eye, was observed

(Figure 15). Twenty one days after the initial lesion when

6H144 was again the S all measures in the experimental eye

were still absent. Cat 7B9 was recorded from again 55 days

after the lesion. The on and off LMSP were severely reduced

and the b- to a-wave ratio was still at about 10% as seen in

Figure 15. Vascular lesion experiment suam~ary figure. The
effect of lesioning the retinal vascular supply on the b- to
a-wave ratio of the ERG, Figure 15A, the ligjht-on LFEP,
Figure 15B and light-off TDLSP, Figure 15C as a function of
time after the lesion in days. For comparison purposes each
of these dependent variables has been normalized with the
control eye value recorded in the same session equaling 100%
(if more than one response was collected the rrean was used
as 100%). Recordings from control eyes are indicated by an
and from experimental eyes with a The data presented
for each day wa~s recorded from a single subject. The S for
each experiment is indicated above each day.

789 SH~44


I) ~O

I **I_ a /

S6 12 #* /
TIME (da ys)

SH44 1BS




Figure 15. The records from this session are presented in

Figure 16 and 17. Figure 16 presents the on and off LMSPs

and Figure 17 presents the ERGs.

As a further summary means, standard deviations and

sample sizes were calculated for each dependent variable

i.e. b/a wave ratio of the ERG, the on LMSP, and the off

LMSP (Table 2). Data from all the cats recorded from after

day I were pooled (Table 2b); the data from 7139 recorded from

on day one are presented separately in Table 2a since the

LMSPs of that experiment were clearly different from other

postlesion LMSPs. From the data recorded from all the

animals it is apparent that by 6 days after the occlusion of

the retinal blood supply the LMSP is absent or tremendously

reduced. The greatest reduction in b-wave occurs by one day

after the lesion and at this time the LMSP is reduced by

about 50%. By 6 days post-lesion the on and off LMSPs are

further reduced or absent.

Figure 18 presents photographs of the normal and

lesioned retina embedded in celloidin. Figure 18a presents

a photograph of a section from the control eye of cat 7B9

cut at 20p and stained with cresyl violet. All layers of

the retina are present and labeled. Figure 18b presents

the retina of the experimental eye which was lesioned 55

days before. Only the receptors remain. Their nuclei are

normal in appearance and number. All central and peripheral

areas surveyed in this cat (789) had receptors present

except for the disc region. The PE was normal throughout.

Figure 16. Figure 16A, B, is the light-on IFISP
recorded respectively from the control eye and
experimnwtal eyes of cat 7B9 55 days after th~e
retinal vasculature was photocoagulated. Traces
C and D, present the light-off response of control
and experimental eyes.

10 MIN



Figure 17. ERG recorded from the control (A) and experimental (B)
eyes of cat 7B9 51 days after lesioning the optic disc.

Table 2: Means, standard deviations and sample sizes
for the B/A-wave ratio of the ERG, the light-on and
the light-off LMSP in the control and exp~erinental eyes.
Table 2A is the data recorded from cat 7B9 on day 1.
Table 2B3 is the pooled data recorded from all cats run
on day 6 through day 55.

B/A Ratio Light-On Light-Off

nean-control ee3.8 1. 52 mV .63 mV

standard deviation-
control eye .28 .12 .06

sarple size-control eye 2 2 2

mean-lesioned eye |.33 .91 .32

standard deviation-
lesioned ee.01 .16 .05

suggle size-lesioned eye 4 3 3

B All Other SS, Days 6 to 55

B/A Ratio Light-On Light-Off

nsean-control eye 2. 93 1. 51 mV .84 mV

standard deviation-
control eye .87 .29 .26

sample size-control eye 7 7 6

meuan-lesioned eye .14 .10 .01

standard deviation-
lesioned ee.10 .141 .02

sanp~le size-lesioned eye 8 8 10

A Cat 7B9, Day 1

Figure 18. Tangential sections from the normal (A) and
experimental (B) retinas of cat 7B9. The tissue was embedded
on cellodlen, cut at 20p1 and stained with cresyl violet.

r sclera


outer sognarnts

receptor nuclei
outer plexifonn

ST~-l' ~ :' ~:5-~,'5';r"1~?'gang~lion cell layer


outer segments

Ireceptor nuclei
remnants of inner


To better demonstrate that the outer and inner segments of

the receptors were present in the lesioned eyes tissue from

the lesioned eye of cat 6H44 embedded in plastic is shown in

Figure 19. The receptor inner and outer segments are

present although the retina became detached during proces-

sing. It should also be noted that the magor nonneural glial

cell of the retina, the Mluller cell, also appeared to be

absent. The Muller cells were probably not present since

their nuclei and most of their cytoplasm lies in the inner

retina (Miller and Dowling, 1970).

When S ICH's experimental eye was examined most areas

of the retina had normal appearing receptors but no inner

layers. An area of approximately 10% of the retina, how-

ever, was largely degenerated with only a few receptors

intact. Some small areas of reduced receptor density were

also observed. The PE appeared normal throughout. When the

retina from the experimental eye of S 6iH44 was examined a

fairly large receptorless area was observed (40%). Some

areas of thinned receptors were also present. Areas of

normal appearing receptors without inner retinal layers were

also present. All areas of the PE examined appeared normal.

It should be noted that in all cats there were always

a considerable number of normal appearing receptors remain-

ing. Additionally an a-wave, indicative of functional

receptors, was also always present. There were however

some small differences across cats with respect to the ERGs.

For cat 7B9 where all the receptors were present except

Figure 19: Tangential section from the
lesioned eye of cat 6H44. The tissue was
embedded in plastic and sectioned at 2 p
and stained with toluidine blue.







remnants of
inner layers


those around the optic disc the a-wave was only 14% lower

on day 55 in the experimental eye, than in the control eye

(Figure 16). Cat 6H44 and ICH had a-waves which were re-

duced by approximately 25% (measured at the last experi-

mental session for each cat) which is surprising considering

the extent of the receptorless area in cat 6H44.

In cat 7B9 a larger b-wave was preserved also indi-

cating that the lesion was not as severe. Only a miniscule

b-wave or none at all was observed in the ERGs of the experi-

mental eyes of the other two cats indicating a more severe

lesion than in 7B9.


Lesioning the retinal blood supply eliminating the

inner retina but leaving the receptors intact is a standard

technique in cat and monkey (Brown, 1968; Maffei and Pop-

pele, 1968). Like aspartate, lesioning the inner retina

eliminates the b-wave yet leaves the a-wave intact.

As expected photocoagulating the optic disc of the cat

had the predicted effect on the ERG, reducing or eliminating

the b-wave although the a-wave was still present. As in the

aspartate experiments severe reduction in ERG b-wave to a-

wave ratio occurred in tandem with severe reduction in the

LMSP responses in all lesioned eyes. On the first day after

the lesion the decrement in ERG was more severe than the re-

duction in LMSPs. Between the first and sixth days after

the photocoagulator lesion the LMSP is reduced by BO% or


more as is the b- to a-wave ratio of the ERG. Both signals

remain depressed up to 55 days after the initial lesion, the

longest time at which a measurement was made. The nature of

the lesion was confirmed histologically demonstrating that

the photoreceptors were present but the rest of the neural

retina was absent.

In two cats the lesion was more severe and some photo-

receptor change was observed. This was surprising since

interrupting the retinal circulation of the cat has been

used by other investigators to reportedly produce the same

lesion. Brown (1965) reported the photoreceptors were

intact but the inner retina was absent in the cat following

occlusion of a branch vessel of the retinal arteries.

Maffei and Poppele (1968) photocoagulated the optic disc of

cats and report only the a-wave of the ERG remained but pre-

sented no histological data to confirm the extent of the

lesion. There are two interpretations of the receptor

damage found in two of the Ss in this experiment. It could

be either that the retinal circulation in some cats supplies

parts of the inner retina or that the choroidal circulation

was damaged by the photocoagulation. With respect to the

first hypothesis it should be pointed out that although the

retinal vasculature is reported to supply only the inner

retina in the cat, variability in vascular patterns across

individual Ss is a well known occurrence. The second

hypothesis also cannot be ruled out. Choroidal arteries lie

directly around the disc region of the cat (Wong and Macri,


1964) and either through direct damage or edema may have

been temporarily or permanently occluded. Fortunately in

two of the cats the damage to the receptors was either nil

or not severe. It can be concluded therefore that the

inner retina is necessary for generation of the LMSP. This

experiment therefore confirms the findings of the aspartate

experiment via another totally different technique.



The major finding of this study is that the inner

retina is involved in an essential fashion in the generation

of the LMSP. In addition it was demonstrated that the

LMSPs elicited by a flickering light are not readily dis-

tinguishable from the LMSPs elicited by a steady light of

equal time averaged irradiance. When this result is con-

sidered within the context of the aspartate and vascular

lesion experiments one must conclude that the LMSP is

probably generated by a nonneural element of the inner

retina. This elements) must meet at least two require-

ments. First, it must be situated at least in part vitread

to the receptors on the inner retina. Second, it must be

relatively insensitive to the rates of flicker tested in

this experiment which are known to be psychophysically and

physiologically distinguishable by the cat (Dodt and Enroth,

1954; Kappauf, 1936 cited in Brown, 1965).

A number of authors have suggested that the PE modu-

lated by the receptors is the site of the LMSP (Arden et

al., 1962; Nilsson and Skoog, 1975; Skoog, 1975; Taumer et

al., 1974). The PE Obviously does not meet the first

requirement stated above i.e. it is not located in the inner

retina and there is no evidence that it is in any way re-

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs