Autosomal dominant rod-cone dysplasia in the Rdy cat

Exp. Eye Res. (1991) 53, 489-502

Autosomal

Qominant

Rod-Cone

Dysplasia

2. Electrophysiological A. LEON*, Comparative

Ophthalmology

(Received 27 August

A.A. HUSSAIN

in the Rdy Cat

findings AND

R. CURTIS

Unit, Animal Health Trust, Lanwades Park, Kennett, Newmarket, Suffolk CBB 7PN, U.K. 1990 and accepted in revised form 16 January

1991)

Electroretinography was performed on cats affected with autosomal dominant rod-cone dysplasia (gene symbol Rdy). In normal kittens it was found that retinal sensitivity increased and rod thresholds decreasedas the animals matured. Electroretinogram (ERG)amplitudes were mature by 4.5 weeks and adult timing was attained by 6 weeks of age, consistent with the findings of other workers. In Rdy-affected heterozygous kittens the ERG was absent or barely recordable using conventional cornea1contact lens electrodes.However, the enhanced sensitivity of an intravitreal needle electrode permitted the recording of ERGSfrom affected kittens aged 4.5 weeks and older. The intravitreally recorded scotopic ERGin Rdyaffected kittens was a very low amplitude, largely negative responsewith prolonged a- and b-wave timesto-peak (two to threefold longer than in comparable recordings from an age-matchednormal kitten). The b-wave lacked oscillatory potentials and was relatively small so that the ERGwas a-wave dominated. This was attributed to delayed and defective synaptogenesis in the outer plexiform layer of dystrophic retinas. In contrast to normal kittens, the b-wave threshold was higher than that of the a-wave in affectedkittens. Photopic responseswere unrecordable. The intravitreal ERGwas barely recordable in a 5-month-old Rdyaffectedcat and was apparently extinguished by 7 months of age. In vitro electroretinography permitted a comparison of the photoreceptor responses(fast PIII) from the isolated retinas of 6-week-old control and Rdy-affectedheterozygous kittens. Maximum fast PI11amplitudes were reduced by about 75 % in affected retinas compared with age-matched normal retinas (P < 0.005). The mean fast PI11time-to-peak, at maximum light intensity, in Rdy-affectedretinas was prolonged by about 15 msecand was approximately twofold longer than the time-to-peak of normal retinas (P < 0005). The fast PIII response threshold (light intensity at 5 ,uV response)was increased by 1.0 log unit in affected retinas, and there was also a 0.94 log unit reduction in photoreceptor sensitivity (g), indicated by the shift of the amplitude-response curve (V-log 1function) towards higher light intensities and a corresponding increase in semi-saturation luminance. These findings correlate well with the paucity of photoreceptor outer segment material, and therefore reduced rhodopsin content, seen histologically in affected retinas. The altered slope of the fast PI11V-log I function of Rdy-affectedretinas indicated a smaller increase in relative fast PIII amplitude per log unit light intensity in dystrophic photoreceptors. The fast PIII time-to-peak vs. light intensity curve (T-log I function) of Rdy-affectedretinas had a steeperslope with relatively greater prolongations in timeto-peak at lower luminances compared with normal retinas (P < 0.025). These changes in temporal characteristics may be explained either by severe disorganization of photoreceptor outer segmentsor by altered phototransduction kinetics. Key words: autosomal dominant rod-cone dysplasia : Rdy cat : electroretinography : fast PI11response. cell layers. By 30 weeks of age only two to five rows of nuclei remain in the outer nuclear layer and in adult

1. Introduction In the preceding paper (Leon and Curtis, 1990) we presented the results of histopathological and autoradiographic studies on retinas from cats with auto-

somal dominant rod-cone dysplasia (gene symbol Rdy). Affected kittens manifest a photoreceptor dysplasia, involving both rods and cones, in which the visual cells show abnormal and retarded development. The abnormal development phase is superseded by a

degenerative phase, beginning at 4.5 weeks of age, which results in progressive loss of the photoreceptor * For correspondence at : Visual Neurosciences Unit, JohnCurtin of Medical Research,Australian National University, Canberra,ACT 2601, Australia. School

00144835/91/100489

+ 13 $03.00/O

cats this layer has been reduced to a discontinuous single row of nuclei. Clinically, heterozygous Rdy-affected kittens have dilated pupils and sluggish pupillary light reflexes compared with normal littermates, and a characteristic pendular nystagmus develops by 4-6 weeks of age. Preliminary electroretinographic studies have shown that the electroretinogram (ERG) of affected kittens is unrecordable using conventional recording methods (Curtis, Barnett and Leon, 1987). However, behavioural studies (using obstacle courses and other tests) and recordings of visual evoked cortical potentials (VEPs)have demonstrated that by 6 weeks of age some useful, but impaired, vision does develop in 0 1991 Academic Press Limited

A. LEON

490

kittens heterozygous for the Rdg gene (Leon, 1989). Vision then deteriorates over several weeks and affected cats are effectively blind by 5-6 months of age. When recordable, VEPs from Rdg kittens were lowamplitude responses, abnormal in waveform, with prolonged times-to-peak (Leon, 1989). In this paper we present an electrophysiological analysis of the retinal dystrophy in Rdg-affected cats using both in vivo and in vitro electroretinographic procedures.

2. Materials and Methods In Vivo Electroretinographg Experimental animals. Table I lists those cats which were used for in vivo electroretinography. ERG recordings were performed on 26 cats representing both Rdg-affected heterozygous and control normal animals at various ages. Recording equipment. ERGS were recorded using a Medelec MS6 system (Medelec Ltd, Old Woking, Surrey) comprising a DF06 main frame. AA6 Mk. III A.C. amplifier with associated PA62 pre-amplifier,

TABLE I

Cats usedfor in vivo electroretinographg

Cat

Age

identity

Breed

Controls : AB8 AB3 XXXIVEN AB6 XXVTH XXVTH ABll XXIVSH AB13

Abyssinian x Abyssinian Abyssinian Abyssinian Abyssinian Abyssinian Abyssinian Abyssinian Abyssinian

Affected : Abyssinian AB9 Abyssinian AB4 XXXIVED Abyssinian Abyssinian AB5 XXXVFE Abyssinian Abyssinian AB19 Abyssinian AB19 Abyssinian ABlO XXVIIIWYAbyssinian XXXIBL Abyssinian AB26 Abyssinian Abyssinian AB12 AB29 Abyssinian Abyssinian AB30 AB30 Abyssinian AB23 Abyssinian Abyssinian AB18

Sex (weeks)

x x x x x x x x x

M F M

2 3 3

F

4.5

F F F M

4.5 6 6 6 F 11

M M F M M F

2 3 3 4.5 4.5 4.5

F

6

M M F F M M F F M4 F5

6 6 6.5

9 11 22 17 30 years years

Cornea1 contact Intralens vitreal electrode trode

+ + + + + + + + +

+ -

+ + + + + + + + + + + -

+ + + + + + + + + + +

ET AL

DAVh2 digital averager and AX62 averager expander. The signal was viewed on a cathode ray tube oscilloscope display and recorded with an integral ultraviolet light fibre-optic photographic unit on Kodak professional linagraph direct print paper. The instrument was set to a frequency bandpass of 0.X-l 60 Hz and a 200-msec analysis time. Depending on the signal amplitude and quality, either single responses or an average of up to 16 responses I with an interval of at least 10 set between flashes) were recorded. :1 disposable cornea1 contact lens incorporating a gold ring electrode and lid-retracting posts (ERG-jet. Universe S.A.. La Chaux-de-Fonds. Switzerland) was used as the active electrode in cats aged 4 weeks and over. Because of the small size of the palpebral aperture a 9-mm diameter perspex contact lens with a platinum wire electrode was substituted in kittens younger than 4 weeks. Electrical contact with the cornea was achieved by using a mixture (1 : 1 by volume) of 1.0% hydroxypropylmethylcellulose (Isoptoplain. Alcon Laboratories, Watford, Hertfordshire) and normal saline. This method generally failed to record the very low amplitude ERGSof affected kittens and, thus, towards the latter part of the study. an intravitreal MF12 or MF2 5 monopolar insulated stainless steel needle electrode (Teca Corporation. Medelec Ltd, Old Woking, Surrey) was used as the active electrode. This was inserted, with the animal under general anaesthesia, at the superotemporal quadrant through the pars plana ciliaris in the direction of the optic disk, so that the electrode tip was positioned in the posterior vitreous. In all cats in which this procedure was performed there were no incidences of intraocular haemorrhage nor other signs of tissue damage. MF12 monopolar electrodes were inserted subcutaneously adjacent to the lateral canthus and in the pinna of the right ear to act as reference and ground electrodes, respectively. Eye movement was minimized by traction and fixation of the membrana nictitans. Any subsequent eye movements would have had negligible effects on the results because of the very brief duration and large angular size of the flash stimulus. I.ight intensities were measured using a Tektronix ] 16 digital photometer with a Tektronix J6523 1” luminance probe (Tektronix U.K. Ltd, Harpenden, Hertfordshire). White light stimuli of 10 lcsec duration were delivered by a Grass PS22 photostimulator (Grass Instrument Co.. Quincy. MA) with flash intensities ranging, in 0.3 log unit steps. from 1.05 x IO” cd m-” (setting 1) to 18.0x 1O’cd m-‘-’ (setting 16). The photostimulator was positioned 30 cm from the eye, in the cat’s visual axis. In addition to the variable intensity settings on the stimulator, light intensities were varied using neutral density filters (Kodak Wratten Neutral Density Filter, Kodak Ltd, Rochester, NY) interposed in the light path. Filter densities increased in 0.5 log unit steps up to 4.0 log units. Red

ROD-CONE

DYSPLASIA

IN THE Rdy CAT

(KW 26) and blue (KW 47) filters were also used to examine cone and rod function. In vivo ERG recordingprotocol. The pupils of all cats were dilated 2 hr before electroretinography with topical 1 Y0 atropine sulphate (Isoptoatropine, Alcon Laboratories, Watford, Hertfordshire) and 10% phenylephrine hydrochloride (Minims, Smith and Nephew, Romford, Essex)eye-drops.The animals were then dark-adapted for 2 hr. General anaesthesia was induced in dim red light by intravenous injection of 12 mg ml-’ alphaxalone and alphadolone acetate (Saffan, Glaxovet, Uxbridge, Middlesex) to effect, and, after intubation, maintained with halothane (Fluothane, Coopers, Crewe, Cheshire) in oxygen. Electrodes were then positioned as described above. Halothane has been shown to affect the ERG by reducing a- and b-wave amplitudes in a dosedependent manner, whilst times-to-peak remain unchanged (Raitta et al., 1979; Tashiro et al., 1986). Therefore, in the present study the depth of anaesthesia was maintained at a light level, and acceptable ERGSwith b-wave amplitudes over 500 PV were often recorded from control cats. With the cat in total darkness, scotopic photoresponses were recorded to white light flashes with intensities increasing in 0.3 log unit steps,followed by flashes of red and blue light of maximal intensity. The photopic response was unrecordable in Rdy-affected kittens and attempts to record flicker responsesfrom affected kittens were largely unsuccessful due to the low amplitude and prolonged response time of these ERGS. The amplitudes and times-to-peak of the ERGa- and b-waves were recorded digitally from the signal on the oscilloscopedisplay using a strobe marker. The a-wave amplitude was measured from the isoelectric line to the a-wave trough, and b-wave amplitude was measured from the a-wave trough to the b-wave peak. The results were analysed by plotting absolute a- and b-wave amplitudes (V) against log relative light intensity (V-log I function). A computer program was used to determine the probability of these amplituderesponse curves fitting a sigmoidal distribution and to provide a line of best-fit. This program also calculated the maximum photovoltage amplitude (V,,,), since the output of the light source was frequently insufficient to achieve response saturation. The amplitude data were also normalized by plotting V/V,maxagainst log relative light intensity. The temporal characteristics of the a- and b-waves were determined by plotting time-to-peak values against log relative light intensity (T-log I function).

In Vitro Electroretinography Experimental animals. In vitro electroretinography was performed on the retinas from ten kittens and two adult cats. Six kittens were Rdy-affectedheterozygotes.

491

all aged approximately 6 weeks, and four kittens were normal controls (three aged approximately 6 weeks and one aged 11 weeks). Most in vitro ERGSwere recorded from 6-week-old kittens for several reasons. The ERG of the normal cat has attained adult amplitude by this age (Ikeda and Jacobson, 1982 ; Hamasaki and Maguire, 1985). In affected heterozygous kittens at 6 weeks of age there are no ophthalmoscopic signs of retinal degeneration, useful vision can be demonstrated by behavioural tests and recordings of VEPs and, histologically, retinal degeneration is in its early stages (photoreceptor degeneration begins at 4.5 weeks of age). Thus, a reasonable comparison can be made between the photoreceptor responsesfrom affectedand normal cats at this time. The two adult cats used consisted of a 4year-old Rdy-affected heterozygous Abyssinian and a 1 S-year-old normal domestic short-haired animal. Recording equipment. Transretinal photoresponses were recorded from freshly dissectedportions of retina after mounting in a specially constructed chamber [Fig. l(A)]. This chamber was constructed from transparent perspex and consisted of a central recording chamber surrounded by two communicating, water-filled side chambers which were connected in turn to a circulating water bath. The water bath was thermostatically heated and maintained the temperature of the buffer in the recording chamber at 37-38°C. The recording chamber was filled with one of several buffers, depending on the experiment. The initial buffer used was a dissection medium-a slightly modified version of Minimum Essential Medium (MEM, Sigma). The composition of this and the other buffers used are given in Table II. In dim red light a rectangular Teflon holder mounted with a dark-adapted retinal specimen was inserted into a greased,vertical slot in the buffer-filled recording chamber, creating two water-tight compartments on either side of the retina [Fig. l(A)]. Each of the two compartments was aerated with 95 % O,/ 5% CO,, adjusted to a gentle stream so as not to dislodge photoreceptor outer segments. Transretinal photoresponses were recorded by silver/silver chloride electrodes in the two buffer-filled compartments on either side of the retina. The electrodes were connected to a PA62 pre-amplifier which, together with the recording chamber, was encased in a Faraday cage to eliminate electrical interference. The cage was light-tight except for an aperture through which the retina was stimulated by flashes of white light delivered via a fibre-optic light guide from a Kowa SL2 tungsten light source. This source provided a maximal light intensity of 1.5 x lo6 cd rne2. A motor-driven camera shutter system placed between the light source and the fibreoptic light guide yielded 8-msec flash stimuli. The light intensity was varied by interposing neutral density filters (Kodak Wratten Neutral Density Filter, Kodak Ltd, Rochester, NY) between the light source and the

A. LEON

492

ET AL

ilead stage preamp!lfser

(A)

95%0,/W- co;? -

Silm~s~:ver chloride : electrodes

-7)

5% AGAR/O.9%

Side chamber/ wOter both

I I I Retina on millipore

,

Teflon insert dividing recording chamber

Side chamber/ water bath

NaCl

L ____ -.-_--.: Medeiec main-frame

Aperture for fibre optic light guide

Mlllipore filter

(B)

Teflon

insert I

Retina

FIG. 1. In vitro electroretinography chamber. (B) illustrates how the retinal specimen is mounted on to the Teflon holder before insertion into the recording chamber.

TABLE II

Composition of buffers used in in vitro electroretinography Constituent NaCl KC1

CaCl, .2H,O KH,PO, MgSO, .7H,O

NaHCO, Hepes Glucose

Albumin Glutamine Na glutamate Verapamil

Dissection Glutamate buffer medium 120 rnM 2.3 mM 2.0 rnM 1.0 rnM 1.2 rnM 13 rnM 10 rnM 10 rnM

110 rnM

0.01 %

0.01%

2.0 rnM

-

2.3 mM 2.0 rnM

1.0 rnM 1.2 rnM 13 rnM 10 rnM 10 rnM 2.0 rnM

10 rnM

10 w%

fibre-optic guide. Filter densities increased in 0.5 log unit steps. The transretinal electrical response was passed from the pre-amplifier to a Medelec polygraph system where the signal was displayed and recorded as described above. Single responses were recorded using a bandpass of 0.8-160 Hz and an 800 msec analysis time.

Preparation of retina and ERG recording protocol. After 2 hr of dark adaptation, and with the cat under deep general anaesthesia in dim red light, the animal’s left eye was enucleated and placed in an ice-filled lighttight container. Within S-10 min of enucleation, and under dim red light, the eye was placed in a Petri dish containing pre-aerated dissection medium at approximately 37’C and the anterior segment with vitreous removed. After cutting the posterior eyecup into quadrants, the retina was lifted away from the underlying retinal pigment epithelium using fine forceps and floated in the surrounding medium. A piece was cut from the mid-periphery of the retina and placed, photoreceptor side down, on a small millipore filter. The retinal specimen and millipore filter were trimmed to approximately 5 mm in diameter and placed, millipore side down, over a 2.5-mm diameter aperture in the centre of the large rectangular Teflon holder. The edges of the specimen were sealed with a thin layer of high vacuum silicone grease (BDH Chemicals Ltd, Poole, Dorset) and the retina was then fixed in position by overlying with a smaller (15 x 15 mm) Teflon square with a central hole [Fig. l(B)]. The edges of the Teflon square were sealed with silicone grease and the mounting procedure completed by inserting the Teflon holder into the pregreased slot in the buffer-filled recording chamber. Once set up the

ROD-CONE

DYSPLASIA

IN THE Rdy CAT

493

in vitro preparation could be maintained for up to 6 hr with only a slight, gradual reduction in response amplitudes. After mounting, the retina was allowed to stabilize for about 15 min in total darkness. Intermittently during this period the retina was stimulated by an 8 msec flash of white light of maximal intensity and the response amplitude measured. It was found that the transretinal photoresponse amplitude initially increased after mounting, and stabilized at a maximum value after about 1 S-20 min. Responseswere then recorded to white light flashes with intensities increasing in 0.5 log unit steps. A small b-wave was usually present in this iirst series of transretinal responsesand thus, in order to study the photoreceptor response (fast PIII) in the next part of the experiment, the dissection medium was changed to a buffer containing 10 mM sodium glutamate (see Table II). Glutamate and its analogues abolish the b-wave by blocking synaptic transmission between the photoreceptors and second-order neurones and inhibiting post-synaptic neuronal activity (Murakami, Ohtsuka and Shimazaki, 19 75 ; Porciatti et al., 1987). Inner retinal K+ and Ca2+ channels were blocked with verapamil in tetraethylammonium. The time course of the glutamate-isolated responsesrecorded from kitten retinas in the present study was rapid (approximately 16 msec time-to-peak at maximal light intensity in normal retinas) and these clearly represented the fast PI11 component or receptor potential. Fast PI11 responses to flashes of white light of increasing intensity were recorded (as for transretinal responses). Analysis of in vitro ERG recordings.Using a computer program, the fast PI11 amplitude vs. light intensity data were fitted by a method of least squares to the equation V = Vmox/e~(r-zo) + 1,

Maximum photovoltage (V,,,) for the b-wave was estimated at 25 ,uV with a 70 msec time-to-peak. The waveform at this age was immature and no definite oscillatory potentials were discernible. By 3 weeks of age the ERG responses were considerably increased in amplitude (Fig. 2). The amplitudes of the b- and a-waves with a 0 neutral density filter were 146 and 80 ,uV, respectively, with a calculated V,,, of 2 11 ,uV (seeTable III). The times-topeak of the b-wave (74 msec) and a-wave (39 msec) were still long. Retinal sensitivity (the log filter density at 50% V,,,) was 0.66 and rod threshold (the log light intensity which elicited a minimal 5 ,uV response)was -3.16. The ERG waveform was still immature at 3 weeks of age, but small oscillatory potentials could now be seen on responses to high light intensities (greater than 0.3 log units below maximal response luminance). Responsesto red and blue light stimuli were similar and approximated the response to a 1.8 log unit neutral density filter. The ERGSof a normal control kitten at 4.5 weeks of age were markedly different from those of 3-week-old kittens. With a 0 neutral density filter the amplitudes of both b- and a-waves (52 6 and 102 ,uV, respectively, VIncxi= 592 ,uV) were now comparable with those of an 11-week-old cat (Table III and Fig. 2). However, the times-to-peak were still long compared with older cats. Retinal sensitivity was 1.32-greater than at 3 weeks, but not yet at an adult level. The gradual maturation of the V-log I function is illustrated in Fig. 3; the amplitude-response curves shift to the left as retinal sensitivity increases. At 4.5 weeks of age the rod threshold was -4.49. The ERG waveform at 4.5 weeks was virtually of adult proportions (Fig. 2). Oscillatory potentials were more prominent on

where V is the voltage response, V,,, is the saturating voltage, I is the log light intensity, I,, is the log light intensity required to elicit O-5 V,,, (i.e. the semisaturation luminance or r), and p is the slope parameter of the responseat I,,. This equation assumes a sigmoidal distribution of the amplitude vs. light intensity data (V-log I function). Hence, the suitability of the data for this computer analysis was always checked using a Student’s t-test. The relationship between fast PI11time-to-peak and light intensity was also evaluated (T-log I function).

ERG data from normal control kittens

TABLE III

Amplitude at 0 filter density &V)

Maximal response time-to-peak(msec)

Age (weeks)

b-wave

b-wave

a-wave

2 3 4.5 6 11

25 146 526 360 566

70 74 54 33 41

ND

a-wave 6 80

102 130 166

Rod Age Calculated Retinal (weeks) V,,, @V) sensitivity* thresholdt 3. Results In Vivo Electroretinography

(a) Control cats. Table III gives details of the ERG parameters recorded from normal control cats at different ages.In 2-week-old kittens the ERGhad a low amplitude and could only be recorded irregularly in response to light flashes of near maximal intensity.

2 3 4.5 6 11

ND

ND

ND

211

0.66 1.32 2.54 2.18

-3.16

592

386 602

- 4.49

- 6.50 -6.50

* Filter density at 50% V,,. t Log light intensity at minimal (5 pV) response. ND, Not determined.

39

27 12 19

A LEON

494

ET Al.

llweehs

FIG. 2. ERGSrecorded from control normal (A) and Rdy-affected (B) kittens at different ages using cornea1 contact lens electrodes.ERGSwere unrecordable in affectedanimals. Arrows denote onset of 10 ,usecmaximal white light stimulus (0 neutral density filter). Vertical calibration bar = 50 pV. Analysis time was 200 msec.

Log relotlve

FIG. 3. Normalized

b-wave

light

mtensdy

amplitude-response

from control normal kittens aged 3(0-o),

curves

4.5 (~--I-J), 6

(A-A) and II (0-O) weeks showing that the V-log I function shifts to the left with age. There is a corresponding reduction in the semi-saturation luminance (0.5 V,,,)

indicative of increasing retinal sensitivity. recordings at high light intensities (greater than 0.9 log units below maximal response luminance) than in recordings from younger kittens. In a &week-old normal kitten b-wave amplitude with a 0 neutral density filter was high (360 ,uV, Vmax= 386 ,uV), but less than in kittens at 4.5 and 11 weeks of age. The a-wave amplitude was 130 ,uV. The timing of the ERG was now mature with b-wave and a-wave times-to-peak of 33 and 12 msec. respectively, even shorter than in an ll-week-old cat. Retinal sensitivity was 2.54, again higher than in an llweek-old animal. This was also reflected in the extreme left shift of the V-log I function curve at this age (Fig. 3). Rod threshold at 6 weeks of age was reduced to -6.50, equivalent to the threshold of an adult cat. The ERG at 6 weeks was adult in waveform. Oscillatory potentials were prominent on responses recorded from

light intensities greater than 1.5 log units below the maximal response luminance. The ERG of an 11-week-old normal kitten was entirely adult in nature. With a 0 neutral density filter the amplitudes of the b-wave (566 ,uV) and a-wave ( 166 ,uV) were large with a calculated V,,, of 602 ,IIV. The b-wave time-to-peak was 41 msec and a-wave time-to-peak was 19 msec. Retinal sensitivity was 2.18 and the V-log I function curve shifted towards the left (Fig. 3). Rod threshold was - 6.50, the same as in a 6-week-old kitten. The mature ERG waveform at 11 weeks of age is shown in Fig. 2. Oscillatory potential wavelets were detectable on recordings even at low light intensities (greater than 3.6 log units below maximal response luminance). (bj Rdy-affected cuts. ERGS generally could not be recorded from R&-affected cats at any age using conventional cornea1 contact lens electrodes (Fig. 2). Occasionally, a very low-amplitude signal was detected, but this proved difficult to differentiate from artefact even when using high-amplitude gain settings and signal averaging. For this reason an intravitreal needle recording electrode was used. This technique was first tested on a normal 6-week-old kitten in which intravitreally recorded ERGS were increased approximately twofold in amplitude ( x 2.17) compared with ERGSrecorded at the cornea. Latencies and times-to-peak were slightly prolonged ( x 1.3 ). possibly reflecting different tissue resistances and capacitances. ERGS recorded intravitreally from Rdy-affected cats at different ages are shown in Fig. 4. In a !&week-aid Rdy-affected kitten the ERG was unrecordable. In a 4.Sweek-old affected kitten low amplitude ( < 36 /IV). negative responses were recorded at high light intensities. Times-to-peak were greatly prolonged (awave 45 msec, b-wave 130 msec with a 0 neutral density filter).

ROD-CONE

DYSPLASIA

IN THE Rdy CAT

495

(B)

(A)

0

o-3

0 ‘1

0.6 o-9 l-2 l-5

(D)

0.6 0.9 I.2 l-5 I -8 2.1 2.4 2.7 2.7

(E)

(F) O+l 7 months 0

+==

2opv

L

25 msec FIG. 4. ERGSrecorded intravitreally from My-affected cats at different ages: (A) 3 weeks: (B) 4.5 weeks: (C) 6 weeks: (D) 6 weeks: (E) 17 weeks: (F) 5 months. Figures to the left of each recording refer to the neutral density fiiter used. There is relatively poor development of the b-wave so that the ERGis an a-wave-dominated, mostly negative response.Amplitudes are low and the timing is slow compared with normal kittens. Scotopic responsesrecorded after 2 hr of dark adaptation.

Most intravitreal ERGS were recorded from affected kittens at 6-65 weeks of age. Considerable individual variation was noticeable in the ERG waveform, amplitudes and times-to-peak. However, all were negative ERGS, a-wave-dominated with relatively small b-waves, and oscillatory potentials were not observed. In contrast to normal control cat ERGS, at lower light intensities the b-wave of most affected cats was lost leaving only the a-wave. B-wave amplitudes were low, varying from 5 to 45 ,uV with a 0 neutral

density filter. Mean b-wave amplitude ( ~s.D.) was 240 + 19.08 PV (n = 4). B-wave ties-to-peak were prolonged with a mean value of 89.75 + 22.01 msec (n = 4) (range 58-106 msec). A-wave amplitudes varied from 15 to 30 PV with a 0 neutral density filter, with a mean value of 22.75 k6.95 PV (n = 4). A-wave timing was also prolonged with a mean timeto-peak of 45.50-L- 13.92 msec (n = 4) (range 26-56 msec). Responses to the scotopic red and blue light stimuli were often undetectable, but where

496

present they were similar and consisted of very lowamplitude negative responses. In a 17-week-old affected cat the intravitreallyrecorded ERG had higher a-wave amplitudes than at 6 weeks of age (67 ,uV with a 0 neutral density filter), but times-to-peak were longer and the responses were far more variable in waveform than in 6-week-old animals (Fig. 4). In a 22-week-old (5 months) affected cat the intravitreally-recorded ERG was barely detectable, with a very low-amplitude, slightly negative waveform at maximal light intensity (Fig. 4). The intravitreal ERG was unrecordable in a 7-month-old cat and a 5year-old affected adult. It is difficult to make meaningful comparisons and analyses of the ERGSrecorded in vivo from normal and Rdg-affected kittens because of the very different waveforms and the effectively reversed a- and b-wave thresholds in affected individuals. The b-wave latency greatly influences the a-wave amplitude and time-topeak and thus, in order to study the photoreceptor responses, in vitro electroretinography was performed on isolated retinas and the b-wave was abolished by chemical inhibition of post-receptoral responses using glutamate. This technique isolated the photoreceptor (fast PIII) responses in normal and Rdg-affected retinas, permitting a study of the electrophysiological behaviour of the photoreceptor elements-the primary site of the retinal dystrophy.

In Vitro Electroretinographg (a) Trunsretinal responses. Transretinal photoresponses were recorded from the retinas of all cats used except for the retina from a 4-year-old Rdg-affected cat with advanced retinal atrophy in which the response was extinguished. The massed retinal response of the isolated retina to maximal intensity light stimulation appeared as a large negative a-wave with a superimposed small residual b-wave. This small b-wave was found in recordings from both normal control retinas and Rdg-affected retinas (Fig. 5). Responses from affected retinas were considerably smaller than those of control retinas. The amplitude of the b-wave was highly variable, with a range of 28-230 ,uV in normal 6-week-old kittens and 8-22 ,uV in age-matched affected kittens. A-wave amplitudes were greater and ranged from 3 34 to 510 ,uV in normal retinas and from 28 to 91 ,uV in affected retinas. The a-wavedominated nature of the response, even in normal retinas, and the great variation in b-wave amplitude, were attributed to factors such as post-enucleation retinal hypoxia and the trauma of the dissection procedure. The b-wave is more sensitive to such influences. In transretinal photoresponses from normal control kittens the b-wave became more prominent as light intensities were reduced, and a-wave amplitudes

A. LEON

ET AL

decreased (Fig. 5). At very low luminances the a-wave disappeared leaving only the b-wave. This correlates well with the findings of in vivo electroretinography and shows that in normal dark-adapted retinas the awave has a higher threshold than the b-wave. By contrast. in affected retinas the b-wave was lost very early as light intensities decreased, leaving only the awave at lower luminances. Thus in Rdy-affected retinas the b-wave has a higher threshold than the awave. confirming the results of intravitreally-recorded ERGS in vivo. The a-waves of the massed transretinal photoresponse were not studied in detail because the latency of the b-wave strongly influences both a-wave amplitude and time-to-peak. For this reason the fast PI11 response was isolated by glutamate addition as described above. cb) Fast PIII responses. Representative isolated fast PI11 waveforms of both normal and Rdg-affected 6week-old kittens are presented in Fig. 5. In both normal and affected retinas the fast PIII appeared as a long negative wave at low luminances. As stimulus light intensities were increased so the fast PI11 amplitude increased, latency decreased and time-topeak decreased. At higher light intensities a second. faster peak appeared before the receptor potential peak. This fast peak or ’ nose ’ became more prominent. increasing in amplitude and decreasing in time-topeak as light intensity was increased further (Fig. 5). In fast PI11 responses from normal 6-week-old kittens this initial peak first appeared at light intensities 2.0 log units below the maximal response luminance. The fast initial peak in recordings from three affected retinas appeared at light intensities 1.0 log unit below the maximal response luminance, but was not seen on the fast PI11recorded from three other affected retinas. This fast initial peak or nose is thought to represent a secondary stage of excitation within the more vitread portions of the photoreceptors (Arden, 1976). Fast PI11 amplitudes showed a considerable degree of individual variation, but those from affected retinas were markedly less than in control normal retinas. Mean response amplitude ( f s.D.) with a 0 neutral density filter in affected retinas was 74 F 3 7-6 ,uV (n = 6) compared with 294k201.6 PV (n = 3) in agematched normals (P < O*OOS). The fast PI11 amplitude-response curves of both normal control (n = 3) and Rdg-affected (n = 5) 6week-old kittens are shown in Fig. 6. Response amplitudes have been normalized to allow meaningful comparisons to be drawn. The V-log I function of normal kittens shows a higher sensitivity of the photoreceptors than in affected kittens. Fast PIII responses from normal retinas reach saturation (V,,,) at a light intensity 0.84 log units less than in affected retinas. Fast PI11response threshold in affected animals is 1.O log unit higher than in controls. The shift of the V-log I function of affected retinas to the right illustrates the reduced photoreceptor sensitivity with a

ROD-CONE

DYSPLASIA

IN THE My

CAT

497 (B)

(Al

5opv

200 pv L100 msec

I

I 100 msec

50 pv

100 msec

L

100 msec

Representative transretinal responses(A and B) and glutamate-isolated fast PIII responses(C and D) recorded in vitro from retinas of 6-week-old control (A and C) and R&-affectedkittens (B and D). Arrows denote onset of 8-msec white light stimulus.

0.94 log decreasein the cr value (filter density at 50% V,,,) compared with controls. Despitethis reduction in sensitivity, affected photoreceptors still have a dynamic

l.

-2 Log relative

light intensity

FIG. 6. Normalized fast PI11amplitude-response curves (Vlog I function) of control (a = 3: 0-O) and R&-affected (n = 5: a---@) 6-week-old kittens (f s.E.M.).

range not very much lessthan in normal controls with amplitudes saturating 4.3 log units above threshold (Rdy-affected) compared with 4.5 log units above threshold (normal controls). A study of the V-log I function curves reveals that the difference in fast PI11 response amplitude between control and affected kittens is greatest at higher light intensities and is less pronounced at lower luminances. Also notable in Fig. 6 is the altered slope of the V-log I function in affected photoreceptors compared with normal controls. The slope parameter (/?) at the semi-saturation luminance for control retinas is 0.40 whilst for affected retinas /3

is 0.3 5. These slope parameters indicate that, in

498

FIG. 7. Temporal characteristics (T-log 1 function) of the fast PI11responses from control (0-O) and Rdy-affected (0-a) 6-week-old kittens.

affected retinas, the relative increase in fast PI11 amplitude per log unit of light intensity is less than in normal retinas. The time-to-peak of the fast PI11 showed less individual variation than amplitude. The mean fast PI11time-to-peak ( + s.D.) with a 0 neutral density filter in affected retinas was 3 1.3 + 5.4 msec (n = 6) compared with 16.0 + 3.6 msec (n = 3) in age-matched normal controls. Fast PI11response times-to-peak from affected kittens were therefore considerably prolonged (approximately twice as long) compared with controls (P < 0.005 ). Although not directly measured, fast PI11 latencies of affected retinas were also longer than those of control retinas (see Fig. 5). However, fast PI11 responses from Rdy-affected retinas had shorter timesto-peak when compared with responses of equivalent amplitude from control retinas. The temporal characteristics (T-log I function) of the fast PI11 response in both normal and Rdy-affected 6week-old kittens are shown in Fig. 7. Fast PI11 time-topeak decreased with increasing stimulus light intensity in both control and affected retinas. In affected retinas the slope of the fast PI11 T-log I function was markedly steeper than that of control retinas (P < 0.025). Times-to-peak were always longer in Rdy-affected retinas compared with control retinas and this difference became more pronounced at lower light intensities. 4. Discussion In normal control kittens the ERG waveforms in 2and 3-week-old animals were low in amplitude with long times-to-peak. The waveform was adult-like, and the dark-adapted ERG b-wave had attained adult amplitudes, by 4.5 weeks of age, but adult timing was only reached by 6 weeks of age. Rod thresholds decreased as the kittens matured and adult thresholds were attained by 6 weeks of age. Adult retinal sensitivity also was reached at 6 weeks. The oscillatory potentials on the ERG b-wave were first detected at 3 weeks of age at high light intensities. As the kittens’

A LEON

ET AL.

ages increased the oscillatory potentials became apparent at lower light intensities and were most dccreloped at 11 weeks of age. These findings are consistent with those of other workers who studied the development of the ERG in the cat and also found that b-wave amplitudes matured at 4.5 weeks of age with rod b-wave times-to-peak maturing at h-7 weeks (Ikeda and Jacobson, 1982 : Hamasaki and Maguire. 198 5 : Jacobson, Ikeda and Ruddock. 1987). Hamasaki and Maguire (1985) also found comparatively late maturation of the oscillatory potentials in the cat ERG at 18 weeks of age. As was found in previous studies the ERG of Rdyaffected cats was generally unrecordable using standard cornea1 contact lens electrodes (Curtis et al.. 19 8 7). This led to the suspicion that the ERG signal in these animals was of such low amplitude that it was being lost as it traversed the ocular media. This hypothesis was confirmed when the recording technique was altered and the cornea1 contact lens electrode exchanged for an intravitreal monopolar needle electrode. The ERGSof normal kittens, recorded in this manner, were unchanged in waveform, but amplitudes were increased more than twofold and times-to-peak were slightly prolonged ( x 1.3). The enhanced sensitivity of the intravitreal recording technique allowed, for the first time, the recording of a series of ERG responses to light stimuli of different intensities from Rdy-affected kittens. The intravitreally recorded scotopic ERGSfrom Rdyaffected kittens could first be recorded at 4.5 weeks of age at high light intensities. By 6-6.5 weeks of age, the ERG was more developed and could be recorded consistently although there was some individual variation in waveform, amplitudes and timing. It is possible that in one or two individuals some of this variation may have resulted from minor tissue damage to the peripheral superotemporal retina during insertion of the intravitreal recording electrode. Photopic responses were unrecordable. The scotopic ERGSfrom affected kittens were very low amplitude responses with prolonged a- and b-wave times-to-peak. Compared with the intravitreally recorded ERG from an age-matched normal control kitten, the a-wave amplitude in 6-week-old Rdy-affected kittens was reduced by about 90% and the b-wave amplitude was reduced by 97 %. At maximal light intensities the a-wave timeto-peak in affected kittens was prolonged about 29 msec (2.7-fold longer than that of the control kitten) and the b-wave time-to-peak was prolonged about 47 msec (2.1-fold longer than in the control ERG 1. Because the b-waves were particularly small. the ERGS in Rdy-affected kittens were a-wave-dominated, often negative responses. It is probable that the defective, incomplete synaptogenesis which occurs in the outer plexiform layer of affected retinas (Leon and Curtis, 1990) results in correspondingly poor transmission of the visual signal from the photoreceptors to

ROD-CONE

DYSPLASIA

IN THE Rdy CAT

the b-wave-generating inner nuclear layer (the bipolar second-order neurones and Miiller cells). Oscillatory potentials were absent, possibly for similar reasons. In contrast to normal kittens, the b-wave of the ERG in affectedkittens tended to be lost first as light intensities decreased, leaving only the a-wave at lower luminances. Rod thresholds were variable and difficult to determine because of the low-amplitude, shallow, prolonged a-wave. By 5 months of age the ERG was barely detectable at high light intensities as an isolated low-amplitude, monophasic a-wave. The intravitreal ERG was unrecordable in Rdy-affected cats aged 7 months or older. The time-course of the development and subsequent loss of the ERGin Rdy-affectedkittens corresponds well with the histopathological and behavioural findings. Histological studies showed that photoreceptor outer segment material increased most after 3 weeks of age and appeared to have reached its maximum development by 6 weeks of age (Leon, 1989: Leon and Curtis, 1990). There was similar development of visual behaviour in affectedkittens (Leon, 1989). The barely detectable ERGat 5 months of age corresponded to the presence of a small amount of outer segment material at this age. Absence of the ERGat 7 months correlated with absence of outer segment material in the retina, and blind behaviour in kittens at this age. Comparison with ERGS in Other Retinal Dystrophies

The a-wave-dominated, negative ERG waveform, with small b-waves, seen in Rdy-affected kittens resembles that of Norwegian Elkhounds affected with early retinal degeneration (erd: Acland and Aguirre, 198 7). In these Norwegian Elkhounds the b-wave fails to develop normally and the ERGof erd-affecteddogs is a negative a-wave-dominated responsewith a small bwave. In common with Rdy cats, as the retina degenerates the b-wave is lost before the a-wave. Acland and Aguirre (1987) also attributed this waveform to the structural dysgenesis of the dogs’ outer plexiform layer. However, the ERG in the Norwegian Elkhound differs from that of the Rdy cat most notably in that times-to-peak are apparently normal in the first few weeks of life. Acland and Aguirre (1987) showed that the initial a-wave slope developed normally and their ERG traces from erdaffected dogs clearly show that the very small b-wave nevertheless has a normal time-to-peak. Blanks, Adinolfi and Lolley (19 74) reported defective synaptogenesis in rd mice, similar to that observed in Rdy kittens and erd Norwegian Elkhounds, yet the ERG in these mice develops a substantial bwave (Noell, 1958, 1965). The basal contacts which develop between photoreceptor terminals and bipolar processes in rd mice appear to be sufficient for the generation of the b-wave in this species.However, the ERGin young rd mice is similar to that of Rdy kittens in that times-to-peak are prolonged (Noell, 1958).

499

The ERGin dystrophic Irish Setters is unlike that in Rdy cats. Aguirre and Rubin (1975) were unable to

record rod responses from affected Irish Setters, although intravitreal recordings were not performed, but cone responses were present. These were decreased in amplitude and times-to-peak were prolonged. The largely unrecordable cornea1ERGof Rdy kittens resembles the markedly diminished or unrecordable ERGdescribed by West-Hyde and Buyukmihci (1982) in kittens affectedwith a similar dominantly inherited rod-cone dysplasia. However, in Abyssinian cats affected with a recessively inherited progressive rodcone degeneration the ERG is reported to show a progressive reduction of a- and b-wave amplitudes from 1.5-2 years of age with extinction after 4 yr (Narfstrom, Nilsson and Andersson, 1985). Analysis of the V-log I function of the ERG b-wave in these Abyssinian cats showed reductions in response amplitude with only a slight decreasein retinal sensitivity (Narfstriim, Arden and Nilsson, 1989). Unlike Rdyaffectedcats, there were no changes in the slope of the V-log I function, the ERGwaveform or times-to-peak. In children with Leber’s congenital amaurosis the ERG,when recorded by conventional means, is absent or markedly diminished, as in Rdy-affected kittens (Alstrom and Olson, 19 5 7 ; Schappert-Kimmijser, Henkes and van den Bosch, 1959 ; Francois, 1968 : Franceschetti, Francois and Babel, 19 74). In those few cases where the ERG has been recordable, both rod and cone function were severely affected with low amplitudes and, where timing has been described, a prolonged time-to-peak (Alstrom and Olson, 19 5 7 ; Weleber and Tongue, 198 7). Schappert-Kimmijser et al. (1959) examined 56 children with Leber’s congenital amaurosis, and in those caseswhere the ERG could be recorded at high stimulus intensities these were low-amplitude, predominantly negative responses.Most were thought to be photopic responses,but, when recorded, the scotopic response was also a negative a-wave-dominated ERGwith very prolonged times-to-peak-closely resembling the responsein Rdy cats. Transretinal Responses

The transretinal responses recorded from isolated retinas confirmed the results of in vivo electroretinography. In Rdy-affected kittens the responses were small compared with normal kittens and the b-wave threshold was higher than the a-wave threshold-the reverse of the situation in normal retinas. This is consistent with the a-wave dominated nature of the in vivo intravitreal recordings. Post-enucleation hypoxia may well have contributed to the small size of the b-waves observed in transretinal responses of both control and affected retinas. This was one of the reasons why the massed transretinal responseswere not studied in detail and

500

the emphasis of analytical procedures was directed towards the isolated fast PIII responses. In cats the awave (and therefore the fast PIII) is markedly resistant to hypoxia compared with the b-wave (Noell, 1951). Fast PI11 Responses Glutamate-isolated fast PI11responses were recorded from both normal and Rdy-affected retinas. Both showed increasing amplitudes, decreasing latencies and decreasing times-to-peak as stimulus intensities were increased. At higher light intensities a fast initial peak, representing a secondary excitation stage in vitread portions of the photoreceptors (Arden, 1976), was recorded on the fast PI11 of the normal retinas and 50% of the affected retinas. However, the threshold of this initial peak was 1.0 log unit higher in Rdy-affected retinas than in control normal retinas. Fast PI11 amplitudes, although variable, were reduced by about 75% in affected retinas compared with normal age-matched retinas (P < 0.005). The mean fast PI11time-to-peak, at maximal light intensity, in Rdy-affected retinas was prolonged by about 15 msec and was approximately twofold longer when compared with the time-to-peak of normal retinas (P < 0~005). The V-log 1 function curves of the fast PI11 response (Fig. 6) illustrate several features: (1) V,,, in Rdy-affected retinas saturated at light intensities 0.84 log units higher than in control retinas. (2) Fast PI11 response threshold was 1.0 log unit higher in affected retinas compared with normal retinas. (3) There was a 0.94 log unit reduction in photoreceptor sensitivity (gl in affected retinas, indicated by the right shift of the V-log I function curve and the corresponding increase in semi-saturation luminance. (4) The V-log I function curve of the affected retinas had an altered slope (/3 = 0.3 5) compared with normal control retinas (/3 = 0.40), which represents a smaller increase in relative fast PI11 amplitude per log unit light intensity in dystrophic photoreceptors. The reduction in fast PI11 amplitudes, increased fast PI11 thresholds and reduced photoreceptor sensitivity in Rdy-affected retinas are all consistent with the histological findings of considerably reduced amounts of photoreceptor outer segment material, and therefore also rhodopsin, in Rdy kittens at this age when compared with normal kittens. A decreased amount of outer segment material means that there are fewer Na+ channels in affected photoreceptors, with a consequently smaller dark current. These photoreceptors are therefore capable of only small photocurrents and a reduced fast PI11 amplitude at each light intensity, with a reduction in Vmax. Increases in semi-saturation luminance are suggestive of a generalized reduction in retinal quanta1 catch,

A. LEON

ET AL

analogous to placing a filter in front of the eye (Arden. Fox and Bull, 1983; Birch and Fish, 1986). Indeed, the paucity of outer segment material in Rdy-affected retinas means that there is a reduced chance of absorption of quanta. This results in a horizontal shift of the intensity-response function along the intensity axis and a corresponding increase in semi-saturation luminance (C ). The approximately 1.0 log unit threshold elevation and reduction in photoreceptor sensitivity indicate a considerable decrease in the rhodopsin content of Rdyaffected photoreceptors compared with normal photoreceptors. Ernst and Kemp ( 19 72) studied the effects of rhodopsin decomposition on PI11 responses from isolated rat retinas under similar in vitro conditions to those used in the present study. The results of these workers showed that a I.0 log unit loss of sensitivity corresponds to a loss of approximately 50% of rhodopsin. In retinal degenerations in man electroretinographic sensitivity losses are usually determined from the ERG b-wave and it should be remembered in this context that, for a given loss of rhodopsin. the bwave of the ERG shows a much greater drop off in sensitivity than the transretinal PI11 response (Ernst and Kemp. 1972). The markedly prolonged time-to-peak of the fast PI11 from Rdy-affected retinas, the abnormal T-log I function curve, and the altered slope of the V-log ! function curve may reflect abnormal phototransduction kinetics in the photoreceptor outer segments. A prolongation of the time-to-peak could be a result of morphological and/or biochemical abnormalities in the outer segment. Morphologically, the photoreceptor outer segments are highly disorganized and the disk lamellae disoriented such that the distance over which the light-evoked signal must travei, between rhodopsin-containing rod disks and the plasma membrane. is increased. Biochemically, the phototransduction process may be slowed by a defect in several of the proteins involved in the light-induced enzyme cascade. Elevated concentrations of 3’, 5’-cyclic guanosine monophosphate (cGMP) within rod outer segments would mean that upon light-activation cGMPphosphodiesterase would take longer to hydrolyse its cGMP substrate, and therefore the cGMP-mediated ionic channels in the plasma membrane would close at a slower rate. retarding the hyperpolarization phase of the photoresponse. When amplitude-matched fast PI11 responses were compared, those of Rdy-affected retinas had shorter times-to-peak than those of control retinas. However, this point was not pursued further since it was considered not legitimate to compare amplitudematched responses in view of the large standard deviations involved in individual fast PI11 amplitudes, whereas times-to-peak for a given light intensity showed relatively small standard deviations. The abnormal T-log I function curve of Rdy-affected photoreceptors, with relatively greater prolongations

ROD-CONE

DYSPLASIA

IN THE Rdy CAT

ln time-to-peak at lower luminances, is very similar to the T-log I function produced by the effect of isobutylmethylxanthine (IBMX) on the ERG a-wave of the isolated perfused cat eye (Sandberg et al., 198 7). IBMX is a phosphodiesterase inhibitor which elevates the level of cGMP in the retina. Sandberg et al. (1987) showed that perfusion of the isolated cat eye with IBMX (l-0 mM) reduced rod and cone ERG amplitudes, increased the semi-saturation luminance (o), prolonged a-wave times-to-peak and latencies, and prolonged b-wave times-to-peak. These effects were reversible and were shown to be accompanied by a significant elevation in the concentration of retinal cGMP. Other cGMP-phosphodiesterase inhibitors had a similar effect (Pawlyk, Sandberg and Berson, 1990). Similar findings were made by Schneider and Zrenner (1986) who examined the effect of several different phosphodiesterase inhibitors on the isolated PI11 of the cat eye. These authors found a dose-dependent response with low doses increasing PI11 amplitudes and higher doses decreasing PI11 amplitudes, whilst the PI11 time-to-peak was prolonged at all doses. Rod-cone dysplasias in other species, which manifest elevated retinal cGMP levels, are characterized by a sparing of cone photoreceptors relative to rods (Farber and Lolley, 1974 ; Aguirre et al., 1978). However, in R&-affected cats the cone photoreceptors are not or electrospared, as determined histologically physiologically, and raised concentrations of cGMP have not been demonstrated unequivocally in Rdy retinas. A simple loss of visual pigment in otherwise normal photoreceptors would not change the slope of the Tlog I function curve, and loss of photoreceptors per se would not change the ERG time-to-peak (Sandberg et al., 19 8 7). The altered slope of the T-log I function in Rdy-affected kittens may reflect the severely disorganized photoreceptor outer segment structure and/or abnormal phototransduction kinetics therein.

References Acland, G. M. and Aguirre, G. D. (1987). Retinal degenerations in the dog. IV. Early retinal degeneration (erd) in Norwegian Elkhounds. Exp. Eye Res. 44, 491-521. Aguirre, G., Farber, D., Lolley, R.. Fletcher, R. T. and Chader, G. J. (1978). Rod-cone dysplasia in Irish Setters: a defect in cyclic GMP metabolism in visual cells. Science 201.11334. Aguirre, G. D. and Rubin, L. F. (1975). Rod-cone dysplasia (progressive retinal atrophy) in Irish Setters.J. Am. Vet. Med. Assoc. 166, 157-64. Alstrom, C. H. and Olson, 0. A. (1957). Heredoretinopathia congenitalis monohybrida recessiva autosomalis. Hereditas 43, l-l 77.

Arden, G. B. (1976). Voltage gradients across the receptor layer of the isolated rat retina. 1. Physiol. 256, 333-60. Arden, G. B., Fox, B. and Bull, T. (1983). Abnormal photoreceptors in a dog with a delayed progressive retinal atrophy. Trans. Ophthabnol. Sot. U.K. 103, 411-15. 34

501

Birch, D. G. and Fish, G. E. (1986). Rod ERGSin children with hereditary retinal degeneration. 1. Pediatr. Ophthalmol. Strabismus 23, 227-32.

Blanks, J. C., Adinolfi, A. M. and Lolley, R. N. (19 74). Photoreceptor degeneration and synaptogenesis in retinal-degenerative (rd) mice. 1. Comp. Neural. 156, 95-106. Curtis. R., Barnett, K. C. and Leon, A. (198 7). An earlyonset retinal dystrophy with dominant inheritance in the Abyssinian cat: clinical and pathological findings. Invest. Ophthalmol. Vis. Sci. 28, 131-9.

Ernst, W. and Kemp, C. M. (1972). The effects of rhodopsin decomposition on PI11responsesof isolated rat retinae. Vision Res. 12, 193746. Farber, D. B. and Lolley, R. N. (1974). Cyclic guanosine monophosphate : elevation in degenerating photoreceptor cells of the C3H mouse retina. Science 186, 449-51. Franceschetti, A., Francois, J. and Babel, J. (1974). Chorioretinal

Heredodegenerations. Pp. 305-29.

C. C.

Thomas: Springfield, IL. Francois. J. (1968). Leber’s congenital tapetoretinal degeneration. Int. Ophthalmol. Clin. 8, 92947. Hamasaki, D. I. and Maguire, G. W. (1985). Physiological development of the kitten’s retina : an ERGstudy. Vision Res. 25, 153743.

Ikeda, H. and Jacobson, S. G. (1982). Cone and rod electroretinograms during development in the cat. J. Physiol. 329, 21-2.

Jacobson, S. G., Ikeda, H. and Ruddock, K. (1987). Conemediated retinal function in cats during development. Dot. Ophthalmol. 65, 7-14. Leon, A. (1989). An Early-onset Hereditary Retinal Dystrophy in the Cat. PhD thesis, CNAA, London.

Leon, A. and Curtis, R. (1990). Autosomal dominant rodcone dysplasia in the Rdy cat. 1. Light and electron microscopic findings. Exp. Eye Res. 51, 361-81. Murakami, M., Ohtsuka, T. and Shimazaki, H. (1975). EiTectsof aspartate and glutamate on the bipolar cells of the carp retina. Vision Res. 15, 456-9. Narfstrom, K., Arden, G. B. and Nilsson, S. E. G. (1989). Retinal sensitivity in hereditary retinal degeneration in Abyssinian cats: electrophysiological similarities between man and cat. Br. I. Ophthalmol. 73. 516-2 1. Narfstrom, K. L., Nilsson, S. E. and Andersson, B. E. (1985). Progressive retinal atrophy in the Abyssinian cat: studies of the DC-recorded electroretinogram and the standing potential of the eye. Br. 1. Ophthalmol. 69, 618-23. Noell, W. K. (1951). Site of asphyxial block in mammalian retinae. I. Appl. Physiol. 3, 489-500.

Noell. W. K. (1958). Differentiation, metabolic organization, and viability of the visual cell. Arch. OphthalmoI. 60, 702-33.

Noell, W. K. (1965). Aspects of experimental and hereditary retinal degeneration. In Biochemistry of the Retina (Ed. Graymore. C. N.). Pp. 51-72. Academic Press: New York. Pawlyk, B. S.. Sandberg, M. A. and Berson, E. L. (1990). Effects of cGMP-PDE inhibitors on the ERG of the isolated perfused cat eye: antagonism with calcium or background illumination, ARVO Abstracts. Invest. Ophthalmol. Vis. Sci. 31 (Suppl.), 423. Porciatti, V., Bagnoli, P., Alesci, R. and Fontanesi, G. (1987). Pharmacological dissociation of the b-wave and pattern electroretinogram. Dot. Ophthalmol. 65, 377-83. Raitta. C.. Karhunen, U.. Seppalainen, A. M. and Naukkarinen. M. (1979). Changes in the electroretinogram and visual evoked potentials during general anaesthesia. Albrecht Von Graefes Arch. Klin. Exp. Ophthalmol. 211. 13944. EER 53

502 Sandberg, M. A., Pawlyk. B. S.. Crane, W. G.. Schmidt, S. Y. and Berson. E. L. (198 7). Effects of IBMX on the ERG of the isolated perfused cat eye. Vision Res. 27, 1421-30. Schappert-Kimmijser, J.. Henkes, H. E. and Van Den Bosch, J. (1959). Amaurosis congenita (Leber). Arch. Ophthalmol. 61, 211-18. Schneider. T. and Zrenner, E. (1986). The influence of phosphodiesterase inhibitors on ERG and optic nerve response of the cat. Invest. Ophthalmol. Vis. Sci. 27. 1395403.

A. LEON

El AL

Tashiro. C.. Muranishi, K.. Gomyo. 1.. Mashimo, ‘I., Tomi. K. and Yoshiya. I. (1986). Electroretinogram as a possible monitor of anesthetic depth. Graefes Arch. Clin. hp. Ophthalmol. 224. 473-O. Weleber, R. G. and Tongue, A. C. (1987). Congenital stationary night blindness presenting as Leber’s congenital amaurosis. Arch. Ophthnlmol. 105. 360-i. West-Hyde, L. and Buyukmihci, N. (1982). Photoreceptor degeneration in a family of cats. I. Am. Vet. Med. Assoc. 181.243-i.