Experimental Eye Research 81 (2005) 340–349 www.elsevier.com/locate/yexer
Pattern electroretinography in a rat model of ocular hypertension: functional evidence for early detection of inner retinal damage Gil Ben-Shlomoa,1, Sharon Bakalashb,1, George N. Lambrouc, Elisabeth Latourc, William W. Dawsond, Michal Schwartzb, Ron Ofria,* a
Koret School of Veterinary Medicine, Hebrew University of Jerusalem, Herzl Street, P.O. Box 12, Rehovot 76100, Israel b Department of Neurobiology, The Weizmann Institute of Science, Rehovot 76100, Israel c Novartis Institutes for BioMedical Research, Disease Area Ophthalmology, Basel 4002, Switzerland d Department of Ophthalmology, University of Florida, Box 100284 JHMHC, Gainesville, FL 32610, USA Received 12 September 2004; accepted in revised form 16 February 2005 Available online 25 March 2005
Abstract With the increasing use of the rat as an animal model for glaucoma and for the evaluation of neuroprotective treatments, there is a need for a sensitive test of retinal ganglion cell (RGC) function in this species. The aims of this study were to detect functional abnormalities of the inner retina in a rat model of high intraocular pressure (IOP) using the pattern electroretinogram (PERG), and to correlate them with morphometric analysis of RGC survival and the functional integrity of the inner retina. Unilateral ocular hypertension was induced in 17 Lewis rats through laser photocoagulation. Pattern ERGs were recorded prior to lasering and 3 weeks later, using a series of shifting patterns of decreasing spatial frequency projected directly onto the animals’ fundus. IOP was measured at the same intervals, and the number of surviving RGCs estimated. Low amplitude PERG signals could be recorded in response to a narrow grating of 0$368 cycles per degree (cpd), and increased with stimulus size. Lasering caused mean (GS.D.) IOP to increase significantly from 18$3G4$5 (baseline) to 29$8G 8$8 mmHg within 3 weeks (p!0$0001). At this time, PERG amplitudes were significantly reduced (p!0$05), declining an average of 45% compared to the normotensive, control eyes. No outer retinal damage was observed, but the mean number of RGCs decreased significantly (p!0$001), from 2 525$0G372$4 to 1 542$8G333$8 cells per mm2.This decrease in RGC number was significantly (pZ0$03) correlated the decrease in PERG amplitude. The correlation between functional integrity of the inner retina and the rat PERG was further demonstrated by intravitreal tetrodotoxin injections, which temporarily abolished the PERG but did not affect outer retinal activity, reflected in the flash ERG. The evidence for early functional deficits, combined with tonometry and documentation of correlated ganglion cells loss, confirms the sensitivity of this diagnostic tool and the validity and importance of this animal model in glaucoma research. q 2005 Elsevier Ltd. All rights reserved. Keywords: glaucoma; pattern electroretinogram; animal model; ganglion cell
1. Introduction Glaucoma is a group of diseases that constitute a leading cause of blindness affecting millions of patients worldwide. It has been estimated that in the year 2000 nearly 2$5 * Corresponding author. Dr Ron Ofri, Koret School of Veterinary Medicine, Hebrew University of Jerusalem, Herzl Street, P.O. Box 12, Rehovot 76100, Israel. E-mail address:
[email protected] (R. Ofri). 1 These authors contributed equally to this work.
0014-4835/$ - see front matter q 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.exer.2005.02.006
million people in the USA alone were afflicted with primary open angle glaucoma, with an average disease duration of well over 10 years (Quigley and Vitale, 1997). Various animal models, including mice (Bayer et al., 2001a), rabbits (Feghali et al., 1991), dogs (Ofri et al., 1993a,b), and nonhuman primates with induced (Marx et al., 1986, 1988; Glovinsky et al., 1991, 1993; Komaromy et al., 2000; Hare et al., 2001a) or naturally occurring hypertensive (Dawson et al., 1993) and normotensive (Komaromy et al., 1998) glaucoma have been used for a number of years in the study of the disease, with the aim of improving our understanding of various aspects of glaucoma and its treatment. In recent years there is increasing use of ocular hypertensive rats in
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
the investigation of glaucoma (Mittag et al., 2000; Bayer et al., 2001b; Hare et al., 2001b; Schori et al., 2001; Bakalash et al., 2002; Grozdanic et al., 2004). Glaucoma is characterized by decreased retinal ganglion cell (RGC) sensitivity and function, progressive optic nerve and RGC damage and incremental reduction in visual fields (Bathija et al., 1998). Thus, at least in its early stages, it is mostly a disease of the inner retina. Therefore, it is not surprising that loss of RGC has been demonstrated in human patients as well as in many of the animal models of the disease. Reduced RGC numbers have been reported in nonhuman primates (Glovinsky et al., 1991, 1993; Dawson et al., 1993; Hare et al., 2001a,b), dogs (Ofri et al., 1994) and rats (Mittag et al., 2000; Bayer et al., 2001b; Bakalash et al., 2002; Grozdanic et al., 2004) with elevated IOP. In some of the studies, these morphological findings have been supported and augmented by electroretinography, showing reduced retinal function in ocular hypertensive animals, thus reinforcing the validity of these animal models to the study of glaucoma in humans (Ofri et al., 1994; Mittag et al., 2000; Hare et al., 2001a; Bayer et al., 2001a,b; Grozdanic et al., 2004). Electroretinography may be classified according to the type of stimulus used to elicit an electrophysiological response from the retina, or by the generator of that response. In pattern electroretinography, an alternating pattern is used as stimulus, and the response is indicative of inner retinal, mostly ganglion cell, function (Maffei and Fiorentini, 1981). As such, the PERG is an often-used tool in glaucoma research, with numerous studies conducted on changes in the responses of glaucomatous and optic neuropathy patients (Dawson et al., 1982; Trowle et al., 1983; Bodis-Wollner, 1989; Vaegan et al., 1995; Korth, 1997; Bach, 2001; Garway-Heath et al., 2002; Toffoli et al., 2002) and in animal models of glaucoma, including non-human primates (Marx et al., 1986, 1988; Komamromy et al., 2000) and other species (Blondeau et al., 1986; Siliprandi et al., 1988; Feghali et al., 1991; Ofri et al., 1993a,b;). However, to the best of our knowledge, it has not been used to evaluate retinal function in the rat model of chronically elevated IOP. The aim of this study was to test whether early inner retinal functional changes could be detected in what is fast becoming the most frequently used animal model of glaucoma (Mittag et al., 2000; Schori et al., 2001; Bayer et al., 2001b; Hare et al., 2001b; Bakalash et al., 2002; Grozdanic et al., 2004), and to correlate these early changes in the pattern ERG with morphological evidence of RGC loss and loss of functional integrity of the inner retina.
2. Materials and methods 2.1. Animals Seventeen inbred adult male Lewis rats (average weight 300 g) were supplied by the Animal Breeding Center at The Weizmann Institute of Science. The rats were housed in
341
a light- and temperature-controlled room and were matched for age and body weight. All experimental procedures were in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the guidelines of the Institutional Animal Care and Use Committee. 2.2. Induction of high intraocular pressure and tonometry Rats were deeply anaesthetized by intramuscular injection of ketamine hydrochloride 50 mg kgK1 and xylazine hydrochloride 0$5 mg kgK1. Using a Haag-Streit slit lamp emitting blue-green argon laser irradiation, the right eye of the anaesthetized rat was treated by 80–120 applications directed towards three of the four episcleral veins and towards 2708 of the limbal plexus. The laser beam was applied with a power of 1 W for 0$2 sec, producing a spot size of 100 mm at the episcleral veins and 50 mm at the limbal plexus. At a second laser session 1 week later, the same parameters were used except that the spot size was 100 mm for all applications. Irradiation was directed towards all four episcleral veins and 3608 of the limbal plexus. The contralateral eye served as an untreated control. Tonometry was conducted 1 day prior to lasering, and repeated 3 and 6 weeks later. Most anesthetic agents cause a reduction in IOP, thus precluding reliable measurement (Jia et al., 2000). To obtain accurate pressure measurements, rats were sedated with an intraperitoneal injection of acepromazine (10 mg mLK1) 5 min prior to tonometry. We measured IOP using a Tono-Pen XL tonometer (Automated Ophthalmics, Ellicott City, MD, USA), after applying topical anesthesia to the cornea. Due to diurnal fluctuations (Moore et al., 1996) IOP was measured at the same time of the day. Ten measurements were recorded and averaged from each eye. 2.3. Pattern ERG recordings Pattern ERG recordings were conducted in both eyes (sequentially) of all 17 rats 1 day prior to IOP elevation (baseline recordings) and 3 weeks later. The order of the eyes recorded was determined randomly. Rats were anaesthetized using an intramuscular injection of 20 mg ketamine and 1 mg xylazine. Refraction was not carried out as the normal sized pupil of the rat has a relatively long focal length (Hughes, 1977). Furthermore, pattern ERG and pattern visual evoked potential responses in rodents are unaffected by trial lenses (G10 diopters) placed in front of the eye (Porciatti et al., 1996, 1999). Therefore, pupils were untreated, in keeping with similar experimental procedures (Onofrj et al., 1985; Pizzorusso et al., 1997). The rats were then placed in a Faraday cage on a warm water-heating pad. Animals were darkadapted for 10 min prior to the beginning of the recording. This interval was chosen based on preliminary
342
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
experiments (results not shown) that revealed that PERG responses in dark-adapted rats reach a steady state after 10 min, in agreement with other studies (Porciatti and Falsini, 2003). The visual stimulus was generated by commercial software (VikingQuest, Nicolet Biomedical, Inc., Madison, Wisconsin). It consisted of dark and light checkerboards with 50% duty cycle that alternated at a frequency of 6$1 Hz. Contrast was maintained at 99%, and the mean luminance of the projected display was 100 cd mK2. Five consecutive recordings were conducted, using checkerboards of progressively increasing size. The spatial frequency of the initial stimulus was 0$368 cpd. The spatial frequency of each subsequent stimulus was halved, reaching a minimal value of 0$023 cpd for the final and largest stimulus. Animals were kept in the dark for 1 min between each stimulus. The stimuli were presented to the animals using a specially modified, table-mounted Bausch & Lomb direct ophthalmoscope, which was used for viewing and placing an image on the fundus. The modification consisted of the addition of a third optical channel that was used to project the stimulus directly onto the fundus, while allowing for simultaneous visualization of the area being stimulated and for correcting the optical quality of the image. This ‘projecting’ channel was fitted with a 120 W xenon arc lamp that serves as a light source; its voltage is regulated and rectified with a ripple of 5%. Rays from this light source pass through neutral density and monochromatic interference filters. They are then projected through a liquid crystal image plate which displays a computer driven stimulus image (Nicolet commercial software). This image plate is parfocal with the fundus image. A nodal point is in the plane of the pupil, and the double beam system uses Maxwellian view to illuminate a retinal field 308 in diameter, centred around the optic disc. Manual control was used to ensure that the stimulus was projected along the visual axis of the eye, and focused around the optic nerve head. A similar system has been described in detail by Gouras and Niemeyer (1973). Retinal signals were recorded using a corneal contact lens electrode designed for use in rats (Medical Workshops, Groningen, The Netherlands). Subcutaneous needles served as reference and ground electrodes, and were placed at the temporal canthus of the ipsilateral eye and at the base of the ipsilateral ear, respectively. Sweep time was 250 msec. Signals were amplified !500 000 with a 2–250 Hz bandpass (no notch filter was used), and digitized at a rate of 5000 Hz. A total of 350 sweeps, time locked to the shifting of the stimulus, were averaged by the computer and stored for subsequent analysis. This included measurement of the trough-to-peak amplitude of the major cornealnegative component. Responses that were not within 4 standard deviations of the mean amplitude of their group were discarded as artefacts.
2.4. Morphological assessment of retinal damage caused by the increase in IOP Three weeks after the IOP elevation, the number of surviving RGC’s was evaluated in 12 randomly chosen eyes of the 17 experimental animals (5 normotensive, control eyes, and 7 hypertensive eyes). The method was described in detail elsewhere (Bakalash et al., 2003; Schori et al., 2001). Briefly, 1 day after the last PERG recording, the hydrophilic neurotracer dye dextran tetramethylrhodamine (Rhodamine Dextran) (Molecular Probes, Oregon, USA) was unilaterally applied directly into the intra-orbital portion of the optic nerve. Only axons that survive the high IOP and remain functional, and whose cell bodies are still alive, can take up the dye and demonstrate labelled RGCs. The rats were sacrificed 24 hr later and their retinas were excised, whole-mounted, and preserved in 4% paraformaldehyde. The RGCs were counted using a Zeiss fluorescent microscope. From each retina four fields, all with the same diameter (0$076 mm2) and within the same distance from the optic disc, were counted and averaged. The counts were conducted by an investigator who was masked regarding the identity of the experimental and control eyes. Eyes from the remaining animals were submitted for routine histopathological examination. The eyes were fixated in buffered paraformaldehyde/glutaraldehyde solution, paraffin embedded, sectioned and stained with hematoxylin and eosin. Slides were examined by an investigator who was masked regarding the identity of the experimental and control eyes. Particular attention was paid to the morphology of the outer retina, to determine whether damage induced by elevated IOP had spread beyond the inner retina. 2.5. Confirming the inner retinal origin of the PERG The RGC origin of the rat PERG has already been determined by Berardi et al. (1990) who demonstrated its progressive disappearance over 4 months, following sectioning of the optic nerve in this species. Nevertheless, we wanted to provide additional proof for the correlation between functional integrity of retinal ganglion cells and the pattern ERG in the rat, using a different approach. We therefore recorded PERG in three rats following unilateral (right eye) intravitreal injections of tetrodotoxin (TTX), a voltage-gated sodium channel blocker. The drug has been previously shown to ‘silence’ ganglion cell activity in the rat (Thurlow and Cooper, 1989). Injections were performed using a 30 gauge Hamilton microsyringe, inserted 1 mm posterior to the limbus at a 458 angle, as described by Bui and Fortune (2004). Based on an average vitreous chamber volume of 40 mL for the rat, and a 4 mL injected drug volume, the final vitreal concentration of the injected drug was 8 mM. As an additional control for this experiment, two more rats were injected with 4 mL of
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
vehicle alone (balanced salt solution) and served as placebo control animals. Pattern ERG recordings were performed 1 hr and 8 days after injection. In order to confirm that the TTX injection affected only inner retinal function, we also recorded photopic flash ERG (averaged response to five full field flashes presented at 0$1 Hz and an intensity of 2$0 cd mK2 per sec), indicative of outer retinal function, in both eyes of all five animals following the PERG recordings. 2.6. Statistics The effect of the level treatment on amplitude was measured by the ratio between the right (treated) eye and the left (untreated) eye at a given time. Comparison of treatment was performed by paired Student’s t-test, for each grating separately. P-values !0$05 were considered significant. Two competing regression methods were used to model the relationship between RGC counts and ERG amplitude: simple least squares linear regression and robust MM regression. The former is commonly used in regression analysis and assumes error terms are normally distributed with meanZzero and constant variance. The latter method affords the substantial advantage of employing a different analytic weighting scheme that minimizes the leverage of influential (outlying) observations in smaller data sets (which simple least squares linear regression is sensitive to).2 There is also no longer a requirement for normality of error terms in robust regression. Correlation coefficients, and regression coefficientsGstandard errors, were calculated for both regression analyses.
343
averaged retinal responses to five checkerboard stimuli of progressively increasing width. As can be seen, responses could be recorded in response to the narrowest grating that we used, 0$368 cpd, which is in the range of electrophysiological threshold reported in the rat (Berardi et al., 1990; Domenici et al., 1991). There is a trend for progressive elevation in the P1N2 amplitude of the response as the size of the stimulus pattern is increased (Fig. 1A). Mean IOP (calculated for 10 Tono-Pen readings) in this eye, measured 1 day prior to the recordings, was 21$5G5$9 mmHg. Fig. 1B presents the signal array recorded from the same eye 3 weeks after induction of ocular hypertension, at which time the mean IOP in this eye was 30$6G11$6 mmHg. It is evident that the responses of the ocular hypertensive eye are markedly decreased when compared to those of the normal eye. This decrease is apparent in the respective responses to each of the five stimuli. Fig. 2 presents the mean (GS.E.M.) amplitude of the P1N2 response as a function of the spatial frequency of the pattern stimulus. Results are presented for both the experimental (right) and control (left) eyes, prior to lasering (Fig. 2A) and 3 weeks after IOP elevation (Fig. 2B). Responses recorded from the hypertensive eyes were significantly reduced (p!0$05), being, on average, 45% lower than the responses recorded from the normotensive, control eyes (Fig. 2B). Responses of the hypertensive eyes were also significantly reduced (p!0$05, except for the 0$046 cpd stimulus where pZ0$08) compared to the responses of the same eyes prior to elevation of IOP, having declined an average of 37% during these 3 weeks (Fig. 2A,B). 3.3. Loss of RGCs correlated with functional deficits
3. Results 3.1. Tonometry Mean (GS.D.) baseline IOP, 1 day prior to the first lasering session, was 18$3G4$5 mmHg. Three weeks after the lasering of the episcleral veins and limbal plexus, significant unilateral elevation of IOP was recorded in all 17 rats. Mean IOP in the lasered (right) eyes was 29$8G 8$8 mmHg, while mean IOP in the control (left) eyes was 19$2G5$9 mmHg (p!0$0001; Student’s t-test). A more detailed description of the IOP elevation time-course, using an identical procedure, has recently been published (Bakalash et al., 2002). 3.2. PERG recordings in glaucomatous rats Fig. 1A presents a typical signal array recorded from the right eye of an experimental animal prior to lasering (baseline recording). The five signals depicted are the 2
S-PLUS 6 For Windows, Insightful Corporation, Seattle, WA.
A masked investigator estimated the number of surviving RGCs in 12 (7 glaucomatous and 5 normotensive) eyes by applying a dye to the optic nerve and counting the retrogradely labelled RGCs in the excised, whole-mounted retinas 24 hr later. The mean (GS.D.) RGC count in the normotensive, control eyes of seven rats 3 weeks after lasering of the contralateral eye was 2 525$0G372$4 cells per mm2. The mean number of RGCs per mm2 in five hypertensive eyes 3 weeks after IOP elevation was 1 542$8G333$8 (p!0$001). Representative sections of retinal whole-mounts are shown in Fig. 4. The reduction in the number of dextran tetramethylrhodamine-labelled cells in the hypertensive eye (A) compared to the normotensive eye (B) is evident. A substantial loss of RGCs is also apparent in routine histopathological evaluation of hematoxylin and eosin stained section (Fig. 4C,D). The relationship between RGC counts and pattern ERG amplitudes is shown in Fig. 4. Both regression methods demonstrate a positive linear trend in ERG amplitude as retinal cell counts increased, with identical correlation coefficients of 0$39. However, methodologic differences in the leverage given to outlying and extreme observations led to slope coefficients ranging from 0$0035G0$0026 (least
344
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
Fig. 1. Pattern electroretinogram responses recorded from the right eye of a male Lewis rat before (A) and 3 weeks after (B) laser photocoagulation was used to induce unilateral glaucoma. The spatial frequency of the smallest stimulus pattern (bottom trace in each figure) was 0$368 cycles per degree (cpd); this spatial frequency was halved for each of the subsequent stimuli, reaching a maximal size of 0$023 cpd in the largest stimulus (top trace in each figure). There is progressive increase of PERG amplitude with stimulus size. It is evident that the elevation of IOP caused a remarkable reduction in retinal responsiveness to all stimulating patterns. Mean (GS.D.) of 10 IOP readings taken immediately prior to the PERG recordings were 21$5G5$9 (baseline, A) and 30$6G11$6 mmHg (3 weeks post-IOP elevation, B).
We used a pharmacological approach to further demonstrate the correlation between the functional integrity of RGC’s and the PERG signal in the rat. Fig. 5 shows the effect of unilateral intravitreal injection of TTX on the PERG of a representative rat and in a placebo animal. As can be seen, the drug caused the disappearance of the PERG signal within an hour of injection in the right eye (Fig. 5A), while the activity of the left (control) eye was unaffected (Fig. 5B). PERG activity in the injected right eye was restored 8 days post-injection (Fig. 5C) and remained unchanged in the left eye (Fig. 5D). A photopic flash ERG recording in the same animal (Fig. 5E) demonstrates that the TTX injection did not affect outer retinal activity in the injected (top trace) or control (bottom trace) eye. As an additional control, we performed a unilateral injection of the vehicle in the right eye of two more animals. This injection did not have an effect on the PERG signal of either the injected (Fig. 5F) or non-injected (Fig. 5G) eye.
4. Discussion Our results demonstrate significant loss of inner retinal function in rats with chronically elevated IOP within 3 weeks after laser photocoagulation. These deficits were correlated with a significant reduction in the number of RGCs. To the best of our knowledge, this is the first report on the use of pattern electroretinography to demonstrate
Amp. (uV)
3.4. Effect of TTX injection on the rat ERG
A 16.00 14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
0.368
0.184
0.092
0.046
0.023
spatial frequency (cpd) B 16.00 Amp. (uV)
squares regression) to 0$0053G0$0021 (robust MM regression). Although the slope of the least squares regression line was not significantly different from zero (pZ0$21), the slope of the robust MM regression line was significant (pZ0$034).
14.00 12.00 10.00 8.00 6.00 4.00 2.00 0.00
0.368
0.184
0.092
0.046
0.023
spatial frequency (cpd) Fig. 2. Mean P1N2 signal amplitudes (GS.E.M.) of pattern ERG signals recorded from 17 male Lewis rats. Amplitudes are presented as a function of the stimulus spatial frequency, which ranged from 0$368 to 0$023 cpd. Unilateral elevation of IOP was induced in the right eye (blank bars) of the animals, while the left eye (dark bars) served as a normotensive control. Recordings were conducted in both eyes prior to (A), and 3 weeks after (B), elevation of IOP. At these times, mean IOP of the 17 right eyes was 18$3G4$5 and 29$8G8$8 mmHg, respectively. Prior to elevation of IOP, there were no significant differences in pattern ERG amplitudes between the left and right eyes (pO0$10). After 3 weeks of elevated IOP, mean signal amplitudes in the right eyes were significantly lower than the normotensive left eyes (p!0$05) (panel B). The average difference in amplitude between the two eyes was 45%. At this time, mean amplitudes of the signals recorded from the right eyes were also significantly lower (p!0$05, except for the 0$046 cpd stimulus, where pZ0$08) than those recorded in the same eyes prior to elevation of IOP, having declined an average of 37% during these 3 weeks (compare the blank bars in panels A and B). There were no significant differences in the mean signal amplitudes of the left, normotensive eyes, recorded at 0 and 3 weeks (pO0$05) (compare dark bars of panels A and B).
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
early detection of inner retina functional deficits in this commonly used rodent model of chronic IOP elevation, and to correlate these deficits with loss of RGC and inner retinal functional integrity. Previous electroretinographic studies of rats with chronically elevated IOP (Mittag et al., 2000; Bayer et al., 2001b; Grozdanic et al., 2004) or mice with angle closure glaucoma (Bayer et al., 2001a) have been conducted using the flash ERG. However, because the flash ERG is considered to be mostly a response of the outer retina (Gouras, 1970), changes in the signal usually appear at advanced stages of the disease, as the damage spreads to outer layers. In one study, in which IOP in rats was elevated through cauterization and transection of episcleral veins, a significant decrease in a- and b-wave amplitudes was noted after 15 weeks (Mittag et al., 2000). By enhancing the scotopic component of the flash ERG, the same team was able to
345
demonstrate a- and b-wave amplitude reductions of 30 and 15%, respectively, 40 days after episcleral vein cautery (Bayer et al., 2001b). Mean IOP in the glaucomatous eyes was 23$5 mmHg higher than in the control eyes. In our study, a comparable (37%) reduction was noted within 3 weeks, even though the IOP difference between normal and glaucomatous eyes was only 10$6 mmHg. A more moderate, 6$4 mmHg increase in IOP, resulting from hypertonic saline injection in rat episcleral veins, caused a reduction in the scotopic threshold response within 5 weeks (Fortune et al., 2004), compared to our 3-week detection threshold. In a model of laser photocoagulation of the trabecular meshwork, significant flash ERG and pupillary deficits were noted, but histological analysis showed degeneration of all retinal layers, including the outer retina (Grozdanic et al., 2004). In a study of mice with inherited angle-closure glaucoma, changes in the flash ERG responses were
Fig. 3. Ocular hypertension caused substantial reduction in the number of RGC’s in the experimental eyes (A, C) but not in the normotensive, control eyes (B, D). The effect of pressure elevation on the number of RGC is evident when studying both whole-mounts of retinas with dextran tetramethylrhodamine-labelled cells (panels A, B, magnification !10), and in routine histopathology (Panels C, D, hematoxylin and eosin stain, !40). Overall, the mean number of RGCs decreased significantly (p!0$001), from 2 525$0G372$4 to 1 542$8G333$8 cells per mm2.
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
documented at age 7–8 months even though thinning of the retina was observed at age 4 months (Bayer et al., 2001a). Thus, it would seem that in rodents the flash ERG is a rather insensitive tool for early detection of inner retinal damage stemming from elevated IOP. In our experiment the pattern ERG demonstrated functional deficits stemming from moderate IOP elevation within 3 weeks of disease induction (Figs. 1 and 2). This difference between the flash and the pattern ERG likely stems from their different retinal origins. While the precise generators of the PERG signal have yet to be pinpointed, it is generally acknowledged to be of inner retinal, mostly ganglion cell, origin (Holder, 2001). In some species including the domestic cat (Maffei and Fiorentini, 1981), the monkey (Maffei et al., 1985), humans (Dawson et al., 1982), and perhaps most significantly for purposes of this discussion, in the rat (Berardi et al., 1990), it has been shown that the transection of the optic nerve, leading to retrograde degeneration of the RGCs, resulted in abolishment of the PERG response while the flash ERG remained unaffected. Spatial frequency tuning properties of the N95 PERG peak in humans are also indicative of ganglion cell origin of this component (Berninger and Schuurmans, 1985). Consistent with its inner retinal origins, the PERG has been used to evaluate ocular hypertension and glaucoma patients (Dawson et al., 1982; Trowle et al., 1983; BodisWollner, 1989; Vaegan et al., 1995; Korth, 1997; Bach, 2001; Garway-Heath et al., 2002; Toffoli et al., 2002). Some researchers have concluded that the PERG has poor prognostic value because it is reduced in many ocular hypertensives (Vaegan et al., 1995). Others believe it is not selective enough for glaucoma, since the patient’s age increases interindividual variability (Korth, 1997) or because other inner retinal disease may result in abnormal PERG’s (Arden, 1993). However, in recordings from glaucoma patients, PERG changes have been correlated with neuroretinal rim area (Garway-Heath et al., 2002) and nerve fibre layer thickness (Toffoli et al., 2002). Even more important, in patients with early glaucoma the PERG was found to have the best specificity and sensitivity in detecting glaucomatous damage (Graham et al., 1996), and in longitudinal studies it correctly indicated eyes at risk before manifest glaucoma occurred (Bach, 2001). In non-human primate models, changes in PERG may be noted prior to the appearance of any abnormalities in the optic disc or the fundus (Marx et al., 1986, 1988). Thus, the PERG has been described as one of the best electrophysiological tools for assessment of ganglion cell function in clinical and laboratory settings (Bui et al., 2003). Therefore, it is somewhat surprising that to date the PERG has not been used to study the rat glaucoma model. This paucity of PERG studies in the rat stems from the technical difficulties of performing such recordings in this species (Fortune et al., 2004). These challenging difficulties stem from the optical problems involved in the projection of a well-focused pattern stimulus on the rat fundus. These
problems, which arise from the presence of a relatively large lens in a very small eye, were overcome through the use of a specially modified ophthalmoscope that enabled us to project the stimulus directly onto the animals’ fundus. Adjustments in the ophthalmoscope made it possible to define the location of the image on the fundus. Perhaps more importantly, the simultaneous visualization of the projected image on the retinal surface allowed the investigator to constantly assess and refine its optical quality. This is a marked advantage over stimulation through a television monitor placed in front of the animal, a procedure where the investigator cannot determine the optical quality of the stimulus, or its location, on the fundus, and has limited control over these variables. Signals in the hypertensive eyes of our experimental animals were reduced 3 weeks after IOP elevation (Figs. 1 and 2). As noted, these decreases were significant even though the elevation in IOP was relatively moderate, and did not affect the outer retina (Fig. 3). A similar decrease in amplitude was reported by other researchers using the PERG to study inner retinal function following optic nerve transection in this species (Berardi et al., 1990;
15
PERG amplitude (µV)
346
10
5
1200
1400
1600
1800
2000
2200
2400
2600
Number of ganglion cells Fig. 4. The P1N2 amplitude of the response to the largest stimulus (0$023 cycles per degree) as a function of the number of surviving RGCs (per mm2) in 7 glaucomatous (filled circles) and 5 normotensive (empty circles) eyes. Both robust MM regression (solid line) and simple least squares linear regression (dashed line) demonstrate a positive linear trend in ERG amplitude as retinal cell counts increased, with identical correlation coefficients of 0$39. However, methodologic differences in the leverage given to outlying and extreme observations led to slope coefficients ranging from 0$0035G0$0026 (least squares regression) to 0$0053G0$0021 (robust regression). Although the slope of the least squares regression line was not significantly different from zero (pZ0$21), the slope of the robust MM regression line was significant (pZ0$034).
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
347
Fig. 5. PERG recordings were performed following unilateral (right eye) intravitreal injections of tetrodotoxin (TTX), a voltage-gated sodium channel blocker. Injections were performed using a 30 gauge Hamilton microsyringe, inserted 1 mm posterior to the vitreous at a 458 angle. Based on an average limbus chamber volume of 40 mL for the rat, and a 4 mL injected drug volume, the final vitreal concentration of the injected drug was 8 mM. The spatial frequency of the smallest stimulus pattern (bottom trace in each figure) was 0$368 cycles per degree (cpd); this spatial frequency was halved for each of the subsequent stimuli, reaching a maximal size of 0$023 cpd in the largest stimulus (top trace in each figure). The drug caused the disappearance of the PERG signal within an hour of injection in the right eye (A), while the activity of the left (control) eye was unaffected (B). PERG activity in the injected right eye was restored 8 days post-injection (C) and remained unaffected in the left eye (D). A photopic flash ERG recording in the same animal (E) demonstrates that the TTX injection did not affect outer retinal activity in the injected (top trace) or control (bottom trace) eye. As an additional control, we performed a unilateral injection of the vehicle in the right eye of two more animals. This injection did not have an effect on the PERG signal of either the injected (F) or non-injected (G) eye.
348
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349
Domenici et al., 1991). The correlation between functional integrity of the inner retina and the rat PERG was further demonstrated by TTX injections, which temporarily abolished the PERG but did not affect outer retinal activity, reflected in the flash ERG (Fig. 5). In this context, it is interesting to note that the PERG signal disappeared completely when the inner retina was pharmacologically blocked (Fig. 5A), but was only reduced (Fig. 1B) when the inner retina was partially damaged (Fig. 3) by elevated pressure. Furthermore, in our experiment the reduced responsiveness of the inner retina was highly correlated with a significant reduction in the number of RGCs in the hypertensive eyes (Figs. 3 and 4), reflecting again the inner retinal origin of the signal. A similar correlation between reduced RGC counts and PERG responsiveness has been reported in rats following nerve transection (Berardi et al., 1990; Domenici et al., 1991), and in dogs with inherited glaucoma (Ofri et al., 1994). Numerous studies provide morphological evidence of RGC loss in the rat model of chronically elevated IOP (Mittag et al., 2000; Schori et al., 2001; Hare et al., 2001a; Bayer et al., 2001b; Bakalash et al., 2002, 2003; Grozdanic et al., 2004). In many cases, morphological evidence of RGC survival is also used to evaluate the efficacy of various neuroprotective and immunomodulating therapies (Schwartz and Yoles, 1999, 2000; Schori et al., 2001; Bakalash et al., 2002, 2003). However, to date there has been no direct evidence correlating the morphological evidence of increased RGC survival with the function of these cells. Based on our results, which provide evidence for early detection of inner retinal damage, and its correlation with morphological data, we propose that the PERG may be used to functionally evaluate RGC loss, and efficacy of treatment, in this animal model of chronically elevated IOP.
Acknowledgements Funded by Novartis Pharma AG, Switzerland, Proneuron Biotechnologies, Israel, and the Glaucoma Research Foundation, USA. The authors thank Dr Phil H. Kass, College of Veterinary Medicine, University of California, Davis, USA, for his help with statistical analysis of the data.
References Arden, G.B., 1993. Comparison of new psychophysics and perimetry with electrophysiological techniques in the diagnosis of glaucoma. Curr. Opin. Ophthalmol. 4, 14–21. Bach, M., 2001. Electrophysiological approaches for early detection of glaucoma. Eur. J. Ophthalmol. 11, S41–S49. Bakalash, S., Yoles, E., Kipnis, J., Schwartz, M., 2002. Neuronal survival in a rat model of chronically raised intraocular pressure is immune mediated. Invest. Ophthalmol. Vis. Sci. 43, 2648–2653.
Bakalash, S., Kessler, A., Mizrahi, T., Nussenblatt, R., Schwartz, M., 2003. Antigenic specificity of immunoprotective therapeutic vaccination for glaucoma. Invest. Ophthalmol. Vis. Sci. 44, 3374–3381. Bathija, R., Gupta, N., Zangwill, L., Weinreb, R.N., 1998. Changing definition of glaucoma. J. Glaucoma 7, 165–169. Bayer, A.U., Neuhardt, T., May, A.C., et al., 2001a. Retinal morphology and ERG response in the DBA/2Nnia mouse model of angle-closure glaucoma. Invest. Ophthalmol. Vis. Sci. 42, 1258–1265. Bayer, A.U., Danias, J., Brodie, S., et al., 2001b. Electroretinographic abnormalities in a rat glaucoma model with chronic elevated intraocular pressure. Exp. Eye Res. 72, 667–677. Berardi, N., Domenici, L., Gravina, A., Maffei, L., 1990. Pattern ERG in rats following section of the optic nerve. Exp. Brain Res. 79, 539–546. Berninger, T., Schuurmans, R.P., 1985. Spatial tuning of the pattern ERG across temporal frequency. Doc. Ophthalmol. 61, 17–25. Blondeau, P., Olivier, P., Brunette, J.R., Zaharia, M., Lafond, G., 1986. Pattern and flash electroretinogram following increased intraocular pressure in pigeons. Ophthalmic Res. 18, 224–229. Bodis-Wollner, I., 1989. Electrophysiological and psychological testing of vision in glaucoma. Surv. Ophthalmol. 33, S301–S307. Bui, B.V., Fortune, B., 2004. Ganglion cell contributions to the rat full-field electroretinogram. J. Physiol. 555, 153–173. Bui, B.V., Fortune, B., Cull, G., Wang, L., Cioffi, G.A., 2003. Baseline characteristics of the transient pattern electroretinogram in non-human primates: inter-ocular and inter-session variability. Exp. Eye Res. 77, 555–566. Dawson, W.W., Maida, T., Rubin, M., 1982. Human pattern-evoked retinal responses are altered by optic atrophy. Invest. Ophthalmol. Vis. Sci. 22, 796–803. Dawson, W.W., Brooks, D.E., Hope, G.M., Samuelson, D.A., Engel, H., Kessler, M., 1993. Primary open-angle glaucoma in the rhesus monkey. Br. J. Ophthalmol. 77, 302–310. Domenici, L., Gravina, A., Berardi, N., Maffei, L., 1991. Different effects of intracranial and intraorbital section of the optic nerve on the functional responses of rat retinal ganglion cells. Exp. Brain Res. 86, 579–584. Feghali, J.G., Jin, J., Odom, J.V., 1991. Effect of short-term intraocular pressure elevation on the rabbit electroretinogram. Invest. Ophthalmol. Vis. Sci. 32, 2184–2189. Fortune, B., Bui, B.V., Morrison, J.C., et al., 2004. Selective ganglion cell functional loss in rats with experimental glaucoma. Invest. Ophthalmol. Vis. Sci. 45, 1854–1862. Garway-Heath, D.F., Holder, G.E., Fitzke, F.W., Hitchings, R.A., 2002. Relationship between electrophysiological, psychophysical, and anatomical measurements in glaucoma. Invest. Ophthalmol. Vis. Sci. 43, 2213–2220. Glovinsky, Y., Quigley, H.A., Dunkelberger, G.R., 1991. Retinal ganglion cell loss is size dependent in experimental glaucoma. Invest. Ophthalmol. Vis. Sci. 32, 484–491. Glovinsky, Y., Quigley, H.A., Pease, M., 1993. Foveal ganglion cell loss is size dependent in experimental glaucoma. Invest. Ophthalmol. Vis. Sci. 34, 395–400. Gouras, P., 1970. Electroretinography: some basic principles. Invest. Ophthalmol. 9, 557–569. Gouras, P., Niemeyer, G., 1973. Rod and cone signals in S-potentials of the isolated perfused cat eye. Vision Res. 13, 1603–1612. Graham, S.L., Drance, S.M., Chauhan, B.C., et al., 1996. Comparison of psychophysical and electrophysiological testing in early glaucoma. Invest. Ophthalmol. Vis. Sci. 37, 2651–2662. Grozdanic, S.D., Kwon, Y.H., Sakaguchi, D.S., Kardon, R.H., Sonea, I.M., 2004. Functional evaluation of retina and optic nerve in the rat model of chronic ocular hypertension. Exp. Eye Res. 79, 75–83. Hare, W.A., Ton, H., Ruiz, G., Feldmann, B., Wijono, M., WoldeMussie, E., 2001a. Characterization of retinal injury using ERG measures obtained with both conventional and multifocal methods in chronic ocular hypertensive primates. Invest. Ophthalmol. Vis. Sci. 42, 127–136.
G. Ben-Shlomo et al. / Experimental Eye Research 81 (2005) 340–349 Hare, W., WoldeMussie, E., Lai, R., et al., 2001b. Efficacy and safety of memantine, an NMDA-type open-channel blocker, for reduction of retinal injury associated with experimental glaucoma in rat and monkey. Surv. Ophthalmol. 45, S284–S289. Holder, G.E., 2001. Pattern electroretinography (PERG) and an integrated approach to visual pathway diagnosis. Prog. Retin. Eye Res. 20, 531–561. Hughes, A., 1977. The refractive state of the rat eye. Vision Res. 17, 927–939. Jia, L., Cepurna, W.O., Johnson, E.C., Morrison, J.C., 2000. Effect of general anesthetics on IOP in rats with experimental aqueous outflow obstruction. Invest. Ophthalmol. Vis. Sci. 41, 3415–3419. Komaromy, A.M., Brooks, D.E., Kubilis, P.S., et al., 1998. Diurnal intraocular pressure curves in healthy rhesus macaques (Macaca mulatta) and rhesus macaques with normotensive and hypertensive primary open-angle glaucoma. J. Glaucoma 7, 128–131. Komaromy, A.M., Brooks, D.E., Kallberg, M.E., et al., 2000. Relationship between morphological and electrophysiological changes in primate eyes with laser-induced glaucoma. Invest. Ophthalmol. Vis. Sci. 2000 (ARVO Abstract 451). Korth, M., 1997. The value of electrophysiological testing in glaucomatous diseases. J. Glaucoma 6, 331–343. Maffei, L., Fiorentini, A., 1981. Electroretinographic responses to alternating gratings before and after section of the optic nerve. Science 211, 953–955. Maffei, L., Fiorentini, A., Bisti, S., Hollander, H., 1985. Pattern ERG in the monkey after section of the optic nerve. Exp. Brain Res. 59, 423–425. Marx, M.S., Podos, S.M., Bodis-Wollner, I., et al., 1986. Flash and pattern electroretinograms in normal and laser-induced glaucomatous primate eyes. Invest. Ophthalmol. Vis. Sci. 27, 379–386. Marx, M.S., Podos, S.M., Bodis-Wollner, I., et al., 1988. Signs of early damage in glaucomatous monkey eyes: low spatial frequency losses in the pattern ERG and VEP. Exp. Eye Res. 46, 173–184. Mittag, T.W., Danias, J., Pohorenec, G., et al., 2000. Retinal damage after 3 to 4 months of elevated intraocular pressure in a rat glaucoma model. Invest. Ophthalmol. Vis. Sci. 41, 3451–3459. Moore, C.G., Johnson, E.C., Morrison, J.C., 1996. Circadian rhythm of intraocular pressure in the rat. Curr. Eye Res. 15, 185–191. Ofri, R., Dawson, W.W., Foli, K., Gelatt, K.N., 1993a. Chronic ocular hypertension alters local retinal responsiveness. Br. J. Ophthalmol. 77, 502–508. Ofri, R., Dawson, W.W., Foli, K., Gelatt, K.N., 1993b. Primary open angle glaucoma alters retinal recovery from a thiobarbiturate: spatial frequency dependence. Exp. Eye Res. 56, 481–488. Ofri, R., Samuelson, D.A., Strubbe, D.T., Dawson, W.W., Brooks, D.E., Gelatt, K.N., et al., 1994. Altered retinal recovery and optic nerve fiber loss in primary open angle glaucoma in the Beagle. Exp. Eye Res. 58, 245–248.
349
Onofrj, M., Harnois, C., Bodis-Wollner, I., 1985. The hemispheric distribution of the transient rat VEP: a comparison of flash and pattern stimulation. Exp. Brain Res. 59, 427–433. Pizzorusso, T., Fagiolini, M., Porciatti, V., Maffei, L., 1997. Temporal aspects of contrast visual evoked potentials in the pigmented rat: effect of dark rearing. Vision Res. 37, 389–395. Porciatti, V., Falsini, B., 2003. Physiological properties of the mouse pattern electroretinogram. Invest. Ophthalmol. Vis. Sci. 44 (ARVO EAbstract 2705). Porciatti, V., Pizzorusso, T., Cenni, M.C., Maffei, L., 1996. The visual response of retinal ganglion cells is not altered by optic nerve transaction in transgenic mice overexpressing Bcl-2. Proc. Natl. Acad. Sci. USA 93, 14955–14959. Porciatti, V., Pizzorusso, T., Maffei, L., 1999. Vision in mice with neuronal redundancy due to inhibition of developmental cell death. Vis. Neurosci. 16, 721–726. Quigley, H.A., Vitale, S., 1997. Models of open-angle glaucoma prevalence and incidence in the United States. Invest. Ophthalmol. Vis. Sci. 38, 83–91. Schori, H., Kipnis, J., Yoles, E., et al., 2001. Vaccination for protection of retinal ganglion cells against death from glutamate cytotoxicity and ocular hypertension: implications for glaucoma. Proc. Natl Acad. Sci. USA 98, 3398–3403. Schwartz, M., Yoles, E., 1999. Optic nerve degeneration and potential neuroprotection: implications for glaucoma. Eur. J. Ophthalmol. 9, S9–S11. Schwartz, M., Yoles, E., 2000. Self-destructive and self-protective processes in the damaged optic nerve: implications for glaucoma. Invest. Ophthalmol. Vis. Sci. 41, 349–351. Siliprandi, R., Bucci, M.G., Canella, R., Carmignoto, G., 1988. Flash and pattern electroretinograms during and after acute intraocular pressure elevation in cats. Invest. Ophthalmol. Vis. Sci. 29, 558–565. Thurlow, G.A., Cooper, R.M., 1989. Effects of prolonged retinal ganglion cell inactivity on superior colliculus glucose metabolism in the mature hooded rat. Exp. Neurol. 104, 272–278. Toffoli, G., Vattovani, O., Cecchini, P., Pastori, G., Rinaldi, G., Ravalico, G., 2002. Correlation between the retinal nerve fiber layer thickness and the pattern electroretinogram amplitude. Ophthalmologica 216, 159–163. Trowle, V.L., Moskowitz, A., Sokol, S., Schwartz, B., 1983. The visual evoked potential in glaucoma and ocular hypertension: effects of check size, field size and stimulation rate. Invest. Ophthalmol. Vis. Sci. 24, 175–183. Vaegan, Graham, S.L., Goldberg, I., Buckland, L., Hollows, F.C., 1995. Flash and pattern electroretinogram changes with optic atrophy and glaucoma. Exp. Eye Res. 60, 697–706.