Stiles–Crawford effect in focal macular ERGs from macaque monkey

Abstract

Background: To determine whether the focal macular electroretinograms (FMERGs) are affected by the angle of incidence of the stimulating light on the retina, i.e., the Stiles–Crawford effect (SCE). Methods: FMERGs were elicited by focal stimulation of the macula in three light-adapted macaque monkeys. The incidence of the light on the retina was varied from 0 to ±11.7°. The effects of the incidence and wavelengths of the stimulus on the SCE were determined. Results: The amplitudes of the FMERG components were largest when the stimulus beam entered the eye on the visual axis and passed through the center of the pupil. The amplitudes gradually decreased as the stimulus beam passed through the pupil more eccentrically and fell on the retina more obliquely. All components of the FMERGs were decreased with the decrease least for the amplitude of the d-wave. Conclusions: The decrease in the amplitudes of the FMERGs as the angle of incidence of the stimulus beam on the retina increases demonstrates that the SCE can be detected in adult macaque monkeys. This objective method of assessing the SCE suggests that this technique can be used to assess the alignment of cones in humans with different types of macular diseases.

Introduction

Focal macular electroretinograms (FMERGs) have been used to assess the physiological condition of different retinal neuronal cells including the photoreceptors in the macular area (DeLint, Berendschot, & van Norren, 1998; Kondo, Miyake, Horiguchi, Suzuki, and Tanikawa, 1998). In most experimental and clinical studies, FMERGs have been recorded, and the effects of the Stiles–Crawford effect (SCE), a decrease in the luminous efficiency of light entering the edge of the pupil, were not examined. However with focal stimulation, the direction of the incidence beam becomes more important. 

Evidence has been obtained that the directional sensitivity of the cones to light stimuli is responsible for the SCE (Alpern, 1986; Alpern, Kitahara, & Fielder 1987; Alpern & Kitahara, 1983; Stiles & Crawford, 1933). The SCE is generally determined by psychophysical tests (Alpern, 1986; Alpern et al., 1987; Alpern & Kitahara, 1983; DeLint, Vos, Berendoschot, & van Norren, 1997), and the subjects are required to actively participate in the examination. Thus, high level of concentration and good visual acuity are required to perform the tests. 

The SCE has been examined objectively in humans by only a few investigators. DeLint et al. analyzed the SCE by measuring the visual pigment density with different incidences of the bleaching and reflected light (DeLint et al., 1998). Birch et al. stimulated focal macular areas with a steady-state flickering light through a two-channel Maxwellian-view optical system to elicit focal ERGs (Birch, Sandberg, & Berson, 1982). They showed that the SCE could be demonstrated at the fovea in normal subjects. However, they isolated the b-waves by using band-pass filters, and a separation of a-wave and d-wave components was not possible. The FMERGs in our study enabled us to isolate the a-, b-, and d-waves, and we were able to analyze each component to determine whether each showed the SCE effect. 

We have developed an FMERG system that is integrated into a slit lamp that allowed us to stimulate the retina with a spot of light at any incidence (Choshi, Matsumoto, & Nakatsuka, 2003; Yamada, Matsumoto, & Nakatsuka, 2006). This FMERG stimulating and recording system prompted us to assess the SCE objectively with long-duration stimuli. This system allowed the ERG recordings to be performed under direct fundus observation, and each component of the FMERGs could be analyzed separately. 

Thus, the purpose of this study was to determine whether the SCE was present in the foveal area of macaque monkeys. To accomplish this, we elicited FMERGs from three macaque monkeys by light of different incidences on the retina. We shall show that all components of the FMERGs were affected by the SCE with the d-wave least affected. 

Materials and methods

Animals

Focal macular ERGs (FMERGs) were recorded from 3 eyes of 3 adult (ages 6, 6, and 8 years) male macaque monkeys (Macaca fuscata). All experimental and animal care procedures adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care Committee of Oita University. 

FMERG recordings

The monkeys were initially anesthetized by an intramuscular injection of a mixture of ketamine (7 mg/kg), xylazine (0.6 mg/kg), and butorphanol tartrate (0.04 mg/kg) and maintained on an infusion of the same proportions of ketamine, xylazine, and butorphanol tartrate per hour. The depth of the anesthesia was maintained at a level sufficient to inhibit the corneal reflex and eye movements. The pupils were dilated with topical tropicamide (0.5%) and phenylephrine hydrochloride (5%), and the cornea was anesthetized with topical oxybuprocaine hydrochloride (0.4%). The non-stimulated eye was covered with an opaque patch. 

Photopic stimuli

An optical system was built to deliver focal stimuli to the macula under direct fundus observation. The system was designed so that the stimulus light passed through the center of the pupil, i.e., on the visual axis, or through different parts of the pupil from the temporal to nasal edge in 0.5-mm steps. The light sources were from hyper-bright light-emitting diodes (LEDs; NSPW500BS, NICHIA, Tokushima, Japan), and the stimuli positioning unit was a motorized helicoid stage (Sigma Koki, Saitama, Japan) with a telescopic optical system mounted on the stage. The stage moved the telescope so that the stimulating beam entered the pupil from the temporal to nasal sides across the visual axis in 0.5-mm steps. The movement of the stage had an accuracy of 0.05 mm. 

The fundus observation system was composed of a near-infrared CCD camera (Hitachi, Japan) integrated on a customized slit lamp microscope (Carl-Zeiss, Germany). The position of the light spot on the macular area was monitored during all of the recordings. The stimulus spot was 5° in diameter. To examine the effects of the wavelength of the stimulus on the SCE, red (λ _max = 644 nm, half-amplitude bandwidth of 634 to 655 nm, TLRH180P, TOSHIBA, Tokyo, Japan), amber (λ _max = 590 nm, half-amplitude bandwidth of 585 to 596 nm, TLYE260A, TOSHIBA, Tokyo, Japan), green (λ _max = 523 nm, half-amplitude bandwidth of 512 to 545 nm, SLA580EC4T, ROHM, Kyoto, Japan), and blue (λ _max = 470 nm, half-amplitude bandwidth of 460 to 482 nm, NSPB500S, NICHIA, Tokushima, Japan) LEDs were used to elicit the FMERGs. 

The white light stimulus intensity was set to 38 cd/m². The spectral characteristics (bandwidth and λ _max) of the LEDs used in this study were measured with a spectral colorimeter PR-650 SpectraScan and analyzed with SpectraView software (Photoresearch, CA, USA). 

To determine whether the ERGs were focal, the 5° stimulus spot was projected onto the optic nerve head, and FMERGs were elicited by decreasing stimulus intensities. The FMERGs recorded by the stimulus projected on the optic nerve head became non-recordable when the intensity was ≤38 cd/m² indicating that this stimulus intensity would provide a focal response from the macula with negligible effect of stray light (Choshi et al., 2003; Yamada et al., 2006). The intensity of each colored light stimulus was matched by neutral density filters to elicit approximately the criteria amplitude of b-wave (1 μV) as elicited by the white stimulus whose luminance was ≤38 cd/m². The FMERG waveforms were comprised of on and off waves. The adjusted intensity elicited no ERGs when the stimulus spot was projected onto the optic nerve head. 

A gold-foil bipolar contact lens (Mayo, Nagoya, Japan) coupled with a mini-pan fundus lens was placed on the cornea of the examined eye (Figure 1). This provided an inverted real image of the ocular fundus projected approximately 3.5 mm in front of the contact lens unit. The relationship between the angle of incidence of the light beam on the focusing plane of the inverted retinal image and that on the retinal surface was calculated as shown in Figure 1. When the light beam fell on the temporal side of the real image at an angle of θ _a, then the light beam will be projected from the nasal side onto the retinal surface at an angle of θ _b. The relationship between θ _a and θ _b was calculated by 

$[\tan {\theta }_{\mathrm{b}}=C(\tan {\theta }_{\mathrm{a}}\left)\right],$

(1)

with θ _b being the angle of incidence on the retina, θ _a being the angle of incidence on the focusing plane of the inverted retinal image, and C being the lens magnification constant. Thus, θ _b was determined by the telescope angle and C by the lens magnification constant (0.39). 

Figure 1

 

Relationship between the angle of incidence of the light beam on the focusing plane of the inverted retinal image. IF = IF distance, i.e., the axial length minus the corneal thickness and anterior chamber depth and K = distance of the stimulus beam from the visual axis in the iris plane in millimeters.

The distance between the center of the pupil and the stimulus beam was measured as shown in Figure 1. The iris–fovea distance, IF, was calculated by 

$\mathrm{I}\mathrm{F}=K/\left(\tan {\theta }_{\mathrm{b}}\right),$

(2)

with K equal to the distance of the light beam from the visual axis in the iris plane, and IF equal to the distance of the axial length with the subtraction of corneal thickness and the anterior chamber depth. A correction for light transmission through the cornea and the lens was not done. The axial length, corneal thickness, and the anterior chamber depth were measured by A-mode ultrasound echography (Compuscan LT, Stortz, St. Louis, MO, USA). 

The stimulus duration was 100 ms and the stimulus interval was 150 ms, and thus, the frequency of stimulation was 4 Hz. This stimulus pattern of 100 ms on and 150 ms off was used for each wavelength stimulus. A white light of 35 cd/m² for 15 min was used for light adaptation before recording with each wavelength. We believe that the off duration was long enough because the depolarization of the cone is much faster than 150 ms. 

FMERG recordings and analyses

A bipolar contact lens electrode (Mayo, Nagoya, Japan) was used to pick up the FMERGs. The ground electrode was attached to the right ear lobe. After centering and focusing the stimulus spot on the macula, the eye was light adapted with 35 cd/m² for 15 min. Then, the FMERGs were elicited by different stimulus intensities and wavelengths. The responses were amplified with a NeuropackΣ bioamplifier (MEB-5500, Nihon Kohden, Tokyo), A/D converted at 16 bits (PCI-16/16UD, Contec, Japan), and averaged by a customized signal processing program (MTS, Japan). More than forty responses were averaged, and the sampling rate was 10 kHz (every 0.1 ms). The responses were filtered from 0.5 to 200 Hz with a hardwired band-pass filter. With this system, the noise level with the electrode placed on the cornea and no stimulus was less than 0.1 μV. 

The amplitude of the a-wave was measured from the baseline to the trough of the a-wave, and the amplitude of the b-wave was measured from the trough of the a-wave to the following positive peak (Figure 2). The amplitude of the d-wave was measured from the trough just preceding the d-wave to the positive peak after the stimulus offset. 

Figure 2

 

(a) Representative FMERGs elicited by white stimuli entering the pupil at different distances (ordinate) from the visual axis (0) for up to 11.7 degrees. The amplitudes of the a- and b-waves of the FMERGs decrease with increasing distance from the visual axis. (b) The amplitudes of the a-, b-, and d-waves were measured from ERGs such as those from the three monkeys, and the relative amplitudes are plotted on the right. One division of the graph is equal to 25%. K represents the distance of the entrance beam to the visual axis in the iris plane in millimeters; AD represents the mean angle of incidence in degrees of stimulus spot from the axis. The blue LED has a peak at 470 nm with a half-amplitude bandwidth between 442 and 520 nm. The green LED has a peak at 523 nm with a half-amplitude bandwidth between 480 and 620 nm. The yellow–orange LED has a peak at 590 nm with a half-amplitude bandwidth between 470 and 620 nm. The red LED has a peak at 644 nm with a half-amplitude bandwidth between 580 and 675 nm.

All of the results are expressed as means ± standard deviations (SDs). The polynomial fit of the data was made with the Excel program, ver. 12.0. The polynomial fit was at order 2 for all data. 

Results

The amplitudes of the a- and b-waves of the FMERGs were largest when the stimulus beam entered the eye on the visual axis (K = 0 mm, retinal angular incidence = 0 degree), and they decreased progressively as the stimulus beam entered more eccentrically up to the edge of the pupil. For example with white light, the average (n = 3) relative amplitude of the a-wave at 11.7° was approximately one-half of that at 0°. In the same way, the average (n = 3) of the relative amplitude of the b-wave at 11.7° was approximately one-half of that at 0°. Similar changes were found for the d-wave although the degree of change (reduced by 17.7 to 18.6%) was smaller (Figure 2 and Supplementary Table 1). With other wavelengths, the changes depended on the stimulus angle, but the relative amplitude of the d-wave was not evident. 

The relative amplitudes of the FMERGs elicited by different wavelengths with peak transmission at 470 nm, 524 nm, 590 nm, and 644 nm are shown in Figure 3. As with the white stimuli, the amplitudes of the a- and b-waves were largest when the light beam entered the pupil on the visual axis and decreased with greater eccentricities. However, the degree of the SCE was not significantly affected by the wavelength of the stimuli (Figure 3). 

Figure 3

 

Relative amplitudes of the a-, b-, and d-waves elicited by different wavelengths of the stimulus. The λ _max of the stimuli was at 470 nm, 524 nm, 590 nm, and 644 nm. The small symbols (circles, triangles, and asterisks) represent the individual monkeys, the large squares are the averages, and the bars are the standard error of the means. One division of the graph is equal to 25%. *K represents the mean distance of the entrance beam to the visual axis in the iris plane in millimeters; ^†AD represents the mean angle of incidence in degrees of stimulus spot from the axis.

When the relative amplitude difference was compared between the b-wave and d-wave, significant differences were found for each wavelength used. The relative amplitude differences was calculated as the relative amplitude by the stimulus at 0K—that by the stimulus at 3.5K (Supplementary Table 1). The relative amplitude difference between maximal incidence stimuli and minimal incidence stimuli was significantly larger for the b-wave than for the d-wave for all wavelengths (Supplementary Table 1). Thus, the d-wave amplitude at the maximal stimulus incidence did not decrease as much as that of the b-wave suggesting that the SCE was larger for the b-waves than for the d-waves. 

Discussion

Our results showed that the amplitudes of the a-wave and b-waves of the FMERGs were largest when the stimulus beam entered the eye through the center of the pupil and struck the retina perpendicularly. The amplitudes decreased progressively with increasing eccentricities of the stimulus beam. This is comparable to the SCE obtained psychophysically, and our results demonstrated that the SCE can be measured objectively using the FMERGs to assess the response of the retina. 

The SCE is based on the ability of the photoreceptors to absorb photons passing through the outer segments, and the chances of a photon striking a photopigment increase when the beam passes perpendicularly through the entire extent of the outer segments. This explains why the FMERGs are largest when the stimulating beam entered the eye along the visual axis. 

More than 25 years earlier, Birch et al. (1982) showed that the SCE can be demonstrated in the focal ERGs elicited by flickering stimuli obtained from a two-channel Maxwellian-view optical system. Our results confirmed their findings and also provided new information on the characteristics of the different components of the ERG and the influence of the wavelength of stimulating light. 

Our data showed a drop-off in amplitude of the FMERGs when elicited by stimuli entering the edge of the pupil was approximately 50% whereas it was reported that the decrease of sensitivity was 1 log unit psychophysically (Alpern & Kitahara, 1983) or by OCT (Gao, Cense, Zhang, Jonnal, & Donald, 2008). The discrepancy between perimetry or OCT and the FMERGs is most likely due to methodological differences, i.e., the FMERGs were elicited by suprathreshold stimulus intensities from focal areas, whereas perimetry employs near threshold stimuli. The FMERGs sum the activity of all of activated cells in the retina, whereas the psychophysical threshold is determined by the activity of the most sensitive cells. An alternative explanation might be that the differences were due to the differences in the species studied. 

The degree of the SCE was similar for the a- and b-waves but lower for the d-wave. The origin of each wave is thought to be different. The a- and b-waves receive input from postreceptoral cells including off bipolar cells (Bush & Sieving, 1994) and from on bipolar cell (Sieving, Murayama, & Naarendorp, 1994), respectively. The d-wave arises from the activity of both the photoreceptors and inner retinal neurons (Sieving et al., 1994). 

The difference may be due to the distribution of the different types of cones in the macular area (Yamamoto, Gouras, & Lopez, 1995). Another and more likely explanation is that the difference arises from the complexity of the ERG responses with several ERG components interacting. The d-waves of the full-field ERGs result from an interaction of the cone photoreceptor recovery and on bipolar offset and off bipolar cell depolarization. For focal stimuli, the response might be complicated by the photopic negative response (PhNR), because the d-wave is abolished by TTX treatment (Kurimoto et al., 2009; Yamada et al., 2006). Further investigations must be done to determine the reason for the different properties of the d-waves. 

Stiles (1937) showed that the SCE depended on the wavelength of the light stimuli, and Alpern and Kitahara (1983) demonstrated that the directional sensitivity parameter (P) is dependent on the wavelength, namely, P was high at short (400 nm) and long (700 nm) wavelengths and lower at middle (550 nm) wavelengths. Our methods were not sensitive enough to detect these differences for the different wavelengths of the stimuli, probably because the bandwidths of the stimuli were relatively broad in our study. This was a limitation of the FMERGs. We used 4 LEDs with different wavelength and white light, while Alpern and Kitahara used 30 different monochromatic stimuli. Although each LED provides irradiance with relative steep peak, the relative broad bandwidth in our study might reduce the sensitivity to determine the difference depending on the stimulus wavelength. Using shorter, e.g., 400 nm, and longer, e.g., 700 nm, wavelength stimuli might have increased the sensitivity. 

In conclusion, our results showed that the SCE can be determined objectively by the FMERGs. They also indicate that the SCE must be considered in interpreting the results of electrophysiological studies, particularly when focal stimuli are used to assess the macular area. Because the SCE is dependent on the passage of light through the outer segments of the photoreceptors, the FMERGs can be used to examine the alignment and integrity of the photoreceptors of the fovea in different types of macular diseases. 

Supplementary Materials

Supplementary PDF - Supplementary PDF 

Acknowledgments

Support of this study was provided by Researches on Sensory and Communicative Disorders from the Ministry of Health, Labor, and Welfare, Japan. No author has a financial or proprietary interest in any material or method mentioned. 

Commercial relationship: none. 

Correspondence: Kei Shinoda, M.D., Ph.D. 

Email: [e-mail]. 

Address: Department of Ophthalmology, Teikyo University School of Medicine, Kaga 2-11-1, Itabashi-ku, Tokyo 173-8605, Japan.