Effects of Experimentally Induced Ametropia on the Morphology and Optical Quality of the Avian Crystalline Lens

Abstract

purpose. To examine the effects of refractive error on avian lens morphology and optical quality.

methods. Hatchling white leghorn chicks were unilaterally goggled for 7 days with either a form-deprivation goggle (n = 12), a −10 D defocus goggle (n = 12), or a +10 D defocus goggle (n = 12) to induce myopia and hyperopia. Optical quality of lenses (focal length and focal length variability) from treated and contralateral control eyes was assessed using a scanning laser apparatus. Lens morphology was examined by light and electron microscopy.

results. Although the induction of refractive errors did not significantly alter lens size, shape, paraxial focal length, or average focal length, average focal length variability increased. Lenses from eyes goggled with form-deprivation and +10 D defocus goggles demonstrated a twofold increase in average focal length variability, when compared with their contralateral controls. The morphology of the lens is not altered by these experimental manipulations.

conclusions. This study provides evidence that the refractive development of the chick lens is not independent of the refractive development of the ocular globe and that chick lenticular development is influenced by both genetics and visual experience.

The role of the crystalline lens in experimentally induced ametropia has not been fully understood. Some investigators believe that the lens contributes significantly to normal ocular development, particularly to isometropia, the condition of equal refractive state for both eyes. ¹ That study examined the effect of induced ametropia on crystalline lens optical characteristics and found a change in lens focal length in kittens that underwent a radial keratotomy procedure, to induce hyperopia. In an earlier study of tree shrews visually deprived by eyelid suturing, a small but significant decrease in lens weight was observed. ²  

However, the majority of investigations have concluded that despite changes in size, shape, and refractive index distribution, the focal characteristics of the crystalline lens remain constant during ocular development in humans and chicks. ^{3 4 5} Likewise, the lens showed little or no involvement in the induction of refractive errors in cats ⁶ and chicks. ^{7 8} This suggests that the lens is a genetically preprogrammed feature of the eye and is not easily influenced by environmental cues such as altered or absent visual input. 

The study reported here was performed to investigate the effects of inducing myopia and hyperopia on the morphology and optics of the avian crystalline lens. The chick’s visual experience was degraded either by form-deprivation or by defocus-induced blur. Morphologic characteristics of the lens were assessed by light and scanning electron microscopy. The optical quality of the lenses were quantified, in terms of focal length and focal length variability using a scanning laser apparatus. 

Methods

Chicks

This study was approved by the Animal Care Committee of the University of Waterloo in accordance with the guidelines of the Canadian Council on Animal Care. All experimentation was conducted in full compliance with the ARVO Resolution on the Use of Animals in Ophthalmic and Vision Research. To examine the effects of induced refractive error on the optics and morphology of the avian crystalline lens, white leghorn chicks were obtained on the day of hatching from a local hatchery (Big Four Chicks, New Hamburg, Canada). On the day of hatching, the refractive state of both eyes was determined by streak retinoscopy. All retinoscopic measurements were performed on awake chicks without cycloplegia, as in previous studies. ^{9 10} Comparisons between cycloplegic and noncycloplegic chicks’ refractive measurements (e.g., Wallman and Adams ¹¹ ) have shown that there is no significant difference in refractive states. The refractive state of each eye was measured using a Welch Allyn retinoscope (Welch Allyn, Skaneateles Falls, NY), with an estimated accuracy of ±0.5 D, with trial lenses along the nasotemporal axis of the eye. It was assumed that chicks were fixated at a point beyond the retinoscope during the measurements. 

Once the refractive state of the hatchling chicks was determined, each chick was unilaterally fitted with a hook and loop fastener ring and goggle combination as described in previous work. ⁹ Defocus goggles were made up of a 12-mm rigid gas permeable contact lens with a base curve of 8.1 mm, inserted anteriorly into an opening of a clear plastic goggle, which measured 20 mm in diameter. 

Chicks were unilaterally fitted with either concave (−10 D) defocus goggles to induce myopia or convex (+10 D) defocus goggles to induce hyperopia. Additional chicks were fitted with translucent goggles, lacking a contact lens insert, to induce form-deprivation myopia. In all cases, the ungoggled eye served as the contralateral control. Behaviorally, the chicks showed no sign of wearing the goggle and became adjusted to it within 20 minutes. 

Once goggled, the chicks were housed in stainless steel brooders for 7 days and given food and water ad libitum. The ambient temperature was maintained at 32°C. Fluorescent lighting provided a 14-hour light/10-hour dark regimen. The goggles were briefly removed for cleaning on a daily basis. Chicks that lost or removed their goggles before the 7 days had elapsed were not used for further analysis. Three chicks lost goggles, and two lenses were damaged in dissection. Data are presented for completed experiments for 12 chicks in each treatment group. 

In those chicks that kept their goggles on for the course of the experiment, the refractive states of both treated and control eyes were again measured on the 7th day. Chicks were killed by CO₂ asphyxiation, eyes were removed, and globes were measured axially and equatorially. Lenses were carefully excised, weighed, and measured for lens thickness and diameter to estimated accuracies of ±0.001 g and ±0.05 mm, respectively. All dissections were carried out in warmed saline solution (40°C) to maintain the structural integrity of the lens and to prevent the formation of temperature-induced lens opacities. Because accommodation in the case of the avian (i.e., chick) eye results from the direct articulation of the ciliary body (and indirectly, the ciliary muscle) with the lens, all lenses, whether from treated or untreated eyes, were assumed to be in unaccommodated states after they were excised. ⁵ The effect or effects of lens removal from the eye (i.e., the extent to which the lens is in an unaccommodated, or relaxed, state) are not likely to be related to whether the eye was treated or untreated. 

Analysis of Optical Quality

Lens optical quality was quantified and assessed using a scanning laser monitor. ¹² Any lens that was damaged during its removal from the ocular globe was excluded from further analysis. The excised lens was placed, along its equatorial rim, on a bevelled washer, which was suspended in a two-chambered cell made of glass and silicon rubber. ¹³ The chamber was filled with warmed saline solution (40°C) made up of dipotassium phosphate, an inert light-scattering agent, and 0.9% sodium chloride. The lens was oriented such that the anterior surface faced down, toward the incident laser beam, and both lens surfaces were bathed by the saline solution. 

The computer-assisted scanning laser monitor consists of a computer-operated X–Y table, which controls the movement of a low-powered (2 mW) helium–neon laser. Two video cameras (oriented at 90° from each other) and a video-frame digitizer make it possible to visualize the refracted laser beam and to compute the focal length. Initially, the scanning laser monitor locates the lens center as the position at which the laser beam is least deflected for both the X and Y directions. Equivalent focal length is measured from the principal plane (the intersection of the incoming beam and the exiting beam) to the intersection of the beam with the optical axis. Changes in this measure (focal length variability) are essentially a measure of longitudinal spherical aberration, as shown in experiments in which both spherical aberration (change in back vertex distance) and equivalent focal length were measured (Sivak JG and Herbert KL, unpublished observations, June 1999). Previous work has shown that average focal length may not be a sensitive measure of lens change because the multiple factors that determine focal length (e.g., lens shape and/or changes in lens refractive index distribution) may have a neutralizing effect on each other. ⁵ Focal variability is a sensitive measure of lens change. ¹³  

Within minutes after dissection, each lens was scanned across its equatorial dimension, first along a chosen X meridian, and then along a Y meridian, oriented 90° to X. Laser beams were passed through the lens in 22 steps with a distance of 0.1 mm between each beam along both meridia. The refracted beams were recorded and focal lengths quantified. 

Data Analysis

In total, six groups of lenses were studied: form-deprivation treated, form-deprivation control, −10 D defocus treated, −10 D defocus control, +10 D defocus treated, and +10 D defocus control. The equivalent focal length of each of the 22 refracted beams for both X and Y directions was quantified. Because each lens was scanned twice (once each in the X and Y directions), and because each scan consisted of 22 measures of focal length, the optical analysis is based on 44 measurements per lens. The data consisted of 264 measurements for each group of lenses. Data for the whole study (72 lenses) consisted of more than 3100 measurements of focal length. 

For each lens examined, the paraxial focal length (parabolic curves were fitted to the data and paraxial focus measured as the y intercept), average focal length, and focal length variability (the standard error of the average focal lengths) were calculated. For each group of lenses examined, an average paraxial focal length, an average focal length, and average focal length variability were calculated. In addition, for each group, the average refractive state (Fig. 1) of the treated and control eyes was calculated. Averages of lens weight, thickness, and diameter were computed for each of the six groups analyzed. 

All statistical analyses were computed using the SigmaStat (SPSS Science, Chicago, IL) program. To determine whether values measured for treated and control groups were statistically different from one another, a Student’s t-test was used, with a probability level (α) set at 0.05, or 5%. Average lens weight, thickness, and diameter, refractive state, paraxial focal length, average focal length, and focal length variability were tested for differences. 

Lens Morphology

Before measurement of lenses and after analysis by the scanning laser monitor, the lenses were examined under a Zeiss (New York, NY) light microscope equipped with a fiberoptic light and 35 mm camera. Lenses were examined to assess any damage incurred during the dissection. Undamaged lenses were photographed at magnifications of ×10, ×20, and ×25. 

After optical and light microscopic analyses, lenses were fixed in preparation for scanning electron microscopy. Because of their high protein content and avascularity, ^{14 15} lenses underwent a lengthy fixation process. Lenses were initially stored, at room temperature, in a 2.5% glutaraldehyde/0.07 M sodium cacodylate primary fixative solution (pH 7.2) for 5 days, with the fixative changed every day. On the 6th day, the lenses were stored in 0.07 M sodium cacodylate buffer and were maintained at 4°C until they were to be examined by scanning electron microscopy. 

Analysis by scanning electron microscopy was carried out in a masked fashion, without prior knowledge of which lenses were control or treated and which treatment had been applied. In preparation for examination by scanning electron microscopy, the lens capsule, epithelium, and annular pad fibers were removed so that the newest layers of secondary lens fibers were exposed. Lenses were stored in 1% osmium tetroxide overnight, followed by a full-day wash in 0.07 M sodium cacodylate buffer. Lenses were dehydrated in a series of increasing concentrations of ethanol solutions (30%–100% ethanol) and were stored overnight in 100% ethanol. 

A 30-minute exposure to liquid CO₂ followed to allow for the lenses to critical-point dry. Lenses were secured to aluminum mount stubs (Electron Microscopy Sciences, Fort Washington, PA) and sputter-coated with gold in vacuo. Finally, lenses were examined in a JEOL JSM 35C (JEOL, Peabody, MA) scanning electron microscope at 15 kV. Micrographs at various magnifications were taken with a Polaroid camera system (Polaroid Corp., Cambridge, MA). 

Results

Optical Measures

All refractive state findings are presented as the difference between the refractive states measured at days 0 and 7. On the day of hatching, before goggling, chicks were found to be slightly hyperopic with an average refractive error of +1.6 ± 0.4 D. The magnitude of refractive error induced within the 7-day period was dependent on the type of goggle used (Fig. 1) . Concave defocus goggles (−10 D) induced myopic refractive errors averaging −12.8 ± 0.7 D, whereas contralateral control eyes had an average refractive error of− 2.0 ± 0.7 D. Convex defocus goggles (+10 D) induced hyperopia, with average refractive errors of +5.7 ± 0.9 D, whereas the contralateral controls were slightly myopic, with an average refractive state of −0.9 ± 0.5 D. After 1 week of wear, the form-deprivation goggle system (which allowed the passage of light, but did not permit a clear view of any object) induced prominent globe elongation (increase in axial length), resulting in higher amounts of myopia than that observed in the defocus system. Form-deprivation goggles induced −21.6 ± 0.9 D of myopia, whereas average refractive states for contralateral controls were −0.8 ± 0.4 D. 

Measurements of lenticular dimensions revealed no significant changes in weight, thickness, or diameter between control lenses and lenses from eyes treated with defocus and form-deprivation goggles (Table 1) . Unpaired comparisons between groups (e.g., comparisons between the three control groups) were not tested statistically because the groups represented different batches of chicks with potentially differing growth rates. A small contralateral effect by the treated eye on the contralateral control eye has been reported in studies in which ametropia is induced. ¹⁶ However, the contralateral effect is much smaller (e.g., 10%) of the errors induced. In fact, only lens thickness measures show an obvious intergroup difference for the different groups of control eyes (Table 1) , and this is likely due to variations in unaccommodated lens shape after removal of the lenses from the eye. Lens weights vary to a smaller degree across groups, whereas lens diameters hardly vary at all. Analyses of lenses from chicks treated to induce ametropia indicated that after 1 week of defocus, lenses from eyes treated with concave or convex goggles or translucent goggles showed no significant difference in paraxial focal length or average focal length (Table 2) . As noted earlier, changes in lens focal properties may be obscured in the case of average focal length because factors of opposite sign may be neutralized. 

However, average focal length variability was significantly greater in lenses exposed to +10 D defocus goggles and form-deprivation goggles. A twofold increase in average focal length variability between lenses from treated and control eyes (1.13 ± 0.19 versus 0.50 ± 0.14 mm and 0.47 ± 0.05 versus 0.24 ± 0.05 mm, respectively) was observed. On the other hand, lenses from eyes treated with concave (−10 D) goggles showed no significant difference in focal length variability between control and treated lenses. 

Differences in focal length profiles were observed between the control lenses and lenses treated with convex goggles and form-deprivation goggles (Fig. 2) . Although the control lenses displayed an orderly curve indicating negative spherical aberration, as in previous chick lens work, ⁵ treated lenses demonstrated more randomized, nonmonotonic profiles, the effect of which contributes to the increase in focal length variability. Because the morphology of the lens is not affected by the induction of ametropia (see next paragraph), we assume that the effect on focal length profile may be due to change in the refractive index distribution related to change in lens crystallins. This point is referred to again in the Discussion section. 

Lens Morphology

Light and scanning electron microscopic analyses of all lenses revealed that morphology was not affected by the induction of refractive errors (Figs. 3 and 4) . Control lenses and those from eyes treated with defocus or form-deprivation goggles were identical in surface morphology and suture configuration. Lens fibers were regularly arranged in an end-to-end conformation around an axis coincident with the optical axis in all groups. The precision at which lens fibers met at the poles to form the sutures was not altered by the induction of myopia and hyperopia. The size and shape of the point sutures were similar in treated and control lenses. Lens clarity was consistent between control and treated groups. All lenses were clear after the 7-day treatment, and no opacities or cataracts were observed. 

Discussion

In a morphologic study examining the effects of refractive error on dimensions of the myopic eye, Hayes et al., ⁷ using 3-day-old chicks, induced myopia by goggling them monocularly with either a transparent dome or an arch-shaped device affecting the lateral visual field only. Two separate groups were used as controls; one with both eyes left ungoggled, the other group with a monocular ring device (to control for mechanical impediments). Chicks were reared for 55 days. No dimensional changes were observed in the lenses. Likewise, a study by Pickett–Seltner et al. ³ in which chicks were deprived of normal vision from the day of hatching reported similar results. In comparison to their control counterparts, lenses from the form-deprived eyes showed no changes in size, shape, soluble protein content, focal length, and relative transmittance. These results suggest that the lens develops in isolation (i.e., the refractive development of the lens is independent of the refractive development of the globe). A study of pre- and posthatching optics of the chick lens ⁵ found that during embryonic and posthatching development, there is little change in lenticular focal length, despite pronounced changes in lens size and shape. This finding suggests that the refractive power of the lens is predetermined in isolation from the development of other refractive components of the eye and is independent of the visual environment. More recently, Irving et al. ⁹ used defocus goggles ranging in power from −20 D to +30 D to induce myopia and hyperopia in young chicks and demonstrated that there were no differences in lens thickness between control and goggled eyes for any of the inducing lens powers. Thus, the chick lens has been thought of as a static and optically constant refractive feature in a developing, nonstatic ocular system. ⁵  

In experiments in which kittens were subjected to changes in visual environment, either by optical defocus or lid suture, ⁶ no significant changes in lenticular dimensions were observed. Likewise, in American kestrels (Falco sparverius) that were monocularly goggled with a translucent goggle to induce myopia, no significant differences between treated and control lens diameters or thicknesses were noted. ¹⁷ However, embryonic lens growth and development are largely coupled with the development of the ocular globe, ¹⁸ and some connection between the two might be expected during the rapid posthatching development of the chick eye. 

The present study is the first to examine the changes in focal length variability (spherical aberration) of chick lenses subjected to an altered visual environment. Apart from the study undertaken by Pickett–Seltner et al., ³ which demonstrated no change in lens dimensions or focal length between lenses from form-deprived (myopic) eyes and their controls, one other investigation examining lens focal characteristics has been undertaken. ¹ Kittens were administered one of three treatments: radial keratotomy, which reduced corneal power, making the kittens hyperopic; atropinization, which eliminated accommodation; and a combination of radial keratotomy and atropinization. The study revealed that none of the three treatments altered the lens dimensionally, with lens thickness varying from 6.59 to 6.79 mm across groups and slight or no difference within groups (treated and untreated). However, treated lenses from all three groups showed shorter (1.6%–4.9% shorter) focal lengths, a change, which these authors suggest to be an appropriate compensation for the radial keratotomy group, because they become hyperopic. Focal lengths were measured using an interferogrammetric technique, which maintained the lens in saline solution and permitted the examination of the optical pathway through the saline. ¹⁹ Focal lengths were measured to both the anterior and posterior lens vertices. 

Other studies have focused on changes in lens thickness, diameter, curvature, weight, and power. Evidence of decreased lens thickness and a synchronous flattening of the anterior and posterior lens surfaces was documented in studies in which chicks were reared under continuous light ²⁰ or were injected with tetrodotoxin, a voltage-dependent sodium channel blocker that prevents retinal ganglion cell action potentials. ²¹ In both studies, the combination of a thinner and flatter lens is believed to have contributed to the hyperopic state that resulted. Significant decreases in the weights of lenses from treated eyes were observed in lid-suture experiments in which myopia was induced in tree shrews. ^{2 22} Greene, ²³ using ultrasound measurements of lens size and a modified Gullstrand analysis, calculated that the lens of a lid-sutured, myopic rhesus monkey eye is on average 8 D more powerful than its nonsutured control. 

The optical properties of the lens are determined by its size, shape, curvature, and refractive index distribution. A change in any of these parameters can have significant effects on lens focal characteristics. ²⁴ Because changes in lens size, shape, and morphology (including suture morphology) were not observed in lenses subjected to an altered visual environment, the most likely explanation for the change in focal length variability observed in the present study would be that the goggling procedure modified the refractive index distribution of the lens. Additionally, although no obvious change in lens shape was noted in this study, it is possible that subtle changes in lens curvature did take place and this should be examined in more detail. 

The in vitro quantification of lens focal properties used here involves measurements across the full diameter of the crystalline lens. In situ, however, the lens periphery is covered by the iris, and diurnal pupil diameters may be only 1 to 2 mm. Under photopic conditions, the full diameter of the lens is not exposed to incident light. Furthermore, corneal spherical aberration can help neutralize lenticular spherical aberration. In addition, birds accommodate by direct ciliary muscle pressure on the lens. ⁵ Changes in lens surface contours that may occur during accommodation in vivo may change the magnitude and/or sign of spherical aberration. Thus, the large amounts of negative spherical aberration typical of bird lenses and measured here, and in previous work for chicks, ⁵ do not necessarily have a negative effect on the optical quality of the whole (intact) eye. 

Qualitative and quantitative changes in lens proteins during postnatal development have been presumed to affect the refractive index distribution of the lens without any apparent changes in lens size and shape. ³ Lens protein analysis of chick lenses subjected to 2 weeks of monocular form-deprivation ³ revealed that there were no significant changes in soluble protein content between treated and control lenses. However, chicks from the study by Pickett–Seltner and coworkers were not as myopic as those in the present study (−10.7 ± 3.1 versus −21.6 ± 0.9 D) and, therefore, this possibility (change in protein content) cannot be ruled out as a contributing factor to the increase in spherical aberration in lenses from form-deprived and positive defocus eyes. Also, the study by Pickett-Seltner et al. examined overall lens protein content. The contribution(s) of the different crystallins to lens protein content and, consequently, to lens refractive index was not examined, and this is also a point that should be examined. 

The effects of induced myopia and hyperopia on the crystalline lens have been unclear. Previous investigations have led to the belief that the developing chick lens is a stagnant refractive component of the growing eye, one that was genetically “preprogrammed” and not easily influenced by environmental cues. The results of the present study demonstrate that the refractive development of the chick lens is not independent of the refractive development of the globe. Thus, in the chick, lenticular development is influenced by both its genetically predetermined course and visual experience. 

 Supported by a grant from the Natural Sciences and Engineering Research Council of Canada.

 Submitted for publication August 18, 1999; revised January 6 and June 2, 2000; accepted June 16, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

Refractive error change (days 0 and 7) in control eyes and those treated with form deprivation, −10 D defocus, and +10 D defocus goggles for 7 days. The y-axis is measured in diopters. Error bars represent the SEM.

Table 1.

 

Lens Measurements (±SEM) for Defocus and Form-Deprivation Experiments

	−10 D Defocus (n = 12)			+10 D Defocus (n = 12)			Form-Deprivation (n = 12)
		Treated	Control	Treated	Control	Treated	Control
Avg. weight, g	0.019 ± 0.003	0.022 ± 0.003	0.024 ± 0.006	0.020 ± 0.003	0.017 ± 0.001	0.017 ± 0.001
Avg. lens thickness, mm	2.61 ± 0.21	2.66 ± 0.21	2.51 ± 0.19	2.36 ± 0.05	2.17 ± 0.05	2.24 ± 0.06
Avg. lens diameter, mm	3.59 ± 0.06	3.56 ± 0.09	3.58 ± 0.07	3.60 ± 0.16	3.69 ± 0.04	3.61 ± 0.06

 Avg., average.

 No significant differences between treated and control lenses were observed.

Table 2.

 

Focal Characteristics (±SEM) of Lenses from Control Eyes and Their Contralateral Eyes Treated with Form-Deprivation Translucent Goggles and Defocus Goggles of −10 D and +10 D Powers

	+10 D Defocus (n = 12)			−10 D Defocus (n = 12)			Form-Deprivation (n = 12)
		Treated	Control	Treated	Control	Treated	Control
Paraxial focal length, mm	13.12 ± 1.93	13.59 ± 2.35	14.15 ± 2.06	14.39 ± 2.63	12.66 ± 1.90	12.29 ± 1.15
Average focal length, mm	16.50 ± 0.72	16.06 ± 0.75	16.89 ± 0.75	17.00 ± 0.72	15.04 ± 0.51	14.89 ± 0.58
Average focal length variability, mm	1.13 ± 0.19*	0.50 ± 0.14*	0.54 ± 0.13	0.56 ± 0.10	0.47 ± 0.05*	0.24 ± 0.05*

 Significant differences (P < 0.05) between treated and control lenses are indicated by an asterisk (

*  ).

Figure 2.

 

Representative focal length profiles of lenses from chicks goggled to induce form-deprivation myopia. (A) Profile of lens from treated eye. (B) Profile of lens from control eye. The horizontal axis represents the equivalent focal length in millimeters. The vertical axis represents the distance from the optical center (0.0) of the lens in millimeters. Plus signs (+) indicate the focal point for a given refracted laser beam.

Figure 3.

 

Light micrographs of fixed lenses subjected to various treatments. (A) Form-deprivation treated, (B) form-deprivation control, (C) −10 D defocus treated, (D) −10 D defocus control, (E) +10 D defocus treated, and (F) +10 D defocus control. No obvious changes in morphology or transparency were observed.

Figure 4.

 

Scanning electron micrographs of lenses from −10 D defocus–treated (B and D) eyes and untreated (A and C) eyes showing no change in morphology. Magnification, (A) ×400; (B) ×600; (C, D) ×4000. The upper pair of photographs demonstrate the regularity of lens fiber organization, whereas the lower pair show the ball and socket interdigitations between fibers.

The assistance of Kelley Herbert is gratefully acknowledged.