Chick Eyes Under Cycloplegia Compensate for Spectacle Lenses Despite Six-Hydroxy Dopamine Treatment

Transcription

1 Chick Eyes Under Cycloplegia Compensate for Spectacle Lenses Despite Six-Hydroxy Dopamine Treatment Hartmut N. Schwahn* and Frank Schaeffetf Purpose. To test whether eye growth changes produced by spectacle lens wear are mediated by changes in ciliary muscle tonus in chicks. Methods. Because there is evidence that deprivation myopia is based on a local-retinal mechanism in the eye that probably remains functional after cycloplegia as well as after ciliary ganglion or Edinger-Westphal lesions, none of these treatments provides insight into whether accommodation tonus is also important in the control of axial eye growth. Because 6-hydroxy dopamine (6-OHDA) suppresses deprivation myopia, to isolate growth changes mediated by accommodation the authors injected 6-OHDA and paralyzed accommodation in addition (by corneal application of vecuroniumbromide). To quantify the state of cycloplegia, the abnormal pecking responses of cyclopleged chickens were studied. Results. The authors found that cycloplegia could be maintained for 3 hours daily by corneal application of vecuroniumbromide. To ensure that visual exposure was restricted to the time period of cycloplegia, chickens were transferred to a 3-hour light/21-hour dark cycle. Control experiments showed that emmetropization was still functional under the changed light cycle. Strikingly, even with suppressed local-retinal growth control mechanisms (as indicated by the lack of deprivation myopia in a 6-OHDA injected group of chickens with occluders) and paralysis of accommodation, the eyes compensated for the defocus imposed by spectacles by changing their axial growth rates to be similar to those of eyes with functional accommodation. Conclusions. The findings show that the ciliary muscle and the activity of the iris sphincter muscle are not involved in emmetropization in chicks. If accommodation mediates the growth effects with lenses, it must happen via another pathway. Based on previous results, the authors propose that either the choroidal nerves from the ciliary ganglion to the choroid are important or that another yet unknown pathway from the Edinger Westphal nucleus to the eye transmits the necessary information. Invest Ophthalmol Vis Sci. 1994; 35: J. here is no doubt that vision plays an important role in the fine tuning of axial eye length to the focal length of the growing chicken: Shifting the focal plane by placing low-power spectacles in front of the eye results in rapid changes in axial eye growth such that the imposed refractive errors are largely compensated for within days. 1 ' 2 On the other hand, deprivation of high-spatial frequencies in the retinal image 3 by From the *Department of Physiology and Ophthalmology, University of California at San Francisco School of Medicine, and the fdepartment of Pathophysiology of Vision and Neuroophthalmology, Division of Experimental Ophthalmology, University Eye Hospital, San Francisco, California. Supported by the German Research Council (SFB 307, TP A7). Submitted for publication December 13, 1993; revised March 8, 1994; accepted March 10, Proprietary interest category: N. Reprint requests: Dr. Frank Schaeffel, Department of Pathophysiology of Vision and Neuroophthalmology, Division of Experimental Ophthalmology, University Eye Hospital, Ro'ntgeniveg 11, Tubingen, Germany. frosted translucent occluders produces a mismatch of focal length and axial length and axial myopia ("deprivation myopia" 45 ). A striking feature of deprivation myopia is that it is produced by a "local-retinal mechanism" within the eye 6 with no need of a connection between the eye and the brain. 7 Although it is feasible that both lens- and occluder-induced growth changes are triggered by the same visual mechanisms, a number of results indicate that this assumption cannot be unrestrictedly valid. First, we found 8 that deprivation myopia can be blocked both by a neurotoxin of catecholaminergic neurons (6-hydroxy dopamine, 6- OHDA) or by continuous light 9 ; both treatments deplete retinal catecholamine stores. Strikingly, growth responses induced by spectacles remain unchanged in both cases (6-OHDA 10 ; continuous light 9 ). Second, although deprivation myopia can be prevented by short periods of normal visual experience (15 minutes 3516 Investigative Ophthalmology & Visual Science, August 1994, Vol. 35, No. 9 Copyright Association for Research in Vision and Ophthalmology

2 Cycloplegia and Emmetropization in Chicks 3517 per day 11 ), growth changes induced by positive lenses are more resistant to intermittent periods of normal vision. 12 Third, myopia induced by negative lenses is severely reduced after optic nerve section, 13 whereas myopia induced by deprivation is largely unaffected. 7 Fourth, deprivation myopia cannot be stabilized by correction with the appropriate negative lenses, 14 whereas refractions remain stable for at least 6 days with lenses, 15 indicating that different mechanisms might be involved. Fifth, a theoretical argument makes it unlikely that deprivation-induced and lensinduced refractive errors are based on the same mechanisms: Image degradation produced by frosted occluders cannot be corrected by accommodation, hence, accommodation is not triggered by this type of blur. The situation is very different from pure lensinduced defocus, which is successfully cleared if the eye accommodates. With accommodation, there is no longer a consistent local focus error signal that could drive the local-retinal mechanism. Therefore, based on a computer simulation, 16 the involvement of a second feedback loop was proposed that measures the accommodation tonus and interacts with the localretinal feedback loop. In light of numerous investigations showing that induction of refractive errors by changes in visual experience in chickens does not require functional accommodation (deprivation-induced, 6 ' 7 lensinduced 1713 ), it might appear redundant to perform yet another experiment to test whether accommodation plays a role at all. However, the arguments listed above make it worthwhile to reconsider this option. As a working hypothesis, we assume that the motor output signals of the nucleus of the third nerve in the midbrain (Edinger-Westphal nucleus) carry the required information on accommodation tonus to guide axial eye growth. The Edinger-Westphal nucleus receives complex input from several central areas. 18 Its output is relayed in the ciliary ganglion to the short ciliary nerve, which controls both the sphincter pupillae and the ciliary muscle, and to the choroidal nerves, which modulate choroidal blood flow. 19 In the present study, we try to identify one possible target of the second, presumably accommodative, feedback loop for the control of axial eye growth (below referred to as "accommodative feedback loop"): We test whether the growth commands are derived from the tonus of the ciliary muscle. To exclude the operation of local-retinal visual eye growth control, we applied 6-OHDA intravitreally at doses that have been previously shown to suppress development of deprivation myopia. 8 MATERIALS AND METHODS Animals Chickens from a white leghorn strain were obtained from a local supplier located in Suppingen, Germany, at the first day after hatching. A total number of 72 chickens (12 groups of 6) contributed to the present study. Occluders and lenses were used as described previously. 113 The detailed treatment protocols for the different groups of chickens and the time schedules of the experiments are given in Table 1. The treatment of the chickens adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Light Cycles To ensure clearly defined light conditions during the whole experiment, chickens were kept in a light-tight box (0.6 X 0.5 X 0.4 m) that was inside illuminated by a 40-watt bulb. To prevent overheating, a blower provided constant air exchange. Illuminance levels on the floor of the box ranged from 300 to 1000 lux, depending on the direction of measurement. The light-tightness of the box was frequently checked by placing it in a dark room with the light bulb on and observing whether any light could be seen from outside by a dark-adapted experimenter. All chickens were raised under a 12-hour light/12-hour dark cycle from days 1 to 4. From days 5 to 8, the duration of the light period was gradually reduced to 3 hours a day to permit the chickens to adapt to the short period of vision. The short light cycles were maintained until the experiments were completed. Intravitreal Injections A single intravitreal injection of 150 /i,g of 6-OHDA (dissolved in 50-fA saline containing 0.1% L-ascorbate to prevent its oxidation) was performed under deep ether anesthesia on posthatch day 7 as previously described. 810 In all experiments, 150 /ig 6-OHDA were injected because previous results 8 have shown that this dose is efficient but still has no detectable effect on vision. Cycloplegia Paralysis of the ciliary muscle is not easily achieved in chickens 20 because of its striated nature, the nicotinergic receptors, 21 and the fast exchange of aqueous. After several preliminary experiments, it was found that cycloplegia could be most efficiently obtained by corneal application of the following mixture (final concentration: 1.33 mg vecuroniumbromide per ml): 0.5 ml Norcuron (Organon Teknika, Medizinische Produkte, Eppelheim, Germany: 4 mg vecuroniumbromide + buffer/ml Aqua dest) ml saline ml solution of 3% benzalkoniumchloride ml Xylocaine gel (2%; Astra Chemicals, Wendel, Germany) Cycloplegia was induced daily from posthatch day

3 3518 Investigative Ophthalmology 8c Visual Science, August 1994, Vol. 35, No. 9 TABLE l. Treatment Protocols and Experimental Schedules Group Number Purpose of Experiment Light Cycle and Treatment Time Schedule Figure Number Quantification of pecking errors Measurement of body weight under normal light cycle; measurement of refraction and ocular growth Measurement of body weight under changed light cycle; measurement of refraction and ocular growth Effect of bilateral cycloplegia on body weight? Control for group 4 Effects of unilateral cycloplegia on eye development? Effect of intravitreal 6-OHDA and cycloplegia on eye development? Deprivation myopia under (3/ 21 hi) light/dark cycle? Suppression with 6-OHDA? Effect of 6-OHDA + cycloplegia on deprivation myopia? Deprivation myopia under normal light cycle? suppression with 6-OHDA? Suppression of plus lensinduced growth effects by 6- OHDA and cycloplegia? Suppression of minus lensinduced growth effects by 6- OHDA and cycloplegia? (12/12 hr) light/dark, bilateral cycloplegia + videotaping (12/12 hr) light/dark (3/21 hr) light/dark (3/21 hr) light/dark (3/21 hr) light/dark (3/21 hr) light/dark one eye cyclopleged (3/21 hr) light/dark both eyes injected with 6-OHDA (3/21 hr) light/dark occluders, one eye injected with 6- OHDA, other with saline (3/21 hr) light/dark, occluders, both eyes injected with 6- OHDA, one eye cyclopleged (12/12 hr) light/dark, occluders, one eye injected with 6-OHDA, other with saline (3/21 hr) light/dark, both eyes +4 D lenses, both eyes 6- OHDA, one eye cyclopleged (3/21 hr) light/dark, both eyes -4 D lenses, both eyes 6- OHDA, one eye cyclopleged Daily from day Day Day Day 5-15 Day 5-15 Day 5-15 Day 5-15 intravitreal injections day 7 Day 5-15 occluders and injections day 7 Day 5-15, occluders and injections day 7 Day 11-18, occluders and injections day 12 Day 5-15, lenses and injections day 7 Day 5-15, lenses and injections day 7 2 3A 2 3A 2 3A 3B Chickens received 1 to 2 drops of the solution via topical corneal application every 6 minutes for a period of 36 minutes under dim red light (Kodak cutoff filter RG 695; Eastman Kodak, Rochester, NY). There was only a tiny effect of cycloplegia on the measured refraction in the chicks (day 12, 6 chickens: 5.4 ± 0.7 D (before cycloplegia) versus 6.0 ± 0.7 D (cyclopleged), P< 0.02, paired Mest). Quantification of the State of Cycloplegia If accommodation is paralyzed by drugs (Troilo D, personal communication, 1988, and ref. 22) or by lesions of the Edinger-Westphal nucleus, 17 chickens underestimate distances to the grain and peck too short. There is evidence that the accommodation demand is also used to estimate distances in both chameleons 23 and owls. 24 Because an attempt to accommodate is without success after paralysis of the ciliary muscle, the motor output nucleus for accommodation produces exaggerated commands to accommodate which, in turn, result in underestimation of the distance of the target. To measure pecking errors, a narrow tank was made of glass (depth 70 to 120 mm, depending on the age of the chickens) to ensure that the chickens remained at a controlled distance from the camera while they were videotaped during pecking. Behavior analysis was performed off-line in single video frames and was aided by the fact that the back of the tank was covered with graph paper. Pecking errors were recorded as the nearest distance of the beak to the grain during single pecks. To obtain the time course of cycloplegia, chickens were observed for 3 to 4 minutes every 6 minutes for a period of 200 minutes. For each data point (Fig. 1), 20 to 30 pecks were evaluated. Measurement of Ocular Dimensions Anterior chamber depth, lens thickness, and vitreous chamber depth were measured by A-scan ultrasound as previously described. 14 The term axial length as used throughout the paper refers to the distance from the anterior cornea to the vitreo-retinal interface. Data on the anterior chamber depth are only men-

4 Cycloplegia and Emmetropization in Chicks 3519 Statistics Data are presented as mean ± one SD. To evaluate significances, paired f-tests were employed if two differently treated eyes of the same animals were compared. Unpaired Hests were used if different individuals were compared. RESULTS time after begin of drug application (min) FIGURE l. Analysis of pecking errors in four chicks to quantify the time course of cycloplegia. Pecking errors were measured as the minimal distance of the tip of the beak from the target (food) during a peck. Note that pecking errors were maximal between 60 and 150 minutes after termination of vecuroniumbromide application, and they leveled off between 120 and 200 minutes for both 7- and 12-day-old chicks. tioned if there were significant changes with treatment. Optical Techniques Refractions were measured using the latest version of infrared photoretinoscopy, 25 which is not subject to inter-observer variability. In this procedure, infrared light-emitting diodes from several eccentricities of the photoretinoscope are operating simultaneously and create an almost linear brightness profile in the pupil. The slope of the profile was found to be linearly related to refractive error. 25 ' 26 The automated version of infrared photoretinoscopy gave about two diopters more hyperopic readings than the previous nonautomated versions 14 but was found to work very reliably. During measurement of the refractions for Figure 5, a problem with the video equipment caused more than normal scatter and perhaps led to an underestimation of the differences in refractions for the different lenses. Five measurements were taken from each eye and were averaged. Corneal radius of curvature was determined by infrared photokeratometry as previously described. 1 Time Course of Cycloplegia Vecuroniumbromide-treated chicks fixated the grain on the ground from a larger distance (approximately 5 cm) compared to untreated chicks (approximately 3.5 cm); because the length of the pecks appeared to follow a fixed behavioral pattern, 22 cyclopleged animals undershot with pecks that ended midway in the open air. After six initial applications of 1 to 2 drops of vecuroniumbromide (applied at 6-minute intervals), the pecking errors continued to increase during the next hour and reached maximal values, on average, 68 ± 26 minutes after termination of the drugapplication (Fig. 1). In six chickens, pecking errors were quantitatively evaluated on five subsequent days. Data of four individuals (collected on days 7 and 12) are shown in Figure 1. There was a stinking interindividual variability in the magnitude of pecking errors, indicating variable sensitivity to the drug or indicating that individual chicks relied differently on accommodation cues. Without repeated application, the action of the drug leveled off after 120 to 200 minutes (Fig. 1). We tried to extend the period of cycloplegia by using higher doses of vecuroniumbromide; however, side effects such as partial paralysis of lid lever muscles precluded further extension. To ensure that cycloplegia was complete during the 3-hour periods of visual exposure, in the subsequent experiments a few drops were also given at 90 minutes and 150 minutes. Control Experiments With Normal Vision Effect of a Changed Light-Dark Cycle, Cycloplegia, and Intravitreal Injections. Because cycloplegia could be maintained for only about 3 hours, the chickens were transferred to a light/dark cycle with only 3 hours of daily visual exposure. It was necessary to verify that the chickens developed normally under these light cycles. The weight curves were not different from controls raised in a 12-hour light/12-hour dark cycle (Fig. 2; group 2 versus group 3), even if the chicks were cyclopleged in addition (Fig. 2; group 4 versus group 5). Refractive development was not significantly different from a control (groups 2 and 3, Fig. 3A) even if one eye was cyclopleged in addition (group 6, Fig. 3A) or had received a single intravitreal injection of 6-OHDA to suppress the local-retinal growth control mechanism (group 7, Fig. 3B). However, we found that the changed light cycles had some effect on axial

5 3520 Investigative Ophthalmology 8c Visual Science, August 1994, Vol. 35, No sobody weight (g) /21 L/D (group 5) 3/21 L/D cycl (group 4) 3/21 L/D group 3) 12/12 L/D (group 2) FIGURE 2. Effect of different light cycles and additional cycloplegia on development of the chickens. Note that none of the treatments had an effect on body growth. For treatment details, see Table 1. eye growth: Chicks raised under a 12-hour light/12- hour dark cycle had slightly shorter eyes than chicks raised under the 3-hour light/21-hour dark cycle (day 22: 9.84 ± 0.05 mm versus ± 0.14 mm, P < 0.01; Fig. 3A, groups 2 and 3). With the 3-hour light/21- hour dark cycle, daily cycloplegia had no effect on axial eye growth either in the noninjected chicks (Fig. 3A, group 6) or in chicks with additional bilateral intravitreal injections of 6-OHDA (Fig. 3B, group 7). Finally, corneal refractive power increased significantly in chicks raised in the 3-hour light/21-hour dark cycle compared to animals raised in the 12-hour light/ 12-hour dark cycle (day 22: 3.41 ± 0.05 mm versus 3.58 ± 0.06 mm, P < 0.001; Fig. 3A, groups 2 and 3). The anterior chamber depth was slightly reduced (1.45 ± 0.05 mm versus 1.34 ± 0.08 mm, P < 0.003). If the eyes cyclopleged in addition, the cornea became steeper (Fig. 3, group 6). However, the latter effect was at the margin of detectability (3.58 ± 0.06 mm versus 3.43 ± 0.05 mm, P< 0.05). The intravitreal injections of 6-OHDA had no effect on corneal radius of curvature (Fig. 3B, group 7). Development of Deprivation Myopia Under the Different Treatments It was found earlier 8 and confirmed in the present study (Fig. 4, group 10; shaded symbols) that 6-OHDA suppresses the development of deprivation myopia. As described above for groups 2 and 3 (Fig. 3), it was also found in this group of chicks that the 12-hour light/12-hour dark cycles produced shorter axial eye lengths than the 3-hour light/21-hour dark cycles (Fig. 4). Under the 3-hour light/21-hour dark cycle, deprivation myopia developed much more slowly and its amount was significantly reduced ( 2.0 ± 2.7 D versus -9.4 ± 1.1 D, P< 0.001; group 8 versus group 10), but saline-injected eyes became even more significantly myopic than fellow eyes injected with 6-OHDA ( 2.0 ± 2.7 D versus +4.8 ± 1.9 D, P < 0.001; group 8). The differences in refraction correlated with the changes in axial length, although the differences between the 6-OHDA and saline-injected deprived eyes amounted to only about 50% of what was seen in animals raised under the 12-hour light/12-hour dark cycle (Fig. 4, group 10). In 6-OHDA injected and occluded eyes, additional daily cycloplegia had no further effect on the refractive development or on axial eye growth (Fig. 4, group 9). Development of Lens-Induced Refractive Errors In chickens, it has been previously shown that the optical effects of lenses are rapidly corrected because axial eye growth is accelerated or reduced and that these growth effects are not altered by intravitreal injections of 6-OHDA. 10 We tested whether additional cycloplegia can block the growth effects of lenses. Both eyes were injected with 6-OHDA and were treated with the same lenses, but only one eye was cyclopleged. This protocol was assumed to be highly selective for detection of the possibly small changes 10" group 6 groups 2&3 3/21 L/D A 3/21 L/D O 3/21 L/D cycl A- 12/12 L/D axial length (mm) corned radius (mm) axial length (mm) corned radius (mm) group 7 3/21 L/D, 6-OHDA non cycl 3/21 L/D, 6-OHDA cycl FIGURE 3. Control experiments on chickens with normal vision. Four groups of six chickens were used to test the effects of changed light cycles (groups 2 and 3), cycloplegia (group 6), and bilateral 6-OHDA injections with additional cycloplegia (group 7) on refractive development, axial eye growth, and corneal radius of curvature. *P < 0.05; **P < 0.01; ***P < For details, see text and Table 1.

6 Cycloplegia and Emmetropization in Chicks 3521 axial length (mm) 3121 LID: (groups 8&9) saline + Occluder 6-OHDA + occluder 6-OHDA + occluder: no cycl. 6-OHDA + occluder: cycl. necessary to guide emmetropization even if the localretinal growth mechanism is suppressed. The following hypotheses could be raised to explain the observations: 1. Accommodation is not involved in the process of emmetropization; in this case, it is probable that both deprivation-induced and lens-induced growth changes of the eye are produced by local-retinal mechanisms. Furthermore, one must then assume that 6-OHDA has an effect on local-retinal mechanisms only with occluder-induced blur but not with defocus blur. This hypothesis cannot easily explain additional observations listed in the introduction; however, it would readily explain why cycloplegia had no effect on refractive development in the present study with occluders or lenses. 2. Accommodation is involved in emmetropization and drives a second feedback loop of emmetropi- refractions (diopters) FIGURE 4. Control experiments to test the effect of the different treatments (Fig. 3) on deprivation myopia {top graph, refractions; bottom graph, axial lengths). Three groups of six chickens were tested. Both eyes of each chicken were covered with occluders with gaps in the frontal part to permit normal foraging. The change in light cycle severely reduced deprivation myopia to about 50% (gray filled circles versus black filled circles; group 8 versus group 10). A single intravitreal injection of 150 [ig 6-OHDA suppressed deprivation myopia entirely in both groups (gray open circles and black open circles). Cycloplegia had no effect on deprivation myopia in chicks injected bilaterally with 6-OHDA (group 9). Significance levels are indicated in the legend to Figure 3. For treatment details, see Table axial length (mm) (3121 LID, 6-OHDA) (groups 11 & 12) in eye growth. The lenses exerted their normal effect on refractive state and axial eye growth (P < for refractions and P < for axial length; Fig. 5; group 11 versus group 12); only in the refractions after negative lens treatment was there a small change in the cyclopleged eyes (cyclopeged, +3.5 ± 1.7 D; noncyclopeged D;? < 0.05); however, the difference was at the margin of detectability and achieved significance only because of the paired Rest. All the other measurements gave no indication of an effect of cycloplegia on axial eye growth with lenses. DISCUSSION We have found that growing chicken eyes can compensate for defocus imposed by lenses even if they are injected with 6-OHDA and are also cyclopleged. The results indicate that functional accommodation is not FIGURE 5. Effect of cycloplegia on refractions and axial eye growth in chicks injected with 6-OHDA and treated with identical positive lenses (+4 D) in front of both eyes (group 11) or identical negative lenses ( 4 D, group 12). One eye was cyclopleged in each animal. Positive lenses produced more hyperopic refractions (P < 0.01) and shorter eyes (P < 0.001) than negative lenses in cyclopleged and noncyclopleged eyes. The magnitude of the growth effects was not changed even if accommodation was paralyzed by daily cycloplegia and if local-retinal growth control mechanism were lesioned by intravitreal injections of 6-OHDA. Endpoint refractions and axial lengths are illustrated by the inserted column graphs. Significance levels are indicated in the legend to Figure 3. For treatment details, see Table 1.

7 3522 Investigative Ophthalmology & Visual Science, August 1994, Vol. 35, No. 9 zation; in this case, previous observations listed in the introduction can be explained. However, it remains unclear why cycloplegia had no effect in the present experiments. Our hypothesis is that information on accommodation tonus reaches the eye via a pathway different from the ciliary nerve. If this were true, one possibility is that the choroidal nerves are important. In fact, there are also some results in favor of this assumption. The outputs of the Edinger-Westphal nucleus and the ciliary ganglion innervate the ciliary muscle, the iridial muscles, and the choroid. 18 It has been shown that activity of the Edinger-Westphal nucleus also controls choroidal blood flow. 19 Wallman et al 27 have shown that the choroid expands rapidly during ocular compensation of positive lens-induced refractive errors or during recovery from deprivation myopia, 27 and it has been shown that choroidal blood flow is reduced while deprivation myopia develops. 28 It could, therefore, be that the presumed accommodative feedback loop affects eye growth via the choroidal nerves by changing choroidal blood flow. That explains why in this study the activity of the ciliary muscle had no effect on eye growth. The hypothesis could be proven by demonstrating that lenses have no effect in 6-OHDA-injected chicks with lesions in the Edinger- Westphal nucleus or in the ciliary ganglion; in both cases, the accommodation-related inputs to the choroidal nerves are blocked. Possible Experimental Limitations One critical factor in the current study is that cycloplegia might have been incomplete in some cases and that residual accommodation drove the growth effects. There are two arguments against this conclusion. It has been shown by Schmid et al 12 that the growth changes produced by lenses are not very affected if the treatment is interrupted by periods of normal vision. After removal of the lenses, the previous shifts in accommodation tonus are reversed; if working with a short time constant, the accommodation feedback loop should then rapidly correct the growth changes that had already occurred. However, the observations of Schmid et al 12 suggest that the growth-controlling mechanisms integrate over refraction with quite a long time constant. Therefore, even incomplete cycloplegia should have reduced the effect of the lenses in our study, but this result was not found. Another factor that needs attention is that negative lenses induce positive accommodation, whereas positive lenses can be fully corrected only by negative accommodation. Only positive accommodation is blocked by vecuroniumbromide (a nicotinergic cholinergic antagonist). Because the chickens have at least 4 D of negative accommodation, 29 they can also easily compensate the defocus imposed by weak positive lenses. As a result, one limb of the putative accommodation feedback loop is probably not suppressed by vecuroniumbromide; ultimately, only myopia produced by positive accommodation should have been suppressed. In our experiments, there was no hint that negative lenses were less efficient under cycloplegia. We even measured a slight increase in relative myopia (Fig. 5), although this effect was not mirrored by changes in the axial lengths. Finally, it must be taken into account that the average viewing distance of a chick is not at infinity. Therefore, for shorter viewing distances, even positive lenses change the level of positive accommodation. There is some evidence that positive lenses (growth inhibition) and negative lenses (growth promotion) trigger two different systems within the presumed accommodation feedback loop. Positive lenses are more rapidly and more completely compensated than negative lenses. 16 ' 2 The difference may partly result from different possible ranges of choroidal compensation of refractive error. 2 ' Growth changes induced by negative lenses can be more easily suppressed by intermittent periods of normal vision than of positive lenses. 12 The optic nerve affects the compensation of negative lenses much more than of positive lenses. 13 The entire pattern of results is far from being understood, and additional experiments are necessary to separate possible influences of positive and negative accommodation on eye growth. For instance, negative accommodation can be blocked by timolol maleate, 29 and it would be important to test whether the growth responses with positive lenses are changed after its application. Magnitude of Deprivation-Induced and Lens- Induced Refractive Errors Napper et al 30 found that a 12-hour light/12-hour dark cycle reduces deprivation myopia by about 60%. A similar result was found here for the 3-hour light/21- hour dark cycle. On the other hand, the magnitude of changes in axial length with +4 D and 4 D lenses was comparable to previous results (0.33 mm not cyclopleged and 0.45 mm cyclopleged in our study compared with 0.36 mm noncyclopleged previously reported 15 ). Assuming that a change of one diopter in refraction is equivalent to a change in axial length of 60 /im, 32 the difference is equivalent to a difference in refraction of 5.5 D and 7.5 D, respectively; it would, therefore, largely account for optical effects of the lenses. The differences in refraction actually measured were much smaller (2.0 D and 2.7 D, respectively; Fig. 5). Because there were no significant changes measured in corneal refractive power and anterior chamber depth (data not shown), the comparably small changes in refraction remain unexplained. Comparisons to Mammalian Models of Myopia There is some evidence that deprivation myopia is based on a local retinal mechanism also in mammals

8 Cycloplegia and Emmetropization in Chicks 3523 because it can be induced without accommodation (tree shrew, 32 grey squirrel 33 ) locally in restricted retinal areas (tree shrew 34 ) and after lesions of the ganglion cells by TTX (tree shrew 35 ). Experiments with lenses are rare (e.g., tree shrew, 36 cats, 37 Rhesus monkeys 38 ), and, strikingly, sign-dependent compensation of imposed refractive error has not been demonstrated until now. One possible explanation for the lack of this important result is that the treatment was always asymmetrical even though the animals had coupled accommodation. Unfortunately, until the latter result has been obtained, the results of the current study in the chicken cannot be applied to mammals. Key Words accommodation, emmetropization, chicken Acknowledgments The authors thank Gabi Hagel for help with the data processing and Prof. E. Zrenner for general support. References 1. Schaeffel F, Glasser A, Howland HC. Accommodation, refractive error and eye growth in chickens. Vision Res. 1988;28: Irving EL, SivakJG, Callender MG. Refractive plasticity in the developing chick eye. Ophthalmic Physiol Opt. 1992; 12: Bartmann ML, Schaeffel F. A simple mechanism for emmetropization without cues from accommodation or color. Vision Res. 1994; 34: Wallman. J, Turkel J, TrachtmanJN. Extreme myopia produced by modest changes in early visual experience. Science. 1978;201: Hodos W, Kuenzel WJ. Retinal image degradation produces ocular enlargement in chicks. Invest Ophthalmol Vis Sci. 1984;25: Wallman J, Gottlieb MD, Rajaram V, Fugate-Wentzek L. Local retinal regions control local eye growth and myopia. Science. 1987;237: Troilo D, WallmanJ. Visual deprivation causes myopia in chicks with optic nerve section. Curr Eye Res. 1987;6: Li XX, Schaeffel F, Kohler K, Zrenner E. Dose dependent effects of 6-hydroxy dopamine on deprivation myopia, electroretinograms and dopaminergic amacrine cells in the chicken. Vis Neurosci. 1992;9: Bartmann ML, Schaeffel F, Hagel G, Zrenner E. Constant light affects retinal dopamine levels and blocks deprivation myopia but not lens-induced refractive errors in chickens. Vis Neurosci. 1994; 11: Schaeffel F, Hagel G, Bartmann ML, Kohler K, Zrenner E. Six-hydroxy dopamine does not affect lens-induced refractive errors but suppresses deprivation myopia. Vision Res. 1994;34: Gottlieb MD, Nickla D, WallmanJ. The effects of abnormal light/dark cycles in the development of form deprivation myopia. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1992;33: Schmid KL, Wildsoet CF, Pettigrew JD. The effect of daily periods of normal vision on refractive adaptation development in chicks. Invest Ophthalmol Vis Sci. 1993; 34: Wildsoet C, Wallman J. Optic nerve section affects ocular compensation for spectacle lenses. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1992; 33: Schaeffel F, Howland HC. Properties of feedback loops controlling eye growth and refractive state in the chicken. Vision Res. 1991;31: Rohrer B, Schaeffel F, Zrenner E. Longitudinal chromatic aberration and emmetropization: Results from the chicken eye. J Physiol. 1992; 449: Schaeffel F, Howland HC. A mathematical model of emmetropization in the chicken. / Opt Soc Am [A]. 1988; 5: Schaeffel F, Troilo D, WallmanJ, Howland HC. Developing eyes that lack accommodation grow to compensate for imposed defocus. Vis Neurosci. 1990; 4: Gamlin PDR, Reiner A. The Edinger-Westphal Nucleus: Sources of input influencing accommodation, pupilloconstriction, and choroidal blood flow. J Comp Neurol. 1991; 306: Shi YF, Malinda E, Fitzgerald C, Reiner A. Effect of choroidal and ciliary nerve transection on choroidal blood flow, retinal health, and ocular enlargement. Vis Neurosci. 1993; 10: Pettigrew JD, WallmanJ, Wildsoet CF. Saccadic oscillations facilitate ocular perfusion from the avian pecten. Nature. 1990; 343: Pilar G, Numez R, McLennan IS, Meriney SD. Muscarinic and nicotinic synaptic activation of the developing chick iris. /Neurosci. 1987; 7: Stull D, Troilo D, Howland HC. What is the role of accommodation in the perception of target distance and pecking behavior in chicks? Soc Neurosci. 1993; 19:552. Abstract. 23. Harkness L. Chameleons use accommodation cues to judge distance. Nature. 1977;267: Wagner H, Schaeffel F. Barn owls (Tyto alba) use accommodation as a distance cue. / Comp Physiol [A]. 1991;169: Schaeffel F, Hagel G, Eikermann J, Collett T. Lowerfield myopia and astigmatism in amphibians and chickens. / Opt Soc Am [A]. 1994; 11: Schaeffel F, Wilhelm H, Zrenner E. Inter-individual variability in the dynamics of natural accommodation in humans: Relation to age and refractive errors. / Physiol. 1993; 461: WallmanJ, Xu A, Wildsoet C, et al. Moving the retina: A third mechanism of focusing the eye. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1992; 33: Reiner A, Erichsen JT, Carbot JB, Evinger C, Fitzgerald MEC, Karten H. Neurotransmitter organization of the nucleus of Edinger-Westphal and its projection to the avian ciliary ganglion. Vis Neurosci. 1991;6: Troilo D, Glasser A, Li T, Howland HC. Negative accommodation occurs in the chick and may be mediated by sympathetic input. ARVO Abstracts. Invest Ophthalmol Vis Sci. 1993; 34:1310.

9 3524 Investigative Ophthalmology & Visual Science, August 1994, Vol. 35, No Napper GA, Algis JV, Squires MA, Vessey GA, Barrington M, Brennan NA. Influence of continuity of exposure and length of light/dark cycle on occlusioninduced myopia. ARVO Abstracts. Invest Ophthalmol Vis Sri. 1992; 33: Schaeffel F, Howland HC. Visual optics in normal and ametropic chickens. Clin Vis Sri. 1988; 3: McBrien NA, Moghaddam HO, Reeder A. Atropine reduces experimental myopia and eye enlargement via a nonaccommodative mechanism. Invest Ophthalmol Vis Sri. 1993;34: McBrien NA, Moghaddam HO, New R, Williams LR. Experimental myopia in a diurnal mammal (Sciurus carolensis) with no accommodative ability. J Physiol. 1993;469: Norton TT, Siegwart JT. Local myopia produced by partial visual-field deprivation in tree shrew. Soc Neurosri. 1991; 17:558. Abstract. 35. Norton TT. Experimental myopia in tree shrews. In: Myopia and the Control of Eye Growth. Ciba Foundation Symposium 155. Chichester: Wiley; 1990: Siegwart JT, Norton TT. Refractive and ocular changes in tree shrews raised with plus or minus lenses. ARVO Abstracts. Invest Ophthalmol Vis Sri. 1993;34: Ni J, Smith EL III. Effects of optical defocus on the kitten's refractive status. Vision Res. 1989; 29: Crewther SG, Nathans J, Kiely PM, Brennan NA, Crewther DP. The effect of defocusing contact lenses on the refraction in cynomolgus monkeys. Clin Vis Sri. 1988;3: