L-opsin expression in chickens is similarly reduced with diffusers and negative lenses

Previous studies have shown that the expression of L- and M (cid:0) opsins was reduced in chicken retina when eyes were covered with diffusers. The purpose of the current study was to find out whether this is a result of altered spatial processing during development of deprivation myopia or merely a consequence of light attenuation by the diffusers. Therefore, retinal luminances were matched by neutral density filters in fellow eyes that served as controls for diffuser-treated eyes. Furthermore, the effects of negative lenses on opsins expression were studied. Chickens wore diffusers or (cid:0) 7 D lenses for a period of 7 days and refractive state and ocular biometry were measured at the beginning and at the end of the experiment. Retinal tissue was extracted from both eyes to quantify L-, M- and S-opsins expression by qRT-PCR. It was found that L-opsin expression was significantly lower in eyes wearing diffusers, compared to fellow eyes covered with neutral density filters. Interestingly, L-opsin was also reduced in eyes wearing negative lenses. In summary, this study shows that L-opsin expression is reduced due to the loss of high spatial frequencies and general contrast reduction in the retinal image, rather than by a decline in retinal luminance. Moreover, the fact that L-opsin was similarly reduced in eyes treated with negative lenses and diffusers suggests the existence of a common pathway for emmetropization, but it could also be just a consequence of reduced high spatial frequencies and lower contrast.


Introduction
Studies in animal models and human clinical trials have demonstrated that the fine-tuning of eye growth is controlled by retinal image processing, such that refractive state reaches an optimum during postnatal development (Chamberlain et al., 2021;Diether and Schaeffel, 1997;Grosvenor and Scott, 1994;Marsh-Tootle and Norton, 1989;Smith and Hung, 1999;Wallman and Winawer, 2004). The increase in myopia prevalence over the past decades have triggered a large amount of research to find out whether myopia could be prevented or at least inhibited already at young ages (Holden et al., 2016;Williams et al., 2015). The key problem of myopia is that some of the myopic patients develop high myopia where progressive thinning of the choroid may cause retinal pathologies already in the middle adulthood (Fernández et al., 2022;Flitcroft, 2014;Sanz Diez et al., 2019;Saw et al., 2005;Zhao et al., 2002).
Myopia development is based on interactions of environmental and genetic factors (Chen et al., 2011;Morgan et al., 2012). There are gene-gene as well as gene-environment interactions that contribute to the development of myopia, so the genetic component cannot be ignored (Cai et al., 2019). However, the recent rise in myopia cannot be attributed to genetic changes and it is evident that the major reason is the change in the visual environment, linked to more rigorous educational programs and learning requirements (Mirshahi et al., 2014;Morgan and Rose, 2005). During myopia development, signals from the retina lead to choroidal thinning and enhanced growth of the sclera (Chakraborty and Pardue, 2015;Wallman and Winawer, 2004).
We have previously found that cone photoreceptors play an important role in myopia development, beyond that they simply mediate light sensitivity. In the chicken, cone abundancy ratios are genetically determined and have predictive value for later refractive development. The higher the number of L-cones, compared to M− cones, the less hyperopic were the animals. Conversely, more S-cones, compared to Lcones, were associated with more hyperopia when the chickens had normal visual experience (Gisbert et al., 2022). Strikingly, L-opsin expression was reduced when deprivation myopia was induced (Gisbert et al., 2020).
Chicken photopigments are sensitive to light in the wavelength range between 360 and 660 nm (Rohrer et al., 1992). They are synthesized, as in all vertebrates, from all cis-retinal and an opsin protein molecule that determines the spectral absorption (Terakita, 2005). Changes in opsin expression can therefore alter the photoreceptor output and consequently the signals transmitted to downstream retinal neurons (Greenwald et al., 2017;McClements et al., 2013;Neitz and Neitz, 2021).
In a recent study, (Ji et al., 2021) investigated the role of M− opsin on emmetropization in the mouse model. Mice were used with a mutation in M− opsin (Opn lmw -/ -) which made them less sensitive to green light. In wild type mice, green light produced a myopic shift in refractive states which was accompanied by an increase of M− opsin expression. In contrast, mice carrying the mutation developed more hyperopic refractions. These findings suggest that opsin expression may affect the regulation of eye growth.
In line with this conclusion, we found in chickens that L-opsin expression was reduced when deprivation myopia was induced by diffusers (Gisbert et al., 2020). Since diffusers attenuate light, the effect might have been simply light dependent. Therefore, this study aims to determine whether opsin expression is also controlled by changes in spatial information in the retinal image. To control the influence of luminance effects, fellow eyes were covered with neutral density (ND) filters that matched retinal luminances in both eyes. Furthermore, we studied the effects of imposed negative defocus on opsin expression. Since lenses attenuated light only marginally, fellow eyes were simply covered with a clear foil to account for potential minor effects of dust and humidity, and non-visual effects of temperature changes.

Animal research
All experiments were carried out in accordance with the ARVO statement for the use of animals in ophthalmic and vision research, established by the National Institutes of Health Guide for the Care and Use of Laboratory Animals, and approved by the Animal Welfare Commission of the University of Tuebingen. One-day old White Leghorn H&N chickens (Gallus gallus domesticus) were acquired from a chicken farm (A.C Weiss GmbH & Co. KG, Kirchberg, Germany).

Rearing conditions
Chickens were kept in the animal facilities of the Institute for Ophthalmic Research. Animals were provided with food and water supplied ad libitum. Humidity and temperature were controlled, and light was set at 12 h:12 h light/dark cycle, synchronized with the circadian rhythms. Experiments started at day 7 post-hatching, and treatments were continued for 7 days in all experimental groups. Chickens were raised under blue-enriched white light source (BIOLX T8 fluorescent tube, OSRAM GmbH Munich, Germany) which is typically used in animal facilities. The light spectrum ranged between 380 and 780 nm and presented high energy between 425 and 530 nm to achieve a better match with the sunlight spectrum ( Fig. 1 (control)). Illuminance was measured at different points within the cage and was on average 451.21 ± 22.06 lx. Light spectra and illuminances were determined by a photospectrometer (Gossen MAVOSPEC BASE, Nuremberg, Germany).

Refraction and ocular biometry
Ocular parameters were measured in alert animals before and after the treatment period. Refractions were measured by automated infrared photoretinoscopy (Seidemann and Schaeffel, 2002) at low illuminances to limit accommodative stimuli and keep pupils large. Five repeated measurements were taken per eye. No cycloplegia was used. Axial length and vitreous chamber depth were obtained by A-scan ultrasonography (Schaeffel and Howland, 1991). Topical anesthetic drops of 2% Xylocaine solution (Aspen, Munich, Germany) were used to facilitate the measurements. Axial length was defined as the distance between the anterior surface of the cornea and the vitreoretinal interface, while vitreous chamber depth was taken from the posterior lens surface to the vitreoretinal interface. Four repeated measurements were obtained per eye.

Fig. 1.
Spectral energy distribution of the light source in the animal facilities though the different experimental conditions tested. Each color represents a different treatment, with the corresponding legend provided. Black dash line represents the control light spectral profile measured without treatment. The data indicated that the experimental treatments did not affect light spectral transmission. However, they did affect irradiance levels.

Retinal tissue extraction
At the end of the experiment, chickens were sacrificed by an overdose of ether inhalation and eyes were enucleated and cut perpendicular to the sagittal plane. The vitreous humor was stripped away from the posterior eye cup and circular retinal segments of 6 mm diameter were taken from the central area using the pecten as a reference. Retinal pigment epithelium (RPE) and sclera were discarded, and retinal tissue was kept and stored at − 80 • C for opsin quantification by real time polymerase chain reaction (PCR).

Reverse transcription polymerase chain reaction (qRT-PCR)
Reverse transcription (RT) followed by the polymerase chain reaction (PCR) was carried out to quantify changes in the mRNA expression of L-, M-, and S-opsin genes. First, the total RNA was extracted using RNeasy® Plus Kit (Qiagen, Hilden, Germany) and isolated according to the DNA-free™ Kit protocol (Thermo Fisher Scientific, Vilnius, Lithuania). RNA concentration and quality was assessed from 1 to 2 µl of sample using NanoDrop® ND-1000 spectrophotometer (Thermo Fisher Scientific, Wilmington, USA). Later, to initiate the reverse transcription procedure, 500 ng of RNA were taken from each sample to synthetize the complementary strand of DNA (cDNA) using SuperScript VILO cDNA Synthesis kit (Invitrogen, Carlsbad, USA). Forward and reverse primers used for L-, M-and S-opsin genes amplification had the same sequences as those used in our former experiment to keep equal conditions (Gisbert et al., 2020). HPRT and beta-actin were chosen as reference genes (Table 1). PCR was performed using an iCycler device (CFX96 TM System, BioRad) and the protocol temperature was defined as follows: initial denaturation phase of 3 min at 95 • C followed by 40 cycles of 30 s at 95 • C, 15 s at 60 • C for primer annealing, and 15 s at 72 • C to allow the DNA polymerase extends the primers sequencies. Before running the samples, PCR primer efficiencies were calculated using the formula: E = 10 (-1/slope) by making serial dilutions of cDNA. PCR was performed in 96-well transparent plates and 2 ng of cDNA were deposited per well. Three independent replicates were assessed per each gene. SYBR Green (QuantiNova, Hilden, Germany) was used to quantify the amount of PCR product and the fluorescence intensity was analyzed at 72 • C. Opsin expression was quantified using the mean normalized expression (MNE) based on the relative amount of a target gene compared to a reference gene. This equation takes into consideration the mean cycle threshold (CT) value of the target and reference genes, and the efficiency (E) of the PCR (Pfaffl, 2001):

Experimental groups and treatments
Experiment 1. Deprivation myopia induced by diffusers. A group of 12 chickens was monocularly treated with hemispherical diffusers made of heavily frosted plastic foil. Diffusers act as spatial low pass filters and decrease contrast in the retinal image over the entire spatial frequency spectrum. Such filters were used in all our previous studies to induce deprivation myopia. The treatment was continued for 7 days. Fellow eyes were covered with ND filters (Kodak Wratten Neutral Density Filter) to match retinal luminances in both eyes. ND filters are defined by their Optical Density (OD) which indicates the proportion of energy blocked by the filter. In this experiment, the optical density of the filters was chosen to match light attenuation caused by the diffusers. Required optical density was determined by a Minolta luminance meter (LS-100; Minolta Camera Co., LTD, Tokyo, Japan) directed at a white surface 20 cm away. The average luminance from three repeated measurements was 304 ± 12 cd/m 2 . Subsequently, five different diffusers were placed in front of the detector of the candela meter and were found to cut the luminances down to 150.3 ± 7.3 cd/m 2 on average. Accordingly, diffusers attenuated light by about 50%, equivalent to an optical density of a neutral filter of 0.3, as calculated from: OD = -log ( T 100% ). Both, neutral density filters and diffusers, transmitted similarly across the visible wavelength spectrum (Fig. 1). Average illuminance measured with ND filters and diffusers in animal facilities were of 215.1 ± 11.8 lx and 237.05 ± 22.9 lx respectively. Experiment 2. Myopia induced by imposing hyperopic defocus. A group of 10 chickens wore monocularly negative lenses (-7 D) for a period of 7 days to induce myopia. Fellow eyes were covered by transparent plastic foils. Average illuminance measured with negative lenses was of 343.73 ± 30.81 lx, and with transparent plastic foils of 388.81 ± 12.48 lx. Neither negative lenses nor transparent foils affected light spectrum transmission (Fig. 1).

Statistics
Statistical analyses were performed using the SPSS software package, version 22.0 for Windows (SPP, Chicago, Illinois, USA). It was confirmed that data was normally distributed, and T-Tests for related samples were used to assess differences in refractive state, biometry, and opsins gene expression between eyes within the same experimental group. Standard errors of the means (SEMs) were shown in the plots.

Refractive development and eye growth
Experiment 1. Deprivation myopia induced by diffusers. Baseline refractive states were not different between eyes before starting the experiment (right eyes + 2.49 ± 0.08 D and left eyes + 2.41 ± 0.09 D). After one week of wearing diffusers, treated eyes (average of − 4.29 ± 0.84 D) became significantly myopic (p < 0.001) in comparison to their fellow eyes, covered with ND filters, whose refractive states did not change (average + 2.67 ± 0.06 D). Eyes wearing diffusers showed significantly longer vitreous chamber depth (p < 0.001) by + 0.94 ± 0.03 mm and axial length (p < 0.001) by + 1.28 ± 0.05 mm compared to fellow eyes, wearing ND filters, which grew by + 0.40 ± 0.03 mm in vitreous chamber depth and + 0.72 ± 0.04 mm in axial length (data shown in Figs. 2 and 3). This means that in eyes with diffusers, the vitreous chamber showed 0.54 mm more growth and the axial length showed 0.56 mm more growth compared to eyes with ND filters. According to the schematic eye of the chicken described by (Schaeffel and Howland, 1988), an increase of one millimeter in axial elongation corresponds to an additional 16 D of myopia. Therefore, the schematic eye predicts about 9 D more myopia for the observed increase in axial length, but only 6.7 D were measured. Experiment 2. Myopia induced by imposing hyperopic defocus. Similar baseline refractions were measured as in "Experiment 1" (right eyes + 2.73 ± 0.09 D, left eyes + 2.54 ± 0.09 D). Eyes covered with − 7 D lenses developed only a moderate degree of relative myopia (average refractions − 0.16 ± 0.14 D) which was significantly different (p < 0.001) from control eyes (average of + 2.34 ± 0.14 D) after one week. Eyes wearing negative lenses exhibited a significant increase in vitreous chamber depth (p < 0.001) by + 0.75 ± 0.03 mm and in axial length (p < 0.001) by 1.09 ± 0.04 mm, compared to fellow eyes which grew on average + 0.28 ± 0.03 mm and + 0.71 ± 0.04 mm respectively (data shown in Figs. 2 and 3). Therefore, there was 0.47 mm more growth in the vitreous chamber and 0.38 mm more growth in axial length in eyes with lenses, compared to control eyes. The schematic eye predicts about 6 D more myopia, but only 2.5 D were measured. One possible explanation is that the schematic eye used in the study was based on a different chicken strain (Cornell K-strain). Additionally, accurate prediction of refractive errors based on changes in ocular biometry requires accounting changes in both anterior chamber depth and corneal curvature, which were not available in the current study.

Discussion
A key result of this study was that only L-opsin expression was decreased in eyes treated with diffusers, relative to their fellow eyes covered with ND filters. Since there were no differences in retinal luminances between ND filter-treated eyes and those treated with diffusers, the observed decrease in L-opsin expression must trace back to changes in spatial features like reduced contrast or loss of high spatial frequency components in the retinal image. Furthermore, L-opsin expression was also reduced in eyes wearing negative lenses (-7 D), suggesting that there should be a common pathway between deprivation and lens-induced myopia. It was striking that L-opsin expression levels were at least 10 times more abundant than M-and S-cone opsins expression.

Previous findings on changes in opsin expression
Previous studies had also found changes in opsin expression during induction of experimental myopia. Already in 2007, (Li et al., 2007) found that M− opsin expression was increased, and S-opsin expression reduced when deprivation myopia was induced in guinea pigs. Later, (Zou et al., 2018) analyzed changes in the refractive state and opsin expression in guinea pigs that were exposed to different quasimonochromatic light sources. Animals reared under short-wavelength light developed relative hyperopia and exhibited an increase in Sopsin expression and a decrease in M− opsin expression. Conversely, animals reared under mid-wavelength light developed more myopia that was accompanied by an increase in M− opsins, and a decrease in S-opsins. Similar findings were made in the mouse model (Ji et al., 2021). These authors concluded that an excessive M− opsin expression leads to a disruption in normal refractive development and causes myopia. They further proposed that a reduction of M− opsin levels had a protective effect against myopia and caused a shift toward more hyperopic refractions in mice.
In the current study, chickens showed opposite changes in L-opsin expression, compared to guinea pigs and mice. The current results are in line with our previous findings, showing that M-to L-cone ratios were correlated with the amount of deprivation myopia in the chicken model (Gisbert et al., 2022). Lower M-to L-cone ratios (indicating a higher number of L-cones relative to M− cones) were associated with less hyperopic refractions in control eyes exposed to normal visual conditions, while in treated eyes, this correlation was inverted, and lower M-to Lcone ratios were associated with less deprivation myopia. Whereas the number of cones remains unchanged in the retina after treatment with diffusers, we found that L-opsin expression is reduced. This response may be linked to the inverted correlation between M-to L-cone ratios and the refractive state in treated eyes. In contrast to our previous results, no changes in M− opsin expression were found in the current   study. This could be attributed to a signal-to-noise ratio problem caused by the low levels of M− opsin before PCR amplification.
The opposite trends in opsin expression in chickens compared to mammals may trace back to the fact that guinea pigs are dichromates and chickens are tetrachromates. It is unlikely that ultraviolet vision is relevant in this matter since both, chickens and mice have cones at around 365 nm, but guinea pigs do not (Jacobs and Deegan, 1994)nonetheless they still respond like mice.

Why is L-opsin expression so high?
L-opsin expression was much higher than M-and S-opsins expression, consistent with our previous observations (Gisbert et al., 2020). This could result from the presence of single L-cones and double cones in the chicken retina which contain exactly the same opsin type (L-opsin) (Bowmaker et al., 1997). The high L-cone opsin expression rates can be explained by the fact that (1) double cones are the most common photoreceptor type in the avian retina, making up to 40% of the total cone numbers, and (2) that both outer segments of double cones, the principal and the accessory parts, express only L-opsin but no M− opsin (Gunther et al., 2021). Because the quantification of L-opsins is based on both cone photoreceptor types, it is not possible to determine which one undergoes changes during myopia. Single cones mediate color vision, and multiple studies have supported that color vision plays a role in emmetropization (Foulds et al., 2013;Gawne et al., 2017a;Long et al., 2009;Rucker and Wallman, 2009;Smith et al., 2015;Yang et al., 2021). The role of double cones is less clear. One hypothesis indicates that they are involved in motion vision because the spectral sensitivity of motionsensitive cells is similar to the spectral sensitivity of the L-opsin expressed by double and L-cones (Jones and Osorio, 2004;Osorio and Vorobyev, 2005). Moreover, other studies have related double cones to fine pattern recognition (Lind and Kelber, 2011). Additionally, it was suggested that they play a role in luminance detection (Olsson et al., 2015). However (Rucker and Wallman, 2008) found that both, double and L-cones mediate compensatory growth responses to lenses by modulating choroidal thickness. Therefore, both cone types could potentially be involved in the emmetropization response and changes in their L-opsin expression could have an impact on emmetropization.

Possible links between opsin expression and myopia development
Although the role of opsin molecules in emmetropization is not clear, there is evidence that they are involved in the regulation of eye development. Several studies have established a link between opsin expression and retinal dopamine levels (Crewther, 2000;Park et al., 2013). (Ji et al., 2021) found an increase in M− opsin expression in wildtype mice exposed to green light which was associated with a decrease in dopamine content and myopia. Mice lacking functional M− opsins (Opn lmw -/ -) also exhibited lower retinal dopamine levels than wildtype mice. In the chicken model, a causal relationship between reduced L-opsin and decrease in dopamine could be established since myopia development is typically associated with a drop in retinal dopamine (Feldkaemper and Schaeffel, 2013).
Variations in L-and M− opsins expression have also been associated with changes in contrast sensitivity. It was proposed that cones harboring certain mutations that cause either anomalous opsin expression or different levels of opsin expression, may change the susceptibility to myopia (Greenwald et al., 2017;Hagen et al., 2019).  proposed that opsin mutations associated with myopia modify contrast signaling in bipolar cells, triggering ocular elongation. Consequently, the authors suggested that a decrease in contrast could potentially inhibit eye growth. In line with this hypothesis, novel spectacle lenses that reduce retinal contrast in the periphery, have been designed to slow down myopia progression (the "Diffusion Optic Technology" (DOT)). A one-year study has shown that DOT lenses are effective in reducing myopia progression in children, with a reported reduction of up to 74% (Rappon et al., 2022).
In this context, a possible explanation for the results presented in this paper is that L-cones in myopic eyes which express less L-opsins, would catch less photons in comparison to the average absorption of other cone types producing an abnormal retinal contrast and inducing eye growth in chickens. Nevertheless, it is known that photoreceptors can rapidly adapt to variations in the amount of light captured (Burkhardt, 1994), and further research is needed to determine the specific mechanisms underlying the observed effects.
In the current experiment, it is unlikely that the light spectrum used, with high energy presented at 550 nm, had influenced the amount of myopia induced. In our previous study (Gisbert et al., 2022) it was assessed whether different levels of light energy over a spectral range between 380 and 780 nm interact with emmetropization in different groups of chickens. It was observed that no significant differences neither in refraction nor ocular length were found among groups during deprivation myopia. However, it is unclear whether this may have affected opsin expression, as this data was not collected.
Another recently introduced treatment for myopia that may have a potential link to L-opsins is the "Low-level red-light therapy". In this technique, a red laser with a wavelength of 650 nm is directed on the fovea for a duration of 3 min, twice daily. While the mechanism by which this therapy works is still unclear, recent studies describe striking effects (Jiang et al., 2022;Yu et al., 2022). One of the possible explanations could be that red light alters the expression of L-opsins since the local laser energies are high. However, there are a number of studies showing that exposure to low energies of long wavelength light affects emmetropization. While chickens, guinea pigs and fish become more myopic (Kroger and Wagner, 1996;Liu et al., 2011;Seidemann and Schaeffel, 2002), tree shrews and monkeys become consistently more hyperopic under red light exposure (Gawne et al., 2017b;Smith et al., 2015). Further work is needed to determine whether such effects may be related to changes in L-opsin expression.

Similarities between deprivation myopia and negative lens-induced myopia
Previous studies have found that the amount of deprivation myopia induced by diffusers is correlated to the amount of retinal contrast reduction (Bartmann and Schaeffel, 1994;Smith and Hung, 2000). When images are degraded by imposed defocus, spatial contrast is reduced in correlation with the magnitude of defocus (Campbell and Green, 1965). After receiving negative lens treatment, the myopia measured with photorefraction was clearly lower than that measured with the diffusers. Given that the changes in axial eye growth were more similar between the groups (deprivation myopia: +0.56 mm; lensinduced myopia: +0.38 mm), the lower amount of myopia found in lens wearing eyes may be attributed to negative accommodation during the refractions. Schematic eye modeling showed that an increase of 0.38 mm in axial length in 2-week-old chickens would lead to a myopic shift of 6-7 D (Schaeffel and Howland, 1988) which would match the power of the negative lenses. In summary, there is no convincing evidence to suggest that the chickens did not compensate for the − 7 D lenses. Other studies have demonstrated that chicken eyes fully compensated for − 7 D lenses within a week (Diether et al., 2007).
One key difference in the retinal images produced by diffusers and lenses is that only optical defocus, generated by negative lenses, coherently shifts the different focal planes resulting from longitudinal chromatic aberration. Specifically, short-wavelength blue light is brought into better focus than green and red. It is possible that poorer image quality in the red region of the spectrum may result in suboptimal stimulation of the photoreceptor cells responsible for processing longwavelength light, potentially leading to changes in their photopigment expression.
A key finding remains that L-opsin expression was similarly reduced in eyes wearing diffusers and lenses, despite differences in the effects of longitudinal chromatic aberration and the induced amount of myopia. In order to demonstrate differences between both types of myopia, the level of image blur must be matched. The findings of the current study suggest that deprivation myopia and lens-induced myopia share common retinal pathways.

Conclusions
The outcomes of this research have provided a deeper insight of opsin expression when experimental myopia is induced. A key finding was that L-opsin levels are similarly reduced when myopia was induced either by diffusers or negative lenses. Furthermore, it showed that the effect was not due to reduced retinal luminances but rather due to changes in the spatial features of the retinal image. Why changes in the spatial information interact with L-opsin expression is not immediately obvious; it is also not clear why these changes occur only in the L-opsin content but not in M-and S-opsins. Generally, L-opsins were much more abundant than M-and S-opsins because double cones also express Lopsins and they represent up to 40 % of the photoreceptors. L-opsin expression levels may have an effect on the signal strength of L-cones, double cones, or both, potentially interacting with neural signals that regulate eye growth and refractive development. Recent studies in monkeys and tree shrews have shown that red light induces more hyperopic refractions, so it is therefore important to understand the mechanisms underlying the inhibitory effects of long wavelength light on myopia.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability
Data will be made available on request.