Unaltered Retinal Dopamine Levels in a C57BL/6 Mouse Model of Form-Deprivation Myopia

Abstract

Purpose.: Retinal dopamine has been long implicated in the signaling pathway regulating eye growth, as evidenced by its reduced levels in myopic eyes in various species. We examined whether and how retinal dopamine levels were changed in a C57BL/6 mouse model of experimental myopia.

Methods.: Form-deprivation myopia (FDM) was induced in C57BL/6 mice by wearing monocular occluder for 4 weeks. Refractive errors were measured using an infrared photorefractor. Retinal dopamine/DOPAC and vitreal DOPAC levels were assessed by high-performance liquid chromatography (HPLC). Extracellular dopamine concentrations were examined by Western blot analysis of dopamine transporter (DAT) expression levels. The intactness of retinal dopaminergic system was evaluated by counting tyrosine hydroxylase (TH) immunoreactive cells, measuring the areas occupied by processes of these cells, and quantifying retinal TH expression at both protein and transcription levels.

Results.: Form-deprivation myopia was successfully induced in C57BL/6 mice with the refractive status of deprived eyes being significantly different from fellow eyes. Unlike most of the previous results obtained in other myopic animal models, however, no significant changes in retinal dopamine, DOPAC, DAT, and vitreal DOPAC levels were detected in deprived eyes, either in the daytime or at night. Furthermore, neither the number of dopaminergic amacrine cells, the area size occupied by the processes of these cells, nor retinal TH expression, were altered in deprived eyes.

Conclusions.: The retinal dopamine system remains intact in C57BL/6 mice with FDM, and retinal dopamine levels are not associated with the development of FDM in this mouse strain.

Introduction

Myopia is a common ocular disorder, caused by excessive axial growth during visually-guided eye development.^1,2 In order to treat and prevent myopic refractive error effectively, it is important to characterize the signaling cascade carrying information from retina to sclera, which controls the visually-driven ocular elongation. To this end, a wide variety of animal species, including chicks,^3,4 tree shrews,^5–8 guinea pigs,^9–11 rabbits,¹² cats,^13–15 and primates,^16–20 have been used to establish experimental myopia models. Multiple lines of evidence obtained from these animal models implicate dopamine (DA), a neurotransmitter released exclusively by a population of amacrine/interplexiform cells,^21,22 as a key messenger molecule in this cascade.²³ Retinal concentrations of DA and its primary metabolite 3,4-dihydroxyphenylacetic acid (DOPAC) are significantly reduced in various animals with experimental myopia.^24–28 In agreement with it, agents that activate DA receptors, especially the D₂ subtype, when applied either subconjunctivally or intraocullarly, inhibit the development of myopia.^27–37 The fact that outdoor activity might protect children from myopia development seems to support the involvement of DA, since brighter sunlight is thought to promote DA release.^23,38–40 Moreover, pharmacological manipulation lowering retinal DA contents is known to affect refractive development. For instance, reducing retinal DA levels by intravitreal injection of 6-Hydroxydopamine (6-OHDA),^36,41–45 or depleting the retinal DA store by reserpine,^36,46,47 has been found to suppress the progress of experimental myopia in chicks. 

Recently, experimental myopia has been successfully induced in the mouse, by eyelid suture/frosted diffusers (form-deprivation myopia, FDM),^48–53 or by negative spectacle lenses (lens-induced myopia, LIM).^48,52 However, there are very few data concerning the status of dopaminergic system in mice with experimental myopia now available. Specifically, it remains unclear whether retinal DA levels and dopaminergic system also undergo substantial changes in myopic mice, just as observed in other species. In this study, we assessed retinal DA contents of C57BL/6 mice with FDM using high-performance liquid chromatography (HPLC). Although form-deprived eyes displayed a significant myopic shift in refractive error, no changes in retinal DA levels in either the daytime or the nighttime were observed in these eyes. Consistently, vitreal DOPAC levels also remained unchanged, no matter the measurements were made during the daytime or at night. Furthermore, the deprived eyes appeared to maintain normal retinal DA synthesis machineries, as evidenced by the unchanged density of amacrine cells labeled by tyrosine hydroxylase (TH), unchanged areas occupied by the processes of TH-immunoreactive (TH⁺) cells, and unaffected TH expression measured at both protein and mRNA levels. These results suggest that retinal DA levels are not associated with the development of FDM in this mouse strain. 

Materials and Methods

Induction of Form-Deprivation Myopia

All experiments were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and the regulations of Fudan University (Shanghai, China) for animal experimentation. Male C57BL/6 mice (3-weeks old before form-deprivation) were used in this study. Induction of FDM in mice followed the procedures described by Ji et al.⁵³ in detail, with minor modifications. Briefly, right eyes of mice were covered with homemade, white translucent occluders, which were glued onto the fur around the eyes for 4 weeks, with plastic collars fitted around the neck to prevent the mice from removing their occluders. Age-matched naïve control mice were fitted with plastic collars but without occluders. A total of 341 mice were used in this work. The data presented in this paper were only from 299 mice, with the remaining 42 mice not being used due to frequent loss of the occluders (n = 19) or occurrence of cataracts/corneal opacity (n = 23). Mice wearing occluders were housed individually, and naïve control mice were housed in groups of five to six. All mice were reared in a 12-hour light/dark cycle. Illumination was provided by cool, white fluorescent bulbs, which produced an ambient illuminance of approximately 200 lux. 

Refraction Assessment

Mice were refracted in darkness with an automated eccentric infrared photorefractor developed by Schaeffel,⁵⁰ which was regularly calibrated with trial lenses before use. Supplementary Figure S1 shows the calibration curve demonstrating the linear regression relation (r² = 0.989, P < 0.0001) between measured values and the power of trial lenses. The unanesthetized animals were placed on a platform, which could be driven to move slowly, so that one eye was oriented in the direction of the refractor video camera. The data were automatically recorded by the program designed by Schaeffel et al.⁵⁰ The refraction of each eye was obtained by averaging collected three sets of the measurement. To minimize nonspecific effects, mice with an initial interocular refractive difference greater than 3 diopters (D) were not used in the subsequent experiments. 

HPLC Analysis

Because mouse retinal DA contents are light dependent and under diurnal regulation,^54,55 all samples for HPLC analysis were harvested under strictly controlled illumination levels at zeitgeber time (ZT) 1 (1 hour into the light period), or under dim red light at ZT 13 (1 hour into the dark period). At both time points, mouse retinal DA levels are reported to reach their daily peaks.⁵⁴ Mice were euthanized by cervical dislocation. Retinas and vitreous bodies, harvested within 30 seconds under a dissecting microscope, were immediately frozen in liquid nitrogen. Each frozen sample was homogenized into 100 μL (for retina) or 15 μL (for vitreous body) of ice-cold 0.1 M perchloric acid containing 10 μM ascorbic acid, 0.1 mM EDTA disodium salt, and 0.02 μM 3,4-dihydroxybenzylamine. The homogenates were centrifuged at 20,800g for 10 minutes at 4°C, and the supernatants were stored at −70°C until assayed. An HPLC system with electrochemical detection (ECD) (UltiMate 3000 system; Thermo Fisher Scientific, Waltham, MA, USA) was used to measure the levels of DA and DOPAC. Retinal homogenates were thawed on ice and centrifuged for 10 minutes at 29,700g, and the supernatants were injected onto an Acclaim C18 column (2.2 μm, 2.1 × 100 mm; Thermo Fisher Scientific) at 38°C. Separations were performed at a flow rate of 0.2 mL/min using a mobile phase of phosphate buffer, containing (in mM) 0.05 EDTA, 1.7 orthosilicic acid (OSA), 90.0 Na₂HPO₄, 50.0 citric acid, and 5% acetonitrile. Dopamine and DOPAC were detected with an ECD, fitted with a 5041 detection cell set at +350 mV, and with the guard cell set at +360 mV. The data were collected and analyzed by Chromeleon chromatography workstation (Thermo Fisher Scientific). 

Although this HPLC system has worked very well with high reliability and sensitivity, as demonstrated by previous studies,^56,57 we further assessed its reliability for detecting modest changes in mouse retinal/vitreal DA and DOPAC contents. For this purpose, we performed a positive control experiment, testing whether it was sensitive enough for detecting 6-OHDA induced decline in retinal/vitreal DA and DOPAC levels. A total of 50 μg 6-OHDA in 0.8 μL of saline solution containing ascorbic acid (10 μM) was gently injected into the vitreous body twice in 2 successive days, using a Nanoject II microinjector (3-000-205/206; Drummond Scientific Company, Broomall, USA). Retinal DA and DOPAC levels, as well as vitreal DOPAC levels, assessed by the HPLC system, were significantly lower in 6-OHDA–treated eyes, as compared with those obtained in vehicle injected eyes (Supplementary Fig. S2), even though there are several reports showing no significant loss of TH⁺ amacrine cells in the mouse retina after this treatment.^58,59 

The researchers performing HPLC analysis were masked to the treatment group of the animals. All the other analyses involved in this work were carried out by researchers not masked to the treatment groups. 

Immunohistofluorescence

Tissue Preparation.

The retinas were prepared as previously described.⁶⁰ In brief, mice were deeply anesthetized with 25% ethyl carbamate (1 g/kg), and quickly enucleated. When anterior segments of the eyes were removed, the eyecups were immediately fixed in fresh 4% paraformaldehyde in 0.1 M phosphate buffer (PB, pH 7.4) for 20 minutes, and chilled sequentially in 10% (wt/vol), 20%, and 30% sucrose in 0.1 M PB at 4°C. The eyecups embedded in optimal cutting temperature compound (OCT; Miles, Elkhart, IN, USA) were frozen by liquid nitrogen, and then sectioned vertically at 14-μm thickness on a freezing microtome (Leica, Nussloch, Germany). The sections were collected on gelatin chromium–coated slides. For whole-mount experiments, the retinas were dissected from the pigment epithelium, and then attached, ganglion cell side up, to a piece of filter paper (Millipore, Billerica, MA, USA). Four radial cuts were made to flatten the retina. 

Retinal sections and whole-mount retinas were blocked in 0.1 M PBS (pH 7.4) containing 6% normal donkey serum, 1% normal bovine serum albumin, and 0.2% Triton X-100 for 6 hours at 4°C. Mouse anti-TH monoclonal antibody (1: 10000, 3 days, 4°C; Sigma, St. Louis, MO, USA) was used to label dopaminergic amacrine cells. Immunoreactivity was detected with donkey anti-mouse IgG tagged with Alexa Fluor 555 (1: 200, 2 hours, room temperature; Invitrogen, Carlsbad, CA, USA). 

Counting of TH⁺ Cell Bodies.

The number of TH⁺ cell bodies was counted using a fluorescence microscope (Axioskop 40; Carl Zeiss, Inc., Oberkochen, Germany) under a ×20 objective. Eight distinct microscopic fields (520 × 520 μm) of flat-mounted retinas were chosen for counting: two zones in each eye quadrant located at 0.6 (corresponding to central retina) and 1.5 mm (corresponding to peripheral retina) from the optic nerve head. The number of TH⁺ cells in each regional zone was counted, and pooled to calculate the mean cell density of central and peripheral retina. The retinal area was delineated using a computer-based Neurolucida system (Microbrightfield, Williston, VT, USA) with a ×10 objective on an Olympus microscope (Olympus Corporation, Tokyo, Japan), and the data for each retinal area could be obtained from the software directly. The total number of TH⁺ cells was calculated by multiplying the mean density for all eight counting fields by the area examined. 

Quantification of TH⁺ Areas and Grey Level Analysis of TH⁺ Processes.

Images of TH-stained retinal sections were taken by a Leica SP2 confocal laser scanning microscope (Leica, Mannheim, Germany) at ×40 magnification. The measured zones, which were within the inner plexiform layer (IPL), were 250 × 25 μm, containing 500 × 50 pixels, located at 0.4–0.8 mm (central retina) and 1.2–1.6 mm (peripheral retina) from the optic nerve head. To analyze the areas occupied by TH⁺ processes (TH⁺ areas), within the chosen zones, the number of pixels above a fixed fluorescence brightness threshold (6× SD of the background grey level measured in the outer nuclear layer [ONL] in each section, where no TH-specific immunoreactivity was seen) was calculated. To assess the fluorescence intensities of TH⁺ processes quantitatively, grey levels of the photomicrographs obtained from retinal sections of different experimental groups were measured. The nonspecific background was subtracted to provide a real value of TH immunoreactivity. Data obtained in six cross sections were averaged for one sample, and six samples were collected for each differently manipulated group. 

RT-PCR Analysis

The procedures of RNA extraction and reverse transcription, and PCR conditions have been described previously in detail.⁶¹ The total RNA was extracted from retinas using an RNeasy mini kit (Qiagen, Victoria, Australia). One microgram of total RNA was reverse transcribed to cDNA. The following primer sets were used: for TH: 5′-TACGCCACGCTGAAGGGCCTCTAT-3′ (forward) and 5′-AGGTGA GGAGGCATGACGGATGTA-3′ (reverse); for β-actin: 5′-GAGAGGGAAATCGTGCGTGAC-3′ (forward) and 5′-CATCTGCTGGAAGGTGGACA-3′ (reverse). The sizes of the expected products are 255 bp for TH and 455 bp for β-actin. 

Western Blot Analysis

Western blot analysis was performed as previously described.⁶⁰ The protein extractions of mouse retinas were loaded, subjected to 10% SDS-PAGE, and transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked for 2 hours at room temperature in blocking buffer, consisting of 20 mM Tris-HCl (pH 7.4), 137 mM NaCl, 0.1% Tween-20, and 5% non-fat milk. Following incubation in a buffer containing the antibody against TH (1:10000) or rat anti-DAT (1:2000; Chemicon, Temecula, CA, USA) overnight at 4°C, the membranes were treated by horseradish peroxidase-conjugated donkey anti-mouse IgG and donkey anti-rat IgG (1:5000; Jackson Immunoresearch Laboratories, West Grove, PA, USA). The blots were probed with actin (1:45000; Sigma, St. Louis, MO, USA) and glyceraldehyde-3-phosphate dehydrogenase (GAPDH; 1:100000; Kangchen, Shanghai, China) antibody as loading controls for TH and DAT, respectively. They were finally visualized with enhanced chemiluminescence (Amersham Biosciences, Piscataway, NJ, USA). 

Statistics

All data are presented as mean ± SEM. Unless otherwise stated, statistical significance was determined by one-way ANOVA. Values of P less than 0.05 were taken to be significant. 

Results

Myopic Shift of Refractive Error in Form-Deprived Mice

A significant myopic shift in refractive error was induced in mouse eyes, which were form-deprived for 4 weeks by monocular occluder wear. The Table shows the refractive error values in control and form-deprived mice, measured by eccentric infrared photoretinoscopy. The mean interocular difference in refraction in form-deprived mice (n = 30) was −5.099 ± 0.239 D (1.341 ± 0.298 D for deprived eyes versus 6.440 ± 0.292 D for fellow eyes; P < 0.001, paired t-test). The form-deprivation–induced myopic shift of refractive error was comparable to those reported previously in mice^50,51,53 (see Discussion). As expected, the mean interocular difference in refraction was very small (−0.294 ± 0.248 D) in age-matched, normal control animals (5.796 ± 0.392 D for right eyes versus 6.090 ± 0.506 D for left eyes; P > 0.05, paired t-test; n = 20). 

Table

 

Refractive Error Values in Control and Treated Mice Following 4 Weeks of Form-Deprivation (Mean ± SEM)

Refraction (D)
Group	n	Right	Left	Difference	P value
Control	20	5.796 ± 0.392	6.090 ± 0.506	−0.294 ± 0.248	0.214
FDM	30	1.341 ± 0.298	6.440 ± 0.292	−5.099 ± 0.239	<0.001

 Right eye was form-deprived and the left eye was contralateral fellow eye without deprivation.

Unchanged DA and DOPAC Levels

We first explored whether daytime retinal DA levels in eyes form-deprived for 4 weeks were reduced in C57BL/6 mice, just as it occurred in chicks and monkeys.^24,27 Figure 1A shows a comparison of retinal DA levels in normal and form-deprived mice, assessed during the daytime by HPLC assay. Surprisingly, the average DA levels determined in deprived eyes (0.230 ± 0.008 ng/mg) were very close to those in both nondeprived fellow eyes (0.234 ± 0.008 ng/mg) and in normal eyes (i.e., the right eyes of the naïve control mice, 0.260 ± 0.011 ng/mg; P = 0.670). As shown in Figure 1B, the retinal levels of the principal DA metabolite DOPAC in deprived eyes (0.049 ± 0.004 ng/mg) were also very close to those obtained in fellow eyes (0.045 ± 0.003 ng/mg) and in normal eyes (0.046 ± 0.003 ng/mg; P = 0.288). Furthermore, the ratio of DOPAC to DA, which reflects the level of DA turnover,^22,62,63 was rather comparable (P = 0.172) among deprived (0.218 ± 0.017), fellow (0.204 ± 0.016), and normal eyes (0.179 ± 0.008) (Fig. 1C). Since DA release is a direct determinant of the activation strength of DA receptors, thus being a more important parameter for evaluating the intactness of DA-mediated signaling than DA pools, we also measured vitreal levels of DOPAC, a robust index of retinal DA release.⁶⁴ Again, this value was quite similar among deprived (0.057 ± 0.005 ng/μL), fellow (0.061 ± 0.005 ng/μL), and normal (0.070 ± 0.005 ng/μL) eyes (P = 0.383, Fig. 1D). 

Figure 1

 

Daytime levels of DA and DOPAC are not altered in C57BL/6 mouse eyes with 4-week form-deprivation. High-performance liquid chromatography analysis of samples harvested at ZT 1 revealed that the retinal levels of DA (A), DOPAC (B), and the calculated DOPAC/DA ratios (C) were not significantly different among deprived eyes (n = 30), fellow eyes (n = 30), and normal control eyes (n = 20). The concentrations of vitreal DOPAC, an indicator of retinal DA release, were also similar among deprived (n = 21), fellow (n = 21), and normal control eyes (n = 10; [D]). Error bars represent 1 SEM.

To exclude the possibility that form-deprivation induced changes in retinal DA levels might be diurnally related, we also made HPLC analysis of samples harvested at ZT 13, and found that retinal DA levels obtained in deprived eyes (0.295 ± 0.029 ng/mg) were rather close to those in fellow (0.291 ± 0.025 ng/mg) and normal eyes (0.310 ± 0.064 ng/mg; P = 0.937, Fig. 2A). While the levels of retinal DOPAC (0.022 ± 0.004 ng/mg for deprived eyes, 0.030 ± 0.006 ng/mg for fellow eyes, 0.028 ± 0.006 ng/mg for normal eyes, P = 0.537, Fig. 2B) and DOPAC/DA ratio (0.071 ± 0.009 for deprived eyes, 0.100 ± 0.015 for fellow eyes, 0.089 ± 0.009 for normal eyes, P = 0.187, Fig. 2C), determined at ZT 13, were both lower than those at ZT 1, implying diurnal variations, they were comparable among eyes with the three distinct treatments. Furthermore, nighttime retinal DA release also seemed to be unaltered by form-deprivation, because no changes in vitreal DOPAC levels were found in deprived eyes (0.024 ± 0.002 ng/μL vs. 0.027 ± 0.003 ng/μL in fellow eyes and 0.022 ± 0.001 ng/μL in normal eyes, P = 0.232, Fig. 2D). 

Figure 2

 

Nighttime levels of DA and DOPAC are not altered in C57BL/6 mouse eyes with 4-week form-deprivation. Bar charts summarizing the results of retinal DA levels (A), DOPAC levels (B), and the calculated DOPAC/DA ratios (C) yielded by HPLC analysis for samples harvested at ZT 13. No significant difference in these levels were found among deprived (n = 12), fellow (n = 12), and normal control eyes (n = 6). The concentrations of vitreal DOPAC, an indicator of retinal DA release, were also similar among deprived (n = 13), fellow (n = 13), and normal control eyes (n = 10 [D]). Error bars represent 1 SEM.

In all the experiments with HPLC analysis so far described, the samples were harvested from mice, which were monocularly deprived for 4 weeks and underwent dramatic myopic shifts in eye refraction. To rule out the possibility that temporary changes in retinal DA levels might occur during the development of FDM, we did similar analysis in another group of C57BL/6 mice at ZT 1, which were form-deprived only for 2 weeks. As shown in Figures 3A and B, no changes in retinal DA (deprived: 0.278 ± 0.021 ng/mg, fellow: 0.267 ± 0.019 ng/mg, normal: 0.315 ± 0.037 ng/mg, P = 0.426) and DOPAC (deprived: 0.059 ± 0.005 ng/mg, fellow: 0.053 ± 0.005 ng/mg, normal: 0.046 ± 0.007 ng/mg, P = 0.341) were induced by form-deprivation in these mice. Finally, to test whether form-deprivation would cause any extremely transient effects immediately after occluder wearing, we repeated the measurements in a third group of animals form-deprived only for 2 days. Again, neither retinal DA (deprived: 0.581 ± 0.047 ng/mg, fellow: 0.621 ± 0.080 ng/mg, normal: 0.840 ± 0.0.133 ng/mg, P = 0.113) nor DOPAC (deprived: 0.070 ± 0.007 ng/mg, fellow: 0.084 ± 0.007 ng/mg, normal: 0.096 ± 0.019 ng/mg, P = 0.205) levels were found to be significantly influenced by form-deprivation (Figs. 3C, D). 

Figure 3

 

Retinal levels of DA and DOPAC are not altered by 2-week or 2-day form-deprivation in C57BL/6 mice. High-performance liquid chromatography analysis revealed that neither retinal levels of DA (A), nor levels of DOPAC (B) were significantly different among deprived (n = 14), fellow (n = 14), and normal control eyes (n = 6) after 2 weeks of form-deprivation. Similar results were obtained in animals deprived for as short as 2 days (C, D; deprived, n = 10; fellow, n = 10; normal, n = 6). All samples were collected at ZT 1. Error bars represent 1 SEM.

TH⁺ Cell Number and TH⁺ Process Area

We further carried out a series of quantitative analyses to evaluate the intactness of retinal DA synthesis/release system in FDM mice. We first examined whether form-deprivation could influence the number of dopaminergic amacrine cells, the sole neuronal population that synthesizes DA in mouse retina, by counting the number of TH⁺ cells in whole-mount retinal preparations. Figures 4A1–3 shows the micrographs, focusing on the plane of the inner part of the inner nuclear layer (INL), in which the somata of dopaminergic amacrine cells are located, taken from deprived (A1), fellow (A2), and normal eyes (A3). Figure 4B is a schematic drawing of a whole-mount retina, illustrating the eight regions in which the number of dopaminergic amacrine cells was counted. As previously reported,⁶⁵ the distribution of dopaminergic amacrine cells in the whole-mount retina was rather uniform, as indicated by similar densities of dopaminergic amacrine cells in central and peripheral regions (35.66 ± 3.12/mm² vs. 39.62 ± 3.34/mm² for fellow eyes; P = 0.403, paired t-test). Therefore, the data obtained in central and peripheral regions were pooled for each group to calculate the average density for the whole retina. No significant difference in average densities was found among the three groups (39.228 ± 2.498/mm² for deprived eyes, 37.643 ± 2.801 /mm² for fellow eyes, and 34.671 ± 1.091/mm² for normal eyes; P = 0.370; Fig. 4C). 

Figure 4

 

Form-deprivation does not affect dopaminergic amacrine cell density in C57BL/6 mice. (A1–A3) Representative photomicrographs of TH⁺ neurons in retinal whole-mounts of the deprived, fellow, and normal eyes. (B) Schematic diagram of a retina, illustrating the 520 × 520–μm regions where the photomicrographs shown in (A) were acquired. The dashed circle indicates the region of the central retina (radius, 1000 μm). (C) The average density of TH⁺ cells was calculated for the three eye groups respectively, based on the cell counting from the eight regions shown in (B). There was no significant difference in TH⁺ neuron density among the three groups. (D) The retinal areas in the three experimental groups exhibited no significant difference. n = 7 retinas for each group. Error bars represent 1 SEM. D, dorsal; V, ventral; N, nasal; T, temporal.

To test whether FDM may cause changes in retinal size, the area of each retina was measured in deprived, fellow, and normal eyes. The mean retinal areas for these three groups (16.093 ± 0.004 mm² for deprived eyes, 17.183 ± 0.009 mm² for fellow eyes, and 16.944 ± 0.003 mm² for normal eyes) were quite similar (Fig. 4D, P = 0.408). Total numbers of TH⁺ neurons, derived from the cell densities and the retinal areas, were all approximately 600 per retina (627.98 ± 35.76 for deprived eyes, 635.77 ± 31.45 for fellow eyes, and 587.09 ± 19.64 for normal eyes), and no significant difference in these values was detected (P = 0.476). These values are also in good agreement with previous observations in intact mice.^65–69 

Myopia was shown to be associated with abnormal dendritic ramification of dopaminergic cells in rats with oxygen-induced retinopathy.⁷⁰ We tested whether this also holds true for C57 mice, by quantifying the areas occupied by TH⁺ processes in the outermost stratum of the IPL in TH-immunostained retinal cross sections.⁶⁹ Figures 5A and 5B show the confocal photomicrographs of a TH-immunostained C57BL/6 mouse retinal section, obtained with low, nonsaturating intensity (Fig. 5A) and with a bright threshold application (Fig. 5B). Average pixels over the thresholds, obtained in deprived, fellow, and normal eyes, are shown in Figure 5C. No significant difference in these values was found in either central (P = 0.929) or peripheral region (P = 0.875). 

Figure 5

 

Form-deprivation does not affect the area occupied by TH⁺ processes in C57BL/6 mice. The photomicrograph in (A) shows TH⁺ processes spread in the IPL of a retinal cross-section acquired using a confocal microscope with low, nonsaturating intensity; the photomicrograph in (B) shows the same retinal cross-section as in (A) but with a brightness threshold application. (C) Quantitative analysis of the areas occupied by TH⁺ processes shows no significant difference among the three eye groups, either in central or in peripheral retinal regions. n = 6 retinas for each group. Error bars represent 1 SEM.

TH Protein/Transcript Expression Level

Tyrosine hydroxylase catalyzes the conversion of L-tyrosine to L-3,4-dihydroxyphenylalanine (L-DOPA), which is the rate-limiting step in the biosynthesis of DA. Expression of TH therefore reflects the capability of DA synthesis. There is evidence for a marked reduction of retinal TH activity in myopic eyes.^24,27 We quantified the TH immunofluorescence signals in the IPL in the retinal cross-sections by determining the grey levels. Average grey levels obtained in these three cases were similar (P = 0.748 for central retina and P = 0.847 for peripheral retina, Fig. 6). 

Figure 6

 

Form-deprivation does not affect TH immunofluorescence intensity in C57BL/6 mice. (A1–A3) Representative photomicrographs showing TH⁺ processes spread in the IPL of retinal cross-sections from deprived, fellow, and normal eyes. (B) The mean fluorescence intensity of TH⁺ processes in the IPL, quantified by relative grey scale level, shows no significant difference among the three eye groups, either in central or in peripheral retinal regions. All values are normalized against normal eyes. n = 6 retinas for each group. Error bars represent 1 SEM.

Western blot analysis was performed to assess retinal TH protein expression levels after form-deprivation more precisely. A band of approximately 60 kDa, corresponding to the molecular weights of TH,⁷¹ was detected in retinal protein extracts (Fig. 7A). Densitometric analysis did not reveal significant difference in TH protein expression levels among the three experimental groups (P = 0.078; Fig. 7A). 

Figure 7

 

Tyrosine hydroxylase expression is not changed by form-deprivation at either protein or mRNA level in C57BL/6 mice. (A) Western blot analysis. Upper panel, representative example of TH antibody staining on a Western blot of retinal protein extracts from deprived, fellow, and normal eyes. Lower panel, TH expression, normalized as a ratio to β-actin levels, shows no significant difference among the three eye groups. n = 5 for each group. (B) Reverse transcription–PCR analysis of TH-specific mRNA transcript in total retinal RNA isolated from deprived, fellow, and normal eyes, with β-actin used as loading controls (upper). The relative band intensity ratio in relation to β-actin shows no significant difference among the three eye groups (lower). n = 3 for each group. Error bars represent 1 SEM.

Using the primers designed from the coding region of TH gene, RT-PCR analysis was conducted to study effects of form-deprivation on TH expression at transcription level and revealed a 255-bp product, corresponding to the transcripts of TH in the mouse striatum (Fig. 7B).⁷² Quantification of TH mRNA disclosed no significant difference among the samples isolated from deprived, fellow, and normal control eyes (P = 0.575; Fig. 7B). 

Unchanged DAT Expression

Extracellular DA concentrations have been shown to regulate visually-guided eye growth (refractive development) in a more direct way.^35,47 Unchanged total retinal DA levels in myopic eyes may not necessarily reflect unchanged extracellular DA levels. Because DA is taken up from the interstitial space by DA transporters (DATs), the activity of these transporters can be used as an index of extracellular DA levels.⁷³ In chicks with LIM and FDM, elevated retinal levels of DAT labeled by radioactive assay were reported,⁷³ while in form-deprived eyes of guinea pigs, less retinal DAT-labeling was revealed by microimaging and autoradiography.⁷⁴ In this study, DAT expression in the mouse retina was evaluated by Western blotting in form-deprived eyes. As shown in Figure 8, a band of approximately 70 to 80 kDa was detected, which is in line with the molecular weights of DAT in rat and bullfrog retinas.⁷⁵ Densitometric analysis of the blots indicated that there was no significant difference in retinal DAT expression levels among deprived, fellow, and normal eyes (P = 0.762). 

Figure 8

 

Expression levels of retinal DAT are not altered by form-deprivation in C57BL/6 mice. Upper panel, representative example of DAT antibody staining on a Western blot of retinal protein extracts from deprived, fellow, and normal eyes. Lower panel, DAT expression, normalized as a ratio to GAPDH levels, shows no significant difference among the three different conditions. n = 5 for each group. Error bars represent 1 SEM.

Discussion

In this work, we succeeded in inducing myopic shift in refraction in C57BL/6 mice by form-deprivation. Even though the absolute refractive error values after 4 weeks of form-deprivation (1.341 ± 0.298 D) were somewhat different from those previously obtained in FDM mouse (C57BL/6) models (5.14 ± 1.67 D for Ref 53, 4.7 ± 2.4 D for Ref 50, 7.19 ± 2.86 D and 11.87 ± 3.97 D for Ref. 51), the interocular difference in refractive errors (−5.099 ± 0.239 D) caused by form-deprivation was quite comparable (−5.27 ± 1.28 D for Ref. 53, approximately −3.9 D for Ref. 50, −2.09 ± 1.65 D and −6.33 ± 3.89 D for Ref. 51). In the same mouse model, other biometry data of the deprived eyes, including the axial length, was shown to be also significantly changed after experimental myopia development.⁵³ 

A most significant finding in this work is that the retinal levels of DA and DOPAC in this myopic mouse model were unaltered. This result is different from those obtained in experimental myopia models of other species.^24–28 We must emphasize that special attention has been paid to the methodology for determining the retinal levels of DA and DOPAC. First, all samples for HPLC analysis were harvested under strictly controlled illumination levels to exclude any effects caused by ambient light. Secondly, we determined retinal DA and DOPAC levels at ZT 1 and ZT 13, at which retinal DA levels in C57BL/6 mice reach the “day” and “night” peaks respectively.⁵⁴ Even though retinal DOPAC and vitreal DOPAC levels did show diurnal changes (Figs. 1, 2), no form-deprivation induced changes strongly suggest it very unlikely that changes might occur at other time points. 

Retinal levels of DA and DOPAC have been determined by HPLC in normal C57BL/6 mice. In a study concerning circadian rhythms of DA in mouse retina, Doyle et al.⁵⁴ reported that retinal DA and DOPAC concentrations in C57BL/6 mice fluctuate within the range of 1000 to 1200 pg/retina and 100 to 200 pg/retina, respectively across different times of the day. In another study, these values measured at ZT 3 to 4 are 545.33 ± 14.50 pg/retina (DA) and 187.3 ± 10.67 pg/retina (DOPAC).⁷⁶ The data reported in this work for normal eyes (daytime DA: 0.260 ± 0.011 ng/mg, which equals to 771 ± 21.5 pg/retina; nighttime DA: 0.310 ± 0.064 ng/mg, which equals to 700 ± 12.5 pg/retina; daytime DOPAC: 0.046 ± 0.003 ng/mg, which equals to 137 ± 8 pg/retina; nighttime DOPAC: 0.028 ng/mg, which equals to 61.9 ± 11.7 pg/retina) are quite close to these data, with a similar order of variations. No data concerning mouse vitreal DOPAC levels are available, but our results (daytime: 0.070 ± 0.005 ng/μL, which equals to 455.407 ± 50.704 nM; nighttime: 0.022 ± 0.001 ng/μL, which equals to 128.217 ± 7.230 nM) were comparable with those obtained in chicks (approximately 0.019 ± 0.006 ng/μl⁷⁷) and frogs (approximately 311 ± 29 nM⁷⁸). Furthermore, the expression of DAT in the retina was also unchanged, suggesting no changes in extracellular retinal DA concentrations in myopic eyes. There are multiple lines of evidence in favor of the assumption that DAT expression level responds to the level of extracellular DA. Dopamine transporter is known to be the major mechanism for removal of extracellular DA in mammalian brain^79,80 and genetic deletion of DAT leads to 100 times longer dopamine persistence in the extracellular space.^79,81 More importantly, retinal DAT expression is altered in several FDM models that show reduced retinal DA levels.^73,74 All these results suggest that the development of FDM is not related to retinal DA levels in this mouse strain. The notion that retinal levels of DA and DOPAC are unaltered is further strengthened by the intactness of the retinal DA synthesis system in this myopic model, as evidenced by no significant changes in TH expression at protein or mRNA levels, densities of TH⁺ amacrine cells, and areas occupied by processes of these cells. In this context we have noticed that in mice lacking the nyctalopin gene (nob mice), thus leading to a loss of the ON functional pathway and showing markedly lower levels of retinal DA and DOPAC, no significant myopic shift was found in normal visual environment (without form deprivation), even though these mice showed an increased susceptibility to FDM.⁷⁶ Moreover, these mice are more hyperopic than the wild-type C57BL/6 mice.⁷⁶ If reduced levels of DA and DOPAC were directly related to the development of myopia, they should have been myopic in normal visual environment. In other words, these results do suggest that the retinal DA levels may not be a key factor leading to the development of myopia in the nob mouse. Needless to say, extreme caution should be taken for interpreting any data obtained in this strain, given its unexpected behavior and possible developmental reorganization caused by mutation.^82–84 It must also be emphasized that the unchanged retinal levels of DA and the intactness of the DA system in this mouse myopic model should not be considered to oppose to the hypothesis that DA plays an essential role in the development of myopia, which has been proposed based on results obtained in other species. Any conclusion drawn from this work should thereby be restricted to the C57BL/6 mouse at this stage and could be expanded to other species only when further experimental evidence is provided. The unchanged retinal DA levels and the intactness of the DA system also do not necessarily imply that DA plays no role in myopia development of this mouse strain. A speculation is that, instead of DA levels, DA receptor levels (the expression levels of DA receptors) may be somehow be associated with myopic development. Recently, there are some preliminary studies suggesting that genetic and pharmacological manipulation of dopamine D₂ receptors in mice could influence the eye refraction and the development of FDM.⁸⁵ More solid evidence must be provided for strengthening such possibility. 

C57BL/6 mice are genetically incapable of producing melatonin, because they have mutations in two genes encoding enzymes (serotonin–N-acetyltransferase and hydroxyindole–O-methyl-transferase) indispensible for melatonin synthesis, and shows no circadian changes in DA contents,⁵⁴ which is unique and is never seen in animal species with reduced retinal DA levels induced by form-deprivation. This difference raises an intriguing possibility that the changes in retinal DA levels may be a consequence of form-deprivation–caused effects on other systems. Among them, the melatonin system may be a good candidate. Increasing evidence indicates an interplay between DA and melatonin systems.^86–89 Actually, melatonin has been implicated in the development of myopia. Retinal expression of melatonin receptors at both protein and mRNA levels is altered in chicks and guinea pigs with myopia.^90–92 Pharmacological manipulation using melatonin or its analog could interfere with eye elongation in chicks.^36,92,93 If melatonin is not available in the retina, any effects caused by form-deprivation on the retinal melatonin system would not affect DA synthesis and release. This possibility remains to be further explored. 

The fact that myopic shift in refraction could be induced by form-deprivation without changes in retinal DA levels further suggest the involvement of molecules other than DA, at least in C57 mouse. There is plenty of evidence that several other neurotransmitters, such as ACh,^94,95 NO,⁹⁶ retinoic acid,^97–99 may function as key molecules in the signaling pathway regulating myopic eye growth, though most of them are thought of as an indirect effect by interacting with the dopaminergic pathway.²³ 

Note added in revision: While this manuscript was in revision since it was first submitted in September 2013, two other papers,^100,101 also reporting unaltered retinal DA levels in form-deprived mice, have been submitted (August 2014 for Ref. 100, April 2014 for Ref. 101) and published. 

Acknowledgments

The authors thank Xiang-Tian Zhou and Li-Qin Jiang (Wenzhou Medical University, Wenzhou, Zhejiang, China) for helping the establishment of the mouse myopia model, and Dao-Qi Zhang (Oakland University, Rochester, MI, USA) for insightful comments on the manuscript. 

Supported by the Ministry of Science and Technology of China (2011CB504602; Beijing, China) and the National Natural Science Foundation of China (30930034, 31070967, 31171055, 31121061, 31100796, 81430007; Beijing, China). 

Disclosure: X.-H. Wu, None; Y.-Y. Li, None; P.-P. Zhang, None; K.-W. Qian, None; J.-H. Ding, None; G. Hu, None; S.-J. Weng, None; X.-L. Yang, None; Y.-M. Zhong, None