Changes in Intrinsically Photosensitive Retinal Ganglion Cells, Dopaminergic Amacrine Cells, and Their Connectivity in the Retinas of Lid Suture Myopia

Abstract

Purpose: This study investigates alterations in intrinsically photosensitive retinal ganglion cells (ipRGCs) and dopaminergic amacrine cells (DACs) in lid suture myopia (LSM) rats.

Methods: LSM was induced in rats by suturing the right eyes for 4 weeks. Double immunofluorescence staining of ipRGCs and DACs in whole-mount retinas was performed to analyze changes in the density and morphology of control, LSM, and fellow eyes. Real-time quantitative PCR and Western blotting were used to detect related genes and protein expression levels.

Results: Significant myopia was induced in the lid-sutured eye, but the fellow eye was not different to control. Decreased ipRGC density with paradoxically increased overall melanopsin expression and enlarged dendritic beads was observed in both the LSM and fellow eyes of the LSM rat retinas. In contrast, DAC changes occurred only in the LSM eyes, with reduced DAC density and tyrosine hydroxylase (TH) expression, sparser dendritic processes, and fewer varicosities. Interestingly, contacts between ipRGCs and DACs in the inner plexiform layer (IPL) and the expression of pituitary adenylate cyclase-activating polypeptide (PACAP) and vesicular monoamine transporter protein 2 (VMAT2) mRNA were decreased in the LSM eyes.

Conclusions: The ipRGCs and DACs in LSM rat retinas undergo multiple alterations in density, morphology, and related molecule expressions. However, the ipRGC changes alone appear not to be required for the development of myopia, given that myopia is only induced in the lid-sutured eye, and they are unlikely alone to drive the DAC changes. Reduced contacts between ipRGCs and DACs in the LSM eyes may be the structural foundation for the impaired signaling between them. PACAP and VMAT2, strongly associated with ipRGCs and DACs, may play important roles in LSM through complex mechanisms.

Myopia is one of the most common eye diseases worldwide, with a substantial prevalence among children and adolescents. Myopia not only affects visual perception, but also significantly raises the risk of many eye diseases, such as macular degeneration, retinal detachment, and glaucoma, etc.^1,2 A great deal of research has been devoted to exploring the etiology of myopia; however, the complicated mechanisms have not yet been fully elucidated. Recent studies have demonstrated the significance of intrinsically photosensitive retinal ganglion cells (ipRGCs) and dopaminergic amacrine cells (DACs) in the initiation and progression of myopia.^3–6 

The ipRGCs, also known as melanopsin-expressing ganglion cells (mRGCs), are a group of retinal photoreceptors that specifically express the photopigment melanopsin in the retina.^7,8 The ipRGCs receive input from the rods and cones and are also activated by a phototransduction cascade triggered by melanopsin.⁹ They mainly project to non-image forming visual brain regions, including the suprachiasmatic nucleus (SCN) and the olivary pretectal nucleus (OPN), to regulate circadian photoentrainment and pupillary light reflexes.^10–12 A small portion of the signals are transmitted to the image-forming visual regions, such as the dorsal lateral geniculate nucleus (dLGN) and the superior colliculus (SC), to participate in imaging vision.^13–15 In addition, ipRGCs are shown to provide a retrograde visual signaling pathway to interneurons in the outer retina.^16,17 Six ipRGC subtypes (M1–M6) have been identified in mice, each with unique morphological and physiological properties.^18–20 Recent studies have shown that ipRGCs play an important role in ocular refractive development and contribute to myopia.^3,4 

Dopamine (DA) is an important neurotransmitter in the retina, mediating retinal development, visual signaling, and refractive development.²¹ It is proven to be a “stop” signal for myopia,^22–26 although the exact mechanism remains unclear. DACs, which synthesize and release DA in the retina, are important inhibitory interneurons that affect nearly all retinal cells and visual pathways.^21,27 Previous studies have shown that the processes of ipRGCs and DACs form direct contacts in the inner plexiform layer (IPL) of the retina, as demonstrated by anti-melanopsin and anti-tyrosine hydroxylase (TH) double immunofluorescence staining and electron microscopy,^9,28–31 affecting the cellular functions of each other, and working together in visual signaling. 

Pituitary adenylate cyclase activating polypeptide (PACAP) is a pleiotropic biological peptide with neurotrophic and neuroprotective effects, belonging to the vasoactive intestinal peptide (VIP)/secretin/glucagon superfamily.^32–34 PACAP is exclusively observed in ipRGCs in the retina and functions as a neurotransmitter in the retinohypothalamic tract (RHT) to regulate the circadian rhythm with glutamate.^35–38 Numerous studies have confirmed the important neuroprotective effect of PACAP in various retinal diseases, including diabetic retinopathy, ischemia-induced retinal degeneration, excitotoxic retinal injury, and optic nerve severance.^39–43 In addition, research has indicated that PACAP promotes TH activity and has a protective effect on DACs.^39,44–47 Vesicular monoamine transporter protein 2 (VMAT2), another important protein that may be associated with the pathogenesis of myopia, is located in the presynaptic terminals of the DACs and is responsible for transporting DA to synaptic vesicles.^48–51 Furthermore, Vugler et al.²⁹ and Østergaard et al.²⁸ investigated the localization of synaptic markers through triple immunofluorescence labeling and found that VMAT2 was observed in the appositions of ipRGCs and DACs processes in the IPL of the retina. Thus, it may be inferred that PACAP and VMAT2 are 2 proteins that strongly correlate with the functions of ipRGCs and DACs, leading to our hypothesis that they may be affected in myopia. 

In this study, we described the alterations of ipRGCs and DACs in cell density, morphology, and the contacts between them, as well as the expression levels of related genes in lid suture myopia (LSM) rats, and revealed the possible mechanisms involved. 

Methods

Animals

All animal experiments complied with the Association for Research in Vision and Ophthalmology (ARVO) Statement for the Use of Animals in Ophthalmic and Vision Research. For this experiment, 3-week-old male Wistar rats weighing approximately 85 g to 95 g were used. All rats were housed in a suitable environment with a room temperature of 18°C to 25°C, a relative humidity of 40% to 70%, and an illumination of a 12:12-hour light/dark cycle with free access to food and water. 

Induction of Lid Suture Myopia

Wistar rats were randomly divided into the control and LSM groups, and 36 rats were used for this experiment (n = 18 per group). The right eye of the LSM group was deprived of vision by suturing the eyelid closed for 4 weeks, and the left eye served as the fellow eye. The control group was left untreated, and the right eyes were used as controls (Figs. 1A, 1B). The axial length was measured before and after lid suture with an ophthalmic ultrasonographer (ODM-1000A; Maida, Tianjin, China). Five measurements were taken in each eye, and the mean values were calculated for data analysis. Rats with congenital myopia, corneal ulcers, foreign bodies on the eyelid, or loose stitches during lid suture were excluded. 

Figure 1.

 

(A) Timeline of the lid suture and data collection. (B) The pattern of the lid suture myopic rat model: the right eyes of the treatment group served as the LSM eyes, and the left eyes served as the fellow eyes; the right eyes of the control group were taken as the control eyes. (C) Microscopic imaging pattern: 20 images were captured in each whole-mount retina at a 10 times objective lens for the analysis of ipRGC and DAC density, respectively.

Retinal Immunofluorescence

Preparation of Whole-Mount Retinas

Six rats per group were used for immunofluorescence analysis. Both eyes were enucleated immediately after the rats were euthanized by over-anesthesia. Eyeballs were quickly placed into FAS eye fixative for 2 hours at room temperature and subsequently transferred to 4°C for 24 hours. The cornea and lens were removed under a microscope, and the retina was gently detached from the sclera, followed by four radial incisions to flatten the retina into a clover shape. 

Retinal Staining

Specific rabbit anti-melanopsin (ab19383; Abcam, Cambridge, UK) and sheep anti-tyrosine hydroxylase (AB1542; Sigma-Aldrich, St. Louis, MO, USA) primary antibodies were used for double immunofluorescence staining of whole-mounted retinas to co-label ipRGCs and DACs. The retinas were transferred to 24-well plates and rinsed 3 times in 0.1 M PBS to remove residual fixative. After incubating in a blocking solution (10% donkey serum and 0.2% Triton X-100 dissolved in 0.1 M PBS) for 2 hours at room temperature, 250 µL of the primary antibody mixture (1:500) was added and maintained for 24 hours at 4°C, and washed 3 × 15 minutes in 0.1 M PBS. Subsequently, the retinas were incubated with the secondary antibody mixture 250 µL per well (donkey anti-rabbit IgG conjugated with Alexa Fluor 488, A-21206; Invitrogen, Waltham, MA, USA; and donkey anti-sheep IgG conjugated with Alexa Fluor 594, A-11016; Invitrogen, Waltham, MA, USA) for 2 hours at room temperature, protected from light, and washed again. The stained retinal whole mounts were then transferred to a microscope slide, gently flattened, with the ganglion cell layer facing upward, and sealed with glycerol to prevent fluorescence quenching. 

Microscopic Imaging, Cell Counting, and Morphological Analysis

Immunofluorescence images were obtained using a confocal microscope (Leica TCS SP8; Microsystems). Images for cell counting were taken with a 10 × lens following the pattern shown in Figure 1C, with the optic disc as the center. The retina was divided into the peripheral, intermediate, and central regions with a ⅓ spacing, and 20 images were taken separately for ipRGCs and DACs in each retina (1.35 mm² per image). The ipRGCs in the ganglion cell layer (GCL) and inner nuclear layer (INL) were included in the analysis using Z-Stack. Cell counting was performed manually using the ImageJ software (National Institutes of Health, Bethesda, MD, USA), and the densities (cells/mm²) were calculated separately for the peripheral, intermediate, and central regions, as well as for the entire retina. 

Morphology and the contacts between ipRGCs and DACs were analyzed by using images taken with 20 × and 63 × oil lenses. Two images were taken separately for each petal of the retina, for a total of eight images per retina. The diameter of the proximal ipRGC beads was measured using ImageJ software. Two representative ipRGCs were selected from each image, and the diameters of the first three beads of the primary dendrites of the ipRGCs were measured using the “straight” tool in ImageJ software. Finally, the average diameter was calculated for each retina. For the quantitative analysis of the complexity of the DAC plexus, six manually delineated straight lines perpendicular to each other were made on each image. Subsequently, the number of DAC dendritic plexuses intersecting with these straight lines was measured using Plot Profile, a plug-in of ImageJ software, and the average plexus intersection per millimeter in each retina was then calculated. Quantification of the contacts between the ipRGCs and DACs’ processes was conducted in terms of the number of co-localized spots per 100 µm DAC dendrite. 

All images were captured using the Z-Stack option of the confocal microscope. Optical sections of ipRGCs and DACs were collected consecutively at intervals of 0.5 to 1.0 µm to involve all cells in the field of view and then merged for analysis. 

Real-Time Quantitative PCR

Six rats per group were subjected to real-time quantitative PCR (RT-qPCR) analysis. Retinas were quickly stripped off and placed in a lysis solution after the execution of the rats. Total RNA was extracted and reverse-transcribed into cDNA using the commercially available kits EZB-RN001 (EZBioscience, Roseville, MN, USA) and R333 (Vazyme Biotech Co. Ltd., Nanjing, China). TB Green Premix Ex Taq II (RR820A; Takara Bio Inc., Shiga, Japan) was used for the RT-qPCR system, with primers synthesized by Sangon Biotech (sequences are shown in the Table). The relative expression levels of genes were calculated by the 2^−ΔΔCt method. 

Table.

 

Sequences of the Primers

 

Western Blotting

Six rats from each group were subjected to Western blotting to determine relative melanopsin expression. Retinas were collected as previously described, homogenized in RIPA buffer with protease inhibitors (R0020; Solarbio, Beijing, China), and protein concentrations were measured according to the instructions of the BCA Protein Assay Kit (P0011; Beyotime Inc., Nantong, China). Protein samples (80 µg each) were loaded into sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) gels for electrophoresis and subsequently transferred to polyvinylidene fluoride (PVDF) membranes. The membranes were blocked in 5% skimmed milk for 2 hours at room temperature, followed by incubation in a primary antibody solution (anti-melanopsin, Abcam, ab19383, 1:1000) overnight at 4°C, with GAPDH as a loading control. The membranes were then washed 3 times, incubated with the corresponding horseradish peroxidase-conjugated goat anti-rabbit secondary antibodies (Affinity, S0001, 1:4000) for 1 hour at room temperature, and washed again. Proteins were detected using BeyoECL Plus chemiluminescence reagent (P0018M; Beyotime Inc., Nantong, China) in an Automatic Chemiluminescence Image Analysis System (Tanon 5200, Shanghai, China), and the band intensities were quantified using ImageJ software. 

Statistical Analysis

All data are presented as mean ± standard deviation (SD). GraphPad Prism version 9.3.0 (GraphPad Software Inc., San Diego, CA, USA) was used for statistical analysis. The data were first tested for normal distribution, followed by a 1-way analysis of variance (ANOVA) and multiple comparisons among the 3 groups to clarify the differences among the groups. The significant difference was set at P value < 0.05. 

Results

Axial Lengthening in LSM Eyes

There was no significant difference in axial length among the three groups before the treatment (control eyes 5.48 ± 0.18 mm versus LSM eyes 5.44 ± 0.19 mm versus fellow eyes 5.55 ± 0.14 mm, P > 0.1, not significant [ns]). However, after 4 weeks of lid suture, LSM eyes had a significantly longer axial length than that of the control eyes and fellow eyes, with no notable changes between the control eyes and fellow eyes (LSM eyes 7.03 ± 0.22 mm versus control eyes 6.61 ± 0.08 mm and fellow eyes 6.62 ± 0.16 mm, P < 0.0001; control eyes versus fellow eyes, P > 0.5, ns; Fig. 2A). Axial elongation analysis confirmed these results: LSM eyes had significantly greater axial lengthening (LSM eyes 1.59 ± 0.18 mm versus control eyes 1.13 ± 0.18 mm and fellow eyes 1.07 ± 0.21 mm, P < 0.0001; control eyes versus fellow eyes, P > 0.5, ns; Fig. 2B). Therefore, we concluded that 4 weeks of lid suture in rats caused an increase in ocular axes and induced myopia compared with the control eyes and fellow eyes. 

Figure 2.

 

Lid suture (LS) induced excessive axial lengthening in the LSM eyes. (A) There was no significant difference in axial length before LS, but the axial length of the LSM eyes was notably longer than that of the control eyes and the fellow eyes after 4 weeks of LS. (B) The analysis of axial elongation also indicated the overgrowth of the LSM eyes. The data are presented as mean ± SD and were analyzed by 1-way ANOVA. Each point shown in the bar charts indicates an individual sample (n = 18 per group). ****P < 0.0001, ns P > 0.5.

Lid Suture Induced Multifaceted Changes in ipRGCs

Monocular Lid Suture Induced a Binocular Decrease in ipRGC Density

The overall mean density of ipRGC in the retinas of control eyes was 40.03 ± 2.53 cells/mm², with no significant difference between the peripheral (41.91 ± 3.88 cells/mm²) and intermediate zones (41.82 ± 2.81 cells/mm²; P > 0.5). However, the average density of ipRGC in the central region (32.69 ± 4.34 cells/mm²) was significantly less than that in the peripheral (P < 0.005) and intermediate zones (P < 0.005), which is consistent with previous studies.⁵² The overall mean density of ipRGC was significantly reduced in the LSM eyes (26.07 ± 3.59 cells/mm², P < 0.0001) and fellow eyes (23.98 ± 3.83 cells/mm², P < 0.0001), which were about 34.9% and 40.1% less than that of the control eyes, respectively. No significant difference was detected between the LSM eyes and the fellow eyes (P > 0.1; Figs. 3A–D). Additionally, ipRGC reduction was greater in the peripheral and intermediate regions (40.1%; 38.1%) than in the central region (27.7%). This result suggests that monocular lid suture in rats causes a reduction of ipRGC density in both eyes, along with a greater reduction in the peripheral and intermediate retinas. 

Figure 3.

 

Changes in ipRGC density and molecule expression. (A, B, C) Immunofluorescent staining of whole-mount retinas labeled with ipRGCs (green), ipRGCs were more abundant in the control eyes than in the LSM eyes and the fellow eyes. (D) Statistical analysis of ipRGC density: it was significantly decreased in the LSM eyes and the fellow eyes compared with the control eyes, with no marked difference between the LSM eyes and the fellow eyes. (E, F) Western blot of melanopsin. (G, H, I) Relative expression levels of Opn4 (E), PACAP (F), and PAC1R (G) mRNA among the three groups. Monocular lid suture induced increased expression of melanopsin protein and mRNA in the LSM eyes and the fellow eyes, but decreased expression of PACAP mRNA in the LSM eyes. PAC1R remained unchanged. Each point on the bar charts indicates an individual sample. (A, B, C) Scale bar = 100 µm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns P > 0.05.

Altered Expression of Melanopsin and PACAP in LSM Rats

Melanopsin and PACAP are exclusively observed in ipRGCs in the retina. Therefore, we examined the relative gene expressions of melanopsin (Opn4), PACAP (Adcyap1), and PAC1 receptor (Adcyap1r1) in control, LSM, and fellow eyes using RT-qPCR. A significant increase in Opn4 mRNA expression was observed in LSM and fellow eyes compared with the control eyes (P < 0.001), with no significant difference between the two eyes (P > 0.1; Fig. 3G). Western blotting was used to verify the melanopsin expression levels, and the results were consistent with those of RT-qPCR (Figs. 3E, 3F). It seems to be a paradox that the number of melanopsin-expressing ipRGCs decreased in both eyes, whereas the total melanopsin expression increased. The PACAP mRNA expressions were reduced only in LSM eyes (P < 0.001), whereas those in the fellow eyes remained unaltered (P > 0.1; Fig. 3H). PAC1R mRNA expression was not affected by lid suture (P > 0.5; Fig. 3I). 

Morphological Changes of ipRGCs in Both Eyes

We further observed the changes in the morphology of ipRGCs. At high magnification, ipRGC dendrites were found coated with beads, as indicated by the arrows in Figures 4A to 4C, which are beaded enlargements of ipRGC dendrites filled with mitochondria, and may be the sites where ipRGCs communicate with conspecifics or other cells such as bipolar cells and DACs.^9,29,53,54 Morphological parameters, such as soma size, number of primary dendrites, and number of dendritic branching points, did not reveal statistically significant differences among the three groups (Supplementary Material). However, the average diameter of the proximal ipRGC beads (the first 3 beads of the primary dendrites, as shown in Fig. 4D) in the LSM and fellow eyes was significantly larger than that in the control eyes (3.65 ± 0.71 µm and 3.79 ± 0.58 µm versus 2.93 ± 0.55 µm, P < 0.0001; Fig. 4E), with an increase of approximately 24.3% and 29.2%, respectively. The increase in melanopsin expression in these enlarged beads seems to partially explain why the ipRGC density decreased while the total melanopsin expression increased. 

Figure 4.

 

Morphological changes of ipRGCs. (A, B, C) Representative images of the ipRGC beads are indicated by the arrows. The beads in the LSM eyes and the fellow eyes were noticeably larger than those in the control eyes. (D) Schematic of the measurement of the proximal ipRGC beads: the first three beads of the primary dendrites of the ipRGCs were measured using ImageJ, and the average diameter was calculated for each retina. (E) Analysis of the diameters of the proximal ipRGC beads shows significantly enlarged diameters in the LSM eyes and the fellow eyes (n = 6 per group). (A, B, C) Scale bar = 20 µm. ***P < 0.001, ****P < 0.0001, ns P > 0.05.

Alterations in DACs During Lid Suture

Decreased DAC Density in the LSM Eyes

The density of DAC was calculated using the same method as for ipRGC. Results showed that the overall mean density of DAC in the LSM eyes (13.07 ± 0.82 cells/mm²) was significantly lower than that in the control eyes (20.21 ± 2.15 cells/mm²) and the fellow eyes (19.96 ± 1.82 cells/mm²) by about 35.3% and 32.7%, respectively (P < 0.001; Figs. 5A–D), indicating DAC is affected in the LSM eyes. Additionally, DAC densities in the peripheral, intermediate, and central regions of the control retinas were 20.62 ± 2.68 cells/mm², 21.22 ± 2.36 cells/mm², and 17.38 ± 0.26 cells/mm², respectively, with no significant difference among them, P > 0.05. There were also no marked differences in LSM retinas among these regions. 

Figure 5.

 

Alterations in the DAC density and gene expression. (A, B, C) Immunofluorescence staining of whole-mount retinas labeled with DACs (red). DACs were noticeably sparser in the LSM eyes than in the control eyes and the fellow eyes. (D) Statistical analysis of DAC density: it was significantly decreased in the LSM eyes compared with the control eyes and fellow eyes. (E, F) Relative expression levels of TH (E) and VMAT2 (F) mRNA among the three groups, the expression of TH and VMAT2 mRNA decreased in the LSM eyes. Each point on the bar charts indicates an individual sample (n = 6 per group). (A, B, C) Scale bar = 100 µm. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns P > 0.05.

Decreased Expression of TH and VMAT2 in the LSM Eyes

TH is the rate-limiting enzyme for dopamine synthesis in DACs, and VMAT2 acts as a transporter of dopamine to synaptic vesicles, implying their great significance for dopamine synthesis and release. Therefore, we examined the relative expression levels of TH and VMAT2 mRNA in three groups by RT-qPCR. The findings revealed that the expression of TH and VMAT2 mRNA in the LSM eyes was considerably lower than that in the control and fellow eyes (P < 0.001; Figs. 5E, 5F). 

Morphological Alterations of DACs in the LSM Eyes

Under high magnification, the proximal dendrites of DACs showed intense and homogeneous fluorescent staining, whereas the dendritic terminals extending into the IPL branched out into dense varicosities (Fig. 6A),^49,50 which served as sites for DA release and DAC attachment to other cells. Immunofluorescence staining of the whole-mount retina revealed that the dendritic terminals of DACs in the IPL became sparser, with significantly fewer varicosities in the LSM eyes than in the control and fellow eyes (Figs. 6A, 6B). This alteration was confirmed laterally by the decreased expression of VMAT2 mRNA in the LSM eyes, a presynaptic marker of dopamine release in the DACs. 

Figure 6.

 

Morphological change of the DACs. (A) The dendrites of DACs were extensively branched, forming a dense plexus and abundant varicosities in the control and fellow eyes, whereas the dopaminergic plexus became sparser with fewer varicosities in the LSM eyes. (B) Quantitative analysis of the complexity of the DAC plexus showed that the mean number of DAC dendritic plexuses intersecting with the manually delimited straight lines was sparser in the LSM eyes. Scale bar = 40 µm. ****P < 0.0001, ns P > 0.05.

Reduced Contacts Between ipRGCs and DACs’ Processes in the LSM Eyes

Contacts between the processes of the ipRGCs (green) and DACs (red) in the IPL were observed using a 63 × oil lens (arrows in Figs. 7A, 7B). We discovered that these contacts were markedly reduced in the LSM eyes compared with the control and fellow eyes, which may be attributed to the above-described alterations in the morphology of ipRGCs and DACs (sparser dopaminergic plexus and fewer varicosities in LSM eyes), and therefore fewer dopaminergic terminals reaching the ipRGC endings. The quantitative analysis of these contacts demonstrated a decrease of 18.2% and 17.7% in the LSM eyes compared with the control and fellow eyes, respectively, changing from 15.92 contacts per 100 µm in the control eyes and 15.81 contacts per 100 µm in the fellow eyes to 13.02 contacts per 100 µm in the LSM eyes (P < 0.0001; Figs. 7C–F). 

Figure 7.

 

Contacts between ipRGCs and DACs. (A) The appositions between DACs (red) and ipRGCs (green) noted by arrows indicate hypothesized contact sites. (B) Enlarged view of the box in A, arrows indicate contact point sites. (C) Quantitative analysis of the contacts: the number of contacts per 100 µm was counted for comparison. (D) Statistical analysis: the number of contacts in the LSM eyes was less than that in the control and fellow eyes. Each point shown in the bar charts indicates an individual sample (n = 6 per group). (A) Scale bar = 20 µm. (C) Scale bar = 10 µm. **P < 0.01, ns P > 0.05.

Discussion

Lid Suture Induces Multiple Changes in ipRGCs and DACs in the Retinas That May Be Related to the Pathogenesis of Myopia

The main findings of this study are that lid suture, a form of form-deprivation to induce myopia in many studies, induces various alterations in ipRGCs and DACs, including cell density, morphology, and specific gene expression. Monocular lid suture resulted in decreased ipRGC density, increased melanopsin expression, and enlarged proximal ipRGC beads in both the LSM eyes and fellow eyes. In contrast, the changes in DACs were observed only in LSM eyes, showing a significant decrease in DAC density, the expression of TH and VMAT2 mRNA, a sparser dendritic plexus, and fewer varicosities, whereas the fellow eyes remained unaltered. However, a limitation of this study is that the cell density was not examined in specific regions (dorsal/ventral/temporal/nasal) or layers (INL versus GCL), because ipRGCs and DACs are not uniformly distributed in the retina. This should be explored in future experiments, which may provide more information on the roles of ipRGCs and DACs in the pathogenesis of myopia. 

Analysis for the ipRGC Alterations in LSM

The mechanism underlying the decrease in ipRGC density in both the LSM eyes and fellow eyes is unclear, but this change is similar to the degeneration of ipRGCs observed in a variety of ocular and systemic diseases, such as glaucoma, diabetic retinopathy, retinitis pigmentosa, Parkinson’s disease, and Alzheimer’s disease.^53,55–57 The ipRGCs receive input from the rods and cones and are also activated by a phototransduction cascade triggered by melanopsin. In form-deprivation myopia (FDM) rats established by suturing the eyelids, form perception is completely eliminated, but light is also reduced to a certain extent. Diminished form and light perception interfere with rod and cone signaling and also affect melanopsin mRNA and protein levels. Studies have shown that the expression of melanopsin mRNA and protein was rhythmic in albino and colored rat retinas and prolonged exposure to darkness in albino rats induced an increase in melanopsin mRNA and protein expression, showing an increase in melanopsin staining intensity within the individual cells and their processes, whereas exposure to constant light decreased the melanopsin mRNA expression and the melanopsin immunostaining intensity.^58,59 Thus, the elevated melanopsin mRNA and protein levels observed in the study may be related to the attenuation of light induced by eyelid fusion, but whether it is form deprivation or light deprivation that caused this alteration cannot be clarified yet, because form is also deprived in dark environments. Regarding the paradox of reduced ipRGC density and elevated melanopsin expression, we speculate that it is related to ipRGC beads, which are beaded enlargements on ipRGC dendrites filled with mitochondria, and may be the sites where ipRGCs communicate with conspecifics or other cells, such as bipolar cells and DACs.^9,29,53,54 We hypothesized that the enlarged ipRGC beads and increased melanopsin expression observed in this study may be due to a compensatory response of the residual ipRGCs to enhance the expression of melanopsin and intercellular signaling to maintain the function as much as possible when partial ipRGCs are diminished. However, further studies are required to confirm this hypothesis. 

The involvement of ipRGCs and melanopsin in refractive development and progression of myopia has been demonstrated in many studies; however, the underlying mechanisms have not been consistently determined. Liu et al.³ revealed that selective ablation and activation of ipRGCs induced myopic and hyperopic refractive shifts in mice, and noted that melanopsin signals promote ocular axial elongation and are a major factor in ipRGC-mediated myopia progression. Chakraborty et al.⁴ found that melanopsin knockout mice (Opn4^−/−) and ipRGC-ablated mice (Opn4^DTA/DTA) showed a greater myopic shift in response to form deprivation, thus suggesting that disruption of retinal melanopsin signaling alters the rate and magnitude of normal refractive development, creates greater FDM susceptibility, and changes dopamine signaling. Other research noted that melanopsin expression in the experimental eyes of FDM and lens-induced myopia (LIM) guinea pigs significantly reduced compared with the contralateral eyes and recovered after removal of the interference.⁶⁰ It has also been claimed that vitreous injection of melanopsin antagonist caused longer eye axes and greater refractive myopia in LIM guinea pigs and proposed that melanopsin has an inhibitory effect on LIM formation.⁶¹ The results of the present study provide a new insight into the role of ipRGCs and melanopsin in the development of myopia. Elevated melanopsin expression in both the LSM and fellow eyes, but without myopic shift in the fellow eyes, suggests that the effects of ipRGCs and melanopsin in myopia are complex and may not be the direct cause of axial lengthening and myopia formation, there must be other factors working together. 

Another novel finding was that changes in ipRGC density, morphology, and melanopsin expression were binocularly consistent, a phenomenon identified for the first time and difficult to explain. These binocularly similar alterations may be related to the non-image forming functions of ipRGCs, including the control of circadian rhythms and pupillary light reflexes. The ipRGCs transmit environmental light information through the retinohypothalamic tract (RHT) to the suprachiasmatic nucleus (SCN),^62,63 the “master clock” of the circadian clock, which integrates the light information and then synchronizes it to the peripheral circadian clocks, thus, correcting and synchronizing the circadian clock of the whole body.⁶⁴ The circadian clock system is a central and peripheral interconnected oscillator and feedback loop; the binocular changes in ipRGCs may result from SCN modulation from the center to the periphery. In addition, the pupillary light reflex, a response in which light shines on one pupil and elicits bilateral pupil constriction, suggests that the signaling pathway of ipRGCs is mutually influenced by both eyes. 

Analysis for the DAC Alterations in LSM

Previous studies have shown that form deprivation causes a decrease in retinal dopamine levels. However, DACs, which synthesize and release dopamine, have not been studied. The decreased DAC density in the LSM eyes observed in this study may be closely related to the form and light deprivation caused by eyelid fusion. It has been shown that light affects DAC numbers and retinal dopamine levels, of which the rod pathway is a critical modulator, whereas ipRGCs do not drive light-induced dopamine release.^65,66 The significant decrease in DAC numbers and dopamine levels in mice reared in constant darkness, or lacking rod photoreception, seems to be consistent with the results of eyelid fusion in the present study,⁶⁵ but it is hard to conclude that the decrease in DAC number is absolutely due to light deprivation alone. Actually, the establishment of FDM animal models, including suturing the eyelids, wearing diffusers/goggle systems, and eye or face masks, does not completely separate form perception from light⁶⁷; more effective models need to be explored to achieve this goal. A decrease in TH and VMAT2 mRNA expression was also observed in LSM eyes in this study. TH is the rate-limiting enzyme for dopamine synthesis, and VMAT2 is responsible for transporting dopamine to the synaptic vesicles. Therefore, the decrease in dopamine in LSM eyes may be a result of the decrease in DAC numbers and TH and VMAT2, which cause a decline in the synthesis and release of dopamine. Moreover, we also observed alterations in DAC morphology in the LSM eyes, with significantly reduced plexus endings and varicosities, which may affect the dopamine-related conduction pathways and signaling with other cells, potentially being an important factor in myopia. The morphological alterations were laterally confirmed by the decreased expression of VMAT2 mRNA in the LSM eyes. 

Reduced Contacts Between ipRGCs and DACs in the LSM Eyes May Be a Potential Mechanism for the Development of LSM

A portion of ipRGCs forms direct contacts with DAC processes in the IPL.^28,29,68 In this study, we observed the morphology of ipRGCs and DACs, as well as their appositions in the IPL, and quantitatively analyzed the alteration of these appositions. The results showed that ipRGCs made abundant contacts with DACs in the control and fellow eyes, whereas the contacts were significantly reduced in LSM eyes. We hypothesized that this decrease may be due to the alterations of DACs, including significantly reduced plexus endings and varicosities, as well as the expression of VMAT2 mRNA in the LSM eyes, implying that the dopaminergic presynaptic terminals were significantly reduced, and therefore the appositions between ipRGCs and DACs were reduced in the LSM eyes. 

The ipRGCs and DACs are closely connected in structure and function. Although studies have confirmed that ipRGCs do not appear to play a role in DA synthesis and release, it has been revealed that the sustained light response of DACs is driven by melanopsin signaling from ipRGCs.^16,69 M1 ipRGCs provide retrograde excitatory signaling input to DACs,¹⁷ whereas DACs, in turn, inhibit M1 ipRGCs and regulate melanopsin expression.^54,69 A study suggests that DA binds to the D1 receptor on ipRGCs and attenuates the photocurrent of melanopsin,⁷⁰ whereas another study proposes that DA regulates the expression of melanopsin and PACAP mRNA through the D2 receptor.⁷¹ Thus, it is reasonable to believe that the contacts between ipRGCs and DACs in the IPL play an important role in cellular functions and visual signaling. Given the reduction of these important contacts in the LSM eyes, the functions of ipRGCs and DACs and the signaling between them should be disturbed, which may be a potential mechanism for the development of myopia. 

PACAP and VMAT2 May Play a Role in the Development of LSM in Rats

PACAP is a neurotrophic and neuroprotective peptide that exerts a wide range of protective effects in a variety of retinal diseases and has only been observed in ipRGCs in the retina.^32,34,35 PACAP is also important for TH activity and DAC development.^39,47 In addition, PACAP is thought to be a marker of the RHT, functioning with glutamate as the neurotransmitter of the ipRGCs to convey circadian photoentrainment to non-image forming regions of the brain.^38,72 It follows that PACAP exerts significant and complicated effects on retinal neurons and the circadian system. The present study revealed a significant decrease in PACAP mRNA expression in the LSM eyes for the first time, which may be responsible for the development of myopia through a weakened protective effect on retinal neurons, or by affecting the function of ipRGCs and DACs. However, the expression of PACAP is extremely complicated, and studies have shown that PACAP is positively regulated by dopamine,⁷¹ and has a complex self-regulatory mechanism to stimulate its own transcription.⁷³ Therefore, further research is needed to investigate the mechanisms involved. 

The role of VMAT2 in the retina and its changes in the FDM eyes have been elaborated on previously. Sun et al.⁷⁴ demonstrated that VMAT2 expression in the FDM eyes was significantly lower than that in the fellow eyes of guinea pigs using [18F]FP-(+)-DTBZ and Western blotting, and suggested that VMAT2 could be used as a new biomarker for myopia diagnosis. Landis et al.²⁶ found that the VMAT2^LO mice, with underexpression of VMAT2, were the most susceptible to FDM. The present study also noted a decrease in the expression of VMAT2 mRNA in LSM eyes, affecting the transportation of DA to synaptic vesicles and interfering with the communication between DACs and other cells. Therefore, VMAT2 may provide a new direction for studying the molecular mechanisms underlying myopia. 

In summary, it is concluded that lid suture induces multiple alterations in ipRGCs and DACs in rat retinas, which may be related to the pathogenesis of myopia. However, the observed changes in ipRGCs alone appear not to be required for the development of myopia, and the ipRGC changes alone are unlikely to drive the DAC changes. The effects of ipRGCs on myopia are complex, and may not simply promote or inhibit myopia, but rather affect signaling in the retinas through complicated interactions with other cells, in which the interactions with DACs may be one of the important mechanisms in the pathogenesis of myopia. Reduced contacts between ipRGCs and DACs in the LSM eyes may be the structural foundation for the impaired signaling between them. PACAP and VMAT2, two proteins strongly associated with ipRGCs and DACs, may play important roles in LSM via complex mechanisms, providing new directions for research on the etiology of myopia. 

Acknowledgments

The authors thank Peng Hao, Ailing Chen, and Jiafeng Wang for their help with high quality imaging. The authors would also like to thank the associate editor, the editorial board member, and the reviewers for their valuable and insightful comments and suggestions that improved this paper. 

Supported by the National Natural Science Foundation of China (81770956 and 81371049 to X.S., and 31872764 and 82171090 to Y.G.), Project of Tianjin 131 Innovative Talent Team (201936 to X.S. and Y.G.), the Science Fund for Distinguished Young Scholars of Tianjin (17JCJQJC46000 to X.S.), the Science and Technology Planning Project of Tianjin (No. 21JCYBJC00780 to X.S.), Jinmen Medical Talent Project of Tianjin, the Science and Technology Fund for Health of Tianjin (TJWJ2023ZD008 to X.S.), Shanghai Science and Technology Committee Rising-Star Program (19QA1401600 to Y.G.), Shanghai Municipal Science and Technology Major Project (No. 2018SHZDZX01), ZJLab, Shanghai Center for Brain Science and Brain-Inspired Technology, the Lingang Laboratory (LG-QS-202203-12), and Tianjin Key Medical Discipline (Specialty) Construction Project (No. TJYXZDXK‑016A). 

Disclosure: Y. Ling, None; Y. Wang, None; J. Ye, None; C. Luan, None; A. Bi, None; Y. Gu, None; X. Shi, None