Ocular-Component-Specific miRNA Expression in a Murine Model of Lens-Induced Myopia

To identify tissues and molecules involved in refractive myopic shift and axial length elongation in a murine lens-induced myopia model, we performed a comprehensive analysis of microRNA (miRNA) expression. Three weeks after negative 30 diopter lens fixation on three-week-old C57BL/6J mice, total RNA was extracted from individual ocular components including cornea, iris, lens, retina, retinal pigment epithelium (RPE)/choroid, and sclera tissue. The miRNA expression analysis was pooled from three samples and carried out using Agilent Mouse miRNA Microarray (8 × 60 K) miRBase21.0. The expression ratio was calculated, and differentially expressed miRNAs were extracted, using GeneSpring GX 14.5. Myopic induction showed a significant myopic refractive change, axial elongation, and choroidal thinning. Through the comprehensive miRNA analysis, several upregulated miRNAs (56 in cornea tissue, 13 in iris tissue, 6 in lens tissue, 0 in retina tissue, 29 in RPE/choroid tissue, and 30 in sclera tissue) and downregulated miRNAs (7 in cornea tissue, 28 in iris tissue, 17 in lens tissue, 9 in retina tissue, 7 in RPE/choroid tissue, and 40 in sclera tissue) were observed. Overlapping expression changes in miRNAs were also found in different ocular components. Some of this miRNA dysregulation may be functionally involved in refractive myopia shift and axial length elongation.


Introduction
In 2000, there were 1.4 billion myopic people globally without correction of refractive errors [1]. It has been predicted that the population of people with myopia will increase to 4.8 billion by 2050 [1,2]. This expansion in the prevalence of myopia is dramatic; therefore, there is an urgent need to understand the mechanisms underlying the development and progression of myopia and establish a treatment. Previous studies have revealed some of the molecular and cellular mechanisms that underlie myopia development; for example, a large genome-wide analysis and a meta-analysis have been performed [3,4]. However, more data are needed to identify the causative factors and biomarkers of myopia progression.
MicroRNAs (miRNAs) are single-stranded RNAs comprised of non-coding RNAs of 21-24 nucleotides in length that are known to inhibit gene expression after transcription [15]. MiRNAs play an important role in various diseases [16,17], and may be used as biomarkers of disease. MiRNAs an important role in various diseases [16,17], and may be used as biomarkers of disease. MiRNAs may exist in serum and plasma [18,19], and miRNAs from validated blood samples can be used as predictors of myopia progression. Several previous studies focused on miRNAs involved in myopia progression suggested the possibility of using miRNAs as biomarkers of myopia progression [20,21].
In this study, using a murine model of LIM [5], a comprehensive analysis of miRNA was performed in individual ocular tissues, including cornea, iris, lens, retina, retinal pigment epithelium (RPE)/choroid, and sclera tissues, with the aim of investigating the expression pattern of and change in miRNAs during myopia progression.

Refractive Error and Axial Length Changes in the Murine Model of Myopia
Three-week-old C57/BL/6J mice wore a negative 30 diopter (D) lens on the right eye. After three weeks, the changes in refractive error and axial length were evaluated. The naïve right eyes of the mice that were not wearing a lens were used as normal eyes. The eyes with lens-induced myopia showed a significant myopic shift in refractive error (−18.42 D ± 3.98) compared to normal eyes (0.95 D ± 1.85) (p < 0.001, Figure 1a). A significant axial elongation was also observed in eyes with lensinduced myopia (0.273 mm ± 0.009) compared to normal eyes (0.155 mm ± 0.015) (p < 0.001, Figure  1b).

Change in Retinal and Choroidal Thickness
Three-week-old C57/BL/6J mice wore a negative 30 D lens on the right eye. After three weeks, the changes in retinal and choroidal thickness were evaluated. The naïve right eyes of the mice that were not wearing a lens were used as normal eyes. The eyes with lens-induced myopia showed a reduction in retinal thickness (−12.432 µm ± 8.937) compared to normal eyes (−1.887 µm ± 10.417) (Figure 2a). A significantly thinner choroid was also observed in eyes with lens-induced myopia (-4.196 µm ± 1.716) compared to normal eyes (0.405 µm ± 0.995) (p < 0.05, Figure 2b).

Change in Retinal and Choroidal Thickness
Three-week-old C57/BL/6J mice wore a negative 30 D lens on the right eye. After three weeks, the changes in retinal and choroidal thickness were evaluated. The naïve right eyes of the mice that were not wearing a lens were used as normal eyes. The eyes with lens-induced myopia showed a reduction in retinal thickness (−12.432 μm ± 8.937) compared to normal eyes (−1.887 μm ± 10.417) (Figure 2a). A significantly thinner choroid was also observed in eyes with lens-induced myopia (-4.196 μm ± 1.716) compared to normal eyes (0.405 μm ± 0.995) (p < 0.05, Figure 2b).

Differentially Expressed miRNAs in Individual Ocular Components During Myopia Progression
Each eyeball was separated into cornea, iris, lens, retina, RPE/choroid, and sclera ( Figure 3). The miRNA array was evaluated using the tissues of the individual ocular components. A number of upregulated miRNAs (56 in cornea tissue, 13 in iris tissue, 6 in lens tissue, 0 in retina tissue, 29 in RPE/choroid tissue, and 30 in sclera tissue) and downregulated miRNAs (7 in cornea tissue, 28 in iris tissue, 17 in lens tissue, 9 in retina tissue, 7 in RPE/choroid tissue, and 40 in sclera tissue) were found in myopic tissues compared to normal tissues (Table 1, Figures 4-9). . Data are presented as the mean ± SD. * p < 0.05; Student's t-test.

Differentially Expressed miRNAs in Individual Ocular Components During Myopia Progression
Each eyeball was separated into cornea, iris, lens, retina, RPE/choroid, and sclera ( Figure 3). The miRNA array was evaluated using the tissues of the individual ocular components. A number of upregulated miRNAs (56 in cornea tissue, 13 in iris tissue, 6 in lens tissue, 0 in retina tissue, 29 in RPE/choroid tissue, and 30 in sclera tissue) and downregulated miRNAs (7 in cornea tissue, 28 in iris tissue, 17 in lens tissue, 9 in retina tissue, 7 in RPE/choroid tissue, and 40 in sclera tissue) were found in myopic tissues compared to normal tissues (          mmu-miR-7020-5p

Overlapping Expression Changes in miRNAs in Different Ocular Component Tissues
Overlapping expression changes in miRNAs were found in different ocular tissues. Table 2 shows the miRNAs that were upregulated in both corneal and other tissues (4 in iris tissue, 2 in lens tissue, 3 in RPE/choroid tissue, and 15 in sclera tissue), the miRNAs that were upregulated in corneal tissue and downregulated in other tissues (three in iris tissue and two in sclera tissue), the miRNAs that were downregulated in corneal tissue and upregulated in other tissues (two in RPE/choroid tissue), and the miRNAs that were downregulated in both corneal tissue and other tissues (one in iris tissue and three in lens tissue). Table 3 shows the miRNAs that were upregulated in both iris and other tissues (four in RPE/choroid tissue and one in sclera tissue), the miRNAs that were upregulated in iris tissue and downregulated in other tissues (one in lens tissue and three in sclera tissue), the miRNAs that were downregulated in iris tissue and upregulated in other tissues (six in RPE/choroid tissue and one in sclera tissue), and the miRNAs that were downregulated in both iris tissue and other tissues (one in lens tissue, one in RPE/choroid tissue, and four in sclera tissue). Table 4 shows the one miRNA that was upregulated in both lens and RPE/choroid tissues, the three miRNAs that were upregulated in lens tissue and downregulated in sclera tissue, the miRNAs that were downregulated in lens tissue and upregulated in other tissues (one in RPE/choroid tissue and one in sclera tissue), and the miRNAs that were downregulated in both lens and other tissues (two in retina tissue and two in sclera tissue). Table 5 shows the two miRNAs that were downregulated in retina tissue and upregulated in RPE/choroid tissue. Table 6 shows the four miRNAs that were upregulated in both RPE/choroid tissue and sclera tissue, the two miRNAs that were upregulated in RPE/choroid tissue and downregulated in sclera tissue, and the two miRNAs that were downregulated in RPE/choroid tissue and upregulated in sclera tissue. In these expression changes in miRNAs, 18 miRNAs overlapped in more than three different types of ocular component tissue (Table 7).

Overlapping Expression Changes in miRNAs in Different Ocular Component Tissues
Overlapping expression changes in miRNAs were found in different ocular tissues. Table 2 shows the miRNAs that were upregulated in both corneal and other tissues (4 in iris tissue, 2 in lens tissue, 3 in RPE/choroid tissue, and 15 in sclera tissue), the miRNAs that were upregulated in corneal tissue and downregulated in other tissues (three in iris tissue and two in sclera tissue), the miRNAs that were downregulated in corneal tissue and upregulated in other tissues (two in RPE/choroid tissue), and the miRNAs that were downregulated in both corneal tissue and other tissues (one in iris tissue and three in lens tissue). Table 3 shows the miRNAs that were upregulated in both iris and other tissues (four in RPE/choroid tissue and one in sclera tissue), the miRNAs that were upregulated in iris tissue and downregulated in other tissues (one in lens tissue and three in sclera tissue), the miRNAs that were downregulated in iris tissue and upregulated in other tissues (six in RPE/choroid tissue and one in sclera tissue), and the miRNAs that were downregulated in both iris tissue and other tissues (one in lens tissue, one in RPE/choroid tissue, and four in sclera tissue). Table 4 shows the one miRNA that was upregulated in both lens and RPE/choroid tissues, the three miRNAs that were upregulated in lens tissue and downregulated in sclera tissue, the miRNAs that were downregulated in lens tissue and upregulated in other tissues (one in RPE/choroid tissue and one in sclera tissue), and the miRNAs that were downregulated in both lens and other tissues (two in retina tissue and two in sclera tissue). Table 5 shows the two miRNAs that were downregulated in retina tissue and upregulated in RPE/choroid tissue. Table 6 shows the four miRNAs that were upregulated in both RPE/choroid tissue and sclera tissue, the two miRNAs that were upregulated in RPE/choroid tissue and downregulated in sclera tissue, and the two miRNAs that were downregulated in RPE/choroid tissue and upregulated in sclera tissue. In these expression changes in miRNAs, 18 miRNAs overlapped in more than three different types of ocular component tissue (Table 7). mmu-miR-9-3p -mmu-miR-7047-3p mmu-miR-129b-5p mmu-miR-7085-3p Sclera mmu-miR-1927 mmu-miR-676-3p mmu-miR-338-3p mmu-miR-494-3p mmu-miR-2137 mmu-miR-574-5p mmu-miR-3082-5p mmu-miR-7218-5p mmu-miR-7682-3p mmu-miR-1187 mmu-miR-6918-5p mmu-miR-504-3p mmu-miR-129b-5p mmu-miR-466i-5p mmu-miR-6904-5p mmu-miR-7686-5p mmu-miR-1224-5p Table 3. Overlapping expression changes in miRNAs in iris and other ocular tissues.  mmu-miR-188-5p mmu-miR-6965-5p mmu-miR-124-3p mmu-miR-202-3p Table 5. Overlapping expression changes in miRNAs in retina and other ocular tissues.

Up Each Eye-Tissues
DownEach Eye-Tissues

Affected miRNAs and Their Predicted Target mRNA in Different Ocular Component Tissues
Based on the myopia-induced changes in miRNA expression, the target mRNA was predicted using MiRTarBase. Tables 8-12 show the target genes predicted from the change in miRNA expression in the cornea (73 genes from 20 miRNAs), the iris (27 genes from 11 miRNAs), the lens (32 genes from 8 miRNAs), the retina (no detection), the RPE/choroid (22 genes from 5 miRNAs) and the sclera (89 genes from 21 miRNAs). The genes shown in bold were overlapped in each ocular tissue. Overlapping expression changes in miRNAs were found in two and three types of ocular tissue, and their target genes are shown in Tables 13 and 14.     Table 13. The overlapping expression changes in miRNAs found in two ocular tissues and their target genes.

Discussion
In this study, we performed a comprehensive miRNA analysis of ocular component tissues, including cornea, iris, lens, retina, RPE/choroid, and sclera tissues, from an experimental murine model of myopia. A number of differentially expressed miRNAs were observed in each ocular component. Overlapping expression changes were also found in different ocular components.
Several mRNA expression changes in FDM or LIM mice were previously reported, including an increase in fibroblast growth factor (FGF) 10 expression in sclera tissue [22] and a decrease in Wingless (WNT) 2b/Frizzled (FZD) 5/β-catenin expression in retina tissue [23]. An increase in Wnt3/b-catenin gene expression [24] and a decrease in TGF-β/Col1 gene expression [9] were also reported in FDM guinea pigs. Furthermore, previous studies showed that various gene expression changes may be functionally related to myopic phenotypes. A myopic shift in refractive errors and elongation of the axial length were reported in EGR1-deficient [25], M2-deficient [26], LRP2-overexpressed [27], APLP2-deficent [28] and Lumican-overexpressed [29] mice. Furthermore, studies in human cohorts have identified a number of genes related to myopia development [30]. In our study, a range of target mRNAs were predicted from expression changes in miRNA expression. The predicted genes were found to correspond to EGR-1 and TGF-β or be similar to FGF, WNT, and FZD, which were described in the abovementioned studies. Although the predicted genes were different between myopic induction models and species, these genes may play an important role in myopia progression. On the other hand, other genes, such as PTEN in the cornea and VSIVGP2, NOTCH1, STAT3, and CLICS in the sclera, were found to not correspond to previous reports, suggesting that these genes may have a function that is specific to each ocular component.
It has been reported that miR-200a/b/c expression overlaps in a range of tissues with a tubular structure, including kidney tissue (proximal tubule and collecting duct), lung tissue, pancreas tissue (duct cells), small intestine tissue (intestinal villus), bile duct tissue, and exocrine gland tissue (duct cells). Furthermore, miR-200a/b/c expression was found to be increased in plasma from the site of an acute kidney injury, suggesting that miR-200a/b/c may be used as a biomarker for kidney and other tubular structure organ injury [31]. In the current study, we found overlapping changes in miRNA expression in two and three types of ocular tissue (Tables 2-7). These individual ocular components are in close proximity to and functionally connected with each other. Thus, we suggest that overlapping changes in miRNA expression among different ocular components can be used as myopic diagnosis markers.
In a previous study, eight miRNAs were found to be upregulated and to overlap with retinas and whole eyeballs in a murine model of form-deprivation myopia. The authors screened out 1805 target genes for the eight differentially expressed miRNAs, including MAPK-10 [32]. In the present study, we also found a number of overlapping miRNAs in individual ocular components together with predicted target genes (Tables 13 and 14). Although these new target genes were found to not exactly correspond to previous reports, we speculate that these genes may be important factors in the suppression or acceleration of myopia progression.
Comprehensive approaches to the analysis of mRNA and miRNA expression have also been reported. In sclera from an FDM mice model, Let-7a, miR-16-2, Smok4a, Prph2, and Gnat1 expression were found to fluctuate [14]. Fifty-three (53) miRNAs were previously reported to be either upregulated or downregulated in the retina of LIM mice, and mmu-miR-671-5p was identified among them [33]. In this study, 18 miRNAs were identified as being differentially expressed in three different ocular components. These miRNAs may play an important role in myopia progression. Mmu-miR-7047-3p, mmu-miR-7085-3p, and mmu-miR-96-5p showed an opposite change in expression between the anterior (cornea, iris, and lens) and posterior (retina, RPE/choroid, and sclera) ocular components. Although some differences exist between animal species [34], the expression and functions of miRNAs are largely evolutionarily conserved [35]. In the present study, we identified both discrete and overlapping changes in miRNA expression in individual ocular components during myopia progression in a murine model. Further studies will be conducted to explore miRNA profiles in order to understand the molecular pathogenesis of human myopia progression in humans and to establish of biomarkers for its prediction.

Experimental Animals
The experimental protocol used in this study complied with the National Institutes of Health (NIH) guidelines for working with laboratory animals, the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, and the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. The experimental protocol was approved by the Institutional Animal Care and Use Committee at Keio University. C57BL/6J male mice (CLEA Japan, Yokohama, Japan) were maintained by free intake of a standard diet (MF, Oriental Yeast Co., Ltd., Tokyo, Japan) and water, with three mice in one cage. The mice were raised in an environment with a 12 h/12 h light/dark cycle (the dark cycle from 8:00 p.m. to 8:00 a.m.) at 23 ± 3 • C. The light cycle was maintained using a 50-lux background. These conditions were based on a previously reported experimental murine model of myopia [5]. The animal trial was approved by the ethics committee of Keio University (ethics review number: 16017-(1), 25 October 2017).

Myopia Induction
Before and after the myopia induction, the refraction and axial length of all eyes were measured using a refractometer (Steinberis Transfer Center, Tübingen, Germany) and spectral domain optical coherent tomography (SD-OCT, Envisu R4310, Leica, Wetzlar, Germany), respectively, under anesthesia by medetomidine (0.75 mg/kg, Sandoz K.K., Tokyo, Japan), midazolam (4 mg/kg, Domitor ® , Orion Corporation, Espoo, Finland), and butorphanol tartrate (5 mg/kg, Meiji Seika Pharma Co., Ltd., Tokyo, Japan) dissolved in normal saline (MMB). For the myopia induction group, a −30 D lens was fixed onto the right eye at postnatal three weeks old, and the mice were kept for three weeks. For the normal group, mice were prepared and kept without any special treatment. The induction of myopia and the ocular measurement were based on a previously reported experimental murine model of myopia [5]. Three mice were used for the myopia induction group and the normal group, respectively. In accordance with a previous report [36], the thickness of the retina and choroid was captured at points that were ±300 µm and ±400 µm from the optic nerve, respectively, and measured using the NIH ImageJ software.

miRNA Extraction
After euthanasia was performed by intraperitoneal MMB injection of an overdose of anesthesia, right eyes from the myopia induction group and right eyes from the control group were enucleated and separated into cornea, iris, lens, retina, RPE/choroid, and sclera tissues. The separated ocular tissues were put into QIAzol and homogenized. Total RNA extraction was performed using a miRNeasy Micro kit (QIAGEN, Venlo, Netherlands) according to the manufacturer's instructions.

miRNA Microarray
After mixing equal amounts of three samples extracted from the same eye tissues, a template of 100 ng total RNA was applied to Agilent Mouse miRNA Microarray (8 × 60 K) miRBase 21.0 (Agilent, Santa Clara, CA, USA). Cyamin-3 labeling samples were hybridized at 55 • C and 20 rpm for 17 h using a miRNA Complete Labeling and Hyb Kit (Agilent) and an Expression Hybridization Kit (Agilent). After hybridization, the microarrays were scanned using a DNA Microarray Scanner (Agilent). The scanning data were digitized using the Feature Extraction version 10.7.1.1 software (Agilent). The miRNA microarray protocols were applied by following each manufacturer's instructions. The microarray datasets are displayed in the National Center for Biotechnology Information (NCBI) Gene Expression Omnibus with the accession number GSE131831.

Data Analysis
miRNA microarray data were analyzed using the Geneview Data module in GeneSpring GX 14.5 (Agilent). The expression ratio of sample molecules for the denominator was calculated according to the combination described in Table 1. A fold-change of >2 and <0.5 with a detected flag in both the denominator and the numerator or a fold-change of >5 and <0.2 with no detected flag in either the denominator or the numerator was used as criteria to select the differentially expressed miRNAs. The predicted genes from the changes in miRNA expression and miRNA-mRNA interaction was analyzed by MiRTarBase 7.0 (National Chiao Tung University, Hsinchu, Taiwan).

Statistical Analyses
The data in Figure 1 are expressed as the mean ± standard deviation. Statistical significance was assessed using the unpaired Student's t-test (Microsoft Excel 2013). Results with p-values of less than 0.05 were considered statistically significant.

Conclusions
To the best of our knowledge, this study is the first report of a comprehensive analysis of miRNA expression in different ocular component tissues of LIM mice. Further analyses, such as a cluster analysis or a gene ontology (GO) analysis, are required for a full understanding of the function of differentially expressed miRNAs in different tissues. A comparison to changes in mRNA expression in myopia progression is also important to reveal interactions between miRNA and mRNA. Based on the findings in this study, miRNAs that play a critical role in myopia development and progression may be found and adopted for clinical use as therapeutic targets or diagnostic tools in the future.
Funding: This work was supported by Santen Pharmaceutical Co., LTD (Osaka, Japan), and was a collaborative study between Keio University and Santen.