Association Between Pigmentation Heritage and Susceptibility to Experimentally Induced Myopia: Crossbreeding Insights From Albino and Pigmented Guinea Pigs

Abstract

Purpose

To investigate how pigmentation heritage affects susceptibility to form deprivation myopia (FDM) through crossbreeding myopia-susceptible albino and myopia-resistant pigmented guinea pigs (GPs).

Methods

Ninety GPs (albino: n = 29, pigmented: n = 29, F1 crossbred: n = 32) were studied. Eye parameters were measured using retinoscopy and A-scan ultrasound. Scotopic electroretinograms were recorded by full-field electroretinography in 36 naive GPs at 1 and 5 weeks of age. Monocular form deprivation (FD) was applied from 1 to 5 weeks of age in 27 GPs, with 27 littermates as controls. The retina, choroid, and sclera were imaged using optical coherence tomography. From control GPs at 5 weeks, scleral melanin was assessed histologically; choroidal melanin and tyrosinase activity in the choroid and sclera were quantified biochemically.

Results

Crossbreds had a pigmented appearance, and choroidal thickness (ChT, 100 ± 19 µm) was thicker than in albino GPs (69 ± 12 µm, P < 0.001) but thinner than in pigmented GPs (124 ± 19 µm, P = 0.004). Choroidal melanin content and tyrosinase activity followed similar patterns and were significantly different (one-way ANOVA): highest in pigmented GPs, intermediate in crossbreds, and undetectable in albinos. Albinos exhibited the largest a-wave amplitudes across the intensity-response functions among the three breeds. Crossbreds displayed a monotonic b-wave intensity-response function, similar to but with higher amplitudes than pigmented GPs, whereas albinos showed a bell-shaped response pattern before the second rise. FD induced significant myopic shift and axial elongation in albinos (spherical equivalent [SE]: −7.83 ± 4.74 D, P = 0.001; axial length [AL]: 0.20 ± 0.14 mm, P = 0.003) and crossbreds (SE: −5.42 ± 2.90 D, P < 0.001; AL: 0.17 ± 0.09 mm, P = 0.001). Pigmented GPs showed a mild myopic shift (−2.08 ± 1.93 D, P = 0.01) with no significant AL changes (0.03 ± 0.08 mm, P = 0.357). FDM severity differed across breeds (SE: P = 0.006, AL: P = 0.004); thicker baseline ChT was associated with less FDM in crossbreds (P = 0.027) and reduced axial elongation in albinos (P = 0.004); retinal and scleral thicknesses were nonpredictive.

Conclusions

Crossbreeding improved retinal function and choroidal morphology, while preserving susceptibility to FDM, suggesting a partial restoration of myopia-inducible mechanisms inherited from the albino lineage. These findings support the value of the crossbred model for investigating retinal and choroidal regulation in myopia.

Emmetropization is the postnatal process by which the eye's optical power and axial length are optimized for clear vision.¹ This process is regulated by retinal signals and functional downstream signaling cascades through the choroid to the sclera.^2,3 While environmental factors play a significant role, genetic factors can predispose individuals to a higher risk of developing myopia.⁴ For example, emmetropization is impaired in patients with albinism,^5–7 and this is also observed in albino guinea pigs (GPs), which show a similar bimodal distribution of refractive error.⁸ Both albino GPs and albino chicks exhibit a high incidence of spontaneous myopia,⁹ and the hyperopic albino GPs are more susceptible to developing myopia from form deprivation (FD) compared to their pigmented counterparts.⁸

Albinism is a congenital condition characterized by the absence of melanin. In healthy individuals, melanin is synthesized and stored in the melanosomes of melanocytes¹⁰ or taken up by fibroblasts.¹¹ The disrupted emmetropization observed in albinos suggests that pigment cells may play an important role in eye growth and refractive development. Pigmented cells in the eye vary in their roles due to differing embryonic origins.¹² For example, the retinal pigment epithelium (RPE) originates from the outer layer of the optic cup, while melanocytes in the uveal tract (iris, ciliary body, and choroid) originate from the neural crest. The RPE plays a vital role in eye growth regulation and myopia,¹³ while the choroid is also widely accepted as being actively involved in myopic eye growth.¹⁴ The role of melanin in the RPE and uveal melanocytes includes photo-screening protective effects, as well as biophysical and biochemical protective effects.¹⁵ In the choroid, melanocytes typically cluster around blood vessels, and they contribute to choroidal morphogenesis and/or maintain the normal choroidal vasculature structure.¹⁶ Using thickness as an index of choroidal structure, comparative studies have demonstrated that choroidal thickness (ChT) could be a biomarker for myopiagenesis. For example, albino GPs have thin choroids and are highly susceptible to myopia development.¹⁷ Consistently, lighter-colored GPs have thinner choroids and are readily myopiagenic, whereas darker GPs have thicker choroids and are less likely to develop experimentally induced myopia.¹⁸ Hence, melanocytes may play a role in myopia development by regulating ChT.¹⁶ Thus, we speculate that nonocular phenotypes, such as the pigmentation of the sclera or coat color, could serve as another biomarker for myopiagenesis in GPs.

This study investigates whether crossbreeding albino GPs with pigmented GPs influences susceptibility to form deprivation myopia (FDM). Specifically, we aim to determine whether crossbreeding mitigates myopia susceptibility and whether external phenotypes can predict susceptibility to FDM induction.

Methods

Animal Breeding and Housing

Pigmented and albino GP breeders were purchased from Elm Hill Labs (Chelmsford, MA, USA) and bred on-site. Unrelated animals were paired for breeding, and F1 crossbred GPs (referred to throughout as “crossbred pups”) were the first-generation offspring of albino and pigmented parents. Pups remained with their mothers until weaning at week 3 and then were housed in same-sex pairs in transparent plastic cages under a 12-hour light/dark cycle (floor luminance: 160–180 lux; PHILIPS, TL5/28W/840, Shenzhen Guangdong China). Water and vitamin C–supplemented food were provided ad libitum, along with fresh fruits and vegetables three times weekly.

Pups from three breeds (albino, crossbred, pigmented; n = 9 each) were recruited at week 1 for monocular FDM inducement, and age-matched pups (n = 9 each), using littermates of the myopia inducement pups as much as possible, were recruited as normal controls. During the experiment, cage floor luminance was 510 to 570 lux (HT630 Digital Light Meter; Habotest, Estonia, Dongguan, China). Animal care followed the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research, with protocols approved by the Institutional Animal Care and Use Committee of SingHealth (AAALAC Accredited; 2021/SHS/1667).

Myopia Induction

FDM was induced by fitting a diffuser over the right eye, with the left eye as a contralateral control. Following baseline measurements at week 1, 27 pups (albino, crossbred, pigmented, n = 9 per group) were fitted with diffusers following Howlett and McFadden's method.¹⁹ Briefly, Velcro loop arcs were glued above and below the eyelids to secure the diffuser. To prevent the animals from removing the diffuser, “booties” were made by taping all paws. Pups were housed in transparent cages with stainless wire tops for 4 weeks and monitored daily to ensure that the diffusers were always in place.

Biometric Data Collection

Ocular measurements were obtained at the end of week 1 (W1, baseline) and week 5 of life (W5, endline) for both untreated control and FDM animals. Refraction, reported as spherical equivalent (SE) refractive error (diopter, D), was measured in alert animals using streak retinoscopy 30 minutes after administering four drops of 1% cyclopentolate hydrochloride (Bausch & Lomb, Rochester, NY, USA), with each drop applied 5 minutes apart. Axial length (AL) and vitreous chamber depth were measured using A-scan ultrasonography (10 MHz PacScan 300A; Sonomed Escalon, New Hyde Park, NY, USA) under local corneal anesthesia (one drop of 0.5% Alcaine [proparacaine hydrochloride] eye solution; Alcon, Geneva, Switzerland).

Optical coherence tomography (OCT) scans were conducted in the afternoon to minimize variability caused by circadian fluctuations in ChT.²⁰ Animals were anesthetized via intramuscular injection of a cocktail of ketamine (27 mg/kg; Ceva Animal Health, Lenexa, KS, USA) and xylazine (0.6 mg/kg; Troy Laboratories, Glendenning, NSW, Australia). Spectralis Widefield Imaging Module with eye-tracking (TruTrack; Heidelberg Engineering, Heidelberg, Germany) was used to acquire high-resolution posterior segment spectral-domain OCT scans. Each scan had 13 B-scan lines with 1536 axial A-scan pixels per line, averaging 15 frames per B-scan in enhanced depth imaging mode. Scans were taken across the visual streak, within the superior retina²¹—specifically, from the superonasal retina to the optic nerve head (ONH), down to the midpoint of the ONH. Images approximately 2.5 ONH diameters above the ONH were evaluated using Heidelberg Eye Explorer (HEYEX, version 1.7.1.0; Heidelberg Engineering, Heidelberg, Germany). Retinal, choroidal, and scleral thickness were manually measured with built-in digital calipers as previously described.^17,18 In brief, 10 equally spaced lateral locations within the selected B-scan through the visual streak were measured and averaged.

Electroretinogram Recording

A separate cohort of 36 GPs (albino, n = 11; crossbred, n = 14; pigmented, n = 11) at 1 and 5 weeks of age were dark-adapted overnight (≤12 hours) and anesthetized under dim red light with intramuscular injections of a ketamine (28 mg/kg) and xylazine (0.5 mg/kg) mixture. Pupils were dilated with 1% tropicamide and 2.5% phenylephrine (Alcon). Animals were maintained at 37°C on a heating plate. Flash ERG (white light) responses were recorded using an Espion E³ Electrophysiology System Console (Model Number: D315, Serial Number: 00272; Diagnosys LLC, Lowell, MA, USA) connected with a regular color dome with four LEDs. A gold wire electrode (5 mm) was placed on the cornea, with a stainless-steel reference in a neck skinfold and a ground needle electrode in the tail. Scotopic ERG responses to 4-ms white light flashes (6500k) were captured at increasing intensities from 0.0001 to 200 cd⋅s⋅m⁻².

Tissue Collection for Melanin Content and Tyrosinase Activity Test

Control animals at 5 weeks of age were light-adapted and euthanized for tissue collection in the early afternoon (13:00–16:00) under standard room lighting (300 lux). Deep anesthesia was induced by intramuscular injection of ketamine (45 mg/kg) and xylazine (5 mg/kg), followed by a cardiac injection of 0.5 mL valabarb euthanasia solution (300 mg/mL pentobarbitone sodium; Jurox, Rutherford, NSW, Australia). Right eyes were enucleated and hemisected through the equator, and posterior eye cups were prepared on ice. After vitreous removal, the retina was separated, and each eye cup was incised radially into a “tri-petal” shape. The RPE/choroid complex was scraped from the sclera, and both tissues were placed separately into 2-mL Eppendorf tubes.

Melanin Extraction and Quantification

RPE/choroidal samples were weighed; ground in a mixture of RIPA buffer, PPI, and water (150 µL per sample); and homogenized three times (25 Hx, 90 seconds; TissueLyser II; Qiagen, Hilden, Germany), followed by sonication for 60 seconds. The supernatant was removed, and the pellet was washed twice with 1 mL phosphate buffer (50 mM, pH 7.4). The pellet was then incubated in Tris buffer (50 mM, pH 7.4) containing SDS (5 mg/mL) at 37°C for 1 hour and centrifuged (15,000 rpm, 20 minutes), and the supernatant was discarded. This step was repeated with added proteinase K (0.2 mg/mL; New England Biolabs, Ipswich, MA, USA). The pellet was sequentially washed twice with NaCl (0.25 M), distilled water, methanol (0.6 mL), and hexane (0.3 mL), with centrifugation at each step, then dried using the Concentrator plus (Eppendorf, Hamburg, Germany).

Melanin standard (cat. M0418; Sigma-Aldrich, St. Louis, MO, USA) and test samples were dissolved in a mixture of 800 µL DMSO and 200 µL NaOH (0.1 M). In a 96-well plate, 180 µL of sample or standard was added, and absorbance at 325 nm was measured using a Tecan Infinite M200 Pro microplate reader (Tecan, Mannedorf, Switzerland). Concentrations were determined via a standard curve.

Tyrosinase Dihydroxyphenylalanine (DOPA) Oxidase Activity Assay

Tyrosinase activity was tested via dihydroxyphenylalanine (DOPA) oxidation, performed as previously described.²² Briefly, total proteins were extracted separately from choroid and sclera and incubated with 0.5% L-DOPA at 37°C for 60 minutes. Absorbance was measured at 475 nm using a Tecan Infinite M200 pro plate reader.

Histology Examination

Left eyes from control GPs (week 5) were fixed in freshly prepared 2.5% glutaraldehyde and 2% paraformaldehyde (0.1 M phosphate buffer) at 4°C for 24 hours, with a suture at the superior limbus for orientation. Eyeballs were dissected based on ciliary artery positioning, dehydrated, paraffin embedded, and sectioned horizontally above the optic nerve for hematoxylin and eosin (H&E) staining.

Statistical Analysis

One-way ANOVA with Bonferroni correction was used to compare the thickness of each posterior eye wall sublayer (the retina, choroid, and sclera) at week 1, as well as melanin content and tyrosinase activity at week 5, in right eyes across the different breeds. Normality of the data distribution was confirmed by Kolmogorov–Smirnov tests; Welch's ANOVA was applied when variance was unequal.

Repeated-measures ANOVA assessed ERG difference between breeds, with Greenhouse–Geisser correction for sphericity violations. Univariate tests examined breed effects at each light intensity, followed by Bonferroni-corrected post hoc comparisons.

The difference between the two eyes at the end of FDM was evaluated by paired t-test. Myopia progression was evaluated by calculating the change in the intraocular difference at the endpoint (Treated eye – Fellow eye). A one-way ANOVA was used to compare differences in both physiological eye growth and FD-induced myopic eye growth between breeds. Two-way ANOVA was used to compare the differences in changes ([Post-FDM intraocular difference] – [Baseline intraocular difference]) in the thickness of posterior eye wall sublayers between breeds.

To further interpret the potential role of ChT in the development of FDM, linear regression was performed to explore associations between the absolute ChT at W1 and interocular differences of refractive error and axial length at W5, with results displayed in scatterplots.

Data are reported as mean ± SD, except for full-field ERG (ffERG) data (mean ± SEM). Statistical significance was set at P < 0.05.

Results

Under Physiological Growth, Crossbred GPs Displayed Similar Emmetropization as Pure-Bred Albino and Pigmented GPs

Using the right eyes for comparison, pigmented GPs were the most hyperopic (6.83 ± 1.06 D at week 1; 2.47 ± 1.01 D at week 5, n = 9), followed by crossbred (5.56 ± 2.04 D; 0.67 ± 0.63 D, n = 9) and albino GPs (4.81 ± 1.87 D; 0.10 ± 2.12 D, n = 9) (Fig. 1A). While the difference at week 1 was not significant (P = 0.057), it became significant by week 5 (P = 0.004). Albino GPs had the longest axial lengths at both time points (7.24 ± 0.09 mm and 7.94 ± 0.11 mm, n = 9, Fig. 1B). At week 1, pigmented GPs showed slightly longer axial lengths than crossbred GPs (7.17 ± 0.10 mm vs. 7.15 ± 0.13 mm), but by week 5, crossbred GPs had surpassed pigmented GPs (7.91 ± 0.12 mm vs. 7.84 ± 0.09 mm). Despite these numerical trends, the differences were not statistically significant (week 1, P = 0.2; week 5, P = 0.1). In summary, longitudinal growth and refractive shifts were similar across all breeds (axial length: F(2, 16) = 1.841, P = 0.191; refractive error: F(2, 16) = 0.175, P = 0.841).

Figure 1.

The dispersion of refraction (A) and axial length (B) at baseline (week 1 of age) and endline (week 5 of age) as measured in the right eye from all albino, crossbred, and pigmented GPs in this study and the physiological myopic shift (C) and axial elongation (D) after 4 weeks (week 5 of age) in normal control GPs.

Variations of Pigmentation and Tyrosinase Activity Among the Three Breeds

En face OCT images of 1-week-old untreated GPs showed clear choroidal vessels in albino GPs due to minimal RPE pigmentation (Fig. 2A). In crossbred and pigmented GPs, increased pigmentation obscured these vessels (Figs. 2B, 2C). However, vortex veins (white arrows) remained visible in crossbred GPs (Fig. 2B), unlike in pigmented GPs.

Figure 2.

En face (A–C) and cross-sectional optical coherence tomography B-scan images within the visual streak area (a–c) from representative normal, untreated left eyes of guinea pigs at week 1 from the albino, crossbred, and pigmented animal groups. (D) The labels in panels a–c represent the following structures: R = retina, C = choroid, S = sclera. White arrows denote vortex veins. Asterisks indicate statistical significance from post hoc tests with Bonferroni correction following a significant difference by one-way ANOVA. *P < 0.05, **P < 0.01, ***P < 0.001. This notation is consistent across all figures.

At week 1, ChT differed significantly among breeds (F(2, 24) = 23.661, P < 0.001), with albino animals having thinner choroids than crossbred (P = 0.003) and pigmented animals (P < 0.001, Fig. 2D; Table 1). Crossbred GPs had intermediate ChT values. By contrast, no significant differences were observed in the thickness of the retina (F(2, 24) = 1.547, P = 0.233) and sclera (F(2, 24) = 3.224, P = 0.058).

Table 1.

Comparison of Fundal Layer Thickness in Untreated Guinea Pigs

	Retina (µm)			Choroid (µm)			Sclera (µm)
	W1	W5	P Value*	W1	W5	P Value*	W1	W5	P Value*
A	151 ± 5	149 ± 7	0.358	69 ± 12	68 ± 5	0.813	110 ± 12	115 ± 11	0.212
C	154 ± 5	152 ± 6	0.237	100 ± 19	108 ± 14	0.105	100 ± 13	112 ± 6	0.006
P	155 ± 5	154 ± 5	0.701	124 ± 19	138 ± 14	0.015	116 ± 15	128 ± 11	0.008
P value†	0.233	0.259		<0.001	<0.001		0.058	0.004

W1 indicates week 1 of age, and W5 indicates week 5 of age. Bold font denotes P < 0.05. These notations are consistent in Table 2 and across other figures where it is applicable. A, albino; C, crossbred; P, pigmented.

^*Paired t-test between two age points of the same GPs.

^†One-way ANOVA test for three breeds at the same age point.

Table 2.

Comparison of Fundal Layer Thickness in FDM Guinea Pigs

	Retina (µm)			Choroid (µm)			Sclera (µm)
	bfr	aft	P Value*	bfr	aft	P Value*	bfr	aft	P Value*
A	152 ± 4	149 ± 4	0.051	73 ± 12	72 ± 10	0.765	110 ± 15	126 ± 17	0.011
C	156 ± 4	154 ± 3	0.370	102 ± 21	109 ± 23	0.105	114 ± 16	131 ± 14	0.001
P	159 ± 5	152 ± 3	0.002	124 ± 12	128 ± 13	0.150	122 ± 9	139 ± 9	0.003
P value†	0.093	0.181		<0.001	<0.001		0.334	0.164

Comparison among breeds after FDM was performed using two-way ANOVA. Bold font denotes P < 0.05. aft, after 4 weeks of FD; bft, baseline before FD.

^*Paired t-test between two age points of the same GPs.

^†One-way ANOVA test for three breeds at the same age point.

Over time, ChT significantly increased in pigmented GPs (P = 0.015) but not in albino (P = 0.813) or crossbred GPs (P = 0.105). Retinal thickness remained stable across all groups, while sclera thickening was significant in crossbred (P = 0.006) and pigmented (P = 0.008) GPs but not in albinos (P = 0.212) (Table 1).

Upon enucleation at week 5, a gross examination of the sclera revealed no visible pigment in albino GPs (Fig. 3A-a), while pigment was primarily clustered anterior to the equator in crossbred GPs (Fig. 3A-b) and distributed throughout the sclera in pigmented GPs (Fig. 3A-c). H&E staining of sections above the optic nerve head showed that melanin was absent in the sclera of albino GP eyes (Fig. 3A-d), occasionally present in crossbred GP eyes, and prominently visible in pigmented GP eyes (yellow arrows, Fig. 3A-f).

Figure 3.

External gross images and H&E-stained images (A), melanin content in the RPE/choroid (B), and tyrosinase activity in the sclera (C) and RPE/choroid complex (D) from normal control guinea pigs at 5 weeks of life. Red lines indicate the corresponding locations from which histochemistry images were derived. Histochemistry images reveal the presence of melanin in the sclera (yellow arrows in A, subpanels a–f).

Choroidal melanin content from the right eyes of untreated albino, crossbred, and pigmented animals was significantly different between groups (F(2, 14) = 38.572, P < 0.001, Fig. 3B). Pigmented GPs exhibited significantly higher melanin content than both crossbred (0.37 ± 0.14 vs. 0.07 ± 0.04 µg/mL, P < 0.001) and albino GPs (0.37 ± 0.14 vs. −0.20 ± 0.13 µg/mL, P < 0.001), with the latter group exhibiting levels comparable to blank controls. Crossbred animals also displayed greater melanin levels than albino animals (0.07 ± 0.04 vs. −0.20 ± 0.13 µg/mL, P = 0.002). Tyrosinase activity was also different among groups in the RPE/choroid complex (Fig. 3D, P = 0.003) but approached insignificance in the sclera (Fig. 3C, P = 0.052). Pigmented GPs exhibited significantly greater tyrosinase activity in the RPE/choroid than both crossbred (P = 0.027) and albino GPs (P = 0.001).

Functional Differences Between Breeds Reflected by Electroretinography

At both week 1 (Fig. 4A) and week 5 (Fig. 4A′), albino animals displayed significantly greater a-wave amplitudes than crossbred and pigmented animals at flash intensities ranging from −1 to 1 log cd⋅s⋅m⁻². However, a-wave amplitudes plateaued between 1 and 2.3 log cd⋅s⋅m⁻² in albino animals, resulting in no significant difference compared to crossbred or pigmented animals at the brightest flash intensity.

Figure 4.

Scotopic a-wave (A, A′), b-wave (B, B′) intensity-response functions (abscissa: log cd⋅s⋅m⁻²; ordinate: µV) and peak time (C, C′; ordinate: ms) for pigmented (black-filled symbols), crossbred (gray-filled symbols), and albino (unfilled symbols) guinea pigs at week 1 (A–C) and week 5 (A′–C′). Each data point represents the mean ± 1 SEM. Different symbols indicate significant post hoc comparisons when univariate tests were significant at each light stimulus level. £ denotes a significant difference between albino and crossbred GPs, ¥ between albino and pigmented GPs, and $ between crossbred and pigmented GPs.

The relationship between b-wave amplitude and flash intensity differed significantly among breeds at both week 1 (Fig. 4B; F(6.99, 115.27) = 3.67, P = 0.002) and week 5 (Fig. 4B′; F(8.61, 137.76) = 7.74, P < 0.002), primarily driven by albino responses. At week 1, b-wave amplitudes increased from −4 to −2 log cd⋅s⋅m⁻² across all groups, with minor peaks at −2.3 log cd⋅s⋅m⁻², where albino GPs exhibited significantly greater responses than crossbred and pigmented animals. Amplitudes then slightly declined between −1.7 and −0.7 log cd⋅s⋅m⁻², with albinos aligning with crossbred GPs but remaining significantly greater than pigmented GPs. A rapid increase followed from −0.3 to 2.3 log cd⋅s⋅m⁻². At week 5, albino GPs exhibited a more exaggerated version of this pattern, with a pronounced peak at −2.3 log cd⋅s⋅m⁻² and a subsequent reduction between −1.7 and −0.7 log cd⋅s⋅m⁻². Unlike week 1, crossbred and pigmented animals showed a steady increase in b-wave amplitude between −4 and −0. 7 log cd⋅s⋅m⁻². Albino amplitudes remained significantly higher at low intensities (−3 to −2 log cd⋅s⋅m⁻²) but became comparable to crossbred animals at −1.7 and to pigmented animals at −0.7 log cd⋅s⋅m⁻². At bright flash intensities (−0.3 to 2.3 log cd⋅s⋅m⁻²), all groups showed increasing b-wave amplitudes, with the albino group slowing down the increase after 1 log cd⋅s⋅m⁻² and before converging at the highest flash intensity.

The a-wave implicit times were similar across all three breeds at both week 1 and week 5. However, b-wave implicit times differed significantly among breeds (week 1: F(14.18, 234.01) = 2.33, P = 0.05; week 5: F(16.39, 262.33) = 5.57, P = 0.003) (Figs. 4C, 4C′). As expected, these differences were primarily driven by albino GPs, which exhibited significantly shorter response times within the −2.7 to 1 log cd⋅s⋅m⁻² range at week 1 and the −2.3 to 0.7 log cd⋅s⋅m⁻² range at week 5.

Crossbred Pigmented GPs Displayed Increased Susceptibility to FDM and Axial Length Elongation, Compared to Purebred Pigmented GPs

Significant relative myopia and axial elongation (relative to contralateral control left eyes) were observed in albino animals (SE: −7.83 ± 4.74 D, P = 0.001, AL: 0.20 ± 0.14 mm, P = 0.003, Figs. 5A, 5B) and crossbred animals (SE: −5.42 ± 2.90 D, P < 0.001, AL: 0.17 ± 0.09 mm, P = 0.001, Figs. 5A, 5B) following 4 weeks of FD. By comparison, a small, yet significant myopic shift was observed in pigmented GPs, but this was not accompanied by significant axial elongation (SE: −2.08 ± 1.93 D, P = 0.01, AL: 0.03 ± 0.08 mm, P = 0.357, Figs. 5A, 5B). The extent of relative myopia and axial elongation observed was significantly different among animal breeds (F(2, 24) = 6.50, P = 0.006, AL: F(2, 24) = 6.88, P = 0.004, Figs. 5C, 5D). Crossbred animals developed similar levels of myopia to albinos (−5.42 ± 2.90 vs. −7.83 ± 4.74 D, P = 0.433), which was accompanied by similar axial elongation (0.17 ± 0.09 vs. 0.20 ± 0.14 mm, P = 1.000). Both albinos and crossbred animals developed significantly more myopia than pigmented animals (−7.83 ± 4.74 vs. −2.08 ± 1.93 D, P < 0.001 and −5.42 ± 2.90 vs. −2.08 ± 1.93 D, P = 0.004, respectively), accompanied by significantly increased axial elongation (0.20 ± 0.14 vs. 0.03 ± 0.08 mm, P < 0.01 and 0.17 ± 0.09 vs. 0.03 ± 0.08 mm, P = 0.031, respectively).

Figure 5.

Refractive error (A) and axial length (B) of albino (FD-treated [FD4W]–treated eyes, n = 9), crossbred (FD4W-treated eyes, n = 9), and pigmented guinea pigs (FD4W-treated eyes, n = 9) and interocular difference of refraction (C) and axial length (D) at the end of week 5. Asterisks represent paired t-test in A and B, as well as post hoc pairwise comparisons with Bonferroni correction between indicated groups in C and D.

Association of Baseline ChT With FDM in GPs

The thickness of the retina significantly decreased following 4 weeks of myopia induction in pigmented animals (−6 ± 1 µm, P = 0.002); however, no significant difference was observed in crossbred animals (−2 ± 2 µm, P = 0.370), and the apparent thinning observed in albino animals failed to reach statistical significance (−3 ± 1 µm, P = 0.051).

As observed in untreated animals, baseline ChT measures from pigmented GPs entering FD were greater than crossbred GPs, though not statistically significant (22 ± 9 µm, P = 0.109), and significantly greater than albino animals (51 ± 5 um, P < 0.001). Crossbred animals additionally had thicker choroids than albinos (29 ± 9 µm, P = 0.033). Following 4 weeks of myopia induction, ChT was similarly unchanged relative to baseline measures in albino (−1 ± 3 µm, P = 0.765), crossbred (+7 ± 4 µm, P = 0.105), and pigmented animals (5 ± 3 µm, P = 0.150).

The thickness of the sclera significantly increased following 4 weeks of myopia induction in albino (16 ± 5 µm, P = 0.011), crossbred (18 ± 3 µm, P < 0.001), and pigmented animals (17 ± 4 µm, P = 0.003). No significant difference in the extent of scleral thickening was observed among animal groups.

Starting from a similar thickness between the two eyes (paired t-test, all not significant) at week 1, the changes in the relative retinal, choroidal, and scleral thickness (right eye – left eye) from week 1 to week 5 in FD animals were not statistically different from that observed in untreated control animals (Figs. 6A–C). In an exploratory analysis, we observed an association between absolute ChT in the treated eyes and the degree of FDM at week 5, as reflected in both refractive error (Fig. 7A) and axial elongation (Fig. 7B). The regression model predicting refractive error change from baseline ChT was statistically significant (R² = 0.48, P = 0.017), indicating a moderate fit. In contrast, the model predicting axial elongation was more robust (R² = 0.59, P < 0.001), explaining a greater proportion of variance. After accounting for genetic background, it became evident that thicker baseline ChT may protect against myopic axial elongation in albino animals.

Figure 6.

Changes of the thickness of the retina (A), choroid (B), and sclera (C) in FD-treated (FD4W) and normal control (NC) albino (FD4W, n = 9; NC, n = 9), crossbred (FD4W, n = 9; NC, n = 9), and pigmented GPs (FD4W, n = 9; NC, n = 9).

Figure 7.

Association between choroidal thickness at W1 and FDM (A) and axial length elongation (B) at W5 when the three guinea pig breeds were combined (n = 27). Unfilled circles represent albino GPs, gray dots represent crossbred GPs, and dark dots represent pigmented GPs. β, regression coefficient.

Discussion

Form deprivation significantly induced myopia and axial elongation in albino and crossbred GPs, while pigmented animals exhibited mild FDM without significant axial elongation. F1 crossbred GPs showed an intermediate response, suggesting that pigmentation heritage influences myopia susceptibility. Given that melanin plays a role in cell generation and mammalian photoreceptor topography,²³ and tyrosinase enzyme activity affects ERG responses in human albinisms, such that tyrosinase-negative subjects show larger responses than those with tyrosinase-positive albinism,²⁴ we investigated these GPs’ susceptibility to myopia development from two perspectives: the retina function, assessed via scotopic ffERG, and the biochemical properties, evaluated through melanin content and tyrosinase activity in the choroid and sclera.

Retinal signaling plays a crucial role in ocular growth regulation, although ffERG studies on myopia have shown conflicting results.²⁵ Our findings revealed that the more FDM-susceptible breeds exhibited higher ffERG responses. Specifically, albino GPs consistently exhibited exaggerated a-wave responses at both weeks 1 and 5 compared to crossbred and pigmented GPs. This phenomenon may result from reduced ocular resistance due to the lack of melanin,²⁶ or it could be attributed to the hypopigmented nature of albino eyes, where light may enter through the anterior eyewall or reflect from the posterior eyewall, stimulating parts of the anterior retina.^24,27 Interestingly, a transient peak b-wave response was observed before the linearity of the a-wave function at –2.3 log cd·s·m⁻². This maximum rod-driven response likely marks the point at which rod photoreceptors begin to saturate, as suggested in previous studies.²⁸ However, the bell-shaped b-wave responses in albinos within the scotopic (−4 to −2 log cd⋅s⋅m⁻²) and mesopic (−1.7 to −0.7 log cd⋅s⋅m⁻²) regions may reflect rod–cone interactions, or the inhibitory and disinhibitory feedback loops affecting the ON-pathway, mediated by the second- and third-order neurons in the retina, such as dopaminergic, GABAergic, and glycinergic neurons.²⁹ The reduced b-wave amplitude in albino GPs within the mesopic region could be attributed to glycine release,³⁰ as retinal glycine levels were positively associated with the extent of FDM in tricolor GPs.³¹ Notably, crossbreeding appears to rescue retinal function in crossbred GPs, as evidenced by their nearly overlapping a-wave and slightly elevated, yet similarly patterned, b-wave responses compared to their pigmented counterparts.

Results from our tyrosinase activity assays (Fig. 3) indicate lower tyrosinase activity in crossbred GPs compared to pigmented GPs. This may suggest a reduced functionality of melanocytes, given that tyrosinase activity is essential for melanin synthesis and therefore impacts pigmentation levels.^32,33 The partial restoration of choroidal function and structure observed in crossbred GPs may be attributed to this reduction. This restoration is also reflected in the ChT of F1 crossbred GPs, which falls between that of albino and pigmented GPs. Their intermediate susceptibility to myopia development further supports a potential relationship between baseline ChT and myopia susceptibility.¹⁷ However, this relationship may be context-dependent and not directly generalizable across species, as baseline ChT has not been shown to predict myopia development in humans³⁴ or chickens.³⁵

Additionally, a previous study used a tyrosinase inhibitor to promote myopia development in tricolor GPs,²² suggesting that the differing susceptibility to myopia development between crossbred and pigmented GPs may be linked to variations in tyrosinase activity.

Because of the partial retention of melanin synthesis, F1 crossbreds exhibited a pigmented appearance, although pigmentation levels in their fundus were lower than those in purebred pigmented GPs. Although the effect of crossbreeding does not directly implicate melanin, as the case would be if we rescued albino sclera with melanin, the notion of the potential role of melanin and melanocytes in myopia is an interesting one to consider. Melanin content has been linked to biomechanical properties, as seen in other contexts, such as the mechanical strength provided by melanin–protein crosslinks in the jaws of Glycera dibranchiata (bloodworm),³⁶ and the contribution of melanin granules to the stiffness of primary retinal pigment epithelium in porcine eyes.³⁷ Similarly, melanoma cells containing melanin show altered elastic properties, making them less capable of spreading compared to their nonpigmented counterparts.³⁸ These findings suggest that increased total choroidal melanin and tyrosinase activity may strengthen the biomechanical properties of the choroid in crossbred and pigmented strains, potentially offering some protection against ocular elongation in FDM models. If this theory holds, enhancing melanin content could be a strategy to improve the biomechanical strength of ocular tissues and potentially slow myopia development.

Melanocytes are dispersed throughout the choroid and sclera.³⁹ Melanin spots in the sclera may indicate melanin transport between melanocytes and fibroblasts or the extension of melanocyte dendrites. In F1 crossbred GPs, this intercellular transport appears nascent, as indicated by the light pigmentation visible on the external surface of the sclera. In pigmented GPs, it seems more established, with frequent pigmentation spots observed in the outer sclera layer. Postmortem analysis has identified melanocytes in the posterior sclera, with no statistically significant difference in their numbers among light-, medium-, and dark-colored eyes.⁴⁰ This suggests that differences in pigmentation may be more closely linked to variations in tyrosinase activity, reflecting the function of melanocytes rather than their number or distribution.

In addition to their role in melanogenesis, melanocytes are also involved in inflammatory responses.⁴¹ Since inflammation-related signaling is implicated in the pathogenesis of myopia⁴² and melanin has anti-inflammatory properties,⁴³ the functional balance of melanocytes between melanogenesis and immune response regulation may influence the scleral microenvironment and susceptibility to myopia development.

Irrespective of whether crossbreeding effects stem from melanin and melanocytes or other protective traits in pigmented GPs, our findings suggest that pigmentation may be a key marker. Specifically, pigmented GPs show higher levels of posterior eye pigmentation and greater resistance to myopia development than crossbred GPs, and both are more resistant than albino GPs. Thus, lower scleral melanin levels may indicate higher myopia susceptibility.

Limitations

Limitations of our study include a relatively small sample size. Another limitation is that the observed relationship between pigmentation heritage and myopia susceptibility is correlational rather than causal. Future studies using F2 or later-generation crosses would provide greater genetic variation, allowing more robust assessment of this association while minimizing confounding effects such as Simpson's paradox.

Utilizing an emerging polarization-sensitive optical coherence tomography technique to monitor melanin content in vivo⁴⁴ with a larger sample size could better elucidate the association between pigmentation and myopia susceptibility. Additionally, other potential confounders differing between pigmented and albino animals may contribute to the observed differences. For instance, we did not examine specific genetic mutations associated with albinism and altered retinal signaling, which could influence emmetropization. Such genetic factors may affect not only ChT but also choroidal light sensitivity,⁴⁵ retinal function, and overall susceptibility to myopia. Further studies incorporating genetic analysis and functional assessments are warranted to better understand these mechanisms.

Finally, corneal curvature was not measured, which may limit the interpretation of refractive differences. Additionally, extending the observation period for refraction and axial length changes in crossbred animals could provide deeper insights into their susceptibility or resistance to spontaneous and experimentally induced myopia.

Conclusions

Crossbreeding myopia-susceptible albino GPs with myopia-resistant pigmented GPs resulted in pigmented offspring that remained inducible to experimental myopia. The intermediate response to FDM observed in crossbred GPs may be linked to elevated scotopic ffERG response, reduced inhibitory retinal feedback, and partial restoration of tyrosinase activity and melanin content in the choroid and sclera. These findings suggest that crossbreeding may partially restore regulatory mechanisms involved in myopia development and perhaps emmetropization in general. Additionally, lower pigmentation is correlated with higher myopia susceptibility, indicating that visible scleral pigmentation could serve as a biomarker for predicting myopia risk.