Effect of Induced Myopia on Scleral Myofibroblasts and In Vivo Ocular Biomechanical Compliance in the Guinea Pig

Abstract

Purpose.: To examine the effect of induced myopia on scleral myofibroblast populations and in vivo ocular biomechanical compliance.

Methods.: One-week-old guinea pigs were monocularly deprived (MD) of form vision for 2 weeks. Ocular biomechanical compliance was measured in both eyes of anesthetized animals by increasing the intraocular pressure (IOP) to 50 mm Hg for 1 hour, while A-scan ultrasound measures were made every 10 minutes to investigate the change in axial length. The total cell population and myofibroblast subpopulation of the posterior 100° of the sclera was determined with immunohistochemical techniques.

Results.: The vitreous chamber depth (VCD) of MD and contralateral control eyes showed significant elastic expansion on increasing the IOP, compared with that of the nonmanipulated normal eyes. The creep response of the VCD in response to increased IOP was initially greater in the normal eyes until eye length was similar to the MD and control eyes. An unexpectedly high proportion of the scleral cell population were myofibroblasts (63.7% ± 1.7%, average ± SEM; n = 30). MD significantly decreased the total number of cells in the region between the optic nerve and 10° nasal (equivalent to myopic crescent location in humans) compared with the number in control or normal eyes, but no significant effect on myofibroblasts or the total number of cells was found elsewhere.

Conclusions.: A high proportion of scleral cells have contractile potential. This proportion is unaffected by MD. However, there is a significant difference in the in vivo elastic response of the sclera between MD and normal eyes, suggesting that factors other than number of cells have an effect on axial length.

Human myopia primarily results from abnormal elongation of the vitreous chamber of the eye. ¹ Similar eye elongation associated with myopia can readily be induced in developing animal eyes, including those of monkeys, ² tree shrews, ³ marmosets, ⁴ and, more recently, guinea pigs, ⁵ by depriving them of form vision. The induced eye elongation is associated with remodelling of the sclera with decreased proteoglycan synthesis, ⁶ a net loss of scleral tissue through reduced collagen synthesis and increased degradation, ⁷ changes in the composition of the sclera, ⁷ and changes in the biomechanics of the sclera. ⁸ In the tree shrew, experimental myopia is associated in particular with an increase in the scleral creep rate (slow extension of a tissue under constant load over time), and it has been proposed that this accounts for eye elongation by reducing scleral resistance to the expansive forces of normal intraocular pressure (IOP). ^8,9  

Mammalian sclera is a typical connective tissue that comprises a single fibrous layer with an abundant extracellular matrix composed almost entirely of collagen and containing few cells. ¹⁰ The sclera shows a typical viscoelastic biphasic response to applied stress, demonstrating an immediate viscoelastic expansion, followed by slow tissue creep on increasing the IOP in vitro in whole enucleated eyes. ^11,12 In isolated scleral strips in vitro, the scleral creep rate is seen to decrease with age, but to increase rapidly with the induction of myopia. ⁸ However, the raised creep rates cannot be explained by scleral thinning and must be due to some other change(s) in scleral composition. ⁸ Furthermore, a correlation has been shown between the creep rate and axial elongation in form-deprived and control eyes. ⁹  

The mechanical behavior of a curved piece of tissue tested in a uniaxial manner in vitro is likely to be different from the behavior seen in vivo. ¹³ Several studies have been undertaken to investigate the in vivo ocular biomechanical compliance (change in axial length for a given change in IOP) response of ocular tissue. ^{14 –16} Surprisingly, when IOP is increased, normal tree shrew eyes undergo viscoelastic expansion as expected, but when the increased IOP is maintained over 1 hour, the eyes shorten—a decrease in axial length that remains when the IOP is returned to baseline levels. ¹⁶ Chick eyes under the same conditions elongate as predicted by tissue strip studies. ¹⁶ The act of ocular shortening under increased IOP may be due to the activation of some contractile element(s) within the choroid or sclera. 

Although the changes occurring in the extracellular matrix of the sclera in myopia have been extensively studied, changes that may be occurring in the cellular components of the sclera (fibroblasts and myofibroblasts) have received little attention. ¹⁷ Myofibroblasts are a specialized, differentiated form of fibroblast that express α-smooth muscle actin (α-SMA), which provides them with contractile potential. ¹⁸ Characterization of the myofibroblast population in mammalian sclera has thus far been very limited, and little is known of the changes to the myofibroblast population in myopia. ^16,19 It is possible that these cells are responsible for the ocular shortening seen over time in the tree shrew with increased IOP ¹⁶ and the reduction in axial length sometimes observed in tree shrews recovering from induced myopia. ²⁰  

One purpose of this study was to examine the effects of induced myopia on the in vivo stretch characteristics of the sclera induced by acute elevation of IOP. Use of this in vivo method had the advantage that the sclera was stretched multiaxially under physiological conditions with the cell populations, such as myofibroblasts, able to respond naturally within a three dimensional matrix compared with a one-dimensional response with uniaxial stretching. A second equally important purpose was to characterize the effect of induced myopia on the myofibroblast population in the sclera, in particular whether the number and distribution of scleral myofibroblasts were altered in induced myopia. Experiments were conducted in the guinea pig, because this is a viable ⁵ and widely available mammalian model of myopia. 

Methods

Pigmented guinea pigs (Cavia porcellus) were weaned at 1 week of age and housed in groups in a temperature-controlled room in open-topped cages, on a 12-hour light–dark cycle (nominal light level 350 lux) with free access to food and water. Animal procedures complied with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Animal Ethics Committee of The University of Auckland. 

Development of Form Deprivation Myopia

Myopia was induced in 29 guinea pigs by monocular deprivation (MD) of form vision with a translucent occluder (25% transmittance) over one eye for 2 weeks from 1 week of age (i.e., within the active emmetropization period in guinea pigs). ⁵ The fur around the eye was trimmed, a small annulus of self-adhesive tape was attached to the fur, a Velcro ring was attached to the tape, and the diffuser was mounted to a matching Velcro ring. The deprived eye (right or left) was randomized among animals to reduce measurement artifacts. In addition, nine age-matched animals were raised in the same conditions, with both eyes receiving normal visual input. 

Measurement of Refractive Status

Cycloplegic refractions were made in awake animals using an infrared optometer (model RM100; Topcon, Tokyo, Japan), 20 minutes after instilling 2 drops of 1% atropine (Sigma Pharmaceuticals Limited, South Croydon, VIC, Australia) into each eye. The refractive status was evaluated at the 0°, 45°, 90°, and 135° meridians, and the evaluation was repeated four times in each eye. The best-fitting sinusoid was plotted to the data points (Solver function in Excel; Microsoft, Redmond, WA), from which the spherocylindrical refraction was determined and converted to power vectors (represented as the mean sphere, M, and the cylinder expressed as the two components of the Jackson crossed cylinder, J ₀ and J ₄₅), ²¹ for averaging the four refractions in each eye. 

Measurement of Axial Components

Guinea pigs were anesthetized with a modified protocol of Chaib et al., ²² consisting of premedication with intraperitoneal (IP) diazepam (5 mg/kg) 10 minutes before the main anesthetic (52.5 mg/kg ketamine HCl with 3.5 mg/kg xylazine IP). Maintenance doses of ketamine/xylazine were given to maintain areflexia as needed. The head was secured in the normal upright position by a nontraumatic guinea pig head holder (Narishige, Tokyo, Japan). An eyelid retractor that did not contact the cornea was used to allow better access for the ultrasound probe. A-scan ultrasound was performed with a 15-MHz ultrasound probe (Panametrics-NDT V313; Olympus NDT, Inc., Center Valley, PA), driven by an amplifier (Panametrics 5073; Olympus NDT, Inc.). Saline was slowly metered into a 17-mm stand-off attached to the probe to act as a coupling fluid. Anterior chamber depth (ACD; anterior cornea to anterior lens), lens thickness (LT; anterior to posterior lens), vitreous chamber depth (VCD; posterior lens to anterior retina), and axial length (AXL; anterior cornea to anterior retina), were calculated using published speeds for ultrasound of the guinea pig eye. ⁵ Each measurement consisted of the average of at least 10 independent A-scan ultrasound recordings. 

Measurement of Ocular Biomechanical Compliance

In vivo ocular biomechanical compliance measures were successfully completed for both eyes of eight MD animals and four normal animals. Although refractive status measurements were made on 38 animals, the number of animals for which biomechanical results were obtained is small because of the complex nature of the biomechanical measures, coupled with the difficulty in chronically anesthetizing juvenile guinea pigs. After baseline A-scan AXL measures were obtained, the eyelids were removed to expose the globe, and the superior conjunctiva was removed to expose the sclera. A 27-gauge needle, connected via a Ringer's trap to a mercury sphygmomanometer was inserted into the vitreous chamber through the superior equator at an angle of 45° to avoid contact with the lens, as described elsewhere in the tree shrew. ¹⁶ The pressure within the system was set to 15 mm Hg, the average normal IOP in guinea pigs. ²³ Ultrasound measures were taken immediately after the IOP was set to 15 mm Hg and again after a 20-minute stabilization period at 15 mm Hg. The IOP was then rapidly increased to 50 mm Hg over a 1-minute period. This pressure was chosen because pilot experiments showed that pressures above this value (e.g., 100 mm Hg used previously in tree shrew ¹⁶ ) produced widely variable results. The immediate viscoelastic expansion of the eye was examined by taking ultrasound readings as soon as the IOP had reached 50 mm Hg, and the slow-creep response was examined by maintaining the IOP at 50 mm Hg for 1 hour and taking ultrasound readings every 10 minutes. Viscoelastic recovery was observed by rapidly reducing the IOP to 15 mm Hg and immediately making ultrasound recordings, and repeating the recordings after a 20-minute stabilization period at 15 mm Hg. The MD eye was always measured first. The animals were then killed with a single 0.3-mL IP dose of 300 mg/mL pentobarbitone sodium (National Veterinary Supplies, Auckland, New Zealand). 

Tissue Preparation and Immunohistochemistry

The eyes were dissected immediately after death, the superior aspect was marked with a suture, the globe was cleaned of extraneous tissue, and an incision was made just posterior to the limbus, circumferentially from 1 to 11 o'clock. The eyes were fixed in 4% paraformaldehyde in 0.1 M phosphate-buffered saline (PBS) for 1 hour and then washed three times in PBS at room temperature, before overnight cryoprotection in 30% sucrose solution in PBS at 4°C. A transverse scleral strip approximately 3 mm in height, centered on the optic nerve head (ONH) and spanning from the nasal to the temporal ora serrata, was removed from each eye. The strips were frozen in a 1:1 solution of 30% sucrose in PBS and optimal cutting temperature (OCT) compound (Leica Microsystems, Germany). Sections (20 μm thick) were collected on slides (Superfrost Plus; Biolab Scientific, Auckland, New Zealand), using a cryostat (CM3050 S; Leica Microsystems, Germany). 

The collected sections contained the ONH to ensure the same general location for cell counting. Tissue sections were preincubated with 10% normal goat serum in PBS containing 0.1% Triton X-100 for 1 hour at room temperature. Mouse monoclonal [1A4] α-SMA antibody directly conjugated with fluorescein isothiocyanate (FITC) secondary antibody (GeneTex, Irvine, CA) was used to detect myofibroblasts. The α-SMA antibody was diluted 1:500 in PBS with 0.1% Triton X-100 and added to the sections overnight at 4°C in the dark. The sections were then washed three times in PBS at room temperature in the dark. 4′,6-Diamidino-2-phenylindole (DAPI; Invitrogen, Carlsbad, CA) was then used to label DNA in all cell nuclei. The DAPI was diluted 1:300 in PBS with 0.1% Triton X-100 and added to the sections for 2 minutes at room temperature in the dark. The sections were then washed twice in PBS for 5 minutes and coverslipped in glycerol/PBS solution (CitiFluor; Alltech, Auckland, New Zealand). 

Cell Counting

Overlapping photographs were taken of the sclera from the nasal to the temporal ora serrata, with a 40× oil-immersion lens (Leica Microsystems, Wetzlar, Germany). Montages of the photos were made (Photoshop; Adobe Systems Inc., San Jose, CA), and imported into Image J (developed by Wayne Rasband, National Institutes of Health, Bethesda, MD; available at http://rsb.info.nih.gov/ij/index.html) for cell counting. Cell counts of myofibroblasts (α-SMA) and the total scleral cell population (DAPI) were made by hand for the entire scleral area, with the Image J Cell Counter tool. To ensure that the cell count area was proportionally the same for each eye regardless of AXL, only the cells within a 50° angular subtense from the edge of the ONH in both the nasal and temporal directions were analyzed. The sclera was assumed to form the arc of a circle, the radius of which was equivalent to half the AXL of the eye being counted. The calculation of the X,Y origin point (0,0) was made by constructing a line normal to the center of the ONH, extending half the AXL of the eye (Fig. 1). The X,Y coordinates from each cell were used to calculate polar coordinates (r,θ) to determine the cell location within the 50° angular regions. ²⁴ The Image J Line Selection and Measurement tools were used to determine the dehydrated scleral thickness at 10° intervals. The sclera was also divided in half into inner (closest to the choroid) and outer (closest to Tenon's capsule) regions to investigate cell gradients through the sclera. The number of eyes examined was limited to 30 in total (10 per treatment group) because of the significant amount of time required to undertake the anatomic investigations. 

Figure 1.

 

An example of a reconstructed section through the guinea pig sclera. The calculation of the origin point (0,0) was made by constructing a line normal to the center of the ONH extending half the AXL of the eye. Also shown is an example of the polar coordinates (r,θ) for a single cell within the sclera described by the radius, r, and the angular subtense, θ.

Statistical Analysis

Data Analyses.

Linear mixed model analysis (SAS Institute Inc., Cary, NC) using the PROC MIXED function allowed both random effects (animal number and litter number) and fixed effects (treatment: deprived, control, normal) to be incorporated into the analysis. ²⁵ The model also accounted for the use of a single subject for multiple measurements (deprived and control eye of the same subject). The overall effect of treatment (in this case form deprivation, contralateral control, and age-matched normal) was assessed by using the type 3 tests of fixed effects. When these results are presented, the F-statistic and degrees of freedom (df) are given along with the P value. The pair-wise between-treatment comparisons were made in the model by using the differences of least squares means. When these results are presented, the t-statistic and df are given along with the P value. 

Number of Subjects.

As the guinea pig is a relatively new model for myopia development, some of the myopia induction findings are the result of piloting this new method in our laboratory. Refractive status data are presented for all 38 animals in which this was successfully completed. Of these animals, 20 were used to perfect the anesthetic protocol for prolonged deep anesthesia and to determine the ideal ocular biomechanical technique, 6 had biomechanical failures in which data were collected from one eye only because of adverse events, and 12 represent the biomechanical successes presented herein (8 MD animals and 4 age-matched normal animals, giving 8 eyes per treatment group). Fifteen of these animals were also used for the anatomic data collection, providing 10 eyes per treatment group. Statistically, as no within-treatment differences were found, the small sample sizes can be considered to be serial successes. This means that valid predictions about the population as a whole can be made, and in this case, approximately 70% of the population would be expected to share the presented findings at the 0.05 confidence level. ²⁶  

Results

Refractive Status and Axial Components

The average mean sphere refraction in MD eyes (−0.55 ± 0.51 D; n = 29, mean ± SEM) was significantly more myopic than that in either the contralateral control eyes (+3.51 ± 0.30 D; n = 29, t = −13.19, df = 37, P < 0.001) or age-matched normal eyes (+3.21 ± 0.27 D; n = 18, t = −4.48, df = 13.5, P < 0.001). There was no significant difference between control and normal eyes (t = 0.24, df = 13.5, P = 0.817). The average relative MD mean sphere (refraction of deprived eye minus control eye) was −4.06 ± 0.35 D (n = 29), which was significantly more myopic than the average relative normal mean sphere (difference between eyes) of +0.00 ± 0.19 D (n = 9; t = −6.43, df = 36, P < 0.001). Linear regression showed a moderate negative correlation between the mean sphere and the VCD (R ² = 0.430, P < 0.001), implying that the induced myopia was due mainly to VCD elongation (Fig. 2). The VCD of deprived eyes was elongated relative to that in the control eyes by 169 ± 49 μm (t = 3.62, df = 37, P = 0.001; n = 29). The absolute VCDs of control and normal eyes were not significantly different from each other (t = 1.67, df = 16.1, P = 0.115). 

Figure 2.

 

Linear regression correlation between difference in refraction and difference in VCD (deprived eye minus control eye). Squared Pearson correlation (R ²) = 0.43. n = 38, comprising 29 form-deprived and 9 age-matched normal guinea pigs.

Ocular Biomechanical Compliance

Immediate Viscoelastic Expansion.

On increasing the IOP to 50 mm Hg from 15 mm Hg, there was a significant increase in the ACD of all three groups (deprived, control, and normal eyes; Fig. 3A, Table 1). The ACD of the deprived eyes increased significantly more than that of the normal, but not the control, eyes. Conversely, there was a significant decrease in LT (Fig. 3B, Table 1) on increasing the IOP in all three groups, but there were no significant differences in LT response between groups. The absolute change in ACD (deprived, +50 ± 8 μm; control, +41 ± 6 μm; and normal, +31 ± 4 μm) and LT (deprived, −30 ± 4; control, −38 ± 7; and normal, −37 ± 4 μm) showed an approximately equal but opposite response to increased IOP. 

Figure 3.

 

Percentage change in ocular component dimensions with time on increasing IOP from 15 mm Hg (baseline; BL) to 50 mm Hg for 1 hour in form-deprived, contralateral control, and age-matched normal eyes. (A) ACD, (B) LT, (C) VCD, and (D) AXL. IOP was increased to 50 mm Hg at time T0 and was maintained at this pressure for 60 minutes (between vertical dashed lines). IOP was returned to 15 mm Hg at time P0 and was maintained at this pressure for 20 minutes. n = 8 eyes per group. Error bars, SEM. *Normal eyes significantly different from deprived eyes; #normal eyes significantly different from control eyes.

Table 1.

 

Immediate Viscoelastic Expansion Statistics within and between Treatment Groups

	ACD					LT					VCD
	Measure 1	Measure 2	t-Statistic	df	P	Measure 1	Measure 2	t-Statistic	df	P	Measure 1	Measure 2	t-Statistic	df	P
Deprived*	Baseline (0)	3.75 ± 0.64	−7.44	30.4	<0.001	Baseline (0)	−0.77 ± 0.10	5.77	39.7	<0.001	Baseline (0)	2.65 ± 1.24	−2.54	28	0.017
Control*	Baseline (0)	3.34 ± 0.51	−6.62	30.4	<0.001	Baseline (0)	−0.99 ± 0.18	7.45	39.7	<0.001	Baseline (0)	3.07 ± 0.74	−2.93	28	0.007
Normal*	Baseline (0)	2.52 ± 0.32	−5.45	10	<0.001	Baseline (0)	−1.04 ± 0.11	7.90	17.9	<0.001	Baseline (0)	−0.42 ± 1.10	0.40	10	0.696
Dep. vs. Con.†	3.75 ± 0.64	3.34 ± 0.51	0.76	22	0.454	−0.77 ± 0.10	−0.99 ± 0.18	1.66	22	0.110	2.65 ± 1.24	3.07 ± 0.74	−0.39	22	0.701
Dep. vs Nor.†	3.75 ± 0.64	2.52 ± 0.32	2.56	19.5	0.019	−0.77 ± 0.10	−1.04 ± 0.11	1.58	5.49	0.171	2.65 ± 1.24	−0.42 ± 1.10	2.95	17.3	0.009
Con. vs Nor.†	3.34 ± 0.51	2.52 ± 0.32	1.70	19.5	0.105	−0.99 ± 0.18	−1.04 ± 0.11	0.22	5.49	0.836	3.07 ± 0.74	−0.42 ± 1.10	3.35	17.3	0.004

 The percentage changes in ACD, LT, and VCD relative to the 15 mm Hg baseline are given.

*  Within-treatment comparisons of the axial dimensions between the 15-mm Hg baseline (measure 1) and immediately after increasing the IOP to 50 mm Hg (measure 2).

†  Between-treatment comparisons of the axial dimensions immediately after increasing the IOP to 50 mm Hg. Deprived and control, deprived and normal, and control and normal treatment groups are compared as measure 1 and measure 2, respectively.

On increasing the IOP, there was a significant increase in the VCD of deprived (+2.65% ± 1.24%) and control (+3.07% ± 0.74%) eyes (Fig. 3C, Table 1), but normal eyes did not differ significantly from baseline (−0.42% ± 1.10%). The percentage change in VCD was significantly different between the normal eyes and both the deprived and control eyes, whereas there was no significant difference between the deprived and control eyes. The equal but opposite response of ACD and LT to increased IOP meant that most of the AXL changes were accounted for by changes in VCD. As such, the AXL response (Fig. 3D) was similar to the VCD response. 

Slow-Creep Response.

Over the 1 hour of increased IOP, there was a significant increase in the ACD in all treatment groups (Fig. 3A, Table 2). The ACD showed a significant positive slow-creep rate over time (increased depth with time), although the significant changes occurred at a slower rate than the 10-minute measurement intervals used. The effect of treatment on the slow-creep response of the ACD was minimal. LT showed a negative slow-creep response, with a significant decrease in thickness over the hour of increased IOP in all three groups (Fig. 3B, Table 2). There were no significant differences between groups at any time point, indicating that treatment had no effect on the slow-creep response of the lens. 

Unlike the ACD and LT, there was no significant elongation or shortening of the VCD over the 1 hour of increased IOP in the deprived, control, or normal eyes (Table 2). As the normal eyes showed significantly less immediate viscoelastic expansion, they were shorter than the deprived or control eyes at the beginning of the hour of increased IOP (Fig. 3C). Although there were no significant within-treatment creep responses over time, Figure 3C suggests a tendency toward elongation of the VCD in the normal eyes over time. The slow-creep response of the AXL to increased IOP was much the same as for the VCD (Fig. 3D). 

Immediate Viscoelastic Recovery.

There was minimal recovery of ACD on reducing IOP from 50 to 15 mm Hg. The only significant decrease in ACD was in the control eyes between the final 50 mm Hg reading and the 15 mm Hg reading after 20 minutes of decreased IOP (Fig. 3A; t = 2.37, df = 48.8, P = 0.022). LT also showed minimal recovery on reducing the IOP. Again, the only significant increase in LT was in the control eyes between the final 50 and the 15 mm Hg reading after 20 minutes (Fig. 3B; t = −2.81, df = 52.8, P = 0.007). The control eyes thus showed an approximately equal but opposite recovery response between the ACD and LT. There were no significant within- or between-group recovery responses for either the VCD or the AXL. 

Overall Effect of Increased IOP.

To ascertain the overall response of the eye in vivo to increased IOP, the baseline and final ocular biometry measures were compared. The ACD showed significant elongation that did not recover on return to baseline pressure in the deprived (t = −4.96, df = 30.2, P < 0.001), control (t = −3.37, df = 30.2, P = 0.002), or normal (t = −3.78, df = 10, P = 0.004) eyes. In the case of the lens, only the deprived eye showed significant thinning that did not recover on returning the IOP to baseline (t = 2.58, df = 28.1, P = 0.016). There was no significant difference in the VCD or AXL in any of the three groups between the baseline and final measurements. Although there was no significant elongation or shortening of the VCD, the small changes that did occur combined to return the component lengths to baseline levels. The creep and viscoelastic recovery responses in vivo thus appear to be slow responses occurring over longer periods than those used in this study. Furthermore, there was significant variability in the ultrasound measurements that may have masked the true creep properties of the tissue, and the interpretation of the results must therefore be approached with caution. As each ultrasound measurement was the average of at least 10 independent A-scan ultrasound recordings, additional recordings would not have significantly reduced the between recording (within measurement) variability. 

Table 2.

 

Slow Creep Response Statistics within Treatment Groups

	ACD					LT					VCD
	50 mm Hg		t-Statistic	df	P	50 mm Hg		t-Statistic	df	P	50 mm Hg		t-Statistic	df	P
	Initial	60 Minutes				Initial	60 Minutes				Initial	60 Minutes
Deprived	3.75 ± 0.64	7.90 ± 0.70	−6.84	106	<0.001	−0.77 ± 0.10	−2.41 ± 0.22	7.04	116	<0.001	2.65 ± 1.24	2.47 ± 0.95	0.13	132	0.900
Control	3.34 ± 0.51	7.72 ± 0.45	−7.22	106	<0.001	−0.99 ± 0.18	−2.48 ± 0.20	6.38	116	<0.001	3.07 ± 0.74	2.16 ± 1.13	0.64	132	0.526
Normal	2.52 ± 0.32	5.81 ± 0.70	−9.29	60	<0.001	−1.04 ± 0.11	−1.91 ± 0.35	5.54	60	<0.001	−0.42 ± 1.10	−0.25 ± 1.89	−1.17	60	0.245

 Data are the percentage changes in axial dimensions between the initial and 60-minute measurements.

Anatomy

Posterior 100°.

The average total number of cells in the posterior 100° of the scleral sections was not significantly different between groups (deprived, 2209 ± 107 cells; control, 2296 ± 90 cells; and normal, 2212 ± 76 cells, n = 10 in all groups. F = 0.38, df = 4.03, P = 0.704. Fig. 4). A high proportion of the total cell population in all groups was found to be myofibroblasts (deprived, 67.3% ± 2.9%; control, 64.5% ± 3.5%; and normal, 59.2% ± 1.4%; Figs. 4A, 4D, 4G), and these proportions were not significantly different from one another. There was an even division of the total number of cells and myofibroblasts between the nasal and temporal 50° regions in deprived, control, and normal eyes (Table 3). 

Figure 4.

 

Typical immunohistochemical staining patterns for form-deprived (A–C), contralateral control (D–F), and age-matched normal (G–I) guinea pig sclera. Sections are double labeled with DAPI for all cell nuclei (B, E, H) and with α-SMA for myofibroblast stress fibers (A, D, G). Co-localization of DAPI and α-SMA is also shown (C, F, I). Arrows: some of the myofibroblasts in the tissue sections; asterisks: some of the cells that are not myofibroblasts (not positive for α-SMA); isolated arrowheads: the borders of the sclera. BV, choroidal blood vessel. The choroid had separated from the sclera in normal eyes (G–I) and was not present in the images. Scale bar, 50 μm.

Table 3.

 

Total Number of Cells and Myofibroblasts within Treatment Groups

	Total Cells					Myofibroblasts
	Nasal	Temporal	t-Statistic	df	P	Nasal	Temporal	t-Statistic	df	P
Deprived	1075 ± 71	1134 ± 55	−0.95	41	0.346	760 ± 76	732 ± 40	0.43	41	0.669
Control	1136 ± 62	1106 ± 39	−0.39	41	0.700	717 ± 78	776 ± 59	−0.92	41	0.364
Normal	1136 ± 32	1076 ± 58	0.96	41	0.343	678 ± 39	627 ± 31	0.80	41	0.430

 Data are cell counts and differences of least squares means in the 50° regions of the nasal and temporal sections.

Distribution with Eccentricity.

Division of the cell counts into 10° regions of the sclera (Fig. 5) revealed that the temporal sclera between 40° and 50° had fewer total cells (averaged over the three groups) than any other region (from P < 0.001 to P = 0.021). When the eccentricities were analyzed between groups, only the nasal region between the ONH and 10° nasal had significantly fewer cells in deprived eyes than either normal (t = −3.51, df = 11.7, P = 0.004) or control (t = −1.97, df = 257, P = 0.049) eyes (which themselves were not significantly different; t = −2.08, df = 11.7, P = 0.060). Significantly more cells were present in the inner half of the sclera than in the outer half of all three regions examined (posterior 100°, nasal 50°, and temporal 50°) in each of the three groups (P ≤ 0.001 in all cases). Of interest, no significant differences were present for the myofibroblasts. 

Figure 5.

 

Effect of eccentricity on the total cell (A) and myofibroblast (B) populations in the posterior 100° of the sclera. Nasal 10° region is equivalent to human myopic scleral crescent location. Error bars, SEM; n = 10 eyes per group. Significant differences in cell numbers are indicated on the graph (* P = 0.049; ** P = 0.004).

Scleral Thickness.

The 10° to 30° region of the temporal sclera was significantly thicker than the equivalent region of the nasal sclera, but with greater eccentricity, the temporal sclera became significantly thinner (Fig. 6). There was no significant effect of treatment on scleral thickness with eccentricity (F = 1.55, df = 257, P = 0.073). The normal eye nasal 10° and 20° scleral thickness measurements, although appearing thicker, were not significantly different from those in the deprived eyes (t = −1.72, df = 8.06, P = 0.124 and t = −1.79, df = 8.06, P = 0.111) and control eyes (t = −2.05, df = 8.06, P = 0.074 and t = −1.87, df = 8.06, P = 0.098). 

Figure 6.

 

Average dehydrated scleral thickness every 10° nasally and temporally from the ONH in form-deprived, contralateral control, and age-matched normal eyes. n = 10 eyes per group. Error bars, SEM. No differences are significant.

Scleral Cell Densities.

There were no significant differences in total cell densities in the posterior 100° (F = 2.55, df = 4.04, P = 0.192), nasal 50° (F = 4.34, df = 4.18, P = 0.096), or temporal 50° (F = 0.55, df = 27, P = 0.583) regions. Myofibroblast densities likewise showed no significant difference between treatments in any of the three regions examined (F = 0.68, df = 2.93, P = 0.573; F = 1.10, df = 13.5, P = 0.360; and F = 1.34, df = 13.5, P = 0.295, respectively). 

Scleral Tensile Stress.

A theoretical model of the tensile stress within the sclera based on Laplace's law, previously used to model susceptibility to glaucomatous damage, ²⁷ was performed.  

 

where τ is the tensile stress in the sclera (in kilopascals/square millimeter), IOP is the intraocular pressure, r is the radius of the globe, and h is the thickness of the sclera. The assumptions made in the model were that the globe approximated a sphere and that the scleral thickness was uniform throughout the globe. The IOP was taken as 15 mm Hg for each eye, the radius was half the AXL of the eye measured at 15 mm Hg, and the scleral thickness was the average of the 10 measurements made in the eye. The resulting scleral tensile stress modeled for each eye was plotted versus the myofibroblast density (Fig. 7). The scleral tensile stress in normal eyes and control eyes showed good correlation with the number of myofibroblasts. In the form-deprived eyes, however, there was no correlation between scleral tensile stress and either the total number of cells or the number of myofibroblasts. 

Figure 7.

 

Correlation between myofibroblast density in the sclera and scleral tensile stress in deprived (R ² = 0.0003), control (R ² = 0.7937), and normal (R ² = 0.4814) eyes. Correlation over all treatment groups (not shown), R ² = 0.3001; n = 8 eyes per group.

Discussion

Myofibroblasts have been reported to be present in human, monkey, and tree shrew sclera. ^16,19 This study confirms that they are also present in guinea pig sclera. A surprisingly high proportion of the total cell population was found to be myofibroblasts (63.7% ± 1.7% averaged over all three treatments). However, there was no significant difference in the total number of cells and myofibroblasts or the percentage of the population that were myofibroblasts between any of the groups (form-deprived, contralateral control, and age-matched normal) within the posterior 100° of the sclera. 

Of interest, when the number of cells were examined with respect to eccentricity, the nasal region between the ONH and 10° nasal possessed significantly fewer cells in deprived eyes than in normal and control eyes. Because of the temporal insertion of the optic nerve into the guinea pig eye, the nasal sclera adjacent to the ONH in guinea pigs can be considered analogous to the temporal sclera in humans, which is the region where a myopic crescent (peripapillary atrophy) usually forms in human high myopia. ^{28 –30} In human myopic eyes, it has been shown that there is decreased blood flow in the region of myopic crescent formation, ³¹ and it is possible that this decrease in blood supply may be responsible for the decrease in cell numbers. No difference was seen in the number of myofibroblasts with eccentricity between any of the groups. 

The number of myofibroblasts present in the sclera could be expected to correlate with the biomechanical response of the tissue because of their contractile potential. As there were no differences in the number of myofibroblasts between treatments, then there would be no difference expected in the viscoelastic and creep responses of the sclera if this correlation existed. However, on increasing the IOP, there was a significant increase in the VCD of only the form-deprived and control eyes, whereas the normal eyes were not significantly different from baseline. This is in contrast to the normal eyes of both chicks and tree shrews, which showed increased VCD with increases in IOP. ¹⁶ A possible reason for this difference in response is that the pressure used in the chicks and tree shrews was 100 mm Hg compared with only 50 mm Hg in the guinea pigs. In normal guinea pig eyes it is possible that the 50 mm Hg pressure was too low to cause expansion of the posterior sclera to increase the VCD, but was sufficient to expand the equatorial sclera, which has been shown to be more extensible in tree shrews. ³² There were no differences in posterior scleral thickness between treatment groups, and so scleral thickness is unlikely to be a contributing factor to the differential responses in guinea pigs. 

Changes in scleral composition, myofibroblast focal adhesions, or α-SMA expression may account for the difference in response between deprived and normal eyes, as there were no significant differences in number of cells between treatments. That the control eye response was the same as the deprived eye response, despite the control eyes having no significant difference in VCD compared with normal eyes, is interesting. This finding could be taken as evidence of an interocular yoking signal between deprived and control eyes, which has been observed with induced myopia previously. ^{5,33 –35} The control eye biomechanical properties may be altered by a yoked signal, enough to change their response to increased IOP, but not enough to induce significant changes in axial growth compared to normal eyes. 

Although there were no significant within-treatment differences in the VCD over the hour at 50 mm Hg, the deprived and control eyes showed nonsignificant decreases in the VCD and the normal eyes showed nonsignificant increases in the VCD. The VCD of normal eyes were shorter than the deprived and control eyes on increasing the IOP to 50 mm Hg, but after 30 and 50 minutes of increased IOP, the normal eyes were not significantly different from the control or deprived eyes, respectively. These findings suggest that deprived and control guinea pig eyes behaved in a manner similar to normal tree shrew eyes (shortening with time), although not reaching within-treatment significance due possibly to the lower IOP increase used. ¹⁶ The creep response of normal guinea pig eyes appeared similar to that in chicks (increasing with time). ¹⁶ Deprived and control eyes appeared to mainly undergo immediate viscoelastic expansion, whereas normal eyes appeared to mainly undergo slow creep expansion. 

The slower response seen in normal eyes indicates that scleral changes may make the deprived (and control) eyes more susceptible to changes in IOP. Although the control eye did not develop myopia, it had the same biomechanical response as the form-deprived eye, which did develop myopia, which suggests that there may be two signals that combine in the development of myopia. The first may be a signal resulting in, say, reduced contraction of myofibroblasts, with the second signal resulting in increased scleral remodelling. If the first signal was yoked interocularly, whereas the second signal was a purely local message, then the change in biomechanical properties of the control eyes compared with the normal eyes, without myopia also developing, could be explained. 

The formation of myofibroblasts from fibroblasts is reliant on the induction and maintenance of tensile force within the extracellular matrix, with loss of force returning the cells to an undifferentiated form. ^36,37 The normal and control eyes showed good correlation between scleral tensile stress and the density of myofibroblast numbers. This correlation was completely lost in form-deprived eyes. These findings suggest that there is some mechanosensory component based on the stress experienced by the sclera that helps in determining the density of cells in the tissue, and that the mechanosensory signal between the extracellular matrix and the cells may be reduced or absent in form-deprived eyes. 

The mechanosensors in fibroblasts and myofibroblasts are the focal adhesions. ³⁸ The loss in correlation between the stress in the extracellular matrix and the myofibroblasts in deprived eyes may be due to a downregulation or dysfunction of, for example, the focal adhesions themselves, fibronectin, or focal adhesion kinase. ^{39 –41} A reduction in the main component of focal adhesions, integrin subunits, has already been shown after as little as 24 hours of form deprivation in tree shrews. ⁴² Changes to the focal adhesions, fibronectin, or focal adhesion kinase would reduce the signal of increased stress in the tissue getting to the cells, thus reducing the α-SMA recruited to stress fibers in the cells to levels below that required. Moreover, there would be a reduction in the transmission of the force generated by myofibroblasts into the surrounding matrix. Both these factors could lead to the myofibroblasts' inability to maintain the required tension in the tissue to counteract the forces placed on it, leading to a slow creep of the tissue and eventual axial elongation. 

A study in the tree shrew has shown increased levels of α-SMA mRNA expression in eyes with developing myopia, although these increases were not significant, and protein expression levels were not quantified. ⁴³ As myofibroblasts, like fibroblasts, are responsible for the normal turnover of collagen, then to prevent rapid expansion of the tissue on degrading the matrix with matrix metalloproteinases, it is likely that the myofibroblasts act as scaffolding, taking up the role of the collagen until the new matrix is synthesized. ^38,44,45 Reduced contractility of myofibroblasts could lead to an inability to maintain the matrix position, resulting in a small amount of expansion that over time produces axial elongation. This process could account for the active remodelling in developing myopia. 

The dehydrated scleral thickness was not significantly affected by form deprivation, and this is in keeping with ultrasound results that have found no significant in vivo differences in scleral thickness in guinea pigs. ⁵ This finding is contrary to that in monkeys, ⁴⁶ tree shrews, ⁴⁷ and humans. ^28,48 This may represent an interspecies difference in the response of the sclera to form deprivation myopia, or it may be a response time factor, as tree shrew scleral thinning occurs after as little as 12 days ⁴⁷ while slower growing monkeys develop similar degrees of myopia after 23 to 38 months. ⁴⁶  

The fact that the number of myofibroblasts is not altered by the induction of form deprivation myopia suggests that some other change(s) in the scleral cells or tissue occurs in developing myopia. The number of focal adhesions and their differentiation state, α-SMA expression levels, and fibronectin expression would all be of interest in terms of cell anatomy and physiology and myopia development. The effect of atropine and pirenzepine on the cell populations during the inhibition of myopia development would also be of great interest. Myofibroblasts are an enigmatic cell type, and determining their role in the sclera necessitates much greater investigation. 

Footnotes

 Supported by a grant from the New Zealand Optometric Vision Research Foundation.

Footnotes

 Disclosure: S. Backhouse, None; J.R. Phillips, None

The authors thank Keely Bumsted O'Brien, Monica Acosta, Silke Fuchs, and Sarah Ready for help with the immunohistochemistry; Andrew Collins, Nicola Anstice, and Jo Black for help with the biomechanical measurements; and Kerry King for writing the ultrasound image-capturing program.