The lens is an avascular tissue largely composed of elongated lens fiber cells responsible for lens function whose anterior surface is overlaid with a sheet of lens epithelial cells (Cvekl and Tamm, 2004; Cvekl and Duncan, 2007). The lens capsule is a contiguous basement membrane that completely encloses the ocular lens (Zampighi et al., 2000). The capsule is composed of extracellular matrix molecules such as collagen IV, laminin, entactin, and heparan sulfate proteoglycans (Cammarata et al., 1986; Mohan and Spiro, 1986; Timpl, 1989; Yurchenco and Schittny, 1990; Hirsch et al., 2001; Kelley et al., 2002). Interaction of capsular molecules with cell membrane bound integrins found at the basal surface of both lens epithelial and fiber cells provides signals to maintain normal growth, survival, and maintain the epithelial cell phenotype in the anterior region of the lens (Zuk et al., 1989; Lovicu and McAvoy, 2001). Additionally, the capsule provides structural support for the lens and the eye (Sivak, 1980; Koretz and Handelman, 1982; Seland, 1992; Krag and Andreassen, 1996) and allows the transmission of accommodative forces to the lens (Fincham, 1937; Koretz and Handelman, 1982) through interdigitation of the fibrillin components of the zonule fibers throughout its matrix in the equatorial region (Mir et al., 1998). The capsule also serves as a semipermeable barrier that allows the transit of macromolecules into and out of all regions of the lens (Boyle et al., 2002; Sabah et al., 2004, 2005).
The anterior capsule is synthesized by the lens epithelial cells while the posterior capsule seems to be produced by the lens fiber cells (Parmigiani and McAvoy, 1991) and fibrillin within the zonule integration region is produced by the ciliary body (Hanssen et al., 2001). Turnover of matrix molecules in both the anterior and posterior capsules is very slow compared with other basement membranes and is measured in months (Young and Ocumpaugh, 1966) or years (Fisher and Pettet, 1972; Seland, 1976) as compared with the hours observed for glomerular (Beavan et al., 1989) and alveolar (Dunsmore et al., 1995) basement membranes. The capsule is also extremely thick compared with other basement membranes. For example, the adult human lens anterior capsule is over 100 times thicker than the basement membrane of muscle capillary (Siperstein et al., 1968) and over 30 times thicker than the glomerular basement membrane (Ramage et al., 2002). However, in the adult lens capsule the thickness varies along the entire surface of the lens from the anterior to the posterior poles (Fisher and Pettet, 1972; Krag and Andreassen, 2003b; Barraquer et al., 2006). In rat embryos, the posterior capsule is thicker than the anterior (Parmigiani and McAvoy, 1989), whereas human neonatal lenses are surrounded by a capsule of almost uniform thickness (Fisher and Pettet, 1972). The anterior capsule also has been reported to continuously increase in thickness throughout much of a lifetime while the posterior capsule ceases to thicken much earlier and is 5–10 times thinner than the anterior capsule in mature lenses (Young and Ocumpaugh, 1966; Fisher and Pettet, 1972; Parmigiani and McAvoy, 1989; Krag et al., 1997; Barraquer et al., 2006). The varied thickness of the capsule has been hypothesized to play an important role in actively molding the lens during accommodation (Fincham, 1937; Fisher, 1969b; Koretz and Handelman, 1982), although this thickness difference is also seen in rodent lenses (Parmigiani and McAvoy, 1989) which do not accommodate (Sivak and Dovrat, 1983).
Lens capsule thickness has also been an important consideration when studying the mechanical characteristics of the capsule. Investigators require capsule thickness measurements to calculate the effects of disease and age on capsule tensile strength (Krag et al., 1997; Krag and Andreassen, 2003b), elasticity (Fisher, 1969a), thermal and mechanical stability (Krag et al., 1998), and to determine the energy required for the transmission of accommodative forces (Fincham, 1937; Fisher, 1969b). In one study, the thickness of the capsule was also required to calculate the diffusion rates of molecules passing through the capsule (Lee et al., 2006). Additionally, an understanding of tensile strength and capsule elasticity has been essential in the development of cataract surgical techniques (Guthoff et al., 1990; Thim et al., 1991). Thus, accurate measurement of lens capsule thickness is crucial for many practical lens capsule studies.
Ideally, lens capsule thickness should be determined using lenses that have been subjected to minimal manipulation with a method that gives excellent contrast between the lens capsule, the lens cells, and the surrounding medium. We have developed a simple and highly reproducible approach for measuring lens capsule thickness that avoids most of the experimental manipulations that could result in measurement artifacts. Using this approach, it is demonstrated that anterior capsule thickness in mice varies between inbred strains suggesting that lens capsule thickness is a quantitative genetic trait. Our data also confirm that the anterior capsule thickens with age in mice as well as varies between the anterior, equatorial, and posterior poles.
MATERIALS AND METHODS
All experiments on nonhuman samples described in this article conform to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Mice of strain 129X1/SvJ were obtained from The Jackson Laboratory (Bar Harbor, ME) and bred in house, ICR(CD-1), FVB/NHsd, and C57Bl/6NHsd mice were obtained from Harlan Sprague Dawley (Indianapolis, Indiana) and bred in house, BALB/cAnNHsd mice and outbred Sprague Dawley rats were obtained directly from Harlan Sprague Dawley. Fresh eyes obtained from 8- to 12-week-old NZW rabbits and from 56- to 120-week-old cattle were shipped to the laboratory overnight on ice packs and were measured on the day of receipt (Pel-freez Biologicals, Rogers, AR). Bovine lenses used to determine the refractive index of the anterior lens capsule were obtained from a local abattoir.
Human Lens Capsules
All human lens capsules were obtained with informed consent and Christiana Care Institutional Review Board approval as waste tissue generated at the Roxana Cannon Arsht Surgicenter, Wilmington Hospital, Wilmington, Delaware. Twenty-two capsulotomy samples from separate donors were acquired from a wide mixture of ethnic backgrounds and ranged from 55–87 years of age.
Tissue Preparation and Stabilization
Mouse and rat lenses were dissected from eyes immediately after killing the animals and placed in Medium 199 with Earle's Salts and L-glutamine (MediaTech, no. 10-060-CV) which has been earlier reported to be isotonic to the lens (Augusteyn et al., 2006). Rabbit and bovine lenses were treated similarly although the lenses were not dissected until 24–36 h after slaughter. Human anterior capsulotomy specimens were obtained during routine cataract surgeries and immediately transferred into Medium 199 after removal from the patient. These specimens were then transferred to uncoated 35-mm cover slip bottom culture dishes (MatTek Corporation, no. P35G-1.5-14-C) and allowed to equilibrate in Medium 199 supplemented with 1 mg/ml of 40-kDa FITC-dextran (Sigma-Aldrich, St. Louis, MO) and 0.2 μL/mL of the live cell permanent nuclear stain Draq-5 (Biostatus Ltd., Leicestershore, UK) for at least 10 min. Whole lenses were positioned and stabilized using rubber blocks or within plastic rings with appropriate sized holes with either the anterior, equatorial, or posterior capsule pole facing downward on the cover slip surface. Human anterior capsulotomy specimens were laid in the dish on the cover slip and stabilized under a second cover slip. In some experiments in order to compare the dependence of lens capsule thickness on media, one lens from a mouse was placed in Medium 199 while the other was placed in Hank's balanced salt solution, both containing 40-kDa FITC-dextran and Draq-5 and positioned as stated earlier.
Imaging and Capsule Thickness Measurements
Images of lenses or lens capsules loaded with FITC dextran and Draq-5 were acquired in XZ line scanning mode on a Zeiss Axiovert 100 M equipped with a LSM 510 VIS confocal microscope using a C-Apochromat 63× (NA 1.2) water immersion objective lens to minimize refractive index mismatch artifacts while imaging the lens capsule in an aqueous medium. The z-motor used in line scanning is controlled by a harmonic drive with a 50 nm minimum step-size and 100 nm reproducibility. Images were acquired with a pinhole setting of 1.0 airy units and 80-nm Z step. A 30 mW argon krypton laser and a 5mW helium neon laser (Carl Zeiss, Göttingen, Germany) were used with 500–550BP and 650LP emission filters to image FITC-dextran and Draq-5, respectively. The photomultiplier tube detector settings were adjusted as needed to provide sufficient contrast between the lens capsule, epithelial cell layer, and outside the media.
The thickness of each capsule was determined by drawing a straight line extending from the capsule-epithelial cell interface to the capsule-media interface using the Zeiss LSM Image Browser (release 3.2). The length of the line was calculated based on calibrated scaling data within the Zeiss LSM 510 AIM software (Rel. 3.2). The line was drawn no greater than 50 μm away from the anterior pole of whole lenses and 500 μm from the center of human capsulotomy samples.
Fixation and Dehydration of Lenses
Whole living lenses were equilibrated in Medium 199, FITC-dextran, and Draq-5 and capsule thickness measured as earlier. The lenses were then placed in an alcohol-formalin fixative, Pen-Fix (Richard-Allan Scientific, Kalamazoo, MI), overnight then transferred to 70% ethanol (Swiderski et al., 2000). The anterior capsules of the fixed and partially dehydrated lenses were then determined as stated earlier by using XZ confocal imaging in 70% ethanol. These lenses were then paraffin embedded and sectioned using standard protocols. The anterior capsule thickness was determined as stated earlier but from images acquired from the sectioned lenses in the XY frame scanning mode using a LSM 510 VIS confocal microscope.
Determination of the Refractive Index of the Lens Capsule, Z-axis Measurement Correction Factor, and Data Significance
Refractive index mismatch is an optical phenomenon known to cause axial distortion in confocal Z-axis imaging (Bucher et al., 2000; Booth and Wilson, 2001). To correct this, the refractive index of the bovine anterior capsule was determined using a Zeiss bench top Abbe refractometer in accordance to the procedure used to measure the refractive index of liquids in which the material being sampled is placed between two refractometer prisms (Monk, 1963). Diffuse light was passed through an illumination prism at a variety of angles and transmitted through the sample and then to a measuring prism until the boundary between transmission and total reflection was apparent in the middle of the viewing field (Rheims et al., 1997). Using extracted anterior capsules from four bovine lenses 12 measurements were made. The contrast at the boundary was enhanced by compensating for the chromatic dispersion of the prisms. The refractometer was calibrated to provide results for the sodium (D) line (wavelength 589.3 nm) and the clarity of the boundary was sufficient to allow measurements to the third decimal. The refractive index for the bovine anterior capsule was determined to be 1.399 ± 0.003. A correction factor of 1.05 was determined based on the difference of this value and the refractive index of water (1.33) and applied to all Z-axis measurements. All resulting data were then analyzed for significance using single factor analysis of variance (ANOVA).
RESULTS AND OBSERVATIONS
Capsule Measurements from X- to Z-axis Confocal Images
It is desirable to make anatomical measurements on biological tissue that has undergone as little tissue manipulation as possible. We have developed a new method using existing technology to measure anterior lens capsule thickness. The method takes advantage of both the ability of 40-kDa fluorescein labeled dextran to enter the lens capsule of intact lenses but not lens cells and the availability of confocal microscopes with Z-axis imaging capability. Representative images of the anterior capsule of a capsulotomy specimen obtained from a human cataract patient (Fig. 1A) and an intact FVB/N mouse lens (Fig. 1B) showed that the contrast between the loading medium, the lens capsule and the lens cells was quite high. This allowed for the precise discrimination between the capsule, the lens cells, and the surrounding media and the measurement of capsule thickness.
Dependence of Lens Capsule Thickness Measurements on Incubation Media
It is known that maintaining lenses ex vivo in nonoptimal media results in swelling which distorts the structure of the lens and capsule (Augusteyn et al., 2006). Thus, the thickness of the anterior lens capsule of 9-week-old ICR mice was measured after 1-h incubation in Media 199, which has been shown to be isotonic to lenses in long-term culture (Augusteyn et al., 2006), and compared with that of Hank's balanced salt solution (HBSS) following the same incubation period. The average anterior capsule thickness of lenses incubated for 1 h measured in HBSS was 9.5 ± 0.8 μm (N = 9) and was 10.2 ± 0.9 μm (N = 9) in Media 199, a significant difference (P = 0.048). There was no significant difference between lenses incubated for 1 and 24 h in Media 199 (data not shown). Thus, all further measurements were made in Media 199 since it appeared to minimize lens swelling and concomitant thinning of the lens capsule.
Effect of Dehydration and Sectioning on Lens Capsule Thickness
The majority of lens capsule thickness measurements reported in the literature has been derived from images taken from sections prepared from paraffin-embedded tissue (Ruotsalainen and Tarkkanen, 1987; Bleckmann et al., 1989; Straatsma et al., 1991; Schneider et al., 2003; Barraquer et al., 2006) or epoxy sections prepared for transmission electron microscopy (Seland, 1974; Streeten et al., 1987). However, both tissue preparation methods require extensive fixation and dehydration steps, which are likely to shrink and distort the original dimensions of biological tissues (Boyde and Maconnachie, 1979, 1981). Thus, to demonstrate the effect of dehydration and fixation, the anterior capsule thickness of whole ex vivo 12-week-old FVB/N mouse lenses were compared before (Fig. 1B) and after fixation with the alcohol-formalin fixative Pen-Fix and dehydration in 70% ethanol (Fig. 1C), as well as after further standard paraffin embedding processing (Fig. 1D). The average thickness of the anterior capsules measured from the whole unfixed lenses was 12.5 ± 0.7 μm compared with 7.0 μm ± 0.9 from the fixed and dehydrated lens, a 44.0% decrease, P = 2.6 × 10−8. The anterior capsules from hematoxylin- and eosin-stained paraffin sections measured 4.5 ± 0.5 μm, a 63.5% decrease when compared with the whole unfixed lenses, P = 9.3 × 10−8.
Thickness of the Mouse Lens Capsule at the Anterior, Equatorial, and Posterior Poles
The thickness of the capsule has been shown to vary in regions around the lens of most species including rats (Young and Ocumpaugh, 1966; Parmigiani and McAvoy, 1989), rabbits (Fincham, 1937; Ziebarth et al., 2005), and humans (Salzmann, 1912; Fincham, 1937; Ziebarth et al., 2005; Barraquer et al., 2006). To demonstrate the versatility of the technique the thickness of the mouse capsule at the anterior, equatorial, and posterior poles of 10 FVB/N lenses (10 weeks old) were measured. As expected the anterior pole was the thickest measuring 11.4 μm ± 1.5, the equatorial pole measured 9.4 μm ± 1.4, and the posterior pole measured 2.5 μm ± 0.2.
Strain and Age Dependence on Anterior Lens Capsule Thickness in Mice
Mice are the best model to study the effect of genetics and disease on capsule thickness, however, no comprehensive assessment of the variation of lens capsule thickness between mouse strains has ever been reported. Here the thickness of anterior lens capsules of young mice (7–9 weeks old) and old mice (24–29 weeks old) from four inbred strains, BALB/c, FVB/N, C57BL/6, and 129X1, and the outbred strain ICR (Fig. 2) were measured. No significant difference in anterior capsular thickness was observed between young BALB/C (9.9 μm ± 0.7), FVB/N (9.9 μm ± 1.3), C57BL/6 (9.9 μm ± 0.9), and ICR mice (10.2 μm ± 0.9), P = 0.81. In contrast, the thickness of the anterior capsule of young 7-week-old 129X1 mice was significantly less (7.0 μm ± 0.6) than BALB/C (P = 2.1 × 10−15), FVB/N (P = 1.1 × 10−21), C57BL/6 (P = 3.0 × 10−24), and ICR mice (P = 1.6 × 10−15).
Several studies have shown thickness of the anterior lens capsule increases throughout life in humans (Fisher, 1969a; Fisher and Pettet, 1972; Krag et al., 1997) and rats (Parmigiani and McAvoy, 1989). To determine if the anterior capsule continues to grow in adult mice as observed in humans and rats, measurements were made of the anterior capsules of older mice from the same five mouse strains. When compared, the capsules measured significantly thicker in the older mice, BALB/C (11.6 μm ± 0.6, P = 6.4 × 10−8), FVB/N (15.9 μm ± 1.0, P = 4.9 × 10−19), C57BL/6 (12.8 μm ± 0.9, P = 4.3 × 10−12), ICR (15.2 μm ± 1.0, P = 2.3 × 10−10), and 129X1 mice (13.0 μm ± 1.0, P = 2.2 × 10−24). Unexpectedly, the anterior capsule of older mice significantly varied between strains (P ≤ 3.7 × 10−4) except between C57BL/6 and 129X1 (P = 0.58) and FVB/N and ICR (P = 0.16) strains.
Anterior Lens Capsule Thickness in Larger Species
Published anterior capsule thickness data of large species have shown significant differences depending on the method used and animal age (Fincham, 1937; Young and Ocumpaugh, 1966; Fisher, 1969b; Fisher and Pettet, 1972; Fisher and Hayes, 1979; Bleckmann et al., 1989; Parmigiani and McAvoy, 1989; Krag et al., 1997; Ziebarth et al., 2005; Barraquer et al., 2006). Thus, the anterior capsule thickness determined for the mouse using Z-axis images were compared with those of larger species. As done previously for the mouse, measurements were made on images taken of whole ex vivo lenses from cattle, rabbits, and rats while human measurements were made on images taken from anterior capsulotomy specimens provided by cataract surgery donors (Table 1). Data significance between interspecies capsule thickness measurements were not considered since it was not possible to age match between species and body size was not controlled for. Overall, the anterior capsule thickness correlated with the size of the animal.
Table 1. Anterior capsule thickness measurements of selected mammals made from confocal Z-axis images
Anterior pole thickness (μm)
Rat (Sprague Dawley)
In a variety of species, the thickness of the lens capsule at the anterior pole has been reported but has ranged dramatically depending on the tissue preparation and measurement method employed (Tables 2 and 3). It was difficult to compare our rat and rabbit anterior capsule thickness measurements, using Z-axis confocal images from intact lenses to previous reports employing other methods since it was not possible to age-matched donor animals. However, the thickness measurements of human anterior capsulotomy specimens were generally higher than previously reported using other techniques on similarly aged donor samples. Specifically, our measurements were much thicker than those made using fixed and sectioned samples. It is well documented that both fixation and dehydration steps can have significant effects on sample volume that is both preparation and tissue dependant (Boyde and Maconnachie, 1979; Crang and Klomparens, 1988; Menache et al., 1997; Jonmarker et al., 2006). In fact, others (Bacallao et al., 1995) have used the XZ imaging capability of the confocal microscope to evaluate fixation effects on cell height and found that methanol caused up to 50% decrease in height when compared with living cells. Basement membranes, such as the lens capsule, have large carbohydrate:protein ratios (Fukushi and Spiro, 1969) as a result of high percentages of heparan sulfate side chains (Winkler et al., 2001) and hyaluronic acid (Chan et al., 1997). These negatively charged carbohydrate side chains have large spheres of hydration (Rossi et al., 2003), providing the capsule with a high water concentration (Popdimitrova and Bettelheim, 1989) and hydrogel-like properties (Li and Graham, 2007). Additionally, the lens capsule is acellular and does not possess any cytoskeleton structure. In cellular tissues, fixation of the cytoskeleton provides stability to the overall tissue structure, perhaps reducing the shrinkage and extraction effects from dehydration. Thus, we believe the primary explanation for the observed thickness discrepancy is the dehydration process necessary for conventional histology results in an extreme collapse of the acellular lens capsule. This assertion is supported by our observation that lens capsule measurements obtained from fixed/dehydrated and from paraffin embedded/sectioned mouse lenses were 44.0% and 63.5% thinner, respectively, than measurements obtained from the same group of nondehydrated lenses imaged as whole lenses in media 199. We believe that the use of nonfixed lens tissue provides measurements that better reflect the capsule in an in vivo environment.
Table 2. Literature review comparison of human anterior pole lens capsule thickness data to measurements made from confocal Z-axis images
In the five commonly used mouse strains assayed, we observed that the thickness of the mouse anterior lens capsule increases with age (Fig. 2). This is consistent with the reported longevity of capsular material (Young and Ocumpaugh, 1966; Wilson et al., 1994a, b) and previous reports of the anterior lens capsule thickening with age in humans (Fisher, 1969a; Fisher and Pettet, 1972; Krag et al., 1997) and rats (Parmigiani and McAvoy, 1989). Thus, increases in capsular thickness with age are probably a typical feature of the mammalian lens. Notably, this has been proposed to contribute to the development of presbyopia in humans (Fisher, 1969a, b; Krag and Andreassen, 2003a).
Unexpectedly, we observed significant differences in anterior capsule thickness between different mouse strains when compared at similar ages. For instance, the lens capsules of young 129X1 mice were significantly thinner than that observed in the other four mouse strains tested. Interestingly, it was recently reported that 129X1 mice express and deposit much lower levels of collagen α6(IV) in lung and bladder smooth muscle basement membranes than C57Bl/6 mice, likely due to a polymorphism in the collagen α6(IV) promoter (Kang et al., 2006). Collagen IV is the major structural protein in basement membranes, providing it with strength and stability (Poschl et al., 2004). The three distinct collagen IV-triple helical networks, α1α1α2:α1α1α2, α1α1α2:α5α5α6, α3α4α5:α3α4α5 (Yurchenco et al., 2004) are known to be differentially expressed in the capsule during lens growth and development (Kelley et al., 2002). The reduced expression of one of the six α subunits, α1(IV)–α6(IV), results in the elimination or reduction of all networks in which it is contained. Previously, we have shown that embryonic and adult mouse lens capsules initially contain only the collagen IV α1α1α2:α1α1α2 and α1α1α2:α5α5α6 networks (Kelley et al., 2002). This suggests that the thin lens capsule observed in young 129X1 mice may result in the relatively low levels of the α1α1α2:α5α5α6 network found in their developing prenatal capsule matrix, a consequence of low collagen α6(IV) expression.
Notably, we found that the thickness of the young 129X1 lens capsule becomes very similar to that of C57Bl/6 mice by 27 weeks of age. The proteomic expression patterns in lens epithelial and fiber cells have been shown to shift dramatically from the developing prenatal lens to the adult lens. Specifically, our previous work also demonstrated that the collagen IV α3α4α5:α3α4α5 network begins to be incorporated into the lens capsule by 2 weeks after birth (Kelley et al., 2002). The inclusion of the α3α4α5:α3α4α5 network into the capsule matrix may be partially responsible for the recovery of the anterior capsule thickness in these older 129X1 mice.
While older 129X1 and C57Bl/6 mice exhibited similar lens capsule thickness, the anterior lens capsules of mice derived from the Swiss (FVB/N and ICR) lineage were significantly thicker than those derived from the Lathrop/Castle (129X1, C57Bl/6 and Balb/C) lineages (Fig. 3) (P = 0.02). These observations suggest that lens capsule thickness is a quantitative trait with a significant genetic component.
In conclusion, we have shown measurements using Z-axis confocal images allow for rapid measurement of lens capsule thickness in diverse animals without the significant tissue manipulation and resulting introduction of artifacts of other methodologies. This led us to re-evaluate the anterior lens capsule thickness of humans, rats, and rabbits and has resulted in the first reported measurements of mouse lens capsules. Most significantly, our data point to genetic variations in lens capsule thickness between mouse strains and emphasize the need to make lens capsule thickness comparisons in age-matched animals. In the future, the contributions to capsule thickness of a variety of genetic variations of molecules composing the lens capsule matrix can be further explored.
The authors thank Frank Warren, Manager of the Office of Laboratory Animal Medicine, University of Delaware, Newark, DE, as well as Qian Chen and Peng Wang, Department of Biological Sciences, University of Delaware, Newark, DE, for providing many of the mice used in this study.