Changes in Mouse Corneal Epithelial Innervation with Age

Abstract

Purpose.: We defined the quantitative changes in corneal epithelial nerve fibers during mouse development.

Methods.: Corneal whole mounts from 36 mice were immunostained with a specific anti-β-III tubulin antibody to label nerve fibers. Epithelial nerve fibers in the whole cornea were scanned and imaged using deconvolution microscopy. Subbasal nerve fiber and nerve terminal densities were calculated based on the length of nerve fibers located in each 1 mm² area of the corneal surface.

Results.: The epithelial nerve terminals and corneal surface areas increased at different rates; therefore, the variations in corneal epithelial nerve terminal density and number were asynchronous. The maximum epithelial nerve terminal density was reached at postnatal day 24, which remained constant until postnatal month 2 and then decreased significantly. However, the epithelial nerve terminal number increased significantly until postnatal month 2 and then subsequently decreased significantly. In contrast to the epithelial nerve terminal, the subbasal nerve fiber density and number increased significantly until postnatal month 2 and then decreased. Since the subbasal nerve vortex was formed at postnatal day 24, its location, morphological appearance, and directionality did not display any further changes with age.

Conclusions.: The changes in the corneal epithelial nerve fibers can be divided into two stages after birth (day 1 to 2 months and 2–15 months). During the first stage, the corneal epithelial nerve fiber density increased significantly with increasing corneal surface area. During the second stage, the density presented a significant downward trend.

Introduction

The cornea is one of the most densely innervated tissues in the body. The dense population of corneal nerves responds to irritation and pain, thus having a critical role in protecting the cornea and the rest of the eye from the external environment. ¹  

Corneal nerves transmit blinking reflexes that maintain proper eye hydration, and they also secrete neuropeptides that have mitogenic effects on the surrounding epithelial cells. ^2–4 In this regard, corneal epithelial nerves, which include the subbasal nerves and nerve endings, are especially important. The nerve ending includes the entire epithelial axon distal to its point of origin from a subbasal nerve, as well as all of its collateral branches and terminal expansions. ⁵  

The absence of corneal nerves results in neurotrophic keratitis, a clinical condition characterized by corneal anesthesia and desiccation, and by abnormal epithelial metabolism. ⁶ Moreover, it has been confirmed that many diseases, such as type II diabetes, keratitis, glaucoma, and dry eye, relate to the corneal nerves. Erdélyi et al. observed corneal subbasal nerves in rabbits with dry eye using confocal laser scanning microscopy, and determined that the density of subbasal nerves was significantly lower compared to the healthy control group. ⁷ Hojel et al. established that the corneal sensitivity observed in patients with dry eyes was significantly lower than that of the normal group. ⁸  

The nerve endings are the portion of the nerve fiber responsible for transducing sensory stimuli into neural signals; therefore, corneal nerve terminal density is directly proportional to corneal sensitivity. ⁹ Corneal sensitivity decreases with increasing age in humans, and this decrease becomes more pronounced in individuals older than 50 years. ^10–13 However, it is unknown whether the age-related changes in corneal subbasal nerve fiber density are related to corneal sensitivity. To date, the reports on the overall trends in subbasal nerve fiber density changes due to age have been inconsistent. An examination of human corneas using in vivo confocal microscopy (IVCM) demonstrated that there was no change in the subbasal nerve fiber density in human corneas from individuals 15–79 years old. ¹⁴ In contrast, other studies have documented greater subbasal nerve fiber density in young versus old corneas. ^15–17 Dvorscak and Marfurt analyzed immunostained rat corneal nerves, and confirmed that nerve terminal density decreased with age, whereas corneal subbasal nerve fiber density increased with time. ¹⁸  

The purpose of our study was to provide a detailed and comprehensive description of the age-dependent changes of the subbasal nerve fiber, and epithelial nerve terminal density and number in mice by analyzing immunostained nerve fibers in corneal whole mounts obtained at various time points. 

Materials and Methods

Animals

This study was performed carefully in accordance with the Association for Research in Vision and Ophthalmology (ARVO) Statement on the Use of Animals in Ophthalmic and Vision Research, and the guidelines of the Animal Experimental Committee at Ji Nan University. Wild-type C57/BL6 mice were purchased from the Animal Experiment Center at Sun Yet-Sen University. Disease-free male and female mice were used, and there were six mice in each group. Using a trephine, the corneas were harvested directly from the whole globes still in the socket. Corneas were harvested in this manner at various ages, including postnatal days 1 and 24, and postnatal months 2, 9, and 15. The corneas were placed in ice-cold Ringer's solution and trimmed at the scleral-limbal region, taking caution to avoid any damage, especially to the corneal epithelium. 

Immunofluorescence Staining and Imaging

Immunostaining of whole-mount tissues was performed according to standard protocols. ¹⁹ The corneas were fixed in freshly prepared 2% paraformaldehyde in PBS for 1 hour at 4°C. Following three washes with 0.1 M PBS, the samples were blocked for 15 minutes at room temperature (RT) in 0.1 M PBS containing 2% BSA (PBS-BSA; GBC, Guangzhou, China). The corneas then were permeabilized with 10% Triton X-100 (Guoao, Guangzhou, China) in PBS-BSA for 15 minutes at RT. All corneas then were incubated with a mouse monoclonal neuron-specific anti-β-III tubulin antibody (NL557; R & D Systems Inc., Minneapolis, MN) for 24 hours at 4°C in 0.1 M PBS-BSA and 0.1% Triton X-100. After incubation, the tissues were washed extensively (3 × 5 minutes) in 0.1 M PBS. Whole corneas were cut into four quadrants, stretched, and mounted using anti-fade mounting medium (Beyotime Institute of Biotechnology, Jiangsu, China) containing 4′,6-diamidino-2-phenylindole (DAPI; BOSTER, Wuhan, China). 

Corneal tissue slides were imaged by a DeltaVision Core imaging system (Applied Precision, Issaquah, WA) using either the 60× objective to estimate nerve fiber density or the 10× objective to visualize the subbasal nerve vortex. The instrument was used to scan the layers of the cornea to observe the subbasal nerves and epithelial nerve terminals. To acquire images of the whole subbasal nerve structure, a fluorescence microscope (Leica, Wetzlar, Germany) was used to obtain images at low magnification (5×), and these images were used to calculate the corneal surface area. 

Quantitative Analysis

The densities of the epithelial nerve terminals and subbasal nerve fibers were determined based on the four corneal quadrants. In addition, nerve fiber density was determined using the length of the nerve fibers in each 1 mm² area of the corneal surface based on 291–554 images in each group. Based on the stereologic principle of measuring linear structures, ²⁰ the normalized parameter (L_A ) is the ratio of the total length of all of the features in the specimen divided by the area that the specimen occupies. This perimeter length per unit area can be estimated by probing the structure with a set of line probes represented by eight horizontal curved lines. The point of interest in the measurement is the “line intersects boundary” event. A simple count of these events forms the basis for estimating L_A through the fundamental stereologic formula: $L_A=\pi /2\times P_L$. In this equation P_L represents the “line intercept count,” which is the ratio of the number of intersections counted to the total length of curved line probe sampled (only the horizontal curved lines in the square formed by dotted lines were used). Each of the curved lines in Figure 1 was determined by calibration to be 6.943 mm long. The total length of the curved line probe sampled in this grid placement was 8 × 6.943 = 55.546 mm. These calculations are performed using Image-Pro Plus 5.1 software. The corneal surface area also was measured using this 5.1 software, and the number of nerve fibers in one cornea was calculated as the product of density and surface area. All data were expressed as the mean ± SEM. Differences in corneal nerve fiber density or number between two consecutive groups of mice were compared by an analysis of variance (ANOVA) using the SPSS 17.0 software package, with P < 0.05 and P < 0.01 considered to be statistically significant differences. 

Figure 1. 

 

Use of the line intercept count to estimate the total length of corneal nerve fibers per unit area (mm/mm²). Green circles: represent the intersections, but not all intersections were marked.

Results

Architecture of Mouse Corneal Epithelial Innervation

The immunohistochemical protocol used for the corneal whole mounts resulted in complete staining of the corneal nerve fibers, including the epithelial subbasal nerves and nerve endings. The descriptions that follow present details of the prominent anatomic features of mouse corneal epithelial innervation, and are the first descriptions of the qualitative and quantitative age-related changes. 

Subbasal Nerves.

Subbasal innervation displayed an even pattern with parallel distributed nerve fibers (Fig. 2A). The term “epithelial leash” was defined as a group of subbasal nerves deriving from the same parent anterior stromal nerve. ^9,21,22 Epithelial leashes were less apparent in the peripheral cornea than in the central cornea. When viewed along with all subbasal nerves, some long and curvilinear subbasal nerve fibers oriented centrally and eventually assembled into a gentle, whorl-swirl pattern at postnatal day 24 that became more pronounced at postnatal month 2 (Fig. 3). No correlation was observed between subbasal nerve morphology and age (Figs. 4A–4D). 

Figure 2. 

 

Mouse corneal subbasal nerves and nerve terminals located in the epithelial cells at postnatal month 2. DAPI was used to stain the corneal nuclei (blue). The corneal nerve fibers were stained with a specific anti-β-III tubulin antibody (red). (A) Some subbasal nerves branched into several short, small nerve fibers. They were distributed in a nearly parallel arrangement. Arrow: subbasal nerve fiber. *Sub-epithelial cell. (B) Some nerve terminals presented bead-like patterns, and the ends of some nerve terminals were inflated as nodules or remained as rings. The innervation was irregular. Arrow: nerve ending. *Epithelial cell.

Figure 3. 

 

The subbasal nerve vortices in four different age groups. Red: nerve fibers labeled with a specific anti-β-III tubulin antibody. Subbasal nerves formed a whorl-like pattern at postnatal day 24 (A), and this became more pronounced at postnatal month 2 (B). However, the morphologic appearance and location of the subbasal nerve vortices remained unchanged with age (C, D).

Figure 4. 

 

Subbasal nerves (A–D) and nerve terminals (E–H) in mice at 24 days, and 2, 9, and 15 months after birth. Red: corneal nerve fibers labeled with a specific anti-β-III tubulin antibody. At postnatal day 24 (A), corneal subbasal nerves were less dense, and the innervation was more irregular than in the other age groups (B–D). The density of subbasal nerve fibers increased until postnatal month 2 (B) when the innervation was most dense compared to the other age groups (A, C, D), while the subbasal nerve fiber density began to decrease since postnatal month 9. The innervation of nerve terminals was most dense at postnatal day 24 (E). However, the corneal nerve terminal density decreased from postnatal day 24 to postnatal month 15 (E–H). These images were representative of 291–554 images for each group.

Epithelial Nerve Terminals.

Each subbasal nerve fiber contained varying numbers of nerve terminals. The endings were distributed abundantly across all layers of the epithelium, and these endings had complex morphologies and variable orientations. In mice, most corneal nerve terminals branch once or twice before terminating, and their distal tips typically display slightly bulbous varicosities, similar to those observed previously in rats. Nerve ending fibers were shorter and smaller than the subbasal nerve fibers, and were not visible in some corneal regions (Fig. 2B). It was clear that the innervation of epithelial nerve endings became increasingly sparse with age (Figs. 4E–4H). 

Age-Related Changes in Corneal Surface Area

The corneal surface areas in mice were calculated in all age groups. The corneal surface area increased linearly from birth until postnatal month 2 (3.4 ± 0.04 mm²) and then remained fairly constant without significant differences observed between the age groups (Fig. 5). 

Figure 5. 

 

Change in corneal surface area with age. * P < 0.05; ** P < 0.01. (A) The mouse corneal surface area increased significantly from birth to postnatal month 2. There was no significant change in corneal surface area from postnatal months 2–15. (B) Representative images depicting the age-related change in corneal surface area.

Age-Dependent Changes in Epithelial Nerve Fiber Density and Number

There were very few nerve terminals at postnatal day 1; however, the density of nerve terminals gradually increased after birth. The maximum density of the nerve terminals was reached at approximately postnatal day 24 (9.9 ± 1.4 mm/mm²), remained unchanged until postnatal month 2 (9.7 ± 1.4 mm/mm²) and decreased thereafter (Fig. 6A). The density of the nerve terminals increased by 46.0% from postnatal day 13 (5.3 ± 1.3 mm/mm²) to postnatal day 24 (9.9 ± 1.4 mm/mm²). The nerve terminal density did not differ significantly between postnatal day 24 and postnatal month 2, but it declined in a linear fashion from postnatal months 2–15. The nerve terminal density was reduced by an average of 54.5% from postnatal day 24 (9.9 ± 1.4 mm/mm²) to postnatal day 15 (4.5 ± 1.0 mm/mm²). The maximum reduction in density was 35.5% from postnatal months 2–9. However, the maximum number of nerve endings was reached at postnatal month 2. The number of nerve terminals was 9.6 ± 3.6, 24.4 ± 5.2, 32.6 ± 6.8, 21.3 ± 2.7, and 15.0 ± 4.1 mm at postnatal days 13 and 24, and postnatal months 2, 9, and 15, respectively (Fig. 6B). The number of nerve terminals increased by an average of 70.7% from postnatal day 13 to postnatal month 2, and decreased by an average of 53.9% from postnatal months 2–15. 

Figure 6. 

 

Age-dependent changes in nerve terminal density and number in the corneal epithelium of mice. * P < 0.05, ** P < 0.01. (A) Changes in nerve terminal density with age. From birth to postnatal day 24, the density significantly increased and then decreased from postnatal day 24 to postnatal month 15. However, there was no significant difference in the nerve terminal density between postnatal day 24 and postnatal month 2. (B) Changes in nerve terminal number with age. From birth to postnatal month 2, the nerve terminal number significantly increased and then significantly decreased after postnatal month 2.

In contrast to the changes in nerve terminal density and number, the changes with age in subbasal nerve fiber density and number were similar. At postnatal day 1, the subbasal nerve fiber density and number were 6.4 ± 1.6 mm/mm² and 3.0 ± 0.6 mm, respectively. The maximum subbasal nerve fiber density and number was reached at postnatal month 2 (26.6 ± 2.1 mm/mm² and 89.1 ± 4.8 mm, respectively) and then decreased gradually (Figs. 7A, 7B). From postnatal day 1 to postnatal month 2, the subbasal nerve fiber density and number increased by 76.0% and 96.6%, respectively. However, from postnatal months 2–15, the subbasal nerve fiber density and number decreased by 28.3%. It was obvious that the declining rates of the subbasal nerve fiber density and number were slower than that of the nerve terminal density and number. 

Figure 7. 

 

Effect of aging on subbasal nerve fiber density and number in the corneal epithelium of mice. * P < 0.05, ** P < 0.01. (A) Changes in subbasal nerve fiber density with age. From birth to postnatal month 2, the density increased significantly and subsequently decreased significantly from postnatal months 2–15. (B) Changes in subbasal nerve fiber number with age. These data displayed a similar trend to the subbasal nerve fiber density.

The morphological appearance and location of the subbasal nerve vortex did not correlate with age (Fig. 3). The direction of whorl-like rotation, and the degree of subbasal nerve fiber curving varied considerably and randomly among corneas within the same age group. Our study determined that, in mice, most vortices were located on the dorsal-nasal side of the cornea. 

Discussion

Animal Model and Methodological Considerations

Although the somatotype of mice is remarkably smaller than humans, their ocular architectures are similar and their genomes have many orthologs. Moreover, the mouse is one of the most important animal models for studying the physiologic and pathologic processes of corneal nerve regeneration. The study in the age-related changes of mouse corneal epithelial innervation was lacking. Thus, we chose to use the C57/BL6 mouse to analyze corneal epithelial innervation during development. According to a previous study, ²³ the six age groups in our study (postnatal days 1, 13, and 24, and postnatal months 2, 9, and 15) are respectively equal to the equivalent of newborn, child, teenager, adult, middle-aged, and quinquagenarian humans, respectively. 

It is important to note that age-associated changes in the subbasal nerve fiber and nerve ending density, whether assessed by immunohistochemistry or DeltaVision Core, must be interpreted with caution because of the inherent resolution limitations of these techniques and instruments. β-III tubulin is a pan-neuronal marker that stains all nerve fibers in the cornea. Positive staining in the images mapped all corneal nerve structures, regardless of their origin or phenotype. Therefore, this marker was optimal for studying corneal innervation in mice. Whole mount preparations were used to permit quantification of corneal epithelial nerve fibers. It was known that nerve branches and terminals with diameters <0.5 mm cannot be imaged using IVCM. ²⁴ However, the use of DeltaVision Core compensated the disadvantage of IVCM in our study. In addition, by changing the plane of focus, it was possible to assess independently the age-dependent changes in the subbasal nerve fiber, and in the nerve terminal density and number. 

Density is a relative magnitude, while number is an absolute value. The corneal surface area and corneal epithelial nerve fiber number may change with time at different rates. Therefore, nerve fiber density is not able to assess objectively the quantitative changes in the corneal epithelial nerve fibers during development, which was why nerve fiber number was analyzed simultaneously in this study. 

Age-Dependent Changes in Corneal Nerve Terminal Density and Sensitivity

There was anatomic evidence for a significant decrease in corneal nerve terminal density as a function of age in the rat. ¹⁸ However, before our study, it was not known how mouse corneal nerve terminal density and number changed during development. Our current study showed that corneal nerve terminal density reached its peak at postnatal day 24 in mice. In contrast, the maximum number of corneal nerve terminals was reached at approximately postnatal month 2. To our knowledge, this is the first report of this phenomenon, and the underlying reason for this phenomenon may be straightforward. The explanation for this discrepancy is that the number of corneal nerve terminals and the corneal surface area increase at asynchronous rates. There was no significant decrease in corneal nerve terminal density between postnatal day 24 and postnatal month 2 (9.9 ± 1.4 mm/mm² and 9.7 ± 1.4 mm/mm², respectively). However, the corneal surface area increased by 26.1% from postnatal day 24 to postnatal month 2. Therefore, the maximum corneal nerve terminal density was reached at postnatal day 24, whereas the corneal nerve terminal number in all groups peaked at postnatal month 2. After postnatal month 2, the corneal nerve terminal density and number decreased in a nearly linear fashion. 

Nerve terminals or free endings are responsible for transducing sensory stimuli into nerve signals, and their numbers are directly proportional to corneal sensitivity. ²⁵ In our study, corneal nerve terminal density and number decreased significantly with age. These results may have implications for understanding changes in corneal sensitivity in human subjects. Normal corneal sensitivity varies depending on age, sex, pregnancy, iris color, and contact lens use. ²⁶ Corneal sensitivity decreases occasionally during the initial phase of development in some human subjects, and the single most important reason for this is likely to be defective nerve ending development. In addition, age-related functional loss in epithelial nerve endings is likely to be the most important cause of corneal sensitivity loss in elderly patients. However, other factors may contribute to this phenomenon, including altered synaptic processing of corneal sensory information in the central nervous system, ¹⁴ age-related decreases in the level of patient alertness, ¹⁰ and reduced nerve responsiveness resulting from altered tear film composition. ¹¹  

In adult rats, a significant decrease in nerve terminal density was observed between 6 and 12 months of age, or roughly the equivalent of 15 and 30 years of age in humans. ¹⁸ However, Boberg-Ans ¹² and Millodot ¹³ measured human cornea sensitivity with a mechanical esthesiometer and concluded that sensitivity remained relatively unchanged until the age of 40. Sensitivity then decreased significantly in a nonlinear (parabolic) fashion to approximately half of this level by age 65–70. In contrast, Murphy et al. used a noncontact corneal esthesiometer to stimulate central corneal thermoreceptors. ¹⁰ They concluded that central corneal sensitivity declined in a linear fashion beginning at approximately age 20, and decreased by almost half between 20 and 50 years of age. Our results agreed most closely with the findings of Murphy et al. ¹⁰ Their study and our current study both revealed a significant decrease in sensitivity (in humans) or nerve terminal density and number (in mice) that begins early in life. 

Corneal sensitivity is associated clinically with the density and number of nerve endings, and clinical studies have shown that the incidence of dry eye increases significantly with age. ²⁷ In our study, we observed changes in corneal epithelial nerve endings in mice from birth (postnatal day 1) to quinquagenarian (postnatal month 15). Interestingly, a reduction in the number of nerve terminals occurred in adult mice (postnatal month 2). This result indicated that the decrease in nerve terminals does not provide a complete explanation for the diagnosis of dry eye, but may have a role in the pathogenesis of dry eye. 

Changes in Subbasal Nerve Fiber Density and Number with Age

Our study demonstrated that subbasal nerve fiber density and number in mouse corneas increased significantly from postnatal day 1 to 2 months of age, and then decreased significantly from 2–15 months of age. These data supported the findings of McKenna and Lwigale, who reported that subbasal nerve density and number increased with age during the embryonic period in mice. ²⁸ The effect of age on corneal subbasal nerves remains uncertain. A study in humans noted a significant decrease in nerve density in the subbasal plexus with time that may correlate with the decrease in corneal sensitivity with age. ¹⁵ However, a more recent study did not identify a significant correlation between age and subbasal nerve density. ¹⁴ In contrast, corneal subbasal nerve density increased significantly from postnatal months 6–24 in the rat. ¹⁸ The reasons for the disparities in the findings between species are unknown, but may include differences in animal models, instrument resolution, light intensity settings, location and size of the corneal surface areas examined, and the methods used to calculate nerve fiber density. 

We determined that subbasal nerve fiber density and number decrease significantly after postnatal month 2. Sensory axons enter the cornea in radially-oriented nerve bundles that contain varying numbers of axons and branch extensively to form the subbasal plexus below the basal epithelial cells. ²⁷ Thus, age-associated nerve terminal and subbasal nerve loss may contribute to the corneal sensitivity decrease observed with aging. However, this hypothesis has not been demonstrated conclusively to our knowledge, and further study is required. 

Human subbasal nerves, not nerve terminals, are imaged readily by IVCM, and the density and morphologic appearance of the subbasal nerves often are assessed to determine the integrity of corneal innervation in normal and diseased eyes. ^17,29 Mouse subbasal nerve fiber density and number increased, and then decreased in our study, as did the nerve terminals. These observations suggested that corneal epithelial innervation is a complex process and that clinicians should exercise caution when using changes in subbasal nerve fiber density or nerve terminal density to evaluate corneal innervation status. The decrease in corneal epithelial innervation disturbs the reflex circuits that drive blinking and lacrimation, and also depletes essential factors supplied by corneal nerves, leading to decreased corneal sensitivity and dry eye. It would be beneficial to make clinicians aware of our results to inform diagnoses of dry eye and facilitate future research. 

Subbasal Nerve Vortex

In our study, the whorl-like patterns of the subbasal nerves are identified in the dorsal-nasal quarter of the mouse cornea. This observation confirmed and expanded on previous reports of the whorl-like assemblages of subbasal nerves in similar locations in murine corneas, and the periapical inferonasal region of human corneas. ^24,30 In mice, epithelial nerve fibers are arranged in patterns that project towards the center of the cornea. These nerves form a whorl three weeks after birth that becomes more pronounced in adulthood. ²⁹ The mechanisms that control the formation of the whorl-like assemblages remain unclear; however, it has been postulated that basal epithelial cells and subbasal nerves may migrate centripetally in tandem. ¹⁴ Chemotrophic guidance, electromagnetic cues, and population pressures have been suggested as possible mechanisms for the whorl-like patterns of basal epithelial cells observed near the mouse corneal apex. ^31,32 Our study documented the presence of corneal subbasal nerve vortices in mice from different age groups, and showed that the morphologic appearance and location of the subbasal nerve vortices do not change with age. These findings are identical to those reported by Dvorscak and Marfurt, ¹⁸ and Erie et al. ¹⁴  

In conclusion, we used a mouse model to demonstrate that there is a significant increase and subsequent decrease in corneal nerve terminal density, and a concurrent alteration in subbasal nerve fiber density as a function of age. In addition, the decrease in nerve terminal density begins early in life. Based on these results, it was tempting to deduce that similar age-related changes may occur in human corneas. If true, there may be implications for the clinical management of dry eye. 

Footnotes

 Supported by the National Natural Science Foundation of China Grants 39970250 (ZL), 30772387 (ZL), and 81070703 (ZL).

Footnotes

²  These authors contributed equally to the manuscript as first authors.

Footnotes

 Disclosure: C. Wang, None; T. Fu, None; C. Xia, None; Z. Li, None