Impact of the Environment on the Mammalian Corneal Epithelium

Abstract

purpose. To evaluate whether the content of ascorbic acid in the corneal epithelium and aqueous humor reflects seasonal fluctuations in parallel with environmental changes.

methods. Reindeer, cattle, rabbits, and humans were examined, to cover a broad spectrum of overlapping habitats. Ascorbic acid was determined by high-performance liquid chromatography. The thickness of the corneal epithelium was measured, and the number of cells was counted in the tissue sections.

results. Three groups of reindeer eyes were used, two of them collected during summer, the third group during winter. Ascorbate content did not show seasonal variation in either the corneal epithelium or the aqueous humor, whereas epithelial thickness and number of cells decreased significantly from summer to winter. In cattle, ascorbate content, thickness of the epithelium, and number of cells were lower in animals tended indoors compared with those tended outdoors, whereas ascorbate level in the aqueous humor remained similar in both cases. The rabbit showed significantly reduced ascorbate content in the corneal epithelium but not in the aqueous humor in tarsorrhaphy-treated eyes. This procedure did not change epithelial thickness, but the number of cells was slightly increased. The mean epithelial thickness in human corneas successively decreased with increasing latitude and decreasing radiation exposure from the summer season in Oslo to the midnight sun, polar night, conditions in Tromsø, 10° far north, although the differences did not reach statistical significance.

conclusions. Ambient radiation is needed to sustain high ascorbic acid concentration in the corneal epithelium. Corneal epithelial thickness and number of cells are prone to seasonal fluctuations regulated by ambient radiation. In contrast, ascorbate content of the aqueous humor is uninfluenced by environmental change. It is suggested that seasonal adaptation of mammalian corneal epithelium in response to variation in ambient radiation may be nature’s strategy for countering radiation damage to the eye.

Photokeratitis (snowblindness) in skiers, especially during the spring season, is not uncommon. The tolerance level within a group of persons equally exposed to it varies strikingly from one individual to the next. It is also noteworthy that reindeer, living in environments where humans regularly experience photokeratitis apparently remain unaffected. Photokeratitis, although reported in the Andes in South America, ¹ does not seem to be a major problem among people who live constantly at altitudes up to 4000 m in tropical mountains. It may be deduced from these observations that the corneal epithelium responds to UV exposure by activating certain protection mechanisms, the nature of which is unknown. 

Early experiments showed morphologic changes such as nuclear fragmentation and loosening of the corneal epithelium after UV exposure. ² The most reliable sign of photokeratitis, however, is tiny intraepithelial granules seen by slit lamp examination. Based on this criterion, the action spectrum of photokeratitis has been obtained by recording threshold values at given wavelength intervals. ³ It turned out that peak sensitivity curves differed from one species to the next ⁴ and that the action spectrum for photokeratitis was incongruent with the absorption spectrum of the corresponding corneal epithelium. This seems to reflect that the action spectrum represents the resultant curve of several UV-absorbing components within the epithelium and that the question as to whether a particular component may act as a target substance is assumed to depend on qualities such as its molar absorptivity, subcellular concentration, as well as local intracellular conditions. Ascorbic acid is of particular interest in this connection because of its high molar absorptivity for UV radiation (UVR). ⁵  

The high content of ascorbic acid in the anterior eye is well known, and sites for concentration mechanisms have been indicated in the ciliary body, lens, and corneal endo- and epithelium. ^{6 7 8 9 10} The ascorbate content is higher in diurnal than in nocturnal aqueous humor ^{11 12} and corneal epithelium, ¹³ and, quite recently, a family of mammalian ascorbate transporters has been identified in the ciliary body and the corneal epithelium of diurnal species. ¹⁴  

In contrast to our detailed knowledge of the ascorbate distribution in the eye, its function remains to be clarified in several respects, as can be deduced from the numerous hypotheses in this field. For example, ascorbate has been claimed to be part of an electron transfer system, a free radical scavenger, a cofactor in collagen fibrillogenesis, a redox coupler in the hexose monophosphate shunt, a regulator of the intraocular pressure, and a radiation absorber. ^{15 16} Experimentally induced high levels of ascorbic acid in aqueous humor of the guinea pig has a protective effect against UV-induced DNA damage to lens epithelium, ¹⁷ and ascorbate as well as tocopherol in the aqueous humor increases the viability of such cells after UVR in a dose-dependent manner. ¹⁸ There is also some evidence that systemic ascorbate supplementation used prophylactically may reduce the prevalence of hazy vision after photorefractive keratectomy. ¹⁹ As a whole, these observations support the hypothesis that ascorbate may act as a UV absorber in the eye, and that the high ascorbate content in the anterior diurnal eye has evolved in response to deleterious environmental exposure, be it the UV part of the spectrum only ¹¹ or solar irradiation in general. ²⁰  

The intensity of ambient radiation varies considerably on a yearly basis, and, accordingly, also the need for protection from it. In cases in which ascorbate is part of this protection strategy, its concentration may show short-term fluctuations covarying with the seasonal environmental strain. The present work has been designed to evaluate whether such short-term fluctuations really occur and to unveil environmental factors that may act as regulators. 

Ambient conditions include, in addition to radiation, factors such as temperature and precipitation, and so these parameters have been considered as well. 

Methods

Samples from four different species were examined (reindeer, cattle, rabbits, and humans), and apart from the rabbit only one eye from each individual was used. 

Reindeer

The animals were living in the mountains all year round, and specimens were drawn from three separate areas in southern Norway (Jotunheimen/Ole, Fosen/Follafoss, and Røros) on three different dates. All animals were killed by the same company (Stensaas Rensdyrslakteri, Røros, Norway). The eyes were enucleated immediately after death, and paired specimens from corneal epithelium and aqueous humor were collected on site. Individual blood samples were taken from the same herds, albeit not necessarily from the animals used for eye sampling. 

Cattle

Two different groups were examined, one of which had lived outdoors 24 hours a day (Veierland, a small island, below 50 m altitude, just off the coast in Oslofjord) for at least 3 months when the samples were collected in September. The animals in the other group had been kept indoors 24 hours a day (in a cowhouse in Akershus) for at least the 5 months before death. These specimens were collected in January to minimize the influence of light coming through the windows. Indoor illumination had been switched off regularly for 8 hours during the night. In the following, these groups are termed “outdoor” and “indoor” animals, respectively. Paired sampling was performed for corneal epithelium and aqueous humor. Individual blood samples were collected from the same animals, but without paired sampling. Eyes and blood were transported to the laboratory in a cooler and processed for storage within 1 to 3 hours. 

Rabbit

Young albino rabbits (2–3 kg) of unspecified sex were used. They were fed conventional pellets (B & K Universal Ltd., Nittedal, Norway) and water ad libitum. The stable was equipped with lighting (Lumilux de luxe, Biolux lightcolor 20 coolwhite, L 36 W/20, 1999, covering wavelengths from 700 nm to just below 400 nm with blue and yellow peaks; Osram Sylvania, Munich, Germany), and light was provided on a 12-hour on–off cycle. Illuminance was measured to 470 lux (LX 93 luxmeter; BEHA, Porsgrunn, Norway) just in front of the cages. Humidity was set at 52% to 55% and temperature at 18.5°C. Two groups were used, each with six animals. The animals were acclimatized for 1 week before a left-side tarsorrhaphy was performed in animals under general anesthesia (0.15 mL fluanisone; 0.1% Hypnorm; Janssen Pharmaceutica, Titusville, NJ; combined with 0.3 mL thiopentone sodium per kilogram intravenously). Two weeks later the animals were killed with intravenous infusion of 5 mL pentobarbital sodium (100 mg/mL) containing 0.1 mL fluanisone, followed by enucleation of the eyes. Eyes from one group were used for biochemical studies and from the other for morphologic studies. Corneal epithelium, aqueous humor, and blood for ascorbate analyses were collected as paired samples. All procedures conformed to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Humans

Specimens were collected from two different areas, Tromsø and Oslo. Tromsø is located north of the polar circle, and the sun is below the horizon (polar night) from November 26 to January 17; conversely, the period of the midnight sun occurs between May 18 and July 26. To ensure the maximal difference in radiation exposure, two groups of specimens were collected in Tromsø, just at the end of the dark and the light periods. For comparison, a similar group of specimens were collected in Oslo in August. The patients were between 34 and 83 years of age and of both sexes. Whole bulbi were obtained at autopsy 24 to 48 hours after death. However, only a limited number of human specimens were available within the selected time spans, and those provided were regarded as less suitable for ascorbate analyses. Human bulbi were therefore used for morphologic studies only. 

Ascorbate Analysis

Corneal epithelium, aqueous humor, and serum were analyzed for ascorbic acid in reindeer, cattle, and rabbits. The total corneal epithelium was scraped off with a glass knife and immediately put into a defined volume of precooled metaphosphoric acid (100 g/L). The glasses with solution were weighed with and without specimens before homogenization with a polytetrafluoroethylene pestle, to determine the native corneal epithelial weight. After removal of the corneal epithelium, the aqueous humor was aspirated from the anterior chamber and mixed with an equal volume of metaphosphoric acid (50 g/L). Blood samples were centrifuged and the serum mixed with an equal volume of metaphosphoric acid (100 g/L). Specimens bound for biochemistry were stored at −35°C for up to 10 days before analysis. 

The specimens were thawed to room temperature and centrifuged at 12,000g for 10 minutes, and the supernatants were injected in triplicate without further dilution into a high-performance liquid chromatography apparatus (LC-10 system: SPD-M10AVP detector, SIL-10A autoinjector, LC-10AS pump, degasser, and LC-10 software; Shimadzu, Kyoto, Japan). The column (Supelcosil LC-18) was 250 × 4.6 mm with a 5-μm particle size. A water mobile phase acidified to pH 2.2 with sulfuric acid, a flow rate of 1.0 mL/min, and detection at 243 nm were used. Standard curves were obtained after triplicate injection of 0.01, 0.1, 0.25, and 0.5 mg/mL of analytical grade ascorbic acid dissolved in 10% metaphosphoric acid. The correlation coefficient of the standard curves was always greater than 0.999, and linearity was shown to be between 0.005 and 0.5 mg/mL. Minimum detection limit of the standard was found to be 10 ng/mL. The variation coefficient of injection repeatability was less than 1%, and interday variation coefficient of identical bovine corneal samples was 13%. ¹³  

Morphology

Corneal specimens for morphologic studies were taken from all four species. Whole bulbi were fixed in 2.5% glutaraldehyde and 0.1 M phosphate buffer (pH 7.3). The corneas were cut out along the limbus, subdivided into two equal pieces, and processed for light microscopy by paraffin embedding and hematoxylin-eosin staining. One section through the central corneal area from each of the bulbi was digitized, magnified 500×, and used for estimating the epithelial thickness and number of nuclei. The thickness was measured perpendicularly to the surface at four regularly interspersed lines (100 μm apart) on every section, and these four values were used to calculate a mean value for each eye. In reindeer and cattle, both central and peripheral thicknesses were estimated (central denotes the top of the corneal curvature, peripheral is 100 μm off the limbus and inward), whereas only the central thickness was recorded in the rabbit. In the case of the human specimens, some sections had epithelium-free regions, and so the epithelial thickness had to be estimated wherever the epithelium was intact, regardless of location on the surface. For all species, the mean ± SD of results in the individual groups was based on the six separate mean measurements in each eye. The number of nuclei was estimated within a 300-μm long area centrally in one section from all species except humans. 

Geophysical Data

Global irradiance means irradiation in the range of a 300- to 3000-nm wavelength measured on a horizontal surface (pyrheliometer). Because most specimens were collected off site in relation to recording units, the indicated readings were obtained by interpolation. The sites for specimen collection and radiation recording were as follows, respectively: Jotunheimen and Løken, Fosen and Mære, Røros and Alvdal, Veierland and Tjølling, Tromsø and Holt, Oslo and Lier. The data were drawn from the Norwegian Crop Research Institute at Aas. 

Illuminance data (i.e., visible radiation, 400–700 nm) were not available on a yearly basis for our purpose and are therefore presented as calculated average values for a standard normal period 1961 to 1990 in which observed cloud cover data from the respective areas had been considered. This information was obtained from the Geophysical Institute at the University of Bergen. 

The UVR, making up some 5% of the global irradiance, varies considerably from year to year. The data represent observations from the respective years collected by the Norwegian Radiation Protection Authority ²¹ and the Norwegian Institute for Air Research monitoring together a network of nine separate recording units (multiband-radiometers GUV 541 and GUV 511; Biospherical Instruments, Inc., San Diego, CA, and spectroradiometer Bentham DM 150; Bentham Instruments, Ltd., Reading, UK, radiation measured on a horizontal surface) on a yearly basis covering all of Norway. The meteorological measurements of biologically important UVR are conventionally presented as International Commission on Illumination (CIE) erythemally weighted irradiance or radiant exposure, termed “effective erythemal dose.” In the absence of available measurements of effective dose for photokeratitis, this spectral weighting is the most relevant unit. The sites for specimen collection and radiation recording were, respectively: Jotunheimen and Kise, Fosen and Trondheim, Røros and Trondheim, Veierland and Landvik, Tromsø and Tromsø, Oslo and Oslo. It should be noted that the data presented herein do not include possible reflectance from the terrain, which is significant when snow covers the landscape. Ground reflectance from a snow field is 50% to 80%. ²² Similarly, possible differences in thickness of the ozone layer on observation versus specimen sites have not been considered. The impact of latitude on UVR was estimated from the August 1999 data for Oslo versus Tromsø. August was chosen to ensure a snow-free landscape in both areas. These cities, lying 10° apart and both at sea level, showed an observed monthly UVR of 1700 and 900 CIE-weighted doses, respectively. This indicates that the amount of UVR decreased roughly 80 CIE/degree latitude northward. The significance of altitude was evaluated from the August 1999 data for Oslo versus Finse, which means two sites at roughly the same latitude with a height difference of 1200 m. The two stations lie approximately 200 km apart. Reflectance from snow-covered terrain is negligible at that time. The observed values were 1700 and 1900 CIE, respectively. Accordingly, the UVR increased 0.17 CIE/m with increasing altitude. 

Precipitation and temperature data were provided by the Norwegian Meteorological Institute in Oslo. The presented values are observations from the respective years, recorded less than 50 km from the collection sites. 

Statistics

Concentrations of ascorbate and morphometric parameters are presented as mean ± SD. A two-sided Student’s t-test was applied to check for significant differences, (i.e., P < 0.05). 

Results

Geophysical data related to the various groups of specimens are listed in Table 1 . Because the design of the study was focused on exploring whether exposure to light and UVR has an impact on the concentration of ascorbic acid in the anterior eye and on corneal morphology, the exposure is classified on an ordinal data scale shown in the rightmost column. Data from the various groups of the separate species were then ordered from low to high exposure; Tables 2 and 3 are based on this classification. 

Reindeer

The low-exposure specimens (from Røros) were collected during the darkest period of the year (December), whereas those exposed to medium or high doses (from Fosen and Jotunheimen, respectively) were collected at the end of the bright summer season (Table 1) . Previous results on ascorbate analyses in animals from Jotunheimen (collected in 1997) have been included for comparison, despite the lower aqueous and serum sample numbers in this group (n = 3 and 2, respectively). Any of the differences in ascorbate content of the corneal epithelium in the three groups were statistically significant (Table 2) . In contrast, the ascorbate concentration was slightly higher in aqueous humor from specimens with low exposure than in those with medium exposure (P = 0.02), whereas the corneal epithelium-to-aqueous ratio was of the same order in all three regions. Concerning the thickness of the corneal epithelium, samples for morphologic studies were not collected in Jotunheimen in 1997. Hence, the 3-year delay between the ascorbate results in Table 2 and the epithelial thicknesses in Table 3 . High-and medium-exposure specimens did not differ significantly from each other in epithelial thickness and number of nuclei (Table 3) . Samples with low exposure, however, showed a significantly thinner epithelium and lower number of nuclei than specimens in either of the other two exposure groups. 

Cattle

Indoor animals were exposed to light 16 hours a day and from a measured intensity of 470 lux in the rabbit stable, the illuminance in the cowhouse was estimated at less than 500 lux (Table 1) . The amount of ascorbate (Table 2) was significantly higher in the corneal epithelium and serum from animals with high exposure living outdoors 24 hours a day during the summer compared with those closed up in the cowhouse during winter. However, the ascorbate content of the aqueous humor did not differ between the two groups. As to the thickness of the corneal epithelium and nuclei number (Table 3) , the animals with low exposure reflected significantly lower figures versus those living outdoors. The central thickness was always higher than the matching peripheral thickness as was true in reindeers, though none of the differences in either species was statistically significant. 

Rabbit

As in the cowhouse, illuminance in the rabbit stable was low and UVR close to zero. In addition, light exposure was further reduced unilaterally by tarsorrhaphy. The ascorbate concentration was significantly reduced in the corneal epithelium of the closed eyes, in contrast to the aqueous humor, showing similar ascorbate content in the paired eyes (Table 2) . No significant difference was observed in corneal epithelial thickness between open and closed eyes, although the number of nuclei was slightly increased (P = 0.04) in the closed eyes, perhaps because of less cell shedding when blinking was prevented (Table 3) . It should be noted that the test period was only 2 weeks in rabbits, as opposed to months in reindeer and cattle. 

Humans

The epithelial thickness in samples with low exposure versus those with higher exposure of the corneal epithelium (specimens collected at the end of the polar night versus the period of midnight sun) in Tromsø were 30.8 and 32.0 μm, respectively (Table 3) . Compared with this, the thickness was 36.6 μm in the reference population at 10° lower latitude in Oslo during roughly the same summer period. None of these differences reached statistical significance. 

Discussion

Our goal was primarily to evaluate the impact of environmental factors on the ascorbate content of the anterior eye. Because nocturnal animals in general were unfit for our purpose because of a normally low concentration of ascorbate in the anterior eye, we had to rely on diurnal animals throughout. Reindeer, cattle, and rabbit were chosen as models to cover a broad spectrum of overlapping habitats from high mountain areas, through open fields in the lowland, to shady or darker milieus mimicking night conditions in more protected surroundings. Also human specimens were added, although for obvious reasons their quality was not optimal. The discussion of how radiation, precipitation, and temperature may influence the cornea and aqueous humor will be presented sequentially. 

As indicated in Table 2 , the results from the two bovine groups of animals are the easiest to discuss. One group had been kept indoors 24 hours a day for at least 3 months before the animals were killed in January—that is, these eyes had been exposed to conventional indoor illumination only. Although the exact specifications of the light sources are unknown, the amounts of UVR were probably fairly low. In contrast, the other group of eyes was derived from animals that had been cared for outdoors for 24 hours a day for at least 3 months before death in September. These eyes had been long-term adapted to bright, sustained outdoor irradiation interrupted only by the short summer nights. Thus, this latter group had been exposed to a significant dose of both light and UVR (Table 1) . 

It is reasonable to suggest that there had been some differences in food supplies between these two cattle groups, as reflected in the statistically significant difference in serum concentrations of ascorbate (Table 2) . It is noteworthy that the two livestock groups still showed equal concentrations of ascorbic acid in the aqueous humor. The saturation curve for ascorbate in the aqueous humor in cattle is unknown. However, this observation indicates that aqueous saturation is reached at a serum concentration of 0.002 mg/mL or less. 

Potential sources of ascorbate for the corneal epithelium include the aqueous humor, the capillaries of the palpebral conjunctivae, the limbal capillaries, and the tear fluid. ²³ However, the fact that the aqueous contains some 30 times the plasma and tear concentrations makes it probable that the aqueous is the principal source of ascorbate for the corneal epithelium. It can be seen that despite a similar ascorbate supply from the anterior chamber to the cornea in the two cattle groups, the epithelial concentration was significantly higher in outdoor than in indoor specimens. Also, the epithelial thickness and number of cell nuclei varied accordingly (Table 3) . These were all differences that occurred when changing one and the same species from indoor to outdoor conditions. 

In contrast, the reindeer groups reflected various outdoor conditions. How do the observations of reindeer fit in with the interpretation of the bovine results? Despite a rather marked decrease in total radiation dose from summer (in Jotunheimen and Fosen) to winter (in Røros) conditions (Table 1) , there was no statistically significant difference between the ascorbate concentrations in the corneal epithelium of the three reindeer groups (Table 2) . This was in contrast to epithelial thickness and number of cell nuclei (Table 3) . The readings from the two summer samples (from Jotunheimen and Fosen) were not mutually significantly different, but compared with the winter sample (from Røros) each was higher, in correlation with the increasing difference in radiation dose (Table 1) . This shows that the increase in epithelial thickness was not merely an osmotic effect. It seems, therefore, that even the low-level outdoor radiation acting during the short winter day had been sufficient to sustain a high ascorbate concentration in the epithelium, but too low for strong stimulation of mitotic activity. The latter finding agrees with previous observations of varying epidermal thickness in response to changing radiation exposure. ²⁴ As to the ascorbate content of the reindeer aqueous humor, the winter-specimen (Røros) showed a slightly higher level than the other two (P = 0.02). However, the corneal-to-aqueous ratios were similar in all three groups, and so the biological significance of this observation, if any, is hard to interpret. 

Similar to one of the cattle groups, the rabbits had been cared for indoors at conventional illumination. In addition, light exposure had been further restricted unilaterally by closure of the eyelids in all animals. This procedure induced a significant reduction of the epithelial ascorbate content on the surgically altered side, leaving both epithelial thickness and aqueous ascorbate content essentially unchanged compared with the control eye. The reduced epithelial ascorbate level was obviously a consequence of the tarsorrhaphy. However, whether the effect was due to lower radiation per se or to a lid–cornea interaction (for instance, hypoxia) is unknown. 

Although all three species have a mainly diurnal behavior in common, they showed striking variations in the ascorbate content of the corneal epithelium. Based on mean values (Table 2) , and setting the level in cattle to 1, a species difference for cattle, rabbit, and reindeer of 1, 1.1, and 1.9, respectively, was found. This fits well with the idea that reindeer have a highly effective protection against UVR in the cornea provided by a high concentration of ascorbate. 

Taking all findings together, the following conclusions are valid: Ambient radiation is needed to sustain a high concentration of ascorbic acid in the corneal epithelium. The corneal epithelial thickness and the number of cells are prone to seasonal fluctuations regulated by ambient radiation. The ascorbate content of the aqueous humor is uninfluenced by environmental changes. 

As for the human specimens, the interpretation of our findings is both easy and difficult: easy because none of the differences between the three groups reached statistical significance, and thus the question became one of whether we should bother to interpret the results. However, some of these people may have been forced to stay indoors for a long period before death, giving rise to heterogeneity of the material, with marked standard deviations. In this way, real differences may have been masked. The interrelationships between the observed mean values were just as expected from findings in the other three species. The corneal epithelium showed highest thickness at the lowest latitude in Oslo, followed by specimens representing the periods of midnight sun and polar night in Tromsø. 

The finding that corneal ascorbate content and epithelial thickness are dependent on ambient radiation is perhaps not surprising, located as these cells are in the doorway of entry of the rays. It follows as a logical consequence that those cells responsible for the high concentration of ascorbate in the aqueous humor, a process hidden behind the pigmented iris, were uninfluenced by changes in ambient radiation. This divergent response to external influence is contrary to the synonymous response of the levels of ascorbate in aqueous humor and the epithelium to an increase in the serum concentration of ascorbate. 

Which wavelengths are responsible for triggering the ascorbate pump and mitotic activity in the corneal epithelium is hard to evaluate with the present evidence. However, the winter reindeer series (from Røros) shows that a moderate dose of radiant energy containing a limited amount of UVR was sufficient to sustain an ascorbate level matching the two summer series (from Jotunheimen and Fosen). Low radiation exposure, in contrast, was clearly linked to thinner epithelium with a low number of cells, both in the Røros specimens and in cattle kept indoors. This could be due to reduced UV exposure of these groups, in accordance with the fact that the mitotic activity is known to be influenced by UVB radiation. ^{2 25} It seems, therefore, that the two processes, the mechanism regulating the concentration of ascorbate and the mitotic response of the corneal epithelium, show different sensitivities to light and UVR. Perhaps the former is mainly stimulated by light and the latter mainly by UVR. 

From a theoretical point of view, increasing tissue concentrations of ascorbate in the cornea are not necessarily synonymous with increased transport to the tissue, but may be explained by increased stability of the agent in the tissue due to confounding factors, such as food components and changing hormonal activity throughout the year, among others. Thus, the present work should be seen merely as a hypothesis-generating study, with the conclusions still to be validated in separate experiments. Variations in the human corneal epithelium, for instance, could be evaluated by in vivo confocal microscopy in suitable groups. 

The possible implications of these findings in terms of photoprotection are intriguing. Photokeratitis involves tissue damage in response to UVB and UVC radiation, and so far it has been the common view that tolerance to UV exposure does not build up in the cornea as it does in the skin. ⁴ This postulate should perhaps be reconsidered. Because of its molecular configuration, ascorbic acid is both a powerful UV absorber ^{5 26} and a potent free radical scavenger. It is perhaps reasonable suppose the maintenance of a high ascorbate concentration in the corneal epithelium by ambient radiation in this context, and also to suggest that the epithelial thickness adapts to the ambient radiation level to minimize damage to the basal cell layer. If so, one practical consequence emerging from this research could be that, in addition to the use of sunglasses, photokeratitis prophylaxis may be improved in two ways: to ensure ascorbate saturation in the anterior human eye, intake of 500 mg of supplemental vitamin C twice daily for at least 10 days is recommended ²⁷ ; and because ambient radiation increases the thickness of the corneal epithelium and sustains its concentration of ascorbate, any planned stay in high-UV environments should be preceded by ample outdoor activity. 

The cornea is, of course, more at risk of environmental damage than are internal eye tissues, and because a number of cofactors have been implicated in cataractogenesis, ²² it seemed appropriate to pay attention to environmental impact beyond radiation. We thus decided to evaluate the possible correlation between the ascorbate level, as a potential safeguard for the anterior eye, and ambient radiation, temperature, and precipitation. In contrast to radiation, neither temperature nor precipitation (Table 1) showed any covariation with the observed fluctuations in ascorbate content and epithelial status of the cornea (Tables 2 3) . 

 Submitted for publication February 19, 2002; revised June 10, 2002; accepted June 26, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Table 1.

 

 

Geophysical Data Related to Site and Time of Specimen Collection for the Various Species

Table 1.

 

Geophysical Data Related to Site and Time of Specimen Collection for the Various Species

Species	Location	Specimen Collection Date	Latitude (°North)	Altitude (m)	Global Irradiance (J/m2)	Illuminance (Lux)	UVR (J/m2)	Precipitation (mm)	Temperature (mean C°)	Exposure to Light/UVR
Reindeer	Røros	12-13-99*	63	800	98.90	641	40	39	−5.3	Low
	Fosen	09-18-00*	64	500	275.64	11643	1100	64	+11.9	Medium
	Jotunheimen	09-01-00*	61	1100	441.74	16976	2100	54	+7.0	High
	Jotunheimen	08-29-97*	61	1100	508.17	17933	2300	84	+9.2	High
Cattle	Cowhouse	01-17-00*				≈500, †	0		+18.0	Low
	Veierland	09-07-00*	59	50	449.44	18161	1550	91	+15.2	High
Rabbit	Stable	11-07-00*				470	0		+18.5	None/low
Human	Tromsø	01-07–14-01	70	0–50	<0.02	0	0	124	−2.6	Low
	Tromsø	07-21–29-01	70	0–50	115.92	18534	1400	90	+10.7	High
	Oslo	08-04–20-01	60	0–50	602.81	21672	2400	84	+17.4	Reference

Table 2.

 

 

Ascorbate Concentration in Corneal Epithelium, Aqueous Humor, and Serum from Reindeer, Cattle, and Rabbits

Table 2.

 

Ascorbate Concentration in Corneal Epithelium, Aqueous Humor, and Serum from Reindeer, Cattle, and Rabbits

Species	Exposure to Light/UVR	Specimen Collection Date	n	Corneal Epithelium (mg/g)	n	Aqueous Humor (mg/mL)	C/A Ratio	n	Serum (mg/mL)	A/S Ratio
Reindeer	Low	12-13-99	6	3.1 ± 0.58	6	0.31 ± 0.03	10.0	6	0.008 ± 0.002	38.8
	Medium	09-18-00	6	2.8 ± 0.57	6	0.26 ± 0.03	10.8	6	0.008 ± 0.003	32.5
	High	08-29-97*	6	2.7 ± 0.23	3	0.27 ± 0.02	10.0	2	0.019/0.014, †
Cattle	Low	01-17-00	6	1.3 ± 0.13, ‡	6	0.19 ± 0.02	6.8	6	0.002 ± 0.0009, ‡	95.0
	High	09-07-00	6	1.7 ± 0.19, ‡	6	0.18 ± 0.02	9.4	6	0.004 ± 0.0008, ‡	45.0
Rabbit		11-07-00						6	0.031 ± 0.007
Closed eyes	None		6	1.4 ± 0.38, ‡	6	0.50 ± 0.002	2.8			16.1
Open eyes	Low		6	2.0 ± 0.40, ‡	6	0.48 ± 0.04	4.2			15.5

Table 3.

 

 

Thickness and Number of Nuclei of the Corneal Epithelium Measured in Tissue Sections from Reindeer, Cattle, Rabbits, and Humans

Table 3.

 

Thickness and Number of Nuclei of the Corneal Epithelium Measured in Tissue Sections from Reindeer, Cattle, Rabbits, and Humans

Species	Exposure to Light/UVR	Specimen Collection Date	n	Corneal Epithelium Thickness (μm)			Epithelial Cell Nuclei (n)
									Central	Peripheral
Reindeer	Low	12-13-99	6	48.0 ± 6.6*	45.0 ± 3.6	118 ± 3*
	Medium	09-18-00	6	59.0 ± 6.0*	57.2 ± 6.2	190 ± 16*
	High	09-01-00	6	62.6 ± 3.8	58.4 ± 4.8	200 ± 36
Cattle	Low	01-17-00	6	56.8 ± 2.6*	53.4 ± 3.8	221 ± 12*
	High	09-07-00	6	68.4 ± 2.8*	67.8 ± 3.2	265 ± 36*
Rabbit		11-07-00
Closed eyes	None		6	19.6 ± 4.0		113 ± 9
Open eyes	Low		6	21.4 ± 2.0		101 ± 8
Human	Low	01-07–14-01	6	30.8 ± 7.2
	High	07-21–29-01	6	32.0 ± 10.0
	Reference	08-04–20-01	6	36.6 ± 3.6

 

The authors thank Roar Prøsch, Magne Strikkert, and Stensaas Rensdyrslakteri in Røros; Thomas Kebely, Bjørnar Jakobsen of the Sentralslakteriet Vestfold, Tønsberg; Vidar Isaksen from the Department of Pathology, University of Tromsø; Endre Skaar of The Norwegian Crop Research Institute, Aas; Jan Asle Olseth of the Geophysical Institute, University of Bergen; Bjørn Johnsen of the The Norwegian Radiation Protection Authority, Oslo; Britt Ann Høyskar of The Norwegian Institute for Air Research, Oslo; Inger Marie Nordin of The Norwegian Meteorological Institute, Oslo; Tor Jacob Eide from the Department of Pathology, Dag Sørensen and Randi Væråmoen, from the Department of Comparative Medicine and Diana de Besche, Eli Gulliksen, and Astrid Østerud from the Department of Ophthalmology, National Hospital, University of Oslo.