Age-Related Changes in the Absorption Characteristics of the Primate Lens

Abstract

purpose. To quantitate aging of the primate lens by changes in the absorption characteristics that are related to the yellowing of lens protein.

methods. The lenses of lower primates and humans were sectioned anterior to posterior every 0.25 mm, and the UV-visible spectrum of each section was measured to determine the cumulative spectra along the visual axis. The ratio of the absorbance at 320 nm (formed with aging) to the absorbance at 365 nm (present in the young lens) was correlated with the age of the lens.

results. In the young primate UV-B is transmitted to the retina, and UV-A is transmitted to the nucleus of the lens. By puberty, changes in the absorption characteristics of the lens that are associated with the yellowing of lens protein prevented most of the UV-B from reaching the retina and by the eighth decade, the transmittances at 320 and 365 nm to the nucleus of the lens were approximately 40% and 79%, respectively. A linear relationship between the ratio of absorbance at 320 to 365 nm and age was found for both lower primates and humans to the age of 80 years. This is surprising, because the maximum life span of the lower primate is approximately 35 years, whereas humans may live 100 years.

conclusions. These data suggest that the observed spectral changes associated with the yellowing of the lens are the result of a chronological process, such as chemical or photochemical modifications, not biological aging.

The human lens has two primary functions: to focus light on the retina and to prevent optical radiation between 295 and 400 nm from reaching the retina. This light-filtering capacity is important, because the retina is almost an order of magnitude more sensitive to damage by UV radiation than by visible light. ¹ The human lens has little or no turnover of protein ² but minimizes photooxidative and oxidative stress by maintaining low oxygen levels ³ and by efficiently dissipating energy from absorbed light through nondestructive photophysical pathways. ^{4 5} Most of the incident light between 295 and 400 nm is absorbed by the low-molecular-weight tryptophan metabolite, the o-β-glucoside of 3-hydroxykynurenine (3-HKG,λ _max = 365 nm), ⁵ which is a relatively inefficient photosensitizer. Absorption of light by 3-HKG prevents UV radiation from reaching the retina, and 3-HKG is relatively benign in the human lens. ⁵  

With aging, the amount of 3-HKG in the human lens decreases as demonstrated by Bando et al. ⁶ and more recently by Stutchbury and Truscott. ⁷ At the same time, the filtering capacity of the human lens actually increases due to a generalized yellowing of lens proteins. ^{5 8 9} Additionally, there is the formation of protein-attached fluorescent material ¹⁰ with both blue (λ_em = approximately 440 nm) and green (λ_em = ∼520 nm) emissions, the creation of higher and lower molecular weight polypeptides, ¹¹ and an increase in the negative charge of the crystallins. ¹²  

Numerous chemical and photochemical processes may account for these changes, such as the photochemical modification of tryptophan, ¹³ the oxidation of lipids, ¹⁴ and the chemical attachment of sugars or ascorbic acid through the Maillard reaction. ^{15 16} Yu et al. ¹⁷ scanned human lenses (cortical to nucleus) and found a decrease in fluorescence caused by 3-HKG with the formation of a new green fluorescent species in the nucleus. They suggested that there is an age-related conversion of 3-HKG to the new fluorophore. We have corroborated these findings ¹⁸ by sectioning lenses and assessing the relative absorptions due to 3-HKG and yellow protein. We found that with age and section of the lens (cortex to nucleus) there is a decrease of 3-HKG relative to yellow lens protein. These initial experiments suggested the possibility that an age-related modification in the human lens is the covalent attachment of 3-HKG to lens protein. Other studies concerning the irradiation of α-crystallin in the presence of 3-HKG have also suggested this possibility. ¹⁹  

The present studies were designed to determine more quantitative information about the changes in lens absorption as a function of age. Lower primate and human lenses of various ages were sectioned from anterior to posterior, and the UV-visible spectrum of each section was measured. The cumulative absorption spectra were calculated from the individual spectra to determine the total absorbance of the lens as a function of depth. With increasing age, the absorption from 3-HKG at 365 nm decreases, and the maximum shifts to 320 nm. A linear relationship is observed between the ratio of absorbance at 320 to 365 nm and the age of the lens. 

Methods

All animals were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Primate Lenses

The Manhattan Eye Bank and Beryl Ortwerth (University of Missouri), supplied human lenses. Two-year-old Rhesus monkey lenses were obtained from the US Food and Drug Administration’s vaccine testing program. Macaque lenses of other ages were obtained from the Regional Primate Research Center at the University of Washington. Other than the tissue from the Manhattan Eye Bank, all lenses were received frozen and kept at −70°C until used. All of the older human lenses that were studied were yellow but noncataractous. The lenses were prepared by one of the following two methods. 

Sectioning of Lenses.

A small amount of optimal cutting temperature compound (Tissue-Tek; Ted Pella, Irvine, CA) was transferred to a spherical plastic dish (diameter of ∼12 mm and depth of ∼7 mm) and placed in a microtome (Lipshaw Manufacturing, Detroit, MI) at −100°C for 15 minutes until the stabilizer was half frozen. The decapsulated lens was then placed on the compound with the anterior side up. The lens was then surrounded by the compound from the bottom to the lens equator and placed in the microtome until the sample was stabilized at −100°C (approximately 30 minutes). A temperature of −100°C provided the optimum conditions to obtain clear and intact segments. 

The mounted lens was placed in the microtome, and the cutter was positioned at 30° to vertical. All of the lenses were sectioned into 0.25-mm sections except for older (> 70 years), more rigid lenses, which were sectioned into 0.5-mm slabs. The resultant slabs were placed in a quartz cuvette with 0.25- or 0.5-mm path length, allowed to thaw, and covered. 

Concentric Analysis of Lenses.

Concentric analysis of lens protein and 3-HKG was achieved by the method of Li et al. ²⁰ The lens was placed in a preweighed test tube with 0.5 ml distilled H₂O and vortexed. The remaining intact part of the lens was removed. This allowed for the separation of approximately the outer 20% of the lens. Ethyl alcohol (2.5 ml; 100%) was added to the test tube, vortexed, and centrifuged (12,000 rpm, 15 minutes). The supernatant was then analyzed spectrally (UV-visible) for its optical density (OD), at 365 nm. The precipitate was dried and placed in 1.25 M Tris-1% sodium dodecyl sulfate (pH 6.8) buffer, and the absorption at 320 nm was determined at the same dilution as for the supernatant. 

Whole Lenses.

Whole lenses were bathed in balanced salt solution at room temperature before all measurements. To record the absorption spectra, each lens was placed in a demountable cell and slightly flattened to minimize focusing. 

Spectral Measurements

Spectrophotometer.

Spectral transmission of the samples was measured using a fiberoptic spectrometer (PC 1000; Ocean Optics, Dunedin, FL). The system uses the continuous output of a deuterium lamp (approximately 200–700 nm) as the excitation source and a high-sensitivity charge-coupled device (CCD) detector mounted on a card in a computer (IBM Pentium 133; IBM, Armonk, NY). The unique feature of this spectrophotometer is its use of fiberoptic cables, with apertures varying from 0.05 to 4 mm, as light guides. The excitation light is led into the sample chamber through a fiberoptic bundle, and the transmitted light is then collected by a second fiberoptic bundle positioned at a 180^o angle to the excitation source. The excitation and transmitted light are collimated by a set of focusing lenses on either side of the sample chamber. The CCD array detector is capable of collecting full-wavelength spectra (200–1000 nm) with good signal-to-noise ratios at integration times as rapid as 1 msec. Thus, once the dark (0% transmittance [T]) and reference (100% T) spectra have been collected and stored, the actual transmission or absorption spectra can be obtained every few milliseconds. 

The rapid data collection rate and the small aperture of the fiberoptics allow the transmittances of different spatial regions of the same tissue. In addition, the flexibility of the spectrophotometer is such that spectra can be acquired with the cuvette in the horizontal position. Most of the spectra from the tissue samples presented here were obtained in this orientation. An additional important advantage of this apparatus is that the experiment is performed in the open. Therefore, the exact position of the beam impinging on the sample is determined visually. 

Data Analysis.

Studies on guanidine HCl solutions of old human lens protein ⁵ have shown that the proteins do not absorb at wavelengths longer than 550 nm. When working on intact sections it is likely that the spectra will exhibit apparent absorption at longer wavelengths caused by light scattering. This is the result of both Rayleigh and Tyndall scattering and can be subtracted from the observed absorption spectrum by first fitting the scattering portion of the spectrum (e.g., 600–700 nm) to the formula A = aλ ⁿ where A is the absorbance, a is a constant, λ is the wavelength, and n is the order of the relationship between absorbance and wavelength. ²¹ The background absorbance due to scattering for all other wavelengths is then estimated using the coefficients determined from the fit. These values are then subtracted out of the spectrum to give the true absorption spectrum. All calculations were performed with commercial software (Origin 5.0; Microcal, Northampton, MA). 

Results

The absorption spectra ¹⁸ of 0.25-mm-thick sections at 1 mm into the anterior of the lens from a young (2-year-old) monkey and an old (70-year-old) human are presented in Figure 1 . The spectrum of the young primate lens was dominated by an absorption with a maximum at 365 nm, which was attributable to 3-HKG, the major chromophore of the young lens. The spectrum of the young lens had an absorption minimum at 320 nm. In comparison, the spectrum of the older whole lens exhibited a blue-shifted absorption maximum at approximately 330 nm, dramatically increased absorption at 320 nm, and increased absorption to wavelengths as long as 550 nm. 

To understand the progression of these changes, whole lenses of several different ages from both humans and lower primates were sectioned from anterior to posterior, and the absorption spectra were measured at the center of each segment. The spectrum of each progressive segment was added to the spectra of previous sections to generate the cumulative absorbance at specific depths within the lens. Figure 2 shows the results for a 2-year-old macaque lens. The final spectrum (spectrum 8) is the total absorption at 2 mm from the anterior side, which is approximately 75% of the total distance between the anterior and posterior of the lens. According to the Beer–Lambert law for the absorption of radiation, an absorbance of 2 (OD = 2) corresponds to 1% of the incident light being transmitted to the next section, whereas an absorbance of 1 (OD = 1) corresponds to 10% of the incident radiation being transmitted through the sample. The data in Figure 2 indicate that approximately 10% of 320-nm radiation is transmitted to the nucleus of the lens, and at least 1% of 320-nm radiation is transmitted through the posterior side of the lens. Thus, it is clear that both UV-B (290–320 nm) and UV-A (320–340 nm) are transmitted not only to the nucleus of the lens but also to the retina in the young primate. 

Figure 3 is the observed absorption spectrum of a whole intact lens from a 2-year-old macaque. Although the spectral shape is identical, the total OD at 365 nm and below approximately 300 nm is less than the value from the cumulative spectra. The response of the detector in all spectrophotometers becomes nonlinear at large ODs and is underestimated out of the linear region. With the CCD array used in these measurements we observed nonlinear behavior at ODs larger than approximately 2. Therefore, the observed value of OD₃₆₅ equal to 2.6 for the whole lens probably has little quantitative meaning. This illustrates the utility of measuring spectra from thin sections and then summing them, rather than measuring spectra from whole lenses. 

Figure 4 shows an analogous study for a 12-year-old macaque. In comparison with the observations made for the 2-year-old macaque lens (Fig. 2) , the 12-year-old lens (Fig. 4) demonstrated a slight blue shift (approximately 5 nm) in the absorption maximum, an increase in the total absorption at 320 nm, and a decrease in the total absorption at 365 nm. By this age the window of transmission to the retina, centered at 320 nm, has essentially closed (T = 0.5% at 2.25 mm), but UV-B still reaches the nucleus of the lens. 

The individual UV-visible spectra from the first five segments of a 48-year-old human lens are shown in Figure 5 . The outermost section still apparently contains some 3-HKG, ⁵ evidenced by the absorption centered at 365 nm but there is a clear shift to shorter wavelengths with each successive slice into the cortex, indicating its increasingly advanced age. To extend these same experiments to older lenses, a 78-year-old human lens was sectioned. Figure 6 includes the first two 0.5-mm sections of a 78-year-old lens. These data show that by the eighth decade, the absorption maximum has shifted to 320 nm with an increase in absorption in the visible region. There is no evidence for the presence of 3-HKG, and significant amounts of UV-R are transmitted into the inner cortex. The spectrum at a 1-mm depth shows typical characteristics of isolated, aged lens proteins. 

It is clear that as the lens ages there is a progressive shift of the absorption maximum from 365 to 320 nm as well as an increase in absorbance above 400 nm out to 550 nm. To quantify this age-related change, the ratio of the OD at 320 nm (increasing with age) to the OD at 365 nm (present in the young lens) of the cumulative spectrum at 1.0 mm depth was plotted as a function of age for both lower primates and humans (Fig. 7) . Values obtained for solutions of lens protein (approximately 20%) were also included. Given the remarkable linearity of the plot, it appears that the ratio OD₃₂₀/OD₃₆₅ is a spectroscopic indicator of age. The linear fit to all of the data points (P < 0.0001) yielded a slope of 0.021. If the data were treated separately, the linear fit to the points from human lenses had essentially the same slope as the combined values (0.022), but the fit to the data for the lower primate lenses had a slightly lower value (0.0143, P = 0.0037). 

Discussion

The main absorbing species in the young primate lens is 3-HKG (λ_max = 365 nm), which filters out most UV radiation before it reaches the retina. This chromophore may function to protect the retina, which is an order of magnitude more sensitive to damage by UV light than visible radiation. ²² The function, if any, of the small window of transmission at approximately 320 nm in the young primate lens is not known. This window closes by puberty as the lens ages and the lens proteins yellow. 

We have previously reported that some UV-B and, more significantly, UV-A from ambient radiation are incident on the human lens. ¹⁸ Inspection of the cumulative absorption spectra of the young primate lens (Fig. 2) indicates that both UV-B and UV-A penetrate to the nucleus in the young lens and are absorbed fairly efficiently. The effect of this absorption in the young lens may result in initial damage to the nucleus, rendering it more susceptible to cataract formation later in life. Several recently reported epidemiologic studies concerning nutrition and nuclear cataract formation have been reviewed by Beebe. ²³ One of these studies found that infant weight at 1 year is negatively correlated with nuclear opacity later in life. Although it is not possible to unambiguously exclude other factors, this study suggested that poor dietary circumstances such as a deficiency of antioxidants may be a causative factor. If this is the case, then reduced amounts of antioxidants available to the infant eye could cause UV injury to be more severe. An analogous observation has been observed for dermal melanoma ²⁴ in which exposure to excessive amounts of UV-B early in life predispose to melanoma later in life. 

The presence of 3-HKG in the primate lens is unique and does not appear in the lenses of any other species studies so far except for a homologue in the diurnal squirrel lens ²⁵ and in some shallow-swimming marine vertebrates. ²⁶ The synthesis ^{27 28} and the mechanism of the age-related loss ^{19 29 30} of 3-HKG is an active area of study. However, the location of synthesis and any age-related changes in the metabolism of 3-HKG are still not known. Attempts to detect 3-HKG in isolated epithelial cells were not successful, suggesting that it is most likely synthesized in the outer cortex, probably in the equatorial region. ²⁸ The presence of it in the first section of the 46-year-old lens (Fig. 5) supports this. The metabolism of 3-HKG apparently decreases with age, because it was not detected in old (∼60 years) lenses. 

As the human lens ages, there is a progressive loss of 3-HKG ^{6 7} and a concomitant yellowing of lens proteins ⁵ resulting in a broad absorption maximum near 320 to 330 nm, as well as absorbance extending out to 550 nm. In Figure 5 these changes are apparent within a single lens from cortex (younger) to nucleus (older). In the younger part of the lens, absorption attributable to 3-HKG is present but disappears from the spectra collected in the older regions. Modified lens protein, which imparts the yellow color to the lens, appears to be uniformly distributed. Attempts were made to quantify this aging process by correlating the ratio of OD₃₂₀/OD₃₆₅ nm from the cumulative spectrum at 1 mm depth, representing an older portion of the lens, with age. The OD₃₂₀, which increases with age, and the OD₃₆₅, which is attributable to absorption by 3-HKG and decreases with age, may be considered to be a qualitative estimate of UV-B and UV-A, respectively, that penetrates the lens. There is a good linear relationship between these parameters, indicating that the loss of 3-HKG (λ_max = 365 nm) and the increase in absorption at 320 nm are proportional to each other and that both are related to the total number of years that the animal has lived. 

This correlation is particularly interesting, because the data in Figure 7 include values from both lower primates and humans. The linear relationship indicates that similar aging processes occur in both species. However, the life expectancies of the two species are very different; monkeys live to a maximum of 35 years whereas humans may live to a maximum of 100 years. Therefore, the observed spectral changes associated with the yellowing of the lens are the result of a chronological process related to the number of years lived and not to percentage of lifetime. The most obvious explanation for this phenomenon is that the yellowing of lens proteins is the result of photochemical reactions or other environmentally induced chemical reactions rather than normal aging. (i.e., failure of systems, enzymes, and repair). 

Additionally, if the data in Figure 7 are fit separately, it is observed that the rate of change of OD₃₂₀/OD₃₆₅ versus years for the lower primates is slightly slower than the comparable line for humans. This again argues in support of a photochemical mechanism for the yellowing of the lens, because the monkeys were raised indoors and had received a lower total light dose than a human. Clarification of this hypothesis will be further studied by careful examination of the spectral changes for individual sections from the anterior to the posterior of human lenses to determine whether the changes are uniform with depth. If the changes are caused by a photochemical mechanism, then it is anticipated that the rate of change will be somewhat accelerated in the anterior portion of the lens. 

 Supported in part by National Eye Institute Grant EY-02283, Research to Prevent Blindness and RR00166 (Regional Primate Research Center at the University of Washington).

 Submitted for publication August 30, 1999; revised December 17, 1999; accepted December 30, 1999.

 Commercial relationships policy: N.

 

Figure 1.

 

Absorption spectra from a young monkey (2-year-old) and an old human (70-year-old) lens. Absorptions in the young lens are due primarily to 3-HKG (inset), whereas the absorptions in older lenses are from yellow lens proteins.

Figure 2.

 

Cumulative absorption spectra from the central portion of a 2-year-old macaque lens. The spectra are calculated from the individual absorption spectra of 0.25-mm sections made from the anterior to posterior of the lens. The final spectrum (spectrum 8) represents a section approximately 2 mm into the lens.

Figure 3.

 

The absorption spectrum from an intact 2-year-old macaque lens.

Figure 4.

 

Cumulative absorption spectra from the central portion of a 12-year-old macaque lens. The spectra are calculated from the individual absorption spectra of 0.25 mm sections made from the anterior to posterior of the lens. The final spectrum (spectrum 8) represents a section approximately 2 mm into the lens.

Figure 5.

 

Individual absorption spectra from the first five 0.25-mm sections of a 48-year-old human lens.

Figure 6.

 

Individual absorption spectra of the first two 0.5-mm sections of a 78-year-old human lens.

Figure 7.

 

Plot of OD₃₂₀/OD₃₆₅ as a function of age of the lens. The ratios were calculated with the values from the cumulative spectrum at 1-mm depth of each lens or from spectra measured for solutions of 20% of the outer portion of the lens.