Radiation cataracts: mechanisms involved in their long delayed occurrence but then rapid progression

Animals and tissues

The first set of 12 irradiated and 12 control C57BL/6 female mice were two months old at the time of the primary radiation experiment and were followed for timing and extent of cataract development by monthly slit lamp examinations for 11 months after the irradiation. A second group of 12 irradiated plus 12 control animals that were also two months of age, similarly irradiated, and of same sex and strain was used to provide irradiated and control lenses, respectively, for cell culture of LEC, half for clonal studies and half for metaphase spreads. A third group of such mice were used for antibody studies of 8-hydroxyguanosine (8-OHG) and DNA repair antibody presence 72 h post-irradiation. Lenses were isolated by posterior excision from the extirpated eyeball 2 min after euthanasia of the mice by cervical dislocation. Only females were used to avoid the stress and fighting damage often seen in separately weaned and caged males of this strain. These specific pathogen-free animals were purchased from Harlan Laboratories, Inc (Indianapolis, IN) and housed four per cage and fed the standard maintenance Purina mouse diet 5015 (Purina Labs, St. Louis, MO). Humane protocols for maintenance and euthanasia from the University of Washington IACUC committee and the national AALAC organization were followed.

Lens treatments

When used for the comet assay, the quickly extirpated lenses were placed immediately over ice to prevent enzymatic activity. Lenses for immunohistochemistry were immediately placed in Carnoy’s fixative. Lenses for confocal examination were placed sequentially in the Hoechst 33342 and the dihydrorhodamine dyes, which were diluted in Eagle’s minimal essential medium (MEM) immediately after removal, with the total period in these dyes not exceeding 2 h. Controls in which the dyes were followed from the first to the fourth hour showed no change in the dye penetration or in the final confocal readout. Further, the differences seen in dye presence between the previously irradiated lenses and the non-irradiated control lenses did not change over the length of time that they were allowed to remain in the dye solution.

Radiation source, dosage, and effects

A cabinet X-ray machine (Picker Company, Cleveland, OH; model 43855A, Series 209, cycles 50/60, Pre volts 115 with PKV 110 and Beryllium window) was used to irradiate mice placed 13 inches from the source. Delivery of 11 Grays (Gy) of X-rays was at 5 mA, 100 kV over 5.5 min. This is a low energy delivery compared to energy delivered by a 300 kV therapy machine or a Cesium source. Five mice at a time were placed on a circular rotating base with heads to the center and the remainder of the body including the tail covered with 3 mm of lead. Thus, only the heads were exposed to the x-radiation. The settings and delivery were measured by specialists from the University of Washington Environmental Health Division using a Victoreen meter. The mice were anesthetized with a mixture of ketamine (100 mg/ml) and xylazine (20 mg/ml) diluted with saline 12.6 fold to deliver anesthesia at 0.02 ml/gm of bodyweight. This was injected intraperitoneally (I.P.). just before the irradiation. They remained anesthetized for approximately 30 min. None of these animals in either the control or the head-irradiated groups showed any signs of distress, reduced physical activity, or reduced food intake during the study period from the beginning to the conclusion of the experiment 11 months following the irradiation. There was a general hair loss from the head that was obvious two weeks after irradiation with re-growth of melanin-free hair over the subsequent three to four weeks. Two mice developed sufficient subsequent scar tissue near one eye in each animal that while not covering the eye, the scar tissue prevented an accurate slit lamp reading of cataract development on that side of the face.

Cataract examinations in the living animals

The irradiated mice and non-irradiated controls in group 1 were examined every 30 days through the 11^th month post-irradiation by slit lamp using a Kowa SL14 hand held slit lamp (Kowa, Tokyo, Japan) after dilation with a 3:1 volume mixture, respectively, of tropicanamide (Acorn, Buffalo Grove, IL) and phenyl hydrochloride (Acorn, Buffalo Grove, IL), human use ophthalmic solutions that achieved full dilation. The viewer was blinded as to the animal’s previous treatment. The degree of lens opacity was rated by half steps from 0 (completely clear) to 4 (complete opacity of a mature cataract). When lenses were removed from the euthanized animals for the confocal studies, they were first scored for opacity using a dissecting microscope. Thus, in each sacrifice of two mice each in the X-ray and the control groups at each time point at one, three, five, and seven months, the slit lamp cataract readings increased in severity in the irradiated mice as noted in the Results section. The surgically removed lenses were immediately examined under a dissecting microscope and rated separately by two persons without knowledge of the irradiated or control status of the donor mouse. There was no more than a 0.5 difference on the 0 to 4 point rating scale between the live animal slit lamp scoring and that made under a dissecting microscope for those lenses removed surgically for confocal study (notations of 1+ to 4+ indicate that the rating level had reached that stage in either type of examination). Confocal, fluorescent dye readings were done immediately after the dissecting microscope reading. These confocal readings rated the corresponding statuses of the LEC nuclear DNA clearance, the internal lens DNA distribution (both fragmented and in free form), and the ROS presence. These, in turn, corresponded with the live animal slit lamp and the post-sacrifice dissecting microscope readings for increasing lens opacity with all three factors increasing with time beginning at five months post-irradiation. It is noted that the appearance of the individual cataracts increased over time but was nevertheless sporadic with some not appearing until the seventh month or beyond. Thus, the selection of donors for confocal studies was made from among those mice presenting with cataracts as determined by slit lamp examination. However, as no cataracts were present at one month and three months post-irradiation, random animals from each group were selected for confocal study at these times.

Sacrifice times for confocal studies of living whole lens and for Carnoy’s-fixed paraffin-sectioned, antibody studies

Groupings of two mice each from the irradiated group and their non-irradiated controls were sacrificed at three, five, and seven months post-irradiation for living lens collections and for immediate exposure to the fluorescent dyes listed below followed by dual laser confocal examination. Sacrifices 72 h post-irradiation were from the third group of mice and were fixed in Carnoy’s fixative and used to prepare 10 μM lens tissue sections using fluorescence antibody studies of the DNA adduct, 8-OHG, and the DNA repair protein, XRCC1. Sacrifice of mice for LEC in vitro culture studies was at 72 h post-irradiation followed by culture as noted below. Mice were first anesthetized then sacrificed by cervical dislocation.

Comet assay for DNA strand breaks

The integrated comet alkaline gel migration assay as described in Singh et al. [16] integrates the values for length, breadth, and intensity of the DNA stained comet tail. The alkaline gel conditions report single strand breaks and alkaline sensitive sites that allow breaks to occur. It was run on the LEC freed from the eight lenses of four mice per each preparation at 5 min, 15 min, 30 min, and 4 h post-irradiation. Removal of the lenses required 2 min. The lenses were placed in small tubes immersed in ice immediately upon extirpation to stop enzyme activity. The LEC were obtained from the lens within 1 min by dispersal in a device as previously published [16].

Use of fluorescent dyes and dual laser confocal microscope scanning of whole living lenses

Briefly, a computer-assisted Zeiss 2-photon LSCM (model 510 NLO, Carl Ziess, MicroImaging Inc., Thornwood, NY) was used on whole lenses immediately after extraction and subsequent 60 min immersion in MEM media containing 10 uM Hoechst 33342 dye for DNA detection and quantitation and 25 uM dihydrorhodamine (DHR) for ROS detection and quantitation. For LSCM settings and dye dilutions see Pendergrass et al. [29]. The source of both dyes was Molecular Probes Inc. (Eugene, OR). Only one fluorochrome was excited at a time to assure that only the desired probe was visualized. Analyses were performed at both the anterior and posterior poles of the lenses using a 10X objective and scanning 31 frames from an area of the lens 1.3 mm in diameter and to a depth of 120 μm deep into the cortex. Deeper scans were collated at a succession of 8 μm depths in a staggered series from the lens surface to a maximal depth of 120 μm and were made to assess the extent of the regions affected. These were also extended to determine whether cataracts extended into the nucleus of the lens, which the larger cataracts often did. For assurance concerning the presence of DNA in both nuclear fragments and free DNA in the lens cortex, some paired sections were first fixed with 5% paraformaldehyde and then treated with DNase. These showed complete removal of the Hoechst stained entities. From the first group of 12 irradiated plus 12 control mice, two mice from each group were sacrificed at three, five, and seven months post-irradiation for the confocal lens examinations. Additional LSCM settings and procedures information is available in references, [28,29].

Antibody staining techniques

All antibody staining was by primary and secondary antibody technique of Carnoy’s-fixed, paraffin-embedded serial sections with viewing and photography using a Bio-Rad (Hercules, CA) computerized microscope with camera attachment. The antibody used for 8-hydroxyguanosine staining (8-OHG; Alpha Diagnostics Inc., San Antonio, TX) was rabbit polyclonal anti-mouse. The primary antibody used for XRCC1 staining was also rabbit polyclonal anti-mouse from the same source. All secondary antibodies were fluoresceinated goat-anti rabbit, which was also attained from the same source. Control slides substituted non-specific rabbit IgG for the primary antibody also from Alpha Diagnostics. Paraffin block preparations fixed in Carnoy’s fixative were found to be superior for sections and antibody stains when compared to frozen sections. Coomassie blue dye was used on a Carnoy’s-fixed tissue section to illustrate the aggregated protein status in a 4+ cataract.

Chromosome damage studies

From the second similarly irradiated group, noted above, the lenses from six irradiated mice and an equal number of non-irradiated controls were removed 72 h post-irradiation; the LEC were collected by stripping them as a layer from the surface of the lens with forceps and placing them in a 1% collagenase/dispase and 3% bovine serum albumin (BSA) solution for 30 min at 37 °C. Completion of LEC separation was accomplished with forceps followed by collagenase/dispase dispersal. A control group consisted of mice of the same age and sex that did not receive the irradiation but were otherwise treated similarly. The LEC were then passed through a 70 μm filter, plated in DMEM+10% fetal calf serum, and cultured for a total of 12 days in 60 mm plates (Biocoat; Becton Dickenson, San Jose, CA) into which glass coverslips has been placed. Metaphase spreads were made from cultured LEC of irradiated mouse lenses removed 72 h post-irradiation and cultured for a total of 12 days under 3% O₂. Twenty-four h before fixation in Carnoy’s fixative, the cells growing on the coverslips were exposed to 0.01 µg/ml colchicine overnight to accumulate metaphases. The coverslips were mounted on glass slides and the resulting metaphase spreads that were treated by Wright-trypsin G-banding procedures were read by one of us (K.S.) for chromosomal aberrations. To increase the number of metaphase cells scored, we have included in Table 1 a small number from old mice that were treated in the same manner as the young mice. This is the only entry for old mice in this study.

Clonal lens epithelial cell growth in vitro

LEC from six animals, each collected from the irradiated and the non-irradiated subgroups of the second group of irradiated and control mice, as noted above, were separately pooled after sacrifice of the mice at 72 h post-irradiation. The cells were collected and dispersed using collagenase/dispace and grown in an MEM medium containing 10% fetal bovine serum (FBS). After seven days in the original mass plating, they were then freed with collagenase/dispase, washed in phosphate buffered saline (PBS), and singly and sparsely replated in triplicate in a similar medium on 35 mm collagen coated plates for individual cell clonal growth. After nine days in the second plating, the size of the individual clones formed was assayed by counting individual cells in each clone after crystal violet staining. The clone counts were then assayed for the percent of clones in large or small cell content groups using several size classification sets (1–4 cells, 30–50 cells, 50–75 cells, >100 cells per clone). Several combinations of these clone number classification groups were used to determine whether significant differences in the percentage of the present large or small clones existed between the collections for the irradiated donors and the non-irradiated controls. Total numbers of clones of any size were also recorded.

Statistical analysis

The degree of DNA presence and DHR in the cortex was compared between irradiated and control lenses for fluorescence intensity by confocal readings from several lens depths using the Mann–Whitney non-parametric test.

Delayed but rapid appearance of X-ray induced cataracts

The typical cataracts that appeared over a 30 day period but only 5-11 months after radiation are shown in Figure 1 as a level 4 cataract that changed from 1.5+ to 4+ opacity within 30 days at seven months post-irradiation when examined by slit lamp in vivo. In Figure 2, the typical lens status is shown immediately after lens removal at one, three, and five months post–irradiation. Lens sections from cataractous lenses stained with the protein dye Coomassie Blue revealed extensive protein aggregation (Figure 3). The status of all lenses of both the treated and control animals remained at minimal opacity (0–0.5+ rating on a 4 point scale) early after irradiation and then progressed only modestly (to a maximum of 2+) with advancing age at each 30 day interval reading until the fifth post-irradiation month. At that time, complete opacity (3.5−4+) suddenly appeared in the first two individual eyes of two individual irradiated mice. In the following months, there were additional random appearances sometimes in one eye and sometimes in both eyes. The accumulation of 3.5–4+ lens readings is shown for individual eyes in Figure 4. By the 11^th month post-irradiation, 80% of the irradiated eyes were positive for advanced cataract at which time the experiment was ended. In all instances and at all observational time points, the individual 3.5–4+ cataracts developed within the 30-day observation periods from previous readings of 1.5–2+. Only opacities scored as 3.5+ to 4+ were considered to be cataracts with 3.5+ representing an advanced opacity including the entire lens as viewed by slip lamp and 4+ representing a lens opacity with 100% light reflection (complete opalescence). As seen in Figure 2 and Figure 4, the non-irradiated control mice changed their lens opacities minimally and gradually, only reaching levels 1.5–2+ by the end of the experiment, at which time all of the mice were 13 months old. No abnormalities of LEC migration, DNA, or ROS presence in confocal studies were seen in the controls that accompanied the irradiated mice at each sacrifice time.

Recovery from the initial radiation-induced DNA breakage

The comet assay measures primarily single strand breaks and alkali labile sites if run under alkaline conditions as it was here [16]. Although the LEC from lenses of the mice sacrificed at 5 min post-irradiation showed large amounts of DNA breaks, essentially complete recovery was apparent in lenses taken 30 min post-irradiation (Figure 5).

In spite of the initial recovery of DNA strand breaks, a dysfunctional lens status increased over time as confirmed by confocal viewing of living lenses using fluorescent dyes for nuclear fragments, free DNA presence, and localized ROS presence as well as noting aberrant LEC migration within the lens. These changes increased progressively in the irradiated mice when viewed in this manner. They were not present in the first month but developed minimally by the third month and were present as marked changes in cataractous, irradiated lenses at five and seven months post-irradiation (Figure 6) with aberrant LEC migration (Figure 7). Thus, the slit lamp and confocal microscopic sequential examinations were usually in agreement, although the confocal examinations at three months showed some minimal early loss of surface LEC while the slit lamp examinations did not yet reveal any evidence of cataract. We selected only those pairs of irradiated mice with obvious opacities when observed by slit lamp examination and pairs of control mice for the confocal studies at five and seven months post-irradiation to illustrate and measure by degree of respective dye fluorescence the internal and surface LEC lens changes. Thus, the most likely candidates were chosen from the slit lamp results for use in the confocal examinations. The advanced changes in a cataractous mouse lens at five months post-irradiation (shown in the confocal view in Figure 6) displayed Hoechst 33343 fluorescence for displaced DNA presence, dihydrorhodamine conversion to rhodamine for ROS presence, and light refraction at the affected sites (note the spatial overlap for the each of these). Both the free DNA and the ROS computerized fluorescence measurements differed from those of control mice by p<0.01 at five months post-irradiation and p<0.001 at seven months using the Mann–Whitney non-parametric test. These measurements included all regions of the interior lens as “stacked” confocal readings of 6 μm each to a 120 μm depth in the cortex. The overall cortex readings contained an accumulation of both nuclear fragments and free DNA. In general, the ROS intensity was greatest at internal lens regions where DNA intensity was also highest. Sequential depth confocal views also showed the migration of LEC strands to the lens interior from inappropriate surface sites that were not at the normal subequatorial bow region. This misdirected migration was a prominent feature in the cataractous lens, and the internalized strands of cells present there did not undergo resolution of their nuclei as shown in the stacked depths integrated LSCM readings in Figure 7. Each of the above noted events are quite similar to those seen in age-related cataract in non-irradiated old mice of the same strain and in old rats as we have previously reported [28,29].

In addition to the LSCM studies, we prepared Carnoy’s-fixed paraffin sections in which we used an antibody to 8-hydroxyguanasine (8-OHG) to measure the in situ presence in the lens of this DNA adduct that is a result of oxidative damage [30,31,32]. This particular DNA oxidation adduct antibody, when used with a fluorescent secondary, was reliable and reproducible in our hands. Many sites for 8-OHG were present in nuclear DNA 72 h following the irradiation, which was long after the comet assay results had returned to control values. This antibody staining was absent in the controls (Figure 8). We also found large amounts of the single strand DNA repair protein, XRCC1, at nuclear sites in the nuclei of irradiated LEC at this same time and none in the non-irradiated controls (Figure 9). Metaphase spreads showed several abnormal chromosomes and chromosomal fragments. This was less common and less severe in the non-irradiated controls as shown in Table 1. An example is shown in Figure 10 (however, note one chromosomal trisomy in the control). The small number of complete metaphase spreads present did not allow a statistical comparison (See Table 1). While it appears that greater damage was present in the metaphases from the previously irradiated mice, this system will need additional studies with methods that provide increased numbers of full chromosome complement metaphase spreads.

Finally, we prepared clonal cultures separately for the irradiated mice and their non-irradiated control group in the manner noted in the Methods section. This separation measured relative capacity of individual cells from each of the two groups to replicate by producing clones with a larger or smaller number of daughter cells, i.e., to proceed through repeated cell divisions in a specific time period. However, we found neither significant clone size differences nor differences in the total number of clones formed between the irradiated and non-irradiated group (results not shown), thus LEC replication capacity was not affected in the cells surviving the irradiation event.