p53 Expression in the Normal Murine Eye

Abstract

purpose. The tumor-suppressor gene p53 encodes a phosphoprotein involved in the control of cell growth. Its expression and function have been documented in malignancy, apoptosis, and other abnormal cell proliferation processes. Recently, expression of p53 has been demonstrated in certain normal tissues, including whole eye. The purpose of the study was to map and to characterize expression of p53 in the normal murine eye.

methods. Eyes of adult C57BL/6 mice were enucleated after death by CO₂ narcosis. Expression of p53 in frozen sections of whole cryoprotected eyes was mapped by indirect immunofluorescence microscopy using the anti-p53 monoclonal antibodies 248 and 421 and the polyclonal antibody FL-393. Additionally, eyes were freshly dissected to separate the various ocular tissues. In these ocular tissues, expression of p53 was quantitated with ELISA and Western blot analyses.

results. Strong expression of p53 was observed in various normal ocular tissues. The corneal and conjunctival epithelium exhibited very high cytoplasmic p53 protein levels. High nuclear p53 protein staining was seen in the lens epithelial cells of the central and pre-equatorial zones and in the lens fiber nuclear bow, situated posterior to the epithelial germinative zone. Cells of the actual lens germinative zone did not stain for p53 protein. Low levels of p53 protein were expressed in retinal tissue.

conclusions. High levels of p53 protein are found in various normal murine ocular tissues, especially the corneal and conjunctival structures and the lens epithelium. Each of these tissues demonstrate unique patterns of staining.

The p53 gene, situated on chromosome 17p13.1, is a well-defined tumor-suppressor gene. The gene produces a 53-kDa phosphoprotein that plays a role in regulating cell proliferation, ¹ apoptosis in response to DNA injury, ^{2 3} angiogenesis, ⁴ and embryogenesis. ⁵ Mutations and overexpression of the p53 gene have been found in up to 50% of all human malignancies, ³ skin keloid, ⁶ pterygia, and pingueculae. ^{7 8 9 10}  

Expression of p53 and function in normal, noncancerous tissues has not been studied to a significant extent. Its role in normal tissues is still an open question. Almon et al., ¹¹ were the first to demonstrate significant expression of p53 in normal untransformed tissue. They showed high expression of p53 in murine testicular tissue and suggested a role for p53 in meiotic spermatogenesis. Subsequently, p53 was demonstrated in animal nervous tissue in oligodendrocytes, ¹² central nervous system (CNS) neurons, ¹³ sympathetic neurons, ¹⁴ and specific CNS structures. ¹⁵ More recently, high expression of p53 has been demonstrated in normal, adult rodent ocular tissues. ^{15 16 17} Because many of these tissues consist of nondividing cells, it is likely that p53 has additional important functions other than in cell proliferation regulation and apoptosis. 

To gain a better understanding of the distribution and the role of p53 in the eye under normal conditions, we undertook this study to map and to characterize expression of p53 in the normal murine eye. 

Materials and Methods

Animals

The use of animals adhered to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Eyes of adult C57BL/6 mice were enucleated after death by CO₂ narcosis. Whole cryoprotected eyes were frozen at −70°C for frozen section immunohistochemical mapping of p53. For p53 protein quantification, eyes were freshly dissected to separate clear cornea, limbal tissues, conjunctiva, sclera, lens, iris, and retina. Furthermore, the corneal epithelial layer was scraped off the underlying stroma, and the lens was separated into the lens capsule and lens nucleus, for ELISA analysis. Respective ocular tissues were pooled, lysed, homogenized, and assayed for p53 protein concentration by ELISA and Western blot analysis. 

Immunohistochemistry

Whole eyes were cryoprotected in 30% sucrose and PBS at 4°C overnight. Frozen sections 6 μm in thickness from tissue blocks frozen in isopentane precooled in liquid N₂, embedded in optimal cutting temperature (OCT) compound in cryomolds, and stored at −70°C were collected on glass slides and allowed to dry for 15 minutes, postfixed in freshly prepared methanol-acetone (50:50 vol/vol) for 2 minutes at room temperature. p53 staining was performed with the monoclonal antibody (mAb) 248, which binds to an epitope at the N-terminal of the p53 molecule, ¹⁸ and mAb 421, which recognizes an epitope at the COOH-terminal of the p53 protein ¹⁹ (a kind gift from Varda Rotter, Weizman Institute of Science, Rehovot, Israel) and rabbit polyclonal antibody FL-393 (Santa Cruz Biotechnology, Santa Cruz, CA), which has multiple recognition epitopes over the whole p53 molecule. Staining for p53 was performed overnight at 4°C in a humidity chamber, followed by secondary antibody counterstaining, with anti-mouse or anti-rabbit IgG FITC conjugated (Sigma Chemical Co., St. Louis, MO) for 30 minutes at room temperature in a humidity chamber. All control sections were processed in the absence of primary antibody. The slides were washed, mounted with an aqueous mounting medium, and photographed within a few hours under a fluorescence digital microscope camera (Axoscope2 with image processing from version 4.1 software; Carl Zeiss, Oberkochen, Germany). Light intensity and contrast were standardized for a respective tissue with an appropriate control section. 

p53 Pan-ELISA

For the detection of p53 protein, we used a p53 pan-ELISA kit (Roche Molecular Biochemicals, Mannheim, Germany). This assay for the quantification of wild-type and mutant p53 of human, mouse, and rat origin, is based on a quantitative sandwich ELISA principle. The biotin-labeled capture antibody is prebound to the streptavidin-coated microtiter plate. During a single incubation step, the p53-containing sample reacts with the capture antibody and the peroxidase-labeled detection antibody to form a stable immunocomplex. Subsequent to the washing step, the peroxidase bound in the complex is developed by tetramethylbenzidine as a substrate. The resultant absorbance is proportional to the concentration of p53. Tissue homogenates were prepared from tissues, as described in the p53 pan ELISA kit, by detergent lysis (lysis buffer-tissue, 50:50 vol/vol) in lysis buffer (1% Triton X-100 and 0.1% SDS in PBS [pH 7.4] and proteinase inhibitor cocktail; Roche Molecular Biochemicals) and homogenization (10 strokes in a glass hand homogenizer). Collected supernatants (10,000g × 10 minutes, 4°C) were analyzed according to the manufacturer’s instructions. 

Western Blot Analysis

Equal amounts of protein ²⁰ derived from the respective ocular tissues were compared with each other by Western analysis, after resolution of protein samples (50 μg/well) by standard denaturing SDS 7.5% polyacrylamide gel electrophoresis under standard conditions. ²¹ Proteins were transferred to nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany), as described. Bovine serum albumin (BSA)-blocked Western blots were subjected to Western blot analysis with the mAbs 421 and 248 followed by horseradish peroxidase–conjugated anti-mouse IgG (ECL; Amersham, Buckinghamshire, UK). Protein molecular weight markers (Sea-blue; Novartis, San Diego, CA) were used on each gel. Fifty micrograms of ²²p53-M clone 314 cell extract (a kind gift from Varda Rotter, Weizman Institute, Rehovot, Israel) was used as a p53-positive control (p53-M), whereas 50 μg BSA served as a negative control. 

Quantification of Data

Data were collected from multiple replicate experiments. For the relative quantification of p53 protein by ELISA and Western blot analysis, tissues from groups of three to five animals were pooled, and multiple pools were statistically compared. For Western blot analysis, autoradiograms were scanned (Deskscan; Hewlett-Packard, Palo Alto, CA) and the digitized data analyzed by computer (Quantiscan software; Biosoft, Cambridge, UK) installed on an IBM-compatible personal computer. Multiple exposures of the same experiments were analyzed so that band intensities represented real values. For ELISA, results were quantitated spectrophotometrically at 450 nm on a microtiter plate reader (reference wavelength: 690 nm) against blank. Data are expressed as the mean ± SEM. One-way analysis of variance (ANOVA) was performed on the data, followed by the Newman-Keuls post hoc test, when appropriate. 

Results

Immunolocalization of p53 protein using mAbs 421 and 248 and the polyclonal antibody FL-393 is shown in Figures 1 2 and 3 . Corneal epithelium demonstrated strong cytoplasmic expression of p53 with mAb 421 and FL-393, but no reaction with mAb 248 (Fig. 2) . The basal epithelial layer and the outer epithelial layers of the entire cornea all stained strongly. Strong epithelial staining continued into the limbus and conjunctiva. The corneal stroma and endothelium, sclera, and conjunctival lamina propria did not react with any of the three anti-p53 antibodies (Figs. 1 2) . 

The lens epithelium exhibited nuclear staining with the mAb 248 and FL-393 antibodies, but not with mAb 421. Not all lens epithelium nuclei stained positively for p53 protein. The flat nuclei of the central and the pre-equatorial zones, situated anterior to the epithelial germinative zone stained positively (Fig. 2 , Anterior Lens mAb 248 and FL-393). The round nuclei of the actual lens germinative zone did not stain for p53 protein (Fig. 2 , Equatorial Lens mAb 248 and FL-393 and Fig. 3 , mAb 248, GZ). The round nuclei of the lens fiber nuclear bow, situated posterior to the epithelial germinative zone, stained strongly positive (Fig. 2 , Equatorial Lens mAb 248 and FL-393, and Fig. 3 , mAb 248, NB). 

The retina (Fig. 1) did not show strong staining with any of the three antibodies, although we have previously shown weak staining in particular retinal layers. ^{15 17}  

Relative quantification of p53 protein expression in specific eye tissues is shown in Figures 4 and 5 . Figure 4 graphically shows the results of the ELISA assay (n ≥ 5 replicate experiments). Whole cornea extracts, which included both the p53-negative stroma and endothelium substructures (Fig. 2) , had very high total p53 protein levels by immunohistochemistry (>10 times the p53 concentration of sclera; compare Figs. 2 and 4 ). This is reinforced by the positive corneal epithelium data in the ELISA analysis (Fig. 4 ; CEp). The conjunctiva, including the p53-negative lamina propria substructure, also exhibited very strong expression of p53 protein (more than six times scleral levels, Fig. 4 ). The retina and iris showed low expression of p53 (Figs. 1 2) . The lens, which consists of a small amount of p53-positive lens epithelium and a much larger concentration of p53-negative lens fibers, showed a small reproducible difference on ELISA analysis (Fig. 4 , compare lens capsule [LensC] to lens nucleus [LensN]). Nonetheless, the small amount of p53 detected in the lens capsule probably was diluted by other p53-negative lens capsule tissue and substantial levels of lens fiber protein contaminating this fraction. 

Figure 5 graphically shows the results of the Western blot analysis (n ≥ 5 replicate experiments). Western blot analysis, as expected, mirrored the ELISA assay, except that the retina showed very low concentrations of p53 protein. Although mAbs 248 and 421 reacted differently in histologic preparations, Western blot analysis showed similar reaction patterns with both mAbs, because this method studied the expression of denatured proteins. Both p53 mAbs reacted specifically at the 53-kDa band on Western blot analysis, with no evidence found for the expression of other members of the p53 protein family (i.e., p63 and p73). This is significant, because these antibodies can detect other members of the p53 family of proteins. ²³  

Discussion

This study demonstrated high-level expression of the p53 protein in the cytoplasm of the corneal and conjunctival epithelium and in the nuclei of a population of lens cells. In addition to immunohistochemical methods to identify expression of p53, in this study, we also corroborated our results by two quantitative biochemical techniques (ELISA and Western blot analysis) on separated ocular tissues. In this study as well as in our previous studies, we also have shown, using immunohistochemistry, significantly lower levels of expression of p53 protein in the retinal cell bodies, ^{15 17} which was confirmed in the current studies by Western blot analysis and ELISA. 

Lens epithelium nuclei of the pre-equatorial and nuclear bow strongly expressed p53, whereas nuclei of the actual lens germinative zone did not. Nuclei posterior to the germinative zone (lens fiber nuclei) are known to be undergoing terminal differentiation with the loss of their organelles, including their nuclei, ²⁴ whereas the cells of the germinative zone are actively dividing. The germinative zone is protected from UV irradiation by its location behind the iris. On the contrary, the nuclei of the pre-equatorial zone are exposed to UV light, thus showing the possibility of a protective function for expression of p53 in that area. That is, if DNA is damaged by the incoming UV light, an appropriate signal could then cause p53 to induce the cell to enter apoptosis. 

The ELISA and Western blot assays measure the p53 protein concentration as a proportion of total tissue protein content. Because the separated ocular tissues consist of both p53-positive and -negative cell types, the p53-positive cells are effectively diluted by the p53-negative cells. For example, p53-positive lens epithelium contributes little to the total lens protein, the bulk of which is composed of p53-negative lens fibers. Hence, these quantitative assays on whole tissue samples actually underestimate the absolute p53 protein levels present in certain cell types within the tissue. 

Although high levels of p53 protein in the epithelial cytoplasm of pingueculae, pterygia, and limbal eye tumors have recently been demonstrated, ^{7 8 9 10} expression of p53 in normal human conjunctival epithelium has been more controversial. Tan et al. ⁸ reported high expression of p53 in seemingly normal conjunctiva in eyes in biopsy tissue obtained during pterygium surgery. Other investigators ^{7 10} report no expression in normal human conjunctival and corneal epithelium. One of these investigators ¹⁰ suggested that these differences may be explained by the use of different anti-p53 protein mAbs in the respective studies. In the present study, different p53 antibodies reacted differently to native expression of p53 protein in different eye structures, when only immunohistochemical methods are used (Figs. 1 2) . On more quantitative approaches, using denatured protein extracts (Figs. 4 5) , these differing p53 antibody results read similarly. 

We found differential staining of mAbs 421 and 248 on the immunohistochemical preparations and identical staining of these same antisera on Western blot analysis of protein extracts derived from identical eye structures. The antibody mAb 248 binds to a N-terminal epitope of p53, ¹⁸ whereas mAb 421 binds to a C-terminal epitope ¹⁹ (Fig. 6) . The nonreactivity of mAb 248 in native corneal and conjunctival epithelium, as seen by using immunohistochemistry techniques, may indicate the presence of an N-terminal p53 binding protein (Fig. 6) , such as MDM2 ²⁵ a direct modulator of p53, ²⁶ covering the antibody epitope. On denaturation, as performed in the Western blot assay, any protein binding to p53 would most probably be denatured and thus removed from its binding site, so that all the internal (hidden in native protein) p53 epitopes would be exposed. This would explain the differential antibody reactivity seen with immunohistochemistry and equal reactivity of both mAbs on Western analysis. A parallel but inverted possibility could explain why mAb 421, but not mAb 248, was negative on immunostaining of normal lens preparations. Again, on Western analysis, both antisera reacted equally well with denatured p53. In this case, mAb 421, which recognizes a C-terminal p53 epitope, may not bind to p53 in the nuclei of lens epithelia cells, because of the presence of another, previously unknown, p53-binding protein covering this epitope. This pattern of p53-binding proteins covering different parts of the p53 protein in a tissue-specific way may explain the variant results that others have reported. ^{7 9 10}  

The p53 protein is known to be a nuclear transcription factor that plays a role in apoptosis and cell proliferation. There have been reports of an inactive isoform found in the cytoplasm. ²⁷ In this study, we have found large amounts of cytoplasmic p53 in the corneal and conjunctival epithelium as well as nuclear p53 in the lens epithelium. This may represent examples of active and inactive p53, or it may represent a previously unknown additional function for p53. Of particular interest, high expression of p53 was seen particularly in the cytoplasm of corneal and conjunctival epithelium, which are tissues with a high turnover rate. Because of its high turnover rate, a nuclear localization of p53 would have been the expectation in these tissues. 

The high levels of expression of p53 we have seen in the normal murine eye would seem to require some identifiable function. Recently, Ikeda et al. ²⁸ have shown severe ocular abnormalities in one strain of p53-deficient mice but not in a second strain. Furthermore, others ^{29 30} have demonstrated that p53 played a role in apoptosis of pathologic retina. To the best of our knowledge, no p53 role within “normal” cornea, conjunctiva, or lens epithelium has been determined. Obviously, more work is necessary to understand the function of the p53 protein in the various ocular tissues. 

Recently, two proteins very closely related to p53—p73 and p63—have been identified. ^{31 32 33 34 35} Both p73 and p63 can induce cell-cycle arrest and apoptosis, as does p53, suggesting that they also may be tumor suppressors. The p63 protein is produced by the TRP63 gene located at chromosome 3q27 to 29. ³² This protein is highly expressed in the basal region of many epithelial tissues and has been shown to be required for limb, craniofacial, and epidermal development. ^{34 35} Less is known regarding the function and physiology of the p73 protein. ³³ In the current studies, neither of these proteins was seen in any of the eye structures studied. 

The involvement of p53 in pathologic states has been well established; however, much less is known about its function in normal adult tissues. To clearly understand how p53 functions in the regulation of the cell cycle, it is important to identify all its possible functions and all the proteins interacting with it. The present study has demonstrated possible new understandings regarding the expression and thus the function of p53 in normal adult ocular tissues. There is a need for additional investigations in this and other tissue, to obtain a more comprehensive understanding of the role of the p53 protein and its regulation, as part of its normal physiological function. 

 Supported by the Center for Immigrant Scientist Absorption of the Ministry of Absorption, State of Israel (YT, GW), and the Fund for the Promotion of Research at the Technion, Haifa, Israel (GW, OZ).

 Submitted for publication March 30, 2001; revised January 15, 2002; accepted February 1, 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.

 

Figure 1.

 

Whole murine eye immunolocalization of p53 protein by fluorescence immunohistochemistry. The cornea (C), iris (I), lens (L), retina (R), and optic nerve (ON) are shown in these axial sections. Corneal tissue showed strong staining (arrow) with the anti-p53 mAb 421 compared with both control and anti-p53 mAb 248–stained sections.

Figure 2.

 

Immunolocalization of p53 protein in murine eye tissues with the antibodies noted. A hematoxylin-stained slide for each structure is also shown.

Figure 3.

 

Immunolocalization of p53 in murine eye lens. Lens capsule (LC), germinative zone (GZ), and nuclear bow (NB) are labeled in hematoxylin, control, and anti-p53 mAb 248–stained sections. The round nuclei of the lens fiber nuclear bow stained strongly for p53.

Figure 4.

 

Quantification of tissue-specific p53 protein expression in different eye tissues using a pan-p53 antibody in an ELISA. The data are presented as the mean of five or more replicate experiments ± SEM. (ANOVA: P ≤ 0.0001). Cor&Lim, whole cornea and limbus; CEp, corneal epithelium; Conj, conjunctiva; LensC and LensN, lens capsule and lens nucleus, respectively.

Figure 5.

 

Quantification of p53 protein expression in various eye tissues by Western blot analysis. The histogram represents the mean of at least five replicate experiments. Below are representative gel images representing such a comparison. Conj, conjuctiva; NC, negative control; p53-M, p53-positive control cell line ²²(Clone 314).

Figure 6.

 

Schematic of location of p53 antibody epitopes and putative p53-binding proteins. The DO1 antibody was used in a study by Dushku et al., ⁹ while the DO7 antibody was used by Tan et al. ¹⁰

The authors thank Bella Levin for histologic advice.