Severe Ocular Abnormalities in C57BL/6 but Not in 129/Sv p53-Deficient Mice

Abstract

purpose. To demonstrate the importance of genetic background interaction on the development of ocular phenotypes in p53-deficient mice.

methods. Eyes of adult mice, homozygous and heterozygous for the p53 gene disruption in the 129/SvJ and C57BL/6J (B6) genetic backgrounds, and their F1 progeny were examined by indirect ophthalmoscopy and by light microscopy.

results. Indirect ophthalmoscopy revealed unilateral or bilateral vitreal opacities, fibrous retrolental tissue, and retinal folds in adult B6 mice but not in 129/Sv mice homozygous for a p53 null mutation. In B6 p53−/− mice, blood vessels extended from the peripapillary inner retina through the posterior vitreous and into the retrolental membrane. Optic nerves were hypoplastic.

conclusions. These findings indicate that alleles from the B6 background contribute to the aberrant ocular phenotypes observed in p53 deficiency. They also suggest that p53 or the pathway in which it functions may be important for normal eye development.

The Trp53 (p53) gene is best known for its role as a tumor-suppressor gene. More than half of tumors studied exhibit loss of heterozygosity or alterations in p53. It is also involved in the cellular response to stress including hypoxia, ionizing radiation, and teratogens. Cellular p53 response is mediated through transcriptional activation or direct signaling to regulate cell cycle control, differentiation, apoptosis, and/or angiogenesis (reviewed in ^{Refs. 1 2 3 4} ). Yet, despite its apparent central function in a number of important basic cellular processes, earlier gene-targeting studies showed that p53 deficiency in mice leads to tumorigenesis but not to developmental abnormalities. ^{5 6} Later, it was reported that a subset of homozygous p53 null mutants had exencephaly and that these mice most often died during embryogenesis, ⁷ suggesting that p53 was indeed important in normal development. 

In our exploration of the role of p53 in the apoptotic process of retinal degeneration, we found that p53−/− mice congenic on the C57BL/6J (B6) background have severe ocular abnormalities, whereas most 129/Sv p53−/− mice have normal fundus characteristics. In this report, we present a comparison of clinical and histologic effects of the p53 null mutation on ocular phenotypes in two genetic backgrounds, B6 and 129/Sv. 

Materials and Methods

Animals

The p53 null mutation was prepared in strain 129-derived D3 ES cells, which were then microinjected into C57BL/6 blastocysts. ⁶ The germ-line transformants were crossed either to C57BL/6 or a 129/Sv strain (R. Jaenisch); F1 mice were then intercrossed to produce homozygous null mutants. ⁶ The 129/Sv-Trp53 ^tm1Tyj line was imported to The Jackson Laboratory, and the mutation has been moved by repeated backcrossing to three different genetic backgrounds: C57BL/6J (B6), BALB/cJ, and C3H/HeOuJ. The p53 null mutation has been backcrossed to B6 for 10 generations, and these mice are currently being intercrossed. For this study, 129/Sv-Trp53 ^tm1Tyj (129/Sv p53−/−), congenic C57BL/6-Trp53 ^tm1Tyj (B6 p53−/−); F1 homozygous null mutants (B6.129/Sv p53−/−), heterozygous (B6.129/Sv p53+/−) and wild-type (B6.129/Sv p53+/+) controls from a B6 p53+/− X 129/Sv p53+/− cross; and offspring from the backcross, F1(B6.129/Sv p53−/−) X B6 p53+/− or F1(B6.129/Sv p53−/−) X 129/Sv p53+/−, were obtained from The Jackson Laboratory Production Facility or were bred in our research colony. DNA was isolated from tail tips of mice and the IMR013, 5′-CTTGGGTGGAGAGGCTATTC-3′; IMR014, 5′-AGGTGAGATGACAGGAGATC-3′; IMR336, 5′-ATAGGTCGGCGGTTCAT-3′; and IMR337, 5′-CCCGAGTATCTGGAAGACAG-3′ oligonucleotide primers were used in polymerase chain reaction (PCR) amplification to detect mice homozygous for the p53 null mutation. PCR products were resolved on a 3% metaphor/1% agarose gel and visualized with ethidium bromide staining. All mice were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Clinical Examination

Eyes of mice were dilated with atropine before examination by indirect ophthalmoscopy with a 60- or 78-D aspheric lens. Fundus photographs were taken with a Kowa fundus camera (Kowa Company, Tokyo, Japan) using a Volk superfield lens (Volk Optical, Mentor, OH) held 2 in. from the eye. The highest flash intensity was used with 400 ASA film. 

Tissue Preparation

Adult mice were anesthetized with tribromoethanol and perfused with phosphate-buffered saline (PBS) followed by 1% paraformaldehyde-2% glutaraldehyde-0.1 M cacodylate. Enucleated eyes were stored in this ice-cold fixative for 24 hours, embedded in hydroxyethylmethacrylate, and sectioned in a plane to include the ora serrata and optic nerve. Alternatively, enucleated eyes were placed in Bouin’s fixative for 24 hours, embedded in paraffin, and sectioned as above. Sections were stained with hematoxylin and eosin. For optic nerve preparations, adult mice were killed by CO₂ asphyxiation and then after removal of the brain, heads were placed in phosphate-buffered glutaraldehyde-paraformaldehyde ⁸ for 48 hours and transferred to 1 M phosphate buffer. The optic nerves extending from the orbits to the optic chiasm were dissected and postfixed with 1% osmium tetroxide, processed under standard procedures, and embedded in a 1:1 Epon–Araldite mixture. Cross sections cut at 1 μm were stained with 1% paraphenyldiamine in a 1:1 propanol-methanol mixture. 

Results

Incidence of Ocular Abnormalities in 129/Sv p53−/− and B6 p53−/− Mouse Stocks, Assessed by Clinical Examination

Thirteen homozygous null and 10 heterozygous 129/Sv-Trp53 ^tm1Tyj mice examined by indirect ophthalmoscopy were normal. Ten B6 p53+/− and three B6 p53+/+ control mice were also examined, and no abnormalities were observed. However, of 65 B6 p53−/− mutants examined, 57 exhibited pigmented and nonpigmented fibrous retrolental tissue, vitreal opacities and/or retinal folds in both eyes, 5 mice were unilaterally affected, and 2 appeared normal by fundus examination but not by histologic examination. Cataracts were present in 14 of the 130 eyes examined. Of the ocular abnormalities, vitreal opacities and fibrous retrolental tissue were the most commonly observed (Fig. 1 Left and Right). Aberrant ocular phenotypes were observed as early as 14 days of age, the earliest time point examined. 

Histologic Examination

Both 129/Sv and B6 homozygous p53 null mutants had abnormally dilated peripheral retinal blood vessels. B6 p53−/− mice also exhibited abnormal vessels extending from the peripapillary inner retina (most often from vessels on the surface of the optic nerve) through the posterior vitreous and into the retrolental membranes. Both nonpigmented and pigmented membranes were observed. The pigmented retrolental membranes were presumably the result of migration of mobilized retinal pigment epithelial cells. Vitreous traction extending from the retinal surface or optic nerve toward the retrolental membrane was also observed (Fig. 2 A). In severe cases, erosion of the posterior lens capsule with the beginnings of extrusion of lens cortex was often noted (Fig. 2B) . The vitreal opacities revealed by fundus examination appeared to result from the accumulation of fibrous and vascular debris in the vitreous. 

B6 p53−/− mice also exhibit unilateral or bilateral hypoplastic optic nerves. Examination of longitudinal sections of the retrolaminar optic nerve revealed fewer nerve fibers, as evidenced by the narrower columns between pial septae (Fig. 2C) . In addition, the pial septae in many areas were disorganized when compared with similar sections from 129/Sv p53−/− or F1(B6/129Sv p53−/−) mice. Cross-sectional analysis of the optic nerves from B6 p53−/− mice at 10 weeks of age revealed optic nerve atrophy with reduction in the number of myelinated nerve fibers, some of which demonstrated degeneration (Figs. 2D 2E) . The degree of degeneration varied between optic nerves from the same animal and among animals of the same age. 

The eyes of 129/Sv p53−/− mice examined were either normal or exhibited a thin blood vessel extending from the optic nerve head toward the lens. Although most of these vessels extended into the vitreous, a few reached the lens surface. No other abnormalities similar to those in B6 p53−/− mice were observed. 

Genetics of the Ocular Phenotypes in B6-Trp53 ^tm1Tyj Mice

To characterize the mode of inheritance of the ocular phenotypes, a number of crosses between mice homozygous or heterozygous for the p53 null mutation from the 129/Sv and B6 background were carried out. Fourteen of 17 F1 progeny homozygous for the p53 null mutation and all heterozygous progeny from a B6 p53+/− X 129/Sv p53+/− cross, exhibited the near-normal fundus characteristics of the 129/Sv p53−/− parental strain. The remaining three p53−/− F1 progeny exhibited either unilateral vitreal opacities, fibrous retrolental tissue, or retinal folds. 

F1 progeny with no ocular abnormalities were then either backcrossed to the resistant 129/Sv-Trp53 ^tm1Tyj or susceptible B6-Trp53 ^tm1Tyj parental strain. In the F1(B6.129/Sv p53−/−) X 129/Sv p53+/− backcross, all 39 p53−/− backcross mice examined by indirect ophthalmoscopy were normal. However, although none of the heterozygous offspring in the F1(B6.129/Sv p53−/−) X B6 p53+/− backcross were affected, 26 of 30 p53−/− backcross offspring exhibited vitreal opacities, fibrous retrolental tissue, and/or retinal folds. 

Discussion

Ocular Abnormalities

Ocular abnormalities in B6 p53−/− mice are reminiscent of the histopathologic changes reported in human eyes with persistent hyperplastic primary vitreous (PHPV). Unilateral malformations including vitreal vasculature, retrolental tissue, and retinal fold and detachment are usually observed in these patients. In some cases, unilateral optic nerve head hypoplasia is also observed (reviewed in ^{Ref. 9)} . No obvious cause or hereditary influence has been reported for human PHPV. The B6 p53−/− mouse may provide an excellent genetic animal model for human PHPV to study the mechanisms underlying this disease. 

The vitreal vascularization observed in adult p53-deficient mice may in part be remnants of the hyaloid system that vascularizes the embryonic and early postnatal vitreous. This hyaloid vasculature normally regresses by adolescence, presumably through apoptosis and removal by intraocular macrophages known as hyalocytes. ¹⁰ Although p53 is not necessary for all forms of apoptosis, ¹¹ indications are that apoptosis in the retina may be p53 dependent. ^{12 13} p53 is thought to function early in the apoptotic pathway by downregulating bcl2 expression and by upregulating bax expression. ^{14 15} Bcl2 promotes cell survival, whereas bax promotes cell death. Conceivably then, hyaloid vessels persist in p53-deficient mice because they never receive the proper p53-mediated signals to regress. In fact, it was recently reported that a delay in apoptosis of the hyaloid vasculature occurs in BALB/c p53−/− mice. ¹³  

Blood vessels originating from the optic nerve and from the retinal surface (ascertained by examination of serial sections through the eye), the proliferation of fibrous tissue in the vitreous and the hypoplastic optic nerve observed in B6 p53-deficient mice cannot be explained solely by p53’s effects on apoptosis. The centrally oriented vascularization, frequently reported in abnormal vascularization in human and animal eyes, poses the possibility that angiogenic factors normally under the control of p53 may diffuse toward the posterior pole of the eye. ¹⁶ p53 is known to regulate two factors important in the balance of angiogenesis. In vitro, p53 deficiency leads to a reduction in thrombospondin 1, an antiangiogenic factor, and an induction of the potent mitogen, vascular endothelial growth factor (reviewed in ^{ref. 17} ). In vivo, this sequence of events could lead to conditions favorable to endothelial cell proliferation and differentiation. 

The hypoplastic optic nerve observed in B6 p53−/− mice appears to be caused by degeneration of rather than abnormal development of the optic nerve. The focal nature of the degeneration suggests that interaction of p53 with factors contained within these particular cells is necessary for their viability. When a severe degeneration of the optic nerve was observed, fewer cells were found in the peripheral ganglion cell layer in comparison with a normal eye (data not shown). The focal loss of the ganglion cell layer was limited to one well-defined retinal area, whereas the opposite side of the section demonstrated a normal number of cells. This observation is consistent with the focal degeneration of the optic nerve, because the ganglion cell axons originating from one region of the retina remain localized in a corresponding part of the optic nerve. ¹⁸  

A terminal dUTP nick-end labeling assay was performed in these eyes to determine whether the ganglion cells die by apoptosis. In five eyes from B6 p53−/− mice with focal optic nerve degeneration, no positively labeled cells were observed (data not shown). This suggests that cells either die by an alternate mechanism, or that the number of apoptotic nuclei at any given time point are so few that they are undetectable by this method. It is important to note that if cell death occurs through an apoptotic mechanism, it does not occur through the p53 pathway in these p53 null mutants. 

Genetic Interaction and the Ocular Phenotypes in B6-Trp53 ^tm1Tyj Mice

The effect of genetic background on the expression of phenotypes in p53 null mutants is not unprecedented; tumor type ⁵ and incidence of embryonic exencephaly ⁷ have been shown to be influenced by the strain on which the mutation is placed. Because vitreal vascularization, proliferation of fibrous tissue, and hypoplasia of the optic nerve were observed in B6 p53−/− mice but not to the same extent in 129/Sv p53−/− mice, we hypothesized that 129/Sv mice must have alleles that can compensate for the loss of p53 function, whereas B6 mice do not. Furthermore, that severe ocular abnormalities were observed only when the backcross was carried out with the B6 p53 null parent suggests that these phenotypes are caused by gene interactions of recessive B6 susceptibility alleles and that dominant 129/Sv alleles are protective. 

The alternative explanation that the ocular abnormalities in the B6-Trp53 ^tm1Tyj mice are the result of a new spontaneous mutation independent of the p53 null muation is an unlikely one. First, some of the abnormalities, although less severe, are also observed in the congenic stock derived from crossing 129/Sv p53−/− into the BALB/c background (now at backcross five). Second, similar phenotypic ocular abnormalities have been observed in an independently derived BALB/c p53−/− stock. ^{11 13} Finally, if the observed abnormalities were caused by a new single-gene mutation, then we would have expected only 15 of 30 p53−/− mice to be affected in the F1(B6.129/Sv p53−/−) X B6 p53+/− backcross, rather than the 26 observed. The distribution of affected-to-unaffected mice in the B6 backcross suggests the interaction of multiple genes. 

The phenotypic differences observed in p53-deficient mice in different background strains allows us the opportunity to perform a modifier screen to identify susceptibility and resistance alleles that interact with p53 in vivo. Such studies are currently under way. 

 Reprint requests: Patsy M. Nishina, The Jackson Laboratory, 600 Main Street, Bar Harbor, ME 04609.

 Supported by grants from The Foundation for Fighting Blindness and Sequana Therapeutics, and Grants EY11996 and EY07758 from the National Eye Institute. Institutional shared services are supported by Grant CA-34196 from the National Cancer Institute Cancer Center grant.

 Submitted for publication June 17, 1998; revised January 7, 1999; accepted February 24, 1999.

 Proprietary interest category: N.

 

Figure 1.

 

Fundus photograph of B6-Trp53 ^tm1Tyj mice homozygous for the p53 null mutation. (Left) A 21-day-old mouse showing fine “snowflake” vitreal opacities (arrowheads) in the posterior vitreous and two smaller falciform folds that lie temporal to the optic nerve (arrow). (Right) A 4-month-old mouse exhibiting retinal folds and a small retrolental pigmented mass (arrow).

Figure 2.

 

Photomicrographs of vitreal neovascularization, retinal dysplasia, and optic nerve hypoplasia in 8- to 10-week-old B6-Trp53 ^tm1Tyj homozygous null mutants. (A) A prominent retinal fold is evident. A delicate fibrovascular membrane extends from the retina to the posterior aspect of the lens, presumably exerting retinal traction (arrow). This fibrovascular bundle did not arise from the optic nerve. (B) Fibrovascular tissue extends from the optic nerve to the lens. The lens capsule is disrupted, and early cortical lens extrusion has occurred (arrow). In both (A) and (B), there is posterior migration of the lens epithelium, indicating cataract formation. (C) Retrolaminar optic nerve. The pial septae are prominent, resulting from nerve fiber loss, and the width of the optic nerve is diminished compared with that of wild-type mice. (D) Cross section of the optic nerve from a 10-week-old F1(B6/129Sv p53−/−) control mouse showing normal optic nerve structure. (E) Cross section of the optic nerve from a 10-week-old B6 p53−/− mouse showing degeneration of a portion of the optic nerve with loss of axons and myelin sheaths. The dark-staining areas (arrow) represent degeneration of myelin sheaths. Note the smaller diameter of the optic nerve compared with (D). Original magnification, (A, B) ×100; (C, D, E) ×400. Scale bar, (A, B) ×50 μm; (C, D, E) ×100 μm.

The authors thank Terry Maddatu for excellent technical assistance and Greg Cox and Barbara Knowles for careful review of the manuscript.