ΔFosB-Induced Cataract

Abstract

purpose. The objective of this study was to investigate a possible relationship between posterior subcapsular cataract (PSC) formation and expression of the transcription factor ΔFosB.

methods. Western blot analysis was performed on bitransgenic NSE-tTA, TetOp-ΔFosB, and single-transgenic NSE-tTA control mice to determine the pattern of ΔFosB expression within the eye. Light and scanning electron microscopy and biochemical analyses were also performed.

results. In mice expressing ΔFosB, cataract developed that initially appeared to be posterior subcapsular and gradually matured to involve the entire lens. The enlarged posterior ends of developing secondary fibers curved away from the visual axis to form an elevated opaque posterior plaque. As a result, posterior suture formation did not occur. At a later time, the attenuated posterior capsule overlying the plaque ruptured and the lens nucleus subluxated into the vitreous. Retinal damage was also observed but only from postnatal day 65, a time when extensive lens degeneration had already occurred. ΔFosB expression was observed well before the detection of morphologic change in both the lens and the retina. Within the lens, ΔFosB expression was found in both the epithelium and fibers. The development of cataracts was a direct consequence of ΔFosB expression and was not due to the disruption of an endogenous gene by transgene integration since cataracts could be prevented by silencing expression of ΔFosB by feeding bitransgenic animals doxycycline (Dox). Moreover, cataracts were observed in bitransgenic mice derived from two independent TetOp-ΔFosB founder lines but not in single NSE-tTA transgenic controls. Cataractogenesis was not a consequence of abnormal development, because mice conceived and raised on Dox to prevent expression of ΔFosB also were subject to formation of PSC when expression of ΔFosB was turned on in adult animals by removing Dox. Examination of biochemical parameters indicated that the earliest change observed was the disruption of calcium homeostasis with a significant increase in Ca²⁺ influx, followed by a gradual but marked decrease in protein content. Significant changes in certain metabolic parameters and protein composition were also observed.

conclusions. The ΔFosB-induced cataract in which the major morphologic early event was the disruption of normal posterior fiber formation, may be a good model for PSC. By identifying ΔFosB-regulated target genes, it should be possible to achieve a better understanding of the molecular mechanisms through which PSC is formed.

The vertebrate lens is composed of anterior lens epithelial cells, which constitute the lens germinal epithelium, and terminally differentiated lens fiber cells. ^{1 2} Lens fiber cells differentiate from epithelial cells as a continuous process throughout the lifetime of an animal. During this process of differentiation, developing lens fibers exhibit a complex pattern of gene regulation that alters their phenotype to that of the mature fiber, which is metabolically and transcriptionally inactive. Ordered differentiation, maturation of fibers, and growth of the lens are essential for preserving lens transparency. Pathologic states that cause cataract may arise in the lens when the intricate coordination of gene expression is interrupted, either in the lens epithelium or in differentiating lens fibers. 

To date, several basic leucine zipper (bZIP) transcription factors have been shown to play an important role in normal lens development. ^{3 4} Transcription factors of the bZIP superfamily are characterized by one conserved region of basic amino acid residues that mediate binding to DNA and by a second conserved region that contains a heptad repeat of hydrophobic leucine residues (known as the leucine zipper) that mediate dimeric protein–protein interactions. ⁵ The bZIP superfamily includes members of the Jun, Fos, CREB/ATF, and Maf families. Many individual bZIP proteins interact with members of their own subfamily or with other bZIP proteins to form active dimeric DNA-binding complexes. ³  

Expression of the lens-specific transcription factor, L-Maf has been shown to participate in both the induction and differentiation of the vertebrate lens. ⁶ Mice with a targeted disruption of the c-Maf gene display defective lens formation that results from a failure of posterior lens fiber elongation and abnormal crystallin gene expression. ^{7 8} Although mice without ATF4 have normal lenses at embryonic day 14.5, they show a deficiency in lens differentiation, and secondary lens fibers do not form. ⁹  

Induction of Fos and Jun family members has been associated with cataractogenesis. ^{10 11} Oxidative stressors such as H₂O₂ and UV irradiation, which are known to induce cataracts, have been shown to cause a rapid and robust induction of c-Fos and c-Jun in lens epithelial cells. ^{12 13} It has been argued that the resultant induction of Fos and Jun heterodimers in the lens serves as one molecular mechanism through which maladaptive changes in gene expression may occur. ^{10 11}  

In the present study, the role of the bZIP protein ΔFosB in lens development and maturation was investigated. ΔFosB is a member of the Fos family of transcription factors. ^{14 15} Although no member of the Fos family can dimerize with itself or with other Fos proteins, ΔFosB is capable of binding all members of the Jun family of bZIP transcription factors. ¹⁴ In addition, ΔFosB has been shown to form complexes with c-Maf. ¹⁶ Finally, through its interactions with Jun and Maf proteins, overexpression ofΔ FosB is predicted to affect CREB/ATF proteins. In this communication it is reported that in bitransgenic mice expressing ΔFosB, cataracts develop that begin as posterior subcapsular cataracts (PSCs). Moreover, by silencing expression of ΔFosB, cataract formation can be prevented. 

There are relatively few animal models for PSCs. They include the spontaneous appearance of PSC in Wistar rats ¹⁷ after intravitreal injection of docosahexaenoic acid ¹⁸ or bacterial endotoxin, ¹⁹ ionizing radiation, ^{19 20} and concanavalin A–induced cataract. ²¹ Of particular interest is the Royal College of Surgeons (RCS) rat model. This model was initially developed to study cataract. ²² However, it has primarily been investigated as a model for retinal degeneration that occurs initially in the outer segments. ^{23 24} In this model, it has been proposed that the PSC is initiated by toxic lipid peroxides released from the degenerating retina. ²⁵  

Recently, Al-Ghoul et al. ²⁶ have reinvestigated PSC in the RCS rat. Based on correlative scanning and transmission electron microscopic as well as light microscopic analyses, they concluded that the PSC results from malformation of the posterior fibers curving away from the polar axis, causing abnormal ordering and stacking of the fibers in the region of the posterior suture branches. Although the role of the retina is not clear, similar observations are reported in the present communication that lead to PSC, followed by total lens involvement, and, finally, nuclear subluxation. Striking changes in protein concentration are also found, suggesting protein degradation that may result from an observed increase in Ca²⁺ influx. Thus, this communication describes the molecular biology, morphology, and biochemistry of bitransgenic mice that express ΔFosB in the lens and retina. 

Materials and Methods

Transgenic Mice

The DNA constructs and pronuclear injections used to create NSE-tTA and TetOp-ΔFosB transgenic mice have been described elsewhere. ²⁷ NSE-tTA transgenic mice derived from founder line A ²⁷ were maintained on an outbred genetic background that contained approximately 50% ICR, 25% C57Bl6, and 25% SJL. Two different TetOp-ΔFosB founder lines, lines 1 and 11, ²⁸ also on an identical outbred ICR/C57Bl6/SJL background, were used in the present studies. In comparison with bitransgenic progeny of line 11 that carry both the NSE-tTA and the TetOp-ΔFosB transgenes (11A N+Δ+), bitransgenic progeny derived from line 1 (1A N+Δ+) expressed considerably higher levels of ΔFosB in all tissues studied. By 3 weeks of age, bitransgenic mice carrying both the NSE-tTA and TetOp-ΔFosB line 1 transgenes (1A N+Δ+) were visibly distinguishable from their single-transgenic NSE-tTA siblings (1A N+Δ−) because their growth was retarded, and they weighed less. The bitransgenic 11A mice, by contrast, did not show these abnormalities. All animals were screened by polymerase chain reaction (PCR) to detect the presence or absence of the NSE-tTA and TetOp-ΔFosB transgenes with published primers exactly as described previously. ²⁷ All animals were treated in accordance with the ARVO Statement on Use of Animals in Ophthalmic and Vision Research. 

To obtain animals for these studies, both homo- and heterozygous NSE-tTA mice (line A) were mated to heterozygous TetOp-ΔFosB mice (line 1 or line 11). Animals fed doxycycline (Dox) do not expressΔ FosB because the Dox causes a conformational change in tTA so that it does not bind the TetOp. When Dox is removed from the diet, the Dox clears from the system, and ΔFosB is gradually expressed. In initial experiments, breeding was performed in animals not treated with Dox to ensure that all bitransgenic animals expressed the ΔFosB transgene. In some experiments, breeders were treated with 50 or 200 μg/ml Dox (Sigma, St. Louis, MO) that was dissolved in distilled water containing 5% sucrose. At weaning (3–4 weeks of age), half the animals were taken off Dox to activate expression of ΔFosB, whereas sibling controls continued receiving Dox to suppress expression. It was found that retarding the expression of ΔFosB delayed the development of cataract. However, the development of cataract and the reported changes in biochemistry and morphology correspond. Thus, animals never treated with Dox with severe posterior opacities and marked changes in biochemistry at 11 to 12 weeks of age corresponded to animals approximately 22 weeks of age in which Dox was removed from the diet approximately 10 to 11 weeks before experimentation. In all the morphologic and biochemical studies reported in this communication, the animals used were never treated with Dox; when Dox was used in the molecular biological studies, it is specifically mentioned. 

Western Blot Analysis

Mice were killed by cervical dislocation or by CO₂ asphyxiation. The eyes were rapidly enucleated and placed in 50 mM ice-cold phosphate buffer. For whole-eye preparations, the entire eye was placed in 1 ml of 1% sodium dodecyl sulfate (SDS), and homogenized with a sonicator. ²⁷ Samples were centrifuged at 15,000g for 5 minutes; supernatants were then aliquoted and frozen for subsequent analysis. To obtain cornea, lens, retina, and retinal pigmented epithelium (RPE) preparations, tissue was separated under a dissecting microscope, homogenized in 100 to 200 μl of 1% SDS, and centrifuged for 5 minutes, as just described. For epithelial cell preparations, two lenses were used, the capsule-epithelium was removed, combined, and placed in 100 to 200 μl 50 mM phosphate buffer (pH 7.0) containing 0.15% Triton X-100. This preparation was homogenized at 0°C, centrifuged (Model 5402; Eppendorf, Fremont, CA) at 14,000 rpm for 5 minutes at 4°C. After centrifugation, protein concentrations were determined by the Lowry method. 

Aliquots of supernatant containing 50 μg total protein were mixed with SDS stop solution (2% SDS, 10% glycerol, 5%β -mercaptoethanol), boiled for 2 to 3 minutes, and applied to one-dimensional SDS–polyacrylamide gel electrophoresis with 10% acrylamide/0.4% bis-acrylamide or 12.5% acrylamide-0.5% bis-acrylamide in resolving gels. ²⁹ Proteins were transferred electrophoretically onto nitrocellulose, blocked with 2% nonfat dry milk for anti-Fos related antigens (FRA) immunoreactivity or 0.5% nonfat dry milk for anti-FosB immunoreactivity, and incubated in primary antibody followed by horseradish peroxidase–conjugated goat anti-rabbit IgG (1:4000). Immunoreactivity was visualized using enhanced chemiluminescence techniques according to the manufacturer’s protocol (Amersham Life Science; Arlington Heights, IL). The following primary antibodies were used: anti-M-peptide (anti-FRA; 1:4000; ³⁰ ; or anti-N terminus FosB (1:1000; sc048; Santa Cruz Biotechnology, Santa Cruz CA). The two antibodies yielded equivalent results. 

Presence of Cataracts

Litters born to TetOp-ΔFosB transgenic mice crossed to NSE-tTA transgenic mice were checked twice per week for the presence of lens opacities that were grossly visible to the naked eye. Observers were blinded to the genotype of mice. (However, as noted above, 1A bitransgenic animals never treated with Dox were visibly different from their single-transgenic sibling controls) Opacities were later determined to be PSCs. Accuracy of cataract detection was estimated to be no better than 3 days. However, we speculate that early-onset small PSCs were routinely missed by visual inspection and that only mature cataracts that progressed to involve the whole lens were easily detected. The average time until the appearance of grossly visible cataracts is reported as the mean ± SE. 

Preparation of Tissue for Morphologic Studies

Transgenic eyes obtained from bitransgenic 1A mice at varying postnatal ages (n = 3 to 5 animals per group) were used for morphologic studies. Age-matched mouse eyes from single-transgenic animals expressing only tTA (n = 0 to 4 animals per group) were used as controls. Transgenic and control eyes were harvested and processed in the same manner. Briefly, animals were killed by CO₂ asphyxiation, the eyes were enucleated, and the corneas were removed. The remaining orbit was immediately placed into a fixative of 2.5% glutaraldehyde in 0.07 M sodium cacodylate buffer at pH 7.2. Tissue was fixed at room temperature for 2 to 3 days, with fresh fixative changes daily. After overnight washing in 0.2 M sodium cacodylate buffer, lenses were removed from the posterior portion of the eye containing the retina. The axial lens dimensions were measured under a dissecting microscope (Carl Zeiss, Thornwood, NY), and the lenses were photographed. From each animal, one lens was processed for light microscopy (LM), whereas the contralateral lens was processed for scanning electron microscopy (SEM). Retinas were processed for LM. 

Light Microscopy

Lenses and retinas were processed in an identical manner. Tissue was postfixed overnight in 1% aqueous osmium tetroxide at 4°C, then washed in cacodylate buffer, and dehydrated through a graded ethanol series to propylene oxide. Tissue was infiltrated and flat embedded in epoxy resin. Embedded lenses were sectioned along the optic axis with a glass knife. Retinas were sectioned parallel to the sagittal plane through or adjacent to the optic nerve. Sections 1 to 2 μm thick were mounted on glass slides and stained with 1:1 mixture of methylene blue and azure II. Light micrographs were taken on a photographing microscope (Vanox AHBS3; Olympus, Melville, NY) equipped with a 35-mm camera. Color slides were digitized using a scanner (Sprint Scan 35; Polaroid, Bedford, MA) and processed by image analysis software (PhotoShop ver. 5; Adobe, San Jose, CA) on a Pentium PC platform (Intel, Mountain View, CA). 

Scanning Electron Microscopy

To expose the fibers and suture patterns, lenses were dissected as previously described. ²⁶ Briefly, the capsule and superficial fibers were peeled away from the lens around its diameter, resulting in a lens core and several crescent-shaped fiber peels with the capsule on the concave surface. Both the fiber peels and the remaining lens core were collected and processed for SEM. 

Specimens were postfixed in 1% aqueous OsO₄ in 4°C overnight, washed in cacodylate buffer, and dehydrated through a graded ethanol series. After overnight dehydration in absolute ethanol, the alcohol was replaced with a graded ethanol/Freon 113 series (Dupont, Wilmington, DE) to 100% Freon 113. Specimens were dried in Freon 23 in a critical-point drying apparatus (CPD 020; Balzers, Hudson, NH), secured on aluminum stubs with silver paste, sputter coated with gold, and examined in a scanning electron microscope (JSM 35c; JEOL, Peabody, MA) at 15 kV. 

SDS Gel Electrophoresis

Material was prepared for electrophoresis in the following manner. The lens was carefully removed and homogenized at 0°C in 100 to 200μ l of 50 mM phosphate buffer (pH 7.0) containing 0.15% Triton X-100. The preparation was centrifuged (model 5402; Eppendorf) at 14,000 rpm for 5 minutes at 4°C. For epithelial cell preparations used for Western blot analysis, two lenses were used. The capsule-epithelium was removed, combined, and treated as described earlier. After centrifugation, the supernatant was mixed with an equal volume of 2× sample buffer, as described by Smith, ³¹ and boiled for 2 to 3 minutes. Protein (50 μg) was then used for SDS gel electrophoresis, as described by Laemmli ³² and modified by Wang and Spector. ³³  

Enzyme Assays

Glutathione peroxidase (GSHPx), oxidized glutathione reductase (GSSG Red) and catalase were determined as described by Spector et al. ³⁴ Reduced nicotinamide adenine dinucleotide phosphate (NADPH) quinone oxidoreductase was assayed as described previously ³⁵ with the following modifications. The assays were conducted at 37°C in a final volume of 0.5 ml with 25 μM 9,10-phenanthrenequinone as substrate. One mouse lens was homogenized in 100 μl of 0.1 M Tris (pH 7.8) and 0.2 mM EDTA at 0°C and then centrifuged at 14,000 rpm for 8 minutes at 4°C (model 5402; Eppendorf ), and 70 μl of the supernatant was used for assay. 

Glutathione-S-transferase (GSH-S-transferase) was assayed as described by Habig and Jakoby ³⁶ using 1.0 mM 1-chloro-2,4-dinitrobenzene as substrate in a total volume of 0.5 ml at 37°C. The mouse lens was homogenized at 0°C in 100 μl of 0.1 M phosphate buffer (pH 6.5). The homogenate was centrifuged at 14,000 rpm (model 5402; Eppendorf) for 8 minutes at 4°C, and 70 μl was used for assay. 

⁸⁶Rb Uptake

⁸⁶Rb uptake was conducted in the following manner. After careful dissection, the lenses were preincubated for approximately 2 hours in 1.5 ml of medium 199 with Earle’s salts and no phenol red (M3769; Sigma), 25 mM HEPES (initially pH 7.3), 100μ g/ml glutamine, and 0.9 g/l NaHCO₃ (300 ± 3 milliosmoles). Final pH, after equilibration with 5% CO₂-95% air, was 7.0. The lenses were then incubated in 490 μl of the described medium ±1 mM ouabain at 37°C for 5 minutes, and then 10 μl of 0.1 μCi/μl ⁸⁶RbCl was added, and the incubation was continued for 1 hour. After this incubation, the lenses were washed rapidly three times with room temperature isotonic saline, gently blotted, weighed, and homogenized in 0.25 ml of 0.1 M NaOH. The homogenization tube was washed with another 0.25 ml of base. The combined aliquots were then counted in 5 ml Ecolite (ICN, Costa Mesa, CA). 

Ca²⁺ Influx

For Ca²⁺ influx measurements, the lens was preincubated in 1.5 ml of medium 199 with the additions described earlier at 37°C for 1 to 2 hours. It was then incubated in 500 μl of the same buffer containing 1 uCi ⁴⁵CaCl₂ at 37°C for 2 hours. The lens was then prepared for counting as described for ⁸⁶Rb uptake. For inhibition of the Ca-adenosine triphosphatase (ATPase), N-(6-aminohexyl)-5-chloro-1-naphthalenesulfon-amide (W7), a calmodulin antagonist was used. ³⁷ W7 has been shown to inhibit lens Ca-ATPase by approximately 90%. ³⁸  

NP-SH, protein, ¹⁴C-choline uptake, and[ ³H]thymidine incorporation were determined as described by Spector et al. ³⁴  

Results

ΔFosB Expression and Cataract Development

The tetracycline-regulated gene expression system, ³⁹ was used to inducibly express the bZip transcription factor ΔFosB, a truncated splice variant of FosB that is missing most of a putative C-terminal transactivation domain. ⁴⁰ This is a dual-component system that relies on the function of two transgenes (Fig. 1-I ). The first transgene uses a crippled promoter TetOp, which is too weak to drive transcription of the downstream ΔFosB cDNA on its own. The second is a synthetic transcription factor tTA, which was placed downstream of a tissue-specific promoter, the rat neuron-specific enolase (NSE) promoter. In tTA-expressing cells (a restricted subpopulation determined by the tissue specificity of the NSE promoter fragment), the TetOp may be activated to drive transcription of the downstream ΔFosB cDNA after binding of tTA. By feeding mice tetracycline or its more lipophilic analog doxycycline (Dox), it is possible to induce a reversible conformational change in tTA that shifts the protein into a DNA nonbinding state. ⁴¹ When Dox treatment is discontinued and mice are given untreated water, systemic clearance of Dox begins. This process causes tTA to revert to its DNA-binding state, resulting in transcription of the ΔFosB transgene (Fig. 1-I ). 

In 29 (91%) of 32 1A N+Δ+ mice, conceived and raised without Dox to ensure expression of ΔFosB (beginning at embryonic day 10.5, the earliest time that the NSE promoter is active ⁴² ), bilateral PSCs developed (Fig. 1-II ). With microscopy, light scattering could be detected as early as 25 days. None of the 115 single-transgenic or wild-type sibling control mice had cataracts develop in either eye over the 4-month period during which they were studied (Table 1) . Development of cataract was associated with the expression ofΔ FosB. Thus, none of the 20 bitransgenic 1A N+Δ+ mice conceived and raised with Dox treatments (either 50 μg/ml or 200 μg/ml) showed formation of cataracts during the 56 to 168 days in which they were observed. Moreover, cataractogenesis was not dependent on developmental expression of ΔFosB. Cataracts developed in later life (mean onset, 109 ± 4 days) in six of six bitransgenic 1A N+Δ+ mice that were conceived and raised with 200-μg/ml Dox treatments but stopped receiving Dox at weaning. These findings demonstrate that the pathogenesis of cataract formation induced by ΔFosB was not limited to unique events occurring in the embryo, but rather could occur in adult animals. 

ΔFosB-expressing mice derived from a second TetOp-ΔFosB founder line, founder number 11 (11A N+Δ+ mice), also had bilateral development of cataracts. Once again, in 11A mice that were incapable of expressing the ΔFosB transgene, either because they were receiving Dox or because they inherited only a single transgene, cataracts did not form. It should be noted that in 11A N+Δ+ mice, cataractogenesis occurred at a much slower rate than in 1A N+Δ+ mice. The observation that two independent TetOp-ΔFosB founder lines have the potential to form cataracts makes it exceedingly unlikely that cataract formation is related to the site of chromosomal integration of the TetOp-ΔFosB transgene or to the disruption of an endogenous gene. Rather, expression of ΔFosB itself is responsible for the appearance of the cataract phenotype. Cataracts formed in both 11A and 1A mice, but we chose the 1A mice for further study, because they exhibit a more severe phenotype. As a first step in studying the link between ΔFosB expression and cataract formation, expression of the ΔFosB transgene in the eye was examined. Western blot analysis was performed on whole-eye extracts to detect expression of ΔFosB, which migrates as 37-, 35-, 33-, 29-, and 28-kDa isoforms. ⁴³ As shown in Figure 2A , 8-week-old bitransgenic 1A N+Δ+ mice that were never exposed to Dox exhibited a dramatic induction of all ΔFosB isoforms relative to bitransgenic N+Δ+ control mice that were conceived and raised with Dox. In fact, within the eye of bitransgenic animals never exposed to Dox, ΔFosB was induced at all time points studied, the earliest of which was 1 week after birth (data not shown). It should be noted that although no ΔFosB immunoreactivity appears to be present within the whole-eye homogenate of 1A N+Δ+ mice treated with Dox, there was a small but significant leak of ΔFosB that is consistent with low levels of expression of ΔFosB observed in single-transgenic line 1 (N−Δ+) mice. ⁴⁴  

Having formally established that bitransgenic mice express ΔFosB in the eye, the next task was to localize the tissue in which this induction was occurring. The eyes of 8-week-old 1A N+Δ+ bitransgenic and N+Δ− single-transgenic siblings were dissected into cornea, lens, retina, and RPE. Western blot analysis of these tissues (Fig. 2B) revealed that the highly stable 37- and 35-kDa ΔFosB isoforms ^{43 45} were expressed at high levels in the retina, with lower levels observed within the lens, of bitransgenic mice. The 33-kDa and 30- to 28-kDa isoforms were also present in retina and lens. A faint band corresponding to the 35-kDa isoform of ΔFosB was also visible within the RPE. There was no detectable expression of ΔFosB isoforms within the lens, retina, cornea, or RPEs of single-transgenic N+Δ− mice (Fig. 2B) . 

The lens of 6-week-old bitransgenic 1A N+Δ+ mice were either kept intact as a whole entity or separated into lens capsule and lens fiber. Western blot analysis of these tissues determined that ΔFosB expressed in the whole lens was attributable to expression in both lens fibers and lens epithelium (Fig. 2C) . Because mature lens fibers have lost their nuclei and are transcriptionally inert, the persistence ofΔ FosB immunoreactivity probably indicates either that the NSE promoter is active during terminal differentiation or that the protein is made in lens cells before differentiation and persists in that tissue. The continued presence of ΔFosB in mature lens fibers would be consistent with a low rate of protein turnover in the lens and a long half-life of the higher relative molecular mass ΔFosB isoforms, estimated to be greater than what is found at 1 week in vivo in brain and in vitro in neuronal cell culture. ^{43 45} A time course of ΔFosB expression in the lens appears in Figure 2D . As expected, there was no induction of ΔFosB isoforms in the lens of 1A N+Δ− single-transgenic mice. In bitransgenic 1A N+Δ+ never exposed to Dox, there was a significant induction of all ΔFosB isoforms in the lens of 2.5-week-old mice that diminished over time. Lenses from 12-week-old bitransgenic mice still had most of the ΔFosB isoforms, but at much lower levels. Because the lenses selected at 12 weeks were still intact and exhibited normal permeability, the decrease in the level of ΔFosB is probably due to either decreased synthesis, increased degradation, or both of these factors. 

Early States of Morphologic Abnormality

Lenses from control (N+Δ−) animals (i.e., those that carry the tTA gene but not the ΔFosB gene), had no evidence of opacity or structural abnormality at any of the ages evaluated (18, 25, 35, 50, 65, and 95 days) when killed. Under a stereo dissection microscope, wild-type lenses were observed to have typical anterior and posterior Y sutures (Fig. 3A ), and the gross shape of asymmetric, oblate spheroids (Fig. 3C) . Light microscopic (LM) examination of thick sections obtained along the visual axis revealed the characteristic monolayer, low cuboidal lens epithelium fiber differentiation resulting in the formation and elongation of progressively longer and more uniformly shaped fibers organized into concentric shells (data not shown but identical with Figs. 4A 4C ). 

In contrast, lenses from N+Δ+ mice (i.e., those that carry both the tTA gene and the ΔFosB gene) were only observed to have normal structure through postnatal day 18. By postnatal day 25, these lenses had speckled posterior subcapsular opacities (Fig. 3B) and the abnormal gross shape of an almost symmetric, prolate spheroid (Fig. 3D) . Correlative LM examination of thick sections obtained along the visual axis of these lenses revealed a normal anterior epithelium (Fig. 4A) and bow region (Fig. 4C) , but the posterior portions of fibers were atypically enlarged and curved away from the polar axis toward the vitreous (Fig. 4F) . Similarly, at 35 days after birth, the anterior and bow regions remained undisturbed, whereas the abnormal enlargement and curvature of the posterior fiber ends increased (Fig. 4G) . It is important to note that cell nuclei were observed neither within the plaque nor beneath the posterior capsule from the bow region to the posterior pole. This indicates that the plaque was composed of enlarged posterior fiber ends rather than enlarged, bladderlike, or Wedl, cells. 

Scanning electron microscopy (SEM) examination further confirmed that the posterior subcapsular opacity was the result of pathologically enlarged posterior fiber ends and also demonstrated that this malformation precluded typical posterior suture formation from postnatal day 25 (Figs. 5A 5B ). SEM analysis of fiber peels also showed that although the affected posterior ends failed to form posterior sutures, the anterior fiber ends of the same fibers still had proper anterior end curvature and thus overlapped and abutted to form anterior sutures (Figs. 5C 5D) . 

Late Stages of Morphologic Abnormality

At 50 days after birth, the lens nucleus was opaque and was posteriorly dislocated (data not shown). In addition, whereas the anterior epithelium appeared normal, anterior fiber ends were slightly enlarged, and nuclei of elongating fibers were anteriorly displaced (Fig. 4B) . Differentiation and elongation of fibers in the bow region were also markedly less ordered (Fig. 4D) . By postnatal days 65 through 70, the posterior capsule of many, but not all the N+Δ+ lenses had ruptured, resulting in subluxation of the lens nucleus into the vitreous (Fig. 5F) . Fiber morphology in the anterior and bow regions continued to degenerate (Fig. 4E) . 

By postnatal day 95, the structure of the entire lens was completely abnormal. Thus, a large portion of the lens nucleus had subluxated through a breach in the posterior capsule into the vitreous (Fig. 6A ), the anterior fiber ends were enlarged and liquefied (Fig. 6C) , and the bow region showed no semblance of normalcy (Fig. 6D) . 

Morphologic Evidence of Retinal Involvement

Throughout PSC formation and nuclear opacification (postnatal days 25 through 50), the retinas of ΔFosB mouse lenses were indistinguishable from age-matched control lenses (Figs. 7A 7B 7C 7D 7E 7F) . The first evidence of retinal damage was seen at postnatal day 65 (Fig. 7G) when slight infoldings of predominantly the outer nuclear layer and layer of rods and cones were noted. By day 95, these infoldings had become progressively worse to involve all retinal layers (Fig. 6B) . These results suggest that the lens pathology may develop independently of retinal changes. 

Metabolic Parameters at Different Stages of Cataract Development

To determine the effect of ΔFosB expression on key metabolic parameters, enzyme activities were determined in lenses expressing (N+Δ+) and control lenses (N+Δ−) in which only tTA was expressed. Generally, animals expressing ΔFosB, showed development of lens opacity within ±5 days of each other. This variation is probably due to the level of ΔFosB expressed. In all cases in which 4-week-old animals were investigated, the lenses of N+Δ+ animals were chosen at a stage at which a clear indication of PSC was observed. As shown in Table 2 , inspection of a number of enzymes involved with the antioxidative defense of the lens indicated little difference between control andΔ FosB lenses at 4 to 6 weeks except for GSSG reductase activity, which was up 1.4-fold in ΔFosB-expressing lenses. At 12 weeks, when extensive deterioration of the lens occurred, GSHPx activity decreased approximately 40% compared with that in control lenses, but GSSG reductase remained high. No significant change in catalase activity, which was slightly lower in ΔFosB-positive animals, was observed between 6- and 12-week-old lenses. Measurement of NP-SH, representative of GSH, indicated little difference betweenΔ FosB and control lenses at 6 weeks. However, at 12 weeks, there was a marked decline in NP-SH in the N+Δ+ animals, indicating a significant decrease in reductive capability. In contrast, thymidine incorporation, reflecting DNA metabolism, was up 70% in the 6-week-oldΔ FosB lenses and increased to 260% compared with control animals in the 12-week-old lenses. Thus, the effect of ΔFosB expression varied markedly, depending on the parameter but, in all cases, was observed much later than morphologic changes. It should be noted that all lenses used for all biochemical studies were intact even in cases (12 weeks old) in which extensive disorganization and opacification had occurred. 

Lens Wet Weight, Protein Concentration, and SDS Gel Pattern Changes with Cataract Development

Examination of the wet weight of the lenses expressing ΔFosB versus control lenses (Fig. 8) from the same transgenic line, as discussed earlier, indicated a complex pattern. In the 4- to 6-week range, the (N+Δ+) lenses were approximately the same as normal but then increased markedly to approximately 40% above normal at 9 weeks. However, by 10 weeks, theΔ FosB lenses were only slightly above normal and remained the same at 12 weeks. Protein content (Fig. 8) , showed little change up to 7 weeks and then gradually declined, so that at 12 weeks the (N+Δ+) lenses had approximately 60% the level of protein of the control lenses. Thus, 12 week-old (N+Δ+) lenses had a wet weight slightly above normal but only a little more than half the normal protein content. 

SDS gel patterns of equal amounts of protein isolated from the lenses of (N+Δ+) bitransgenic mice and control animals expressing only tTA (N+Δ−) are shown in Figure 9 . The pattern of 2.5 week-old (N+Δ+) lenses was almost the same as the 4-week (N+Δ−) lenses, suggesting no large change in protein composition. However, at 12 weeks, significant differences could be observed. Bands at 30, 28, and approximately 19 kDa decreased markedly in the (N+Δ+) lenses, whereas lesser changes were found in (N+Δ−) lenses. There was also an increase in the intensity of some higher and lower molecular weight bands in the (N+Δ+) preparations. Thus, the loss of protein observed at 12 weeks (approximately 40%) may be due primarily to the selective degradation of certain polypeptides. Examination of overall protein synthesis indicated little difference between control and ΔFosB lenses during this period (data not shown). 

Effect on Transport Function

The loss of protein suggests that proteases may have been activated in the ΔFosB expressed lenses. Because it has been reported that calcium-dependent proteases are present in the lens, it was of interest to determine the ability of the lens to maintain a normal calcium concentration. Lenses were therefore placed in organ culture, and labeled calcium influx was measured. As shown in Figure 10 , at 4 weeks of age, lenses from four ΔFosB-expressing animals showed marked variation, with the lenses of two animals showing little change and those of two others a 1.5- to 2-fold increase in Ca influx above that in control animals. However, at 9 weeks of age, there was a 10-fold increase in calcium influx. 

To obtain a better assessment of the contribution of the calcium pump in maintaining calcium homeostasis, the effect of using W7, which inhibits Ca ATPase, ^{37 38} was examined. It was found that in (N+Δ+) lenses from 9 week-old animals in which a mature PSC was apparent, W7 had little effect on the influx of ⁴⁵Ca. In a 2-hour incubation, 38,500 ± 3,200 counts per minute (cpm)/lens of ⁴⁵Ca entered the lens in the absence of W7 versus 34,000 ± 2,500 cpm/lens with W7. In this case, the Ca ATPase that pumps calcium out of the lens was inactive. In the (N+Δ−) control, an approximately fivefold change was found 1900 ± 200 cpm/lens without W7 and 11,100 ± 970 cpm/lens with W7. Thus, a clear inhibition of the ⁴⁵Ca influx was observed with W7. A preliminary attempt to evaluate the other major system for calcium regulation, the Ca/Na exchanger, was unsuccessful. The present data suggest that the Ca ATPase activity was severely compromised by ΔFosB expression. 

Other indicators of plasma membrane function were also investigated. Examination of 10-week-old lenses indicated choline transport and Na/K ATPase were not affected (Fig. 10) . Such results would not have been expected if the cell membrane had become leaky, indicating that the change in calcium influx was probably caused by a specific ΔFosB effect on an unidentified gene controlling the epithelial cell calcium levels and not a ΔFosB effect on general membrane permeability. Examination of the Ca-ATPase mRNAs (PMCA 1–4) gave little indication of ΔFosB inhibition (data not shown). 

Discussion

In this study it was demonstrated that using a bitransgenic tetracycline repressor system, expression of ΔFosB in the lens, retina and pigment epithelium caused the formation of PSC that developed into complete opacification of the lens. The observation that morphologic changes occur in the lens approximately a month before there was any evidence of retinal damage suggests but does not prove that the lens pathology may arise independently of retinalΔ Fos-induced changes. (However, see later discussion of PSC animal models.) To clarify this important point, we are attempting to produce a transgenic mouse in which ΔFosB is expressed only in the lens. 

The cataract is clearly the result of ΔFosB expression, because the transgenic control N+Δ− animal is normal, and repressing ΔFosB expression with Dox resulted in normal lenses. Furthermore, using the bitransgenic NSE-tTA, TetOp-luciferase (which induces expression of luciferase) did not result in cataract development in mice even at 1 year of age (Kelz MB, Chen J, and Nestler EJ, unpublished observations, 1998). Thus, the expression of a protein, even a foreign one, is not sufficient to cause cataract. The cataract developed only after expression of ΔFosB. ΔFosB can be detected in the bitransgenic eye as early as 7 days (data not shown) and was present at high levels in the lens at all postnatal time points until the lens deteriorated. It is probable that ΔFosB is expressed during embryonic development. In spite of the early appearance ofΔ FosB, morphologic changes were not observed until approximately 25 days, and the biochemical changes observed were found at a later time. In most types of cataract, morphologic changes occurred after modification of biochemical parameters. Obviously, the key gene targets of ΔFosB have not as yet been defined but are likely to be associated with fiber differentiation. 

The delayed onset of the appearance of distinct cataracts–from approximately 52 days in bitransgenic 1A mice never treated with Dox to approximately 109 days in bitransgenic 1A mice conceived and raised on 200 μg/ml Dox treatments, with treatments removed at weaning (21 days)–is consistent with the time required for the systemic clearance of this dose of Dox before expression of transgenes downstream of the TetOp promoter may be initiated. ²⁷ Meanwhile, in 11A N+Δ+ bitransgenic mice that constitutively expressed ΔFosB (never treated with Dox), cataracts took more than twice as long to form as in 1A N+Δ+ mice and appeared to have much lower ΔFosB levels (unpublished observations). 11A mice conceived and raised with Dox treatments, with Dox removed at weaning, showed still slower appearance of cataracts, with no apparent cataracts or visual impairment through 16 weeks of age. In 11A mice in which Dox treatment was maintained without interruption, cataract never developed. 

It has been previously demonstrated that a 1.8-kb fragment of the proximal rat NSE promoter directs expression of tTA (and consequently directs expression of ΔFosB in bitransgenic mice) to forebrain structures such as striatum, cortex, and hippocampus, but not to peripheral tissues such as heart, liver, kidney, lung, or spleen. ²⁷ This 1.8-kb NSE promoter fragment used to create the NSE-tTA transgenic mouse has been shown to direct expression of aβ -galactosidase reporter gene to neural retina where expression is detected in the horizontal, ganglion, and amacrine cell layers. ⁴² Moreover, the 1.8-kb promoter fragment is able to drive a lower level of β-galactosidase expression to RPE. ⁴² The regulatory elements present within the proximal 1.8-kb promoter have been studied extensively ^{46 47} and have been found to target a variety of different transgenes to neural tissue within transgenic mice. ^{27 48 49} However, to the best of our knowledge, only two groups have used the NSE promoter to study the effect of transgene expression in the retina. ^{42 50} Moreover, we are unaware of any other reports that demonstrate the ability of the NSE promoter to target expression of a transgene to the lens. 

It is interesting that ΔFosB appeared considerably before morphologic or biochemical changes were noted and that it required considerable time for ΔFosB to produce cataracts. As a transcription factor,Δ FosB is expected to exert its pathologic effects by causing maladaptive changes in gene expression. Because no member of the Fos family of transcription factors can form homodimers, ΔFosB requires the presence of a dimerization partner to form a functional DNA-binding complex. ⁵¹ ΔFosB expression could lead to cataract formation through a direct effect of ΔFosB on gene expression, where it forms a dimer with a constitutively expressed bZIP transcription factor and thereby regulates expression of specific target genes. Alternatively, ΔFosB could act indirectly by sequestering a required bZIP protein and thereby disrupting patterns of gene expression normally mediated by that bZIP protein and its endogenous partner. The correlation between the level of ΔFosB expression and cataract formation (1A N+Δ+ much greater than 11A N+Δ +, which in turn is much greater than the leak expression seen in line 1 TetOp-ΔFosB mice) indicates that there was a tolerable threshold dose of ΔFosB before pathogenesis occurred. This finding may favor the indirect bZIP partner sequestration model. In any event, determining the genes whose transcription is affected by expression of ΔFosB should advance our understanding of PSC formation. 

The morphologic observations indicate that although lens fiber differentiation was clearly modified, remarkably, the initial compromise involved primarily the posterior ends of the fibers. There was no indication that posterior fiber constituents were abnormal, but rather that the fibers had lost their way and did not extend to their normal posterior positions. Thus, posterior sutures were not formed, and disruption of orderly arrays of the posterior fibers was observed. This did not occur on the anterior side of the lens, where fibers developed normally and formed Y sutures and regular arrays. Such observations raise the question: Are there separate controls for anterior and posterior fiber elongation and migration? Because it is abnormal differentiation of a section of a cell rather than the whole cell that is involved, it is unlikely that the deficiency resides within the cell. Such an argument suggests that some extracellular factor regulating fiber extension may be deficient. It is interesting that ΔFosB has been shown to be expressed in the retina and pigment epithelium as well as the lens. There are reports ^{52 53} that lens epithelial cells are stimulated to differentiate when cultured with retinas or retinal extracts. The retinal factors are believed to include the acidic and basic forms of fibroblast growth factor (FGF). FGF has been shown to stimulate cell elongation. ⁵⁴ Elongation of chick lens epithelial cells has also been reported to be initiated by a factor present in the vitreous called lentropin. ⁵⁵ This compound has been identified as an insulin-like growth factor. ⁵⁶ It is also probable that growth factors involved with lens epithelial development are present in the aqueous and tissues in the anterior segment of the eye. ⁵⁷ Thus, it is possible that ΔFosB expression in the retina inhibits the production of a growth factor that guides the elongating fiber cell to its appropriate posterior target site, whereas a similar factor produced in the anterior segment has a similar role in anterior fiber development. 

It is apparent from the biochemical data that ΔFosB affected more than the elongation of the fiber. The marked decrease in protein content in association with the increase in calcium influx suggests that increasing intracellular calcium concentrations may activate calcium-dependent proteases that then degrade the protein. It is probable that such changes in protein would be sufficient to cause loss of transparency and may account for the spreading of the opacity from the posterior subcapsular region to much of the lens. It is also interesting that biochemical parameters that usually change before detection of morphologic disorders such as choline transport, ⁸⁶Rb uptake, and thiol levels were normal even when severe lens degeneration occurred. 

Two other cataract animal models are of interest with respect to the present work. Recently, a hereditary recessive mouse cataract model was reported in which the opacity begins as a faint cloudiness in the posterior suture region and finally results in a rupture of the lens at the posterior pole. ⁵⁸ Although the mutation causing this cataract has not been defined, the mutated gene has been mapped to chromosome 14. ⁵⁹ The RCS rat model for PSC cataract and retinal degeneration is striking in its similarity to the ΔFosB cataract. Although the cataract develops more slowly thanΔ FosB-induced cataracts, the disruption of the posterior sutures and the elongation of the fibers on the posterior side have remarkable similarity. ²⁶ However, in the ΔFosB model, the retina appeared normal until a late stage in the development of the cataract. In the RCS model, retinal changes are observed initially as early as 12 to 18 days after birth, ²⁴ but initial changes in the lens are not seen by ultrastructural methods until 28 days after birth, ²⁶ corresponding to increased reactivity to thiobarbituric acid in the vitreous. ²⁵ Such observations raise questions about the retinal contribution to cataract development. Further work is needed to determine whether these two models involve the same or similar genes and whether such genes are involved in fiber elongation. The observation of Zigler and Hess ²⁵ argue for the involvement of an oxidative element in the RCS model but do not exclude the possibility that factors controlling fiber elongation have been disrupted. 

 MBK, JRK, and AS contributed equally to this work. ⁴Present address: Institute of Molecular Biology, The University of Hong Kong.

 Supported by grants (EJN) and a predoctoral National Research Service Award fellowship (MBK) from the National Institute on Drug Abuse; the Abraham Ribicoff Research Facilities of the Connecticut Mental Health Center (EJN); and grants from the National Eye Institute (AS, JRK), Research to Prevent Blindness, Foundation to Cure Cataracts, and the Department of Ophthalmology, College of Physicians and Surgeons of Columbia University (AS).

 Submitted for publication September 23, 1999; revised February 14, 2000; accepted February 28, 2000.

 Commercial relationships policy: N.

 

Figure 1.

 

(I) The two transgenes of the tetracycline-inducible gene expression system. The first transgene consists of ΔFosB cDNA placed under the control of a crippled promoter TetOp. The second transgene consists of the tetracycline transactivator tTA placed downstream of a tissue-specific promoter (NSE). The TetOp promoter is too weak to drive transcription of ΔFosB cDNA on its own. However, in the presence of tTA, TetOp is activated, which results in high levels of ΔFosB expression. Dox causes a conformational change in tTA so that it cannot bind the TetOp. (II) Cataract formation in a bitransgenic 1A N+Δ+ mouse. (A) Gross appearance of a bitransgenic mouse with cataract. (B) Posterior cataract. This photograph was taken with a retroillumination camera with Tri-X film and an F-stop setting of II. The degree of opacification required imaging at an angle off the visual axis to avoid glare. The photograph shows a large opacity in the posterior region of the lens of an 8-week-oldΔ FosB-expressing 1A bitransgenic mouse. Results are representative of the cataract seen in both eyes of bitransgenic mice (n = 8).

Table 1.

 

Incidence of Cataract in 1A Mice

Genotype	With Cataract (n)	Without Cataract (n)
1A N+Δ+	29	3
1A N+Δ−	0	47
1A N−Δ+	0	25
1A N−Δ−	0	43
1A N+Δ+ on Dox	0	20

 1A N+Δ+ bitransgenic mice carrying the NSE-tTA transgene (N+) derived from founder line A and the TetOp-ΔFos transgene (Δ+) derived from founder line 1, 1A N+Δ− single transgenic mice carrying NSE-tTA (N+) and without the TetOp-ΔFosB transgene (Δ−), 1A N−Δ+ single transgenic carrying the TetOp-ΔFosB transgene (Δ+) and without the NSE-tTA transgene (N−), and 1A N−Δ− wild type controls without both genes.

Figure 2.

 

Western blot analysis of ΔFosB expression. (A) Expression of ΔFosB in the whole-eye homogenate of 8-week-old bitransgenic 1A N+Δ+ mice never treated with Dox, which prevents expression. Results are representative of three experiments. (B) Western blot demonstrating the localization of ΔFosB expression to the lens, retina, and RPE of 8-week-old bitransgenic 1A N+Δ+ mice but not single-transgenic N+Δ− mice. Results are representative of two experiments. (C) Western blot demonstrating the localization of ΔFosB to various parts of lens of 6-week-old lens bitransgenic 1A N+Δ+ mice. Results are representative of three experiments. W, whole lens; F, lens fibers; E, lens epithelium. (D) Western blot analysis demonstrating the time course of ΔFosB expression in lens of bitransgenic 1A N+Δ+ mice, but not single-transgenic N+Δ− mice. Animals were analyzed at 2.5, 4, 6, and 12 weeks of age. Results are representative of four experiments.

Figure 3.

 

Lenses from 25-day-old N+Δ− (A, C) and ΔFosB N+Δ+ (B, D) mice, seen under a stereo dissecting microscope. (A) Posterior aspect of a whole lens showing the typical inverted Y suture. (B) Posterior aspect of a whole lens showing the atypical, circumscribed, speckled posterior subcapsular plaque. (C) On dissection to remove the epithelium and elongating and superficial fibers, the normal asymmetrical oblate spheroid shape of a mouse lens was apparent. (D) On comparable dissection, the abnormal almost symmetrical prolate spheroid shape of a ΔFosB mouse lens resulted from the posterior ends of fibers having curved up and away from the polar axis to form the posterior subcapsular plaque (curved arrows), rather than having overlapped and abutted within and between concentric growth shells to form sutures.

Figure 4.

 

LM micrographs of thick sections obtained along the visual axis of N+Δ+ lenses at various ages. Although the anterior epithelium (A) and elongating and superficial fibers in the bow region (C) were normal at postnatal day 25, the posterior ends of the same fibers were abnormally enlarged and curved away from the polar axis (F). Although the progressive enlargement and aberrant curvature of posterior fiber ends continued (G; postnatal day 35), the onset of bow region structural compromise did not begin until postnatal day 50 (D). At that time, elongating fibers were noted to be progressively less uniform. This disorder continued with aging (E; postnatal day 65). In addition, by postnatal day 65, although the anterior epithelium was still a normal monolayer, the nuclei of elongating fibers were markedly displaced (B). A, anterior pole; P, posterior pole; EQ, equator.

Figure 5.

 

SEM micrographs of 35- (A through E) and 65 (F)-day-old N+Δ+ lenses. (A) The normal posterior suture was not formed, because the posterior fiber ends were aberrantly curved away from the polar axis to form a subcapsular plaque beneath the posterior pole (PP; curved arrows; AP, anterior pole). At higher magnification (B) the nonuniformity and degree of posterior fiber end (PFEs) enlargement was apparent. Analysis of superficial fiber peels (C through E) demonstrated that the anterior fiber ends (AFEs; D) still overlapped and abutted to form suture branches, irrespective of the pathologic changes occurring posteriorly (E). By 65 days of age, the nuclei of many of these lenses had subluxated in the vitreous (F).

Figure 6.

 

LM micrographs of thick sections from an N+Δ+ mouse eye. When sectioned parallel to the sagittal plane, it was obvious that the lens nucleus and a large portion of the cortex had subluxated into the vitreous (A). In addition, the normal retinal architecture was compromised; marked infolding occurred between layers (A, B). Finally, beneath a slightly stratified epithelium (delineated by black lines), the anterior fiber ends were liquefied (∗; C), and the bow region showed no semblance of normalcy (D). A, anterior pole; EQ, equator; P, posterior pole.

Figure 7.

 

LM micrographs of retinal thick sections N+Δ+ mouse eyes (left), and age-matched control eyes (right) at 25 (A, B), 35 (C, D), 50 (E, F), and 65 (G, H) days of age. Retinal structure was indistinguishable between experimental and control animals until 65 days of age. At that time, infolding was noted to occur between the outer nuclear layer and layer of rods and cones (G). This structural compromise progressively increased to include all retinal layers, as shown in Figures 6A and 6B .

Table 2.

 

Determination of Metabolic Parameters of Control and ΔFosB-Positive Animals at Different Ages

	Age (Wk)	Control (N+Δ−)	ΔFosB (N+Δ+)	ΔFosB/Control
NADPH quinone oxidoreductase (mU/lens)	4	2.7	2.3	0.9
GSH-S-Transferase (mU/lens)	4	170	138	0.8
GSH-Px (mU/lens)	6	48	40	0.8
	12	22.8	13.1	0.6
GSSG Reductase	6	4.8	6.5	1.4
	12	7.5	10.8	1.4
Catalase (10 mM H2O2) (U/lens)	6	0.18	0.15	0.8
	12	0.14	0.12	0.9
NP-Thiol (nmole/lens)	6	18.6	19.9	1.1
	12	34.4	22.8	0.7
Thymidine incorporation (cpm/lens DNA)	6	1402	2372	1.7
	12	707	1772	2.6

Figure 8.

 

Changes in wet weight and protein of ΔFosB and control lenses with age. Wet weight and protein comparison of lenses from bitransgenic (N+Δ+) mice with control (N+Δ−). Wet weight was determined on the lens immediately after dissection. The lenses were first gently rolled on filter paper to remove excess liquid.

Figure 9.

 

SDS gel electrophoresis followed by Coomassie blue staining of protein from N+Δ+ and N+Δ− control lenses at different ages.

Figure 10.

 

Examination of transport systems. (A) Measurement of influx of [⁴⁵Ca]Cl₂ from medium into ΔFosB-expressing N+Δ+ and N+Δ− lenses. (B) Determination of NaKATPase activity.

The authors thank D. L. Stull for technical assistance with retinal dissections, G. Zeng for assistance with screening of the mice, and Allison Beckman for help with the Western blot analysis.