Inhibition of lens fiber cell morphogenesis by expression of a mutant SV40 large T antigen that binds CREB-binding protein/p300 but not pRb.

Simian virus (SV) 40 large T antigen can both induce tumors and inhibit cellular differentiation. It is not clear whether these cellular changes are synonymous, sequential, or distinct responses to the protein. T antigen is known to bind to p53, to the retinoblastoma (Rb) family of tumor suppressor proteins, and to other cellular proteins such as p300 family members. To test whether SV40 large T antigen inhibits cellular differentiation in vivo in the absence of cell cycle induction, we generated transgenic mice that express in the lens a mutant version of the early region of SV40. This mutant, which we term E107KDelta, has a deletion that eliminates synthesis of small t antigen and a point mutation (E107K) that results in loss of the ability to bind to Rb family members. At embryonic day 15.5 (E15.5), the transgenic lenses show dramatic defects in lens fiber cell differentiation. The fiber cells become post-mitotic, but do not elongate properly. The cells show a dramatic reduction in expression of their beta- and gamma-crystallins. Because CBP and p300 are co-activators for crystallin gene expression, we assayed for interactions between E107KDelta and CBP/p300. Our studies demonstrate that cellular differentiation can be inhibited by SV40 large T antigen in the absence of pRb inactivation, and that interaction of large T antigen with CBP/p300 may be enhanced by a mutation that eliminates the binding to pRb.

DNA tumor viruses such as simian virus 40 (SV40) and adenovirus inactivate pivotal checkpoints for cell cycle control, thereby stimulating host cell DNA synthesis and allowing viral DNA replication. The early region of SV40 encodes two proteins: the large T and small t antigens. Large T antigen is a potent 708-amino acid oncoprotein, which has transforming ability and can induce tumors in animal models (1). Relevant targets of T antigen include the retinoblastoma family members, pRb, p107, and p130, which interact with the aminoterminal domain of T antigen (amino acid residues 102-114). The central and carboxyl-terminal regions of T antigen bind to the tumor suppressor p53 and to a variety of other proteins including the transcriptional co-activators CREB 1 -binding protein (CBP), p300 and p400 (1,2). In certain cell types, such as lens fiber cells, preadipocytes, and myoblasts, expression of large T antigen is sufficient to stimulate cell cycle entry and cellular transformation (3)(4)(5)(6)(7)(8).
The ocular lens provides an excellent model system in which to study the relationship between cell cycle exit and differentiation-specific gene expression. The lens is composed of an anterior monolayer of proliferative cuboidal epithelial stem cells overlying a core of terminally differentiated, post-mitotic, elongated fiber cells. At the equatorial region of the lens, epithelial cells are induced to exit from the cell cycle and to differentiate into fiber cells. Initiation of fiber cell differentiation is accompanied by up-regulation of expression of the cyclindependent kinase inhibitor Kip2 (p57) (9), and activation of expression of fiber cell-specific proteins including the ␤and ␥-crystallins (10,11). These changes in gene expression reflect alterations in the activities of specific transcription factors including c-Maf, Sox-1, and Prox-1 (12)(13)(14). Recent in vivo and in vitro studies indicate that transcriptional coactivators CBP/ p300 are required for lens fiber cell differentiation and expression of crystallin genes (15,16). CBP and p300 are distinct but functionally related coactivator proteins with histone acetyltransferase activity (17). Both proteins have been shown to interact with sequence-specific transcription factors, including CREB (18), c-Fos (19), c-Jun (20), c-Myb (21), E2F1 (22), and p53 (23). CBP and p300 also interact with viral proteins such as the adenovirus E1A protein (24), SV40 large T antigen (2), human papillomavirus E6 protein (25), and polyomavirus large T antigen (26). These viral proteins can bind to and inhibit both CBP and p300. In the ocular lens, we have shown that CBP and/or p300 are important coactivators of c-Maf-mediated transcription from various crystallin promoters (16). Although it has been reported that lens-specific expression of wild type SV40 large T antigen causes lens tumors and reduced expression of crystallins (3), little is known about whether inhibition of crystallin expression by SV40 T antigen is solely because of loss of pRb activity and cell cycle re-entry (27), or because of direct inhibition of other essential endogenous proteins. In this study, we have expressed in the lens a mutant version of SV40 large T antigen, E107K⌬, which does not bind to the pRb family members. We show that the E107K⌬ protein can inhibit both fiber cell differentiation and crystallin gene expression without altering cell cycle regulation. We also show that the E107K⌬ protein binds to CBP/p300 and blocks interaction of CBP/p300 with c-Maf.

EXPERIMENTAL PROCEDURES
Generation of the Constructs and Transgenic Mice-The entire early region of SV40, encoding small t and large T antigens, was previously linked to the ␣A-crystallin promoter (Fig. 1A) (3). For the E107K⌬ construct, a BamHI/BsgI fragment from plasmid ␣TruncT⌬268-E107K (4) was substituted for the corresponding region of the wild type clone (Fig. 1C). This fragment contains both a deletion of nucleotides 4586 -4854 (⌬268) to remove the small t splice donor, and the K1 mutation (2,6,28) that eliminates binding of T antigen to pRb. The K1 mutation replaces glutamic acid (E) with lysine (K) at amino acid 107 (E107K) in the LXCXE motif (2). The resulting plasmid was digested with KpnI and EcoRI to release a 2.5-kb fragment for microinjection (Fig. 1C). The truncated T antigen construct (Fig. 1B) has been described previously (4). The fragments used for microinjection were separated by electrophoresis through a 1% agarose gel, and purified using a Geneclean kit (Bio 101 Inc., Vista, CA). Transgenic mice were generated by pronuclear injection of the purified fragments into one-cell stage inbred FVB/N embryos (29,30).
Screening of Transgenic Mice-Tail DNA was isolated as previously described (29). PCR screening was performed using SV40 primers as described (4).
Histology-Embryonic heads at E15.5 were fixed in 10% formalin, paraffin embedded, cut as 5-m thick sections, and stained with hematoxylin and eosin (H&E) by standard techniques.
In Situ Hybridization-In situ hybridizations were performed using 35 S-labeled riboprobes as described in Fromm and Overbeek (31). To test for E107K⌬ transgene expression, a BamHI/PvuII fragment of E107K⌬ was subcloned into pBluescript KS(Ϫ) (Stratagene, La Jolla, CA), and used to generate a specific E107K⌬ riboprobe. Mouse c-Maf cDNA was obtained from Lixing Reneker (University of Missouri, Mason Eye Institute). The probes for genes involved in cell cycle regulation were generated from the following mouse cDNAs: E2F1 from Peggy Farnham (University of Wisconsin), p57 from Stephen Elledge (Baylor College of Medicine, Houston, TX), cyclins A2 and B1 from Debra Wolgemuth (Columbia University, New York, NY); and cyclin E from Julie A. DeLoia (Magee-Womens Research Institute, Pittsburgh, PA). Hybridization signals were captured as dark-field images.
BrdUrd Incorporation-DNA replication was detected by 5-bromo-2Ј-deoxyuridine (BrdUrd) incorporation. BrdUrd (from Sigma, 100 g/g body weight) was injected into pregnant female mice. One hour later, the mice were sacrificed and embryos were analyzed by immunohistochemistry as described previously (4).
Immunohistochemistry and Immunofluorescence-Tissue sections were deparaffinized in xylene and rehydrated through a graded alcohol series, washed with PBS, and incubated with 10% methanol, 3% hydrogen peroxide in PBS to block endogenous peroxidase activity. For detection of ␣-crystallins, rabbit antisera (from Samuel Zigler, National Institutes of Health) were used at 1:1000 dilution in 10% normal horse serum in PBS, with a 24-h incubation at 4°C. After washing with PBS, biotinylated goat anti-rabbit IgG was applied for 45 min at 37°C. The secondary antibody was detected using streptavidin-conjugated horseradish peroxidase, and peroxidase activity was visualized with diaminobenzidine/H 2 O 2 (Vector Laboratories, kit SK-4100). All slides were incubated in diaminobenzidine solution for 1 min. Sections were counterstained with hematoxylin. For detection of transgenic SV40 large T antigen, endogenous c-Maf, and CBP/p300 proteins, slides were boiled in 10 mM sodium citrate buffer (pH 6.0) for 20 min using a microwave oven. Slides were then incubated with 10% normal horse serum for 1 h at 37°C. Mouse monoclonal anti-SV40 large T antigen (1:500) (BD Biosciences), rabbit polyclonal anti-c-Maf (1:500) (Santa Cruz), mouse monoclonal anti-CBP/p300 (1:500) (Sigma), and rabbit or mouse IgG (1:500) (Vector Laboratories) were added and incubated overnight at room temperature. The slides were washed with PBS (4 ϫ 5 min), and a secondary antibody (1:200) (biotin-conjugated anti-rabbit or antimouse IgG, Vector Laboratories) was added for 1 h at room temperature. After washing 4 times with PBS, ExtrAvidin-fluorescein isothiocyanate conjugate (from Sigma) was added and incubated for 1 h at room temperature, followed by PBS washes. The slides were mounted with Aqua-Poly/Mount (Vector Laboratories).
Detection of Apoptosis-DNA fragmentation was detected by TUNEL assay using an in situ apoptosis detection kit (apoTACS TM : In Situ Apoptosis Detection Kit, Trevigen, Inc.). Tissue sections from embryos at E15.5 were treated as described (4,15).
Co-immunoprecipitation-Lens lysates from prenatal non-transgenic and transgenic embryos were prepared as described above. Lens lysates (1000 -2000 g) were pretreated by addition of 80 l of protein G-agarose beads (1:1 dilution in 50 mM Tris-HCl, 1% Nonidet P-40, 1 mM EDTA, 67 mM NaCl). After centrifugation, supernatants were collected and incubated with anti-CBP/p300 antibody (Sigma), anti-Tantigen antibody (BD Biosciences), or rabbit IgG (Santa Cruz) for 2-3 h at 4°C, then with 80 l of protein G beads overnight at 4°C. Bound proteins were eluted with 2ϫ SDS sample buffer, separated by SDS-PAGE, transferred onto polyvinylidene difluoride membranes and assayed by Western blotting with anti-T-antigen or anti-c-Maf antibodies. For loading control, the blot was stripped and re-probed with anti-actin antibody. lens fiber cells induces cell cycle entry dependent upon the ability of T antigen to bind to the tumor suppressor pRb, and that cell cycle re-entry is accompanied by defects in fiber cell elongation and a reduction in ␤and ␥-crystallin expression (4). To analyze whether T antigen can inhibit cell morphogenesis without binding to pRb, a non-pRb binding mutant of T antigen (2), which we term E107K⌬, was linked to the mouse ␣Acrystallin promoter (Fig. 1C), and used to generate transgenic mice. The ␣A-crystallin promoter typically targets transgene expression to the lens fiber cells (32)(33)(34). The E107K⌬ mutant carries a mutation (E107K) in the Rb-binding motif (2) along with a deletion (⌬268) that removes the small t antigen splice donor (4). Two stable transgenic families were characterized (OVE481 and OVE484). Hetero-and homozygous transgenic mice in both families exhibit severe microphthalmia (Fig. 1, D  and E). Because ␣〈-TruncT⌬268-E107K, which encodes just the first 191 amino acids of E107K⌬ does not cause lens defects (4), the dramatic ocular phenotype in the OVE481 and OVE484 mice is likely because of an activity that requires downstream amino acids. Mice lacking p53 do not show lens defects (35), so the E107K⌬ phenotype is not because of p53 inactivation. We compared the E107K⌬ mice to transgenic mice expressing either the intact early region of SV40 (OVE266A, OVE266B) (Fig. 1A), or a truncated T antigen (191 amino acids) (Fig. 1B) that binds specifically to pRb family members (OVE272) (4). A summary of our transgenic studies is shown in Table I Lens Histology-Transgenic lenses expressing E107K⌬ are small (Fig. 2), and are partially hollow because of failure of primary fiber cell elongation. The small lenses show anteriorization of the differentiation zone (arrows in Fig. 2, D and F).  The E107K⌬ phenotype resembles that seen in c-Maf or Sox-1 knockout mice (13,36). Our prediction when we started these experiments was that the mice would have normal lenses. Because the mice exhibited a dramatic phenotype, we have studied them to characterize the pattern of transgene expression, and to assay for alterations in cell cycle regulation, fiber cell differentiation, and expression of crystallin genes.

Microphthalmia in E107K⌬ Transgenic
Expression of E107K⌬ in Lenses of Transgenic Mice-Based on in situ hybridizations (Fig. 3, B and C) and immunostaining using an antibody specifically against SV40 large T antigen (Fig. 3, E and F), expression of the E107K⌬ transgene was lens-specific. Hybridization signals (Fig. 3, B and C) and green nuclear staining for SV40 large T antigen (Fig. 3, E and F) were found in lens epithelial and fiber cells in both transgenic families (OVE481 and OVE484). The expression of the transgene in lens epithelial cells was not typical for the ␣A-crystallin promoter, and might be related to the small size of the lens. In non-transgenic lenses, cell proliferation is normally confined to the epithelial cells that cover the anterior of the lens (Fig. 3G). The epithelial cells are normally induced to become post-mitotic fiber cells at the equatorial region. Lens-specific expression of either wild type or truncated SV40 T antigen, or targeted mutagenesis of the Rb gene, causes the post-mitotic lens fiber cells to re-enter S phase (4,27). In contrast, posterior lens fiber cells expressing E107K⌬ did not incorporate BrdUrd (Fig.  3, H and I). This result is consistent with the prediction that E107K⌬ is unable to bind to pRb. It also implies that E107K⌬ does not disrupt fiber cell differentiation by inducing cell cycle re-entry (Table I). TUNEL assays were performed to test for cell death in the different transgenic lenses (Table I and Fig. 3, J-L). There were no detectable apoptotic nuclei in the lens fiber cells expressing E107K⌬ (Fig. 3K). Thus, the altered differentiation in E107K⌬ transgenic lenses is not because of inappropriate induction of cell death.
Expression of E2F1 and Cyclins-E2F1 is an important transcription factor that can activate expression of genes required for the G 1 /S transition (37). In normal mouse lenses, E2F1 is expressed in both epithelial and fiber cells (38). In contrast, expression of cyclins A2 and B1 in the lens is confined to the lens epithelial cells (Fig. 4, D and G). These cyclins are also expressed in proliferating retinoblasts (Fig. 4) (4, 34). Expression of full-length T antigen in lens fiber cells resulted in induction of expression of these cell cycle genes (Fig. 4, C, F, I, and L) (15,31). By comparison, there was no detectable upregulation of these genes in lens fiber cells expressing E107K⌬ (Fig. 4, B, E, H, and K).
Expression of p57-The expression of p57 (Kip2), a cyclin-dependent kinase inhibitor, is strongly up-regulated when fiber cell differentiation is induced at the equatorial region of the lens (Fig. 4M, arrows). Enhanced p57 expression is required for the cell cycle exit that accompanies fiber cell differentiation (9). In the E107K⌬ transgenic lenses, p57 expression was still up-regulated during fiber cell induction (Fig. 4N, arrows). Induction of p57 also occurred in lens cells expressing full-length T antigen (Fig. 4O). These results suggest that inhibition of fiber cell morphogenesis by E107K⌬ takes place after fiber cell differentiation has been initiated.
Expression of Other Lens Fiber Cell Differentiation Markers-Previous studies have shown that lens fiber cells without functional pRb have reduced expression of ␣-, ␤-, and ␥-crystallins (3,4), markers for lens cell differentiation. We used immunohistochemistry to test for ␣-crystallin expression, and Western blotting to assay for changes in expression of ␣-, ␤-, and ␥-crystallins. We compared non-transgenic lenses to lenses expressing wild type T antigen (OVE 266A), truncated T antigen (OVE 272), or E107K⌬ (OVE 481) (Fig. 5). Expression of ␣-crystallin was detected in lens cells in the presence of wild type T (Fig. 5, B and E), truncated T antigen (Fig. 5, C and E), or E107K⌬ (Fig. 5, D and E), although expression was significantly reduced when compared with non-transgenic FVB lenses (Fig. 5,  A and E). Expression of ␤and ␥-crystallins was also decreased about 30 -40% in the presence of either wild type T or truncated T antigen (Fig. 5, F and G). By comparison, lenses expressing E107K⌬ showed a truly dramatic loss of ␤and ␥-crystallin expression (Fig. 5, F, lanes 10 -12 and G; Table I).
E107K⌬ Does Not Block Expression of c-Maf and CBP/p300 -Recent studies indicate that transcriptional co-activators CBP and/or p300 are required for c-Maf-mediated transactivation of ␤and ␥-crystallin promoters during lens fiber cell differentiation (16). Therefore, the lens defects induced by E107K⌬ expression might be because of inhibition of c-Maf or CBP/p300 expression or function. To test for inhibition of c-Maf expression, in situ hybridizations and Western blots were performed (Fig. 6). In non-transgenic FVB lenses (Fig. 6A), transcripts of c-Maf are up-regulated in post-mitotic lens cells at the equatorial region. Transgenic mice expressing either wild type T antigen or E107K⌬ still showed lens-specific c-Maf mRNA expression (Fig. 6, B and C). Using Western blots, we found that the c-Maf protein was present in the E107K⌬ transgenic lenses at levels comparable with the wild type T antigen lenses (Fig.  6D). To assay for effects of E107K⌬ on c-Maf nuclear translocalization, immunofluorescence using an antibody against c-Maf protein was performed. As shown in Fig. 7, A-D, c-Maf proteins are normally present in cell nuclei of non-transgenic FVB lenses (Fig. 7A). Expression of wild type T antigen ( 7B), truncated T antigen (Fig. 7C), or E107K⌬ (Fig. 7D) did not alter c-Maf protein nuclear localization. These results indicate that the inhibition of lens fiber cell differentiation is not because of interference with c-Maf localization. To assay for effects of E107K⌬ on expression of CBP/p300, immunofluorescence was also performed using rabbit IgG and an antibody specifically against both CBP and p300 (Fig. 7, E-H). There was no nuclear staining when rabbit IgG was used as the primary antibody on eye sections (Fig. 7, I-L). By comparison, CBP/p300 proteins were detected in cell nuclei of nontransgenic FVB (Fig. 7E) and T antigen transgenic lenses (Fig.  7, F-H). This result suggests that disruption of lens differentiation and crystallin expression by E107K⌬ occurs without a dramatic decrease in the level of CBP/p300 proteins.
E107K⌬ Blocks the Binding of c-Maf to CBP/p300 -Previous studies have indicated that wild type T antigen and the E107K mutant can bind to the third C/H-rich domain of CBP or p300, similar to adenoviral E1A (2,39). Because CBP/p300 can coactivate crystallin promoters through direct binding to c-Maf protein (16), we hypothesized that the dramatic lens defects observed in E107K⌬ transgenic mice might be because of inhibition of the binding of CBP/p300 to c-Maf. To test for this, lens lysates from transgenic and non-transgenic mice were used for immunoprecipitation assays (Fig. 6, E and F). A monoclonal antibody directed against both CBP and p300 was able to pull down both wild type T antigen and the E107K⌬ protein, but not truncated T antigen (Fig. 6E). There was no T antigen protein in non-transgenic FVB mice (Fig. 6E). The endogenous c-Maf protein in FVB lenses interacts with CBP/p300 (16), and can be co-precipitated with the CBP/p300 antibody (Fig. 6F). Expression of either wild type or truncated T antigen did not prevent the co-precipitation of CBP/p300 with c-Maf (Fig. 6F). Interestingly, the CBP/p300 antibody did not pull down c-Maf protein in the presence of E107K⌬ (Fig. 6F). These results suggest that E107K⌬ interferes with the binding of c-Maf to CBP/p300 (Table I). DISCUSSION Terminal differentiation in the ocular lens as well as other systems is blocked by expression of the SV40 early genes (3)(4)(5)(6)(7). Suppression of adipocyte differentiation by SV40 large T antigen has previously been shown to occur independent of pRb family binding and small t action (5,6). However, the mechanisms by which T antigen blocks differentiation for different cell types have not been established. We have used transgenic mice to characterize the dramatic biological activity of E107K⌬, a point mutant of SV40 large T antigen that no longer binds to pRb. In this study, we show that E107K⌬ can function as a potent repressor of morphogenesis and of crystallin expression during terminal differentiation of lens cells. Although it is possible that inhibition of lens cell differentiation may involve various activities of T antigen, our results indicate that  1-3), wild type T (lanes 4 -6), truncated T (lanes 7-9), and E107K⌬ transgenic eyes (lanes 10 -12). The amount of protein loaded was: 20 g in lanes 1, 4, 7, and 10; 40 g in lanes 2, 5, 8, and 11; 80 g in lanes 3, 6, 9, and 12. Actin was used as an internal control. G, blots were quantified using an imaging densitometer. Protein levels of ␤or ␥-crystallin were quantified by the ratio of blot density of crystallin to the density of actin. Results are the mean Ϯ S.D. from three individual experiments. Statistical analysis was done using two-tailed Student's t test using Microsoft Excel software. Data were considered to be significantly different for p Ͻ 0.05. E107K⌬ selectively disrupts the interaction of co-activators CBP/p300 with the transcription factor c-Maf. Because a truncated version of E107K does not disrupt this interaction and does not inhibit morphogenesis, and because CBP/p300 activity is required for normal fiber cell differentiation (16), we predict that this function of E107K⌬ plays a central role in its ability to inhibit fiber cell differentiation and morphogenesis.

Crystallin Expression
Requirement for CBP/p300 -The ␤and ␥-crystallin families of proteins normally undergo a dramatic induction during maturation of lens fiber cells. Expression of these genes begins after the fiber cells have permanently exited from the cell cycle. Multiple genes in each family are coordinately regulated. The coordinate regulation and strong levels of expression suggest that a small set of trans-activating nuclear proteins configure the promoters of these genes for maximal expression. Our results indicate that the E107K⌬ mutant can inactivate the pathway or process by which all of these genes are up-regulated. It does so without binding to pRb family members, without inducing cell cycle re-entry, without disrupting the initiation of differentiation, and without inducing apoptosis. Lack of crystallin induction correlates with inhibition of CBP/p300 binding to c-Maf. 〈s a result, CBP and p300 are presumably no longer recruited by c-Maf to the crystallin promoters, and expression of these genes is consequently disrupted.
Requirement for Cell Cycle Exit-When pRb is inactivated in lens fiber cells, either by targeted mutagenesis or by overexpression of SV40 large T antigen, expression of ␣-, ␤-, and ␥-crystallins is reduced in conjunction with inappropriate cell cycle re-entry (3,27,31). The reduced expression of crystallins may simply be a consequence of cell proliferation in the absence of pRb function. However, pRb has been shown to play a direct role in regulation of muscle cell differentiation and muscle- . Expression is still detected in the transgenic lenses. D, Western blots were used to assay for c-Maf proteins and T antigen in non-transgenic lenses (FVB) and transgenic lenses expressing either wild type T antigen (Tag) or E107K⌬. Actin was used as a loading control. E and F, co-immunoprecipitation of CBP/p300 with E107K⌬ and c-Maf. Lens extracts from transgenic and non-transgenic mice were immunoprecipitated (IP) with anti-CBP/p300 antibody (recognizing both CBP and p300 proteins) and the complexes were analyzed by Western blots (WB) using anti-T antigen antibody (recognizing the COOH terminus of both wild type T antigen and the E107K⌬ protein, but not the truncated T antigen) (E) and anti-c-Maf antibody (F). Lane IgG, lens lysates from FVB mice were precipitated with rabbit IgG instead of CBP/p300 antibody. Actin was used as a quality control for the lens extracts (F). indicates that c-Maf and CBP/p300 are present in lens cell nuclei, at levels comparable among non-transgenic and transgenic lenses, suggesting that lens-specific expression of SV40 T antigen did not block nuclear localization of c-Maf or CBP/p300. specific gene expression by acting as a cofactor for MyoD (40). Therefore, an alternative possibility is that pRb is a positive regulator of crystallin expression. In our studies, we discovered that E107K⌬ can significantly inhibit expression of ␤and ␥-crystallins without inhibiting pRb function (Fig. 5). Although not conclusive, this finding would indicate that the reduced expression of crystallins in RbϪ/Ϫ lenses or transgenic lenses expressing truncated T reflects an intrinsic incompatibility between cell cycle progression and the high levels of crystallin gene expression seen in terminally differentiated fiber cells.

E107K⌬: A Repressor of Crystallin Expression
E107K⌬ inhibited fiber cell morphogenesis and crystallin expression without inducing cell cycle re-entry or cell death (Fig. 3). Our previous studies showed that when the same E107K mutation was introduced into a truncated T antigen, which does not encode the COOH-terminal domains of T antigen, no inhibition of lens fiber cell differentiation or crystallin expression occurred (4). Therefore, the COOH terminus of E107K⌬ is essential for its ability to inhibit differentiation and morphogenesis. This region of the protein is known to be required for interaction with the third C/H-rich domain of CBP and p300 (2,39).
Other observations lead us to hypothesize that the E107K⌬ mutation might result in a "gain of function" as well as a "loss of function." First, we noted that the E107K⌬ phenotype was dominant when E107K⌬ mice were mated to full-length T antigen mice (data not shown). Second, we found that E107K⌬ inhibited crystallin expression more efficiently than wild type SV40 large T antigen (see Fig. 5). Third, E107K⌬ blocks the binding of c-Maf to CBP/p300, whereas wild type T antigen does not (Fig. 6F). Previous cell culture studies reported that both wild type T antigen and the K1 mutant (E107K⌬) bind to CBP/p300 at comparable levels (2,39). However, it was also shown that deletion of residues 98 -126 (mutant T50L7) resulted in the simultaneous loss of interaction with pRb and binding to CBP (2) indicating that the binding sites overlap. We hypothesize that the E107K⌬ mutant interacts with CBP/p300 differently than wild-type T antigen.
In summary, our results indicate that the E107K⌬ mutation may cause both a loss of function (loss of binding to pRb) and a gain of function (inhibition of CBP/p300 binding to c-Maf). It is not yet clear whether other functions of E107K⌬ play additional role(s) in disrupting lens fiber cell morphogenesis. Because the molecular control of morphogenesis in vivo is not well understood, our transgenic mice provide a novel model system for future studies.