Microphthalmia due to p53-mediated apoptosis of anterior lens epithelial cells in mice lacking the CREB-2 transcription factor

Developmental Biology 222, 110-123 (2000) doi:10.1006/dbio.2000.9699, available online at http//www idealibrary.com on I ~ [ ~

[®

Microphthalmia Due to p53-Mediated Apoptosis of Anterior Lens Epithelial Cells in Mice Lacking the CREB-2 Transcription Factor Thore Hettmann,* Kevin Barton, t and Jeffrey M. Leiden*'* The Laboratory of Cardiovascular Biology, *Harvard School of Public Health, and ~Harvard Medical School, Bostou, Massachusetts 02115; and tDepartment of Medicine, Loyola University Medical Center, Chicago, IlEnois 60611

CREB-2 (also called ATF4, TAXREB67, or C/ATF) is an evolutionarily conserved member of the CREB/ATF family of basic-leucine zipper transcription factors. CREB-2 is expressed ubiquitously in the adult mouse and can function as both a transcriptional activator and a repressor. However, little was understood about the normal function of CREB-2 in mammalian development or organ physiology. In this report we have used gene targeting to produce CREB-2-deficient (CREB-2 -/-) mice. Adult CREB-2 -t- mice displayed microphthahnia due to the complete absence of a lens. Early embryonic lens development including formation of the optic vesicle, primary lens fibers, and proliferating anterior epithelial cells occurred normally in these mice. However, beginning at ED 14.5 the CREB-2-deficient anterior epithelial lens cells underwent massive and synchronous apoptosis, This was followed by the complete resorption of the developing lens. Consistent with this defect in anterior epithelial cell survival, in situ hybridization studies showed that CREB-2 is expressed at high levels in wild-type anterior epithelial lens cells at ED 14.5. The defect in lens formation seen in the C R E B - 2 -/- mice was not associated with qualitative defects in the expression of Pax-6, ~A-crystallin, c-maf, or PDGF-Rm However, apoptosis of the anterior epithelial cells was mediated by a p53-dependent cell death pathway because ablation of the p53 gene rescued anterior epithelial cell death and allowed the formation of a lens in the absence of CREB-2. Taken together, these results identify CREB-2 as an important regulator of mammalian lens development. © 2000 AcademicPress Key Words: transcription factor; CREB; ATF; eye development; CRE; microphthalmia; aphakia.

INTRODUCTION The vertebrate lens represents an ideal model system for studies of cell differentiation and organ morpbogenesis. The eye is easily visualized in both the embryo and the adult, thereby facilitating phenotypic analysis. Moreover, a great deal is understood about the embryology and molecular biology of lens morphogenesis (Graw, 1996; Piatigorsky, 1981). In the mouse, formation of the lens from surface ectoderm is initiated at ED 9-9.5 by an inductive signal derived from the underlying optic vesicle. This signal results in the formation of the lens placode, a thickened ridge of surface ectoderm, which by ED 11.5 invaginates to form a hollow spherical lens vesicle. Epithelial cells at the posterior pole of the lens vesicle, called primary fiber cells, differentiate and elongate to fill the cavity of the vesicle by ED 13.5. Beginning around ED 14.5, a rim of proliferating anterior epithelial cells begins to give rise to daughter cells

110

that migrate to the equatorial region of the lens where they elongate and differentiate into secondary lens fiber cells. These secondary fibers envelope the primary fiber ceils thereby giving rise to the outer layers of the maturing lens. Proliferation, migration, and differentiation of the anterior epithelial cells continue into adult life and are responsible for the maintenance of the lens. During the past 10 years a great deal has been learned about the molecular pathways involved in lens development (Fini et at., 1997; Graw, 1996). The identification of human mutations associated with lens defects in conjunction with biochemical and genetic studies in mice have allowed the identification of a number of growth factors, transcription factors, and cell cycle regulatory proteins that play important roles in normal lens formation. Thus, for example, BMP-7 (Dudley et aL, 1995; Luo et al., 1995), acidic FGF (Robinson et al., 1998), and IGl~s (Arnold et al., 1993; Beebe et aI., 1987) all appear to regulate the differen0012-1606/00 $35.00 Copyright © 2000 by AcademicPress All rights of reproductionin any form reserved.

111

CREB-2 m?d Lens DevalopmeT?t

tiation and elongation of the lens fiber cells. The p53 and retinoblastoma (Rb) tumor suppressor genes have been shown to play important roles in maintaining the viability of differentiating lens fiber cells (Morgenbesser et aL, 1994). In Rb-deficient mice, posterior lens fiber cells that are normally postmitotic spontaneously enter the cell cycle and subsequently die via a p53-dependent apoptotic pathway. Four transcription factors, Pax-6, Lqnaf, cqnaf, and Soxl, have been shown to play important roles in eye development. Pax-6 (in the mouse) (Grindley et a], 1995; Hanson and Van Heyningen, 19951 and L-mar lin the chicken) (Ogino and Yasuda, 1998) appear to function as master regulators, as animals lacking these factors fail to develop eyes. In contrast c~maf is required for posterior fiber cell elol~gation and early lens formation (Kawauchi et al., 1999; Kim eta]., 1999), while Soxl is required for the expression of 9/-crystallms and for elongation of the primary lens fibers during early lens development (Nishiguchi et ai., 1998). In contrast to our understanding of the transcriptional pathways that regulate posterior lens cell differentiation and survival, relatively little is currently understood about the pathways that regulate the proliferation, survival, and mi~ gration of the anterior lens epithelial cells. CREB-2 (also called ATF4, TAXREB67, or C/ATF) is a member of the CREB/ATF family of mamrnalian transcrip~ tion factors (Hai et al., 1989; Karpinski et a]., 1992; Tsujimoto et aL, 1991; Vallejo et al., 1993). All of the members of this family share an evolutionarily conserved basicleucine zipper (bZip) domain and brad to their octanucleotide recognition sequences (TGANNTCA) either as homodimers or as heterodimers with other CREB/ATF proteins (Hai eta]., 1989; Ivashkiv et aI., 1990). In addition, some family members can heterodimerize and bind to D N A in combination with structurally related AP1 and mar transcription factors (Benbrook and Jones, 1990, 1994; Hal and Curran, 1991; Kerppola and Curran, 1994; Vallejo et al., 1993). Dimerization of CREB/ATF proteins is mediated by their coiled-coil leucine zipper domains, whereas DNA binding is mediated by their structurally related basic domains. Although originally cloned from a T cell tumor cell line, CREB~2 is expressed ubiquitously in the adult mouse and h u m a n (Karpinski et aI., 1992; Tsujimoto et aL, 1991; Vallejo et al., 1993). Transient transfection assays have suggested that CREB-2 can function either as a transcriptional repressor or as an activator depending upon the cell and promoter system being tested (Karpinski et aL, 1992; Reddy et al., 1997; Vallejo et al., 1993). However, the function of CREB-2 in mammalian development and organ physiology remained unclear. Previous studies suggested that CREB-2 might regulate the expression of pituitary hormones (Kato eta/., 1999). In addition, CREB-2 has been shown to bind to ZIP, a serine/threonine kinase that can induce apoptosis in cultured cells (Kawai et al., 1998). Finally, studies of the structurally related aplysia transcrip-

tion factor, ApCREB2, have suggested a role for CREB-2 in long-term facilitation and memory (Bartsch et al., 1995). To better understand the function(s) of CREB-2 i~ vivo, we have used gene targeting to produce CREB-2-deficient mice. Adult mice homozygous for the CREB~2 mutation (CREB~2 I mice) displayed microphtbalmia due to the complete absence of a lens. Studies of embryonic CREBo2deficient mice demonstrated that the initial stages of lens development including formation of the optic vesicle, prio mary lens fibers elongation, and proliferation of anterior epithelial cells occurred normally between EDs 12 and 15. However, beginning at ED 14.5 the CREB-2-defi.cient ante° riot epithelial lens cells underwent massive and synchronous apoptosis that was followed by the complete resorption of the developing lens. The lens defect seen in the CREB-2 / mice was not associated with qualitative defects in the expression of Paxo6, ~A°crystallin, c~maf, or PDGFRR, all of which are known regulators of lens fiber differen~ tiation and survival. In contrast, this defect was p53dependent and could be rescued by breeding the CREB-2 / mice to p53-defieient animals. Taken together, these results identify a novel CREB-2-dependent transcriptional pathway that regulates the survival of anterior epithelial cells and thereby controls the development of the mammalian lens.

MATERIALS A N D METHODS

Generation of CREB-2-Deficient Mice The murine CREB-2 locus (CREB2) was isolated from a murine 129sv genomic library (Stratagene) by hybridization to a radiolabeled human CREB~2 cDNA probe (nucleotide 1 to 940 (Karpinski et aL, 1992)). The locus was mapped with. restriction endonucleases and partially sequenced for alignment with published CREB-2 sequences (data not shown). A 5.5qg/ml) and gancyclovir (1 bcM), Targeted ES cell clones were identified by Southern blot hybridization using a radiolabeled 0.6-kb SstI genomic probe (Probe A, Fig. 1) and injected into C57BL/6 donor blastocysts. Offspring were genotyped by BstXI digestion and Southern blot hybridization using a radiolabeled 1-kb KpnI/SacII fraglrlent (Probe B, Fig. 1).

Southern, Northern, and Western Blot Analyses For Southern blot analysis, high-molecular-weight DNA was extracted from ES cell clones, from mouse tails, or from embryonic yolk sacs as described previously (Hogan et al., 1986). Genomic DNA was digested with XbaI or BstXI, electrophoretically separated, transferred to nylon membranes (Schleicher and Schuell), and hybridized to radiolabeled probes. For Northern blot analysis 10 ixg of Trizolamine (Gibco BRL)-purified RNA from thymocytes or from the EL4 T cell line was resolved by dectrophoresis, transferred to nitrocellulose membranes (Schleicher and Schuell),

Copyright©2000byAcademicPress.All rightsof reproductionin anyformreselwed.

112

H e t t m a n n , Barton, and L e i d e n

A

Probe A E1

1 kb

E2 E3 CREB-2 locus

×

'

×

ATG

!

I !

~

5.5 kb

2.4 kb

K-neo

Targeting construct

Probe B

i 4pG - o--0-11 113

Mutant locus

111

m

2.3 kb targeted BstXl fragment 3.2 kb wild type BstXl fragment

B

C

~.

.JI

D

+

~+

, +-L - L - "~ + + + +

+, "~--,

wt-

CREB-2 ~ - CREB-2

- 3.2 kb - 2.3 kb

t-

~.

+

,.L~+ ~ + +

28S

FIG. 1. Targeted disruption of the CREB-2 gene. (A) Schematic representation of the wild-type routine CREB-2 locus (top), the targeting vector (middle), and the mutant locus (bottom). The locations of selected BstXI sites within the CREB-2 genomic tocus and the neomycin-resistance gene (neo R) are indicated. Exons (El to E3) are represented as boxes, the coding region is shaded, and the 5' UTR is re:shaded. The HSV-tk Irk) and/leo R genes are regulated by the mouse phosphoglycerate kinase (PGK) promoter. The gene targeting construct results in the deletion of the entire ORF of CREB-2. The locations of the external (Probe A) and internal (Probe B}hybridization probes are shown. (B) Southern analyses of tail DNA from the offspring of a CREB~2 +/- × CREBo2 +/- mating. BstXI-digested tail DNA from 2-week-old mice was hybridized to a genomic probe (Probe B) in order to distinguish the wild-type (wt) and targeted (t) alleles. (C) Northern analysis of mRNA from wil&type (+/+) and CREB-2 -I- ( - / - ) thymocytes and from a marine T cell line {EL-4) hybridized to an exon 3 CREB-2-specific probe (see Materials and Methods). Ethidium bromide staining of 28S RNA was used to ensure equal loading of the gel (bottom, 28S). Mutant mice lacked detectable CREB-2 mRNA. (D) Western analysis of thymocyte extracts from wild-type (+/+) and CREB-2-deficient ( - / - ) mice. The CREB-2~deficient mice lacked detectable CREB-2 protein.

and hybridized to a 0.6-kb SstI/KpnI probe spanning the 5' portion of exon 3 of the CREB-2 (E3) (see Fig. 1A). For Western blot analysis, whole-cell extracts were prepared as described previously (Hettmann et aL, 1999), fractionated by SDS-PAGE in 10% polyacryl-

amide gels, and transferred to PVDF membranes (Millipore). Western blots were probed with a mouse monoclonal antibody spccific for GREBe2 (C20; Santa Cruz) (11100 dilution) and developed with a rat anti-mouse secondary antibody (Kirkegaard and Perry Laboratories) (1:5000 dilution) and a commercially available chemiluminescence kit according to the manufacturer's instructions (Pierce Chemical Co.).

TABLE 1

Frequencies of Embryonic and Adult Genotypes

+/+

+/-

G e n e r a t i o n of CREB-2 -j- x p 5 3 -/- M i c e

-/-

~D 10.5

11 (27.5)

~ (S2.5)

s (20)

E D 18.5

20 (29.4)

35 (51.5)

13 (19.1)

Adult

79 (28.9)

173 (63A}

21 (7.7)

p53 j~ mice were obtained from Taconic Laboratories and mated with CREB-2 ÷/- mice. Offspring from these matings were genotyped as described above. Mutant p53 alleles were identified by PCR analysis using primers specific for exon 5 (TACTCTCCTCCCCTCAATAA) and exon 7 (ATAGGTCGGCGGTTCAT)of p53

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

113

GREBe2 and Lens Development

+l+ 7

......

FIG. 2, Microphthalmia in the CREB-2~deficientmice. Photographs of 4-week-old wild-type (+/+) and CREB-2 / ( /~. ) mice demonstrate the microphthalmia and closed eye lids of the CREB-2-deficient animals.

(Donehower at aI,, 1992). p53/CREB-2 doubly deficient mice were obtained by interbreeding CREBo2 ~/ , p53 -/- mice.

RESULTS Generation and Molecular Characterization of CREB-2-Deficient Mice

Histological Analysis Mouse tissues were fixed in 4% paraformaldehyde/phosphatebuffered saline, dehydrated through graded ethanol solutions, embedded in paraffin, and sectioned at 5 /~ln as described previously (Kuo et al., 1997). Paraffin-embedded sections were dewaxed, rehydrated, and used for dn situ hybridization (Morrisey et al., 1996) or stained with hematoxylin and eosin. Sense and antisense gaslabeled cRNA probes were generated from PCR-amplified cDNA fragments that had been suhcloned into pcDNA3 (lnVitrogen), using Sp6 and T7 RNA polymerases according to the manufacturer's instructions (Promega). The following gene-specific primers were used: CREB-2~5', cccaagcttCCTAAGCCATGGCGCTCTTC; CREB-2-3', ggctctagaGCACAAAGCACCTGACTACC; Pax-6-5', cccaagct tAAGGAGA GAGCATGTGATCG; Pax-6-3', ggctctao gaTTTGGAAAACCAACATATAG; c-maf-5', cceaagcttGTGGTG~ GTGGTGATGGCTCT; C-lnaf-3', ggctctagaCTCCAAAAAATTAGTAATAG; c~A-crystallin-5', c c c a a g c t t C T A T T T G GTTGATGCAG; c~A-crystallin-3', ccgctcagaGCGGGTGAGCATTCCAG; PDGF-R~-5', cccaagcttTGTAACTGACACGCTCCGGG; PDGF-Rc~-3', ggctctagaTGGTCTCTTCAGAGGTCCGG; p21-5', cccaagcttCACAGCGACCATGTCCAATCC; and p21-3', ggctctagaCAGAGTGAGGGCTAAGGCCG. Autoradiography was performed for 5 to 7 days. Slides were counterstained with Hoechst 33258 (Sigma) and viewed under dark-field and fluorescence microscopy and photographed with a Zeiss Axioskop using Kodak Ektachrome 160 film.

TUNEL Assays TdTqnediated dUTP nick end labeling (TUNEL) assays were performed on paraffin-embedded sections according to Gavrieli et al. (1992).

Genomic clones encoding m urine CREB.o2 were isolated from a 129sv mouse genomic library by hybridization to a human C R E B - 2 - s p e c i f i c probe. To avoid cloning-related CREB/ATF transcription factors, this probe lacked the sequences encoding the bZip domain. Overlapping genomic clones identified by hybridization were sequenced and found to contain the routine C R E B - 2 gene. The routine C R E B - 2 gene is composed of three exons spanning approximately 2.5 kb of genomic D N A (Fig. 1A). The ORF of C R E B - 2 is contained in exons 2 and 3 and can be conceptually translated into a 351-amino-acid polypeptide. To produce CREB-2-deficient mice, a targeting construct was generated in which the entire ORF of C R E B - 2 was replaced with a neomycin-resistance gene (neo ~) (Fig. 1A). The linearized targeting vector was electroporated into RW ES cells and G418- and gancyclovir-resistant colonies were screened for homologous recombination by Southern blot analysis. All homologous recombinants were also screened by hybridization with a n e o ~ probe to ensure a single integration event (data not shown). Two independently derived C R E B - 2 +/- ES cell lines were injected into C57BL/6 blastocysts and chimeric offspring were tested for their ability to transmit the targeted allele through the germ line. One of these chimeric mice transmitted the targeted allele through the germ line to yield C R E B - 2 +/~ mice. Heterozygous ( C R E B - 2 +/ ) mice were viable, fertile, and phenotypically normal and were intercrossed to generate C R E B - 2 lanimals (Fig. 1B). To confirm that the targeted mutagenesis event had resulted in a null allele, we performed Northern and Western blot analyses using thymocytes from wild-type and C R E B - 2 / mice. Unlike wild-type thymi that ex-

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

114

Hettmann

Barton, a n d L e i d e n

+1+

ED 14.5

ED 16.5

Neonatal

G Adult

FIG. 3. Defective lens development in the CREB-2-deficient mice. Eyes from wild-type (+/+) (A, C, E, G) and CREB-2 ~/- ( - / - ) (B, D, F, H) mice of different ages were sectioned and stained with hematoxylin and eosin. Note the normal formation of the lens, normal elongation of the posterior lens fiber cells, and normal formation of the anterior epithelial cell layer of the lens in the CREB-2 -/ eyes at ED 14.5. At ED 16.5, the anterior ocular chamber of the CREB-2 /~ mice has collapsed and is filled with vacuolated lens fiber cells extending from the lens nucleus. Few if any anterior epithelial cells are visible. At birth (Neonatal), the CREB-2 mutant lenses appear disorganized without evidence of viable epithelial cells. Adult CREB-2-deficient eyes display aphakia, but fail to display abnormalities in the retinal, corneal, and neuroretinal cells.

pressed high levels of CREB-2, t h y m i from the C R E B - 2 -/ m i c e lacked detectable CREB-2 R N A and p r o t e i n (Figs. 1C and 1D).

G e n o t y p i c analyses of the offspring of CREBo2 '/ × C R E B - 2 ~j~ crosses revealed a perinatal lethal p h e n o t y p e associated w i t h c o m p l e t e CREB-2 deficiency. O n l y 21 of

Copyright © 200[) by Academic Press, All rights of reproduction in any form reserved.

CREB~2 aad Leus Development

115

~C2 •

@

,4o

•

• ....

)/,

•

j

~

,~

~c4q

•

m

4. 4. } •

~, ,i~::~ ,,

,?

.... i, ••• ,,~ : } ~•~•i~'j....

I

4. 4.

~

> e~

c~

•

@

G~ ,

iii

,4

~,x C9~

Q ul

Q

Q

C o p y r i g h t @ 2 0 0 0 b y A c a d e m i c P r e s s . A l l r i g h t s of r e p r o d u c t i o n i n a n y l~orm r e s e r v e d .

Hettmann, Barton, and Leiden

116

C G)

c~ !

w i

C i

X q) cn

cn

CO d} (n c I .ti,

um

eLI

O:

c~

C G)

cn


0 ÷

I

+

Copyright © 2000 by Academic Press. All rights of reproductioll in any form reserved,

I

117

CREB-2 and Lens Development

PDGF-Ra E

Sense

Anti-Sense

÷/*

m

5

p21

+/+

-/6 FIG, 5. Expression of CREB-2, Pax-6, c-mar, RA-crystallin, and PDGF-Rc~ in wild-type and CREB-2-deficient mice. Expression of CREB-2 (A), Pax-6 (B), c-maf (C), ~A-crystallin (D), and PDGF-R~ (E) was analyzed by in situ hybridization using eyes from ED 14.5 wild-type (+/+) and CREB-2-deficient ( - / - ) mice. Note the presence of CREB-2 transcripts in all layers of the wild-type eye and its absence in the CREB-2 -/~ eye. Pax-6 was expressed in the inner retinal layers, anterior epithelial cells, and corneal epithelial cells and its expression was indistinguishable in the wild-type and CREB-2 /~ eyes. Expression of c-maf and c~A-crystallin was restricted to the lens fiber cells and was not different in the wild-type and CREB-2 /- eyes. PDGF-Ra expression was restricted to the anterior epithelial cells and was equivalent in eyes from the wild-type and CREB*2 -/- mice. Staining of the pigmented epithelium is artifactual, due to the presence of pigmented granules in these cells. Original magnification × 12.5. FIG. 6, p21 expression in wiid-type and CREB-2~deflcient eyes. p21 expression in ED 15.5 wild-type (+/+) and CREB-2-deficient t - / - J eyes was analyzed by in situ hybridization. Note the elevated levels of p21 transcripts in the anterior epithelial cells from the CREB-2-deficient mouse. Original magnification × 10.

Copyright© 2000by AcademicPress. All rights of reproductionin any formreserved.

118

Hettmann, Barton, and Leiden

273 adult mice analyzed displayed the CREB-2 /~ genotype, i~dicating that only approximately 30% of the homozygous-deficient mice survived until adulthood (Table 1). Further genotyping of pups from timed matings of heterozygous mutant mice revealed the expected Mendelian ratio of wild-type, heterozygous, and homozygousdeficient embryos until ED 17.5. However, the CREB-2 -+ embryos were runted compared to age-matched wild-type or heterozygous-deficient embryos (data not shown). The number of perinatal CREB-2 -/- mice (ED 18.5 to +1) was substantially reduced (Table 1). Surviving CREB-2 ' / mice matured to adulthood, although they appeared runted and displayed defective eye morphology (Fig. 2). The etiology of the partial perinatal lethality remains to be elucidated.

Microphthalmia in CREB-2-Deficient Mice Adult CREB-2 -/- mice had small eyes (microphtbalmia) and dosed eyelids (Fig. 2). Histological analyses demonstrated that this phenotype resulted from the complete absence of a tens (aphakia) in the adult CREB-2-deficient animals (Fig. 3). To better understand the developmental basis of this defect we compared histological sections of eyes from wild-type and CREB-2-deflcient embryos between ED 12.5 and adulthood (Figs. 3 and 4). The CREB-2 -lmice displayed normal lens development between EDs 12.5 and 14.5. Specifically, these mice demonstrated normal formation of the early lens vesicle, normal elongation of posterior primary fiber cells, and normal formation of the anterior lens epithelial cells (Figs. 3A and 3B 1. By ED 14.5, the anterior epithelial cells in both the wild-type and the CREB-2-deflcient eyes had begun to proliferate normally and to migrate to the equatorial region, where they differentiated into secondary lens fibers. However, between EDs 15.5 and 16.5 CREB-2 '/ embryos began to manifest specific defects of the anterior lens (Figs. 3C and 3D). The anterior lens cells of the CREB-2-deficient embryos appeared vacuolated and elongated with condensed nuclei. These degenerating epithelial cells filled the anterior chamber of the eye. At birth, few if any viable anterior epithelial cells could be identified in the CREB-2 °/~ lenses (Figs. 3E and 3F). From ED 16.5 onward, the remaining lens ceils in the CREB-2-deficient mice were seen to undergo degeneration and the entire lens was resorbed such that the adult eye lacked remnant lens tissue (Figs. 3E-3H). These defects appeared to be specific for the lens as no gross abnormalities in the retinal, pigmented epithelial, neuroretinal, and corneal layers could be detected in either embryonic or adult CREB-2 -/- eyes. The iris and ciliary bodies from the CREB2 /~ mice were small and malformed (Fig. 3H). Whether this represented a primary defect or was secondary to the lens degeneration seen in these animals remains unclear. A n t e r i o r E p i t h e l i a l Cells in C R E B - 2 . D e f i c i e n t L e n s e s D i e b y Apoptosis

The appearance of the dying anterior epithelial cells in the CREB-2-deficient lenses suggested that these cells were

undergoing apoptosis. To test this hypothesis, TUNEL assays were performed on eyes from wild-type and CREB2 -/- embryos at EDs 12.5, 14.5, and 16.5 (Fig. 4). No apoptotic cells were detected in eyes from either wild-type or CREB-2 /- mice at ED 12.5 (Figs. 4A and 4B). In contrast, large numbers of TUNEL-positive anterior epithelial cells were observed in eyes from the CREB-2 ;- mice at ED 14.5 /Figs. 4E and 4F/. These cells were not detectable in eyes from the wild-type control animals (Figs. 4C and 4D). By ED 16.5, the majority of the anterior epithelial cells in the CREB-2-deficient lenses had been resorbed. However, TUNEL-positive nuclei were observed in the equatorial lens cells that are the direct descendants of the anterior epithelial cells (Figs. 4I and 4J/. Once again, these apoptotic cells were not detected in the same region of the wild-type control eyes (Figs. 4G and 4H). We were unable to detect apoptotic lens fiber cells at either ED 14.5 or 16.5. Thus, in the absence of CREB-2 the anterior epithelial cells and their direct descendants at the equatorial poles of the lens undergo massive and synchronous apoptosis between EDs 14.5 and 16.5. The complete loss of these calls, which are the renewable precursors of the differentiated lens fiber cells, likely accounts directly for the embryonic lens degen~ eration seen in these animals.

Normal Expression of Lens-Specific

Genes

As described above, a number of genes are known to play important roles in the differentiation and survival of lens fiber cells. These include Pax-6, c-maf, ~A-crystallin, and PDGF and its receptor. We were interested in asking if one or more of these genes might be downstream targets of CREB-2 in the developing eye. In addition, we wished to assess the temporal and spatial patterns of expression of CREB-2 itself during normal eye development. Accordingly, we used in situ hybridization to investigate the patterns of expression of these genes in both wild-type and CREB-2deficient embryos (Fig. 5). Interestingly, CREB-2 was expressed at high levels in both retinal cells and anterior epithelial lens cells of wild-type animals beginning at ED 14.5 (Fig. 5A). Lower level CREB-2 expression was observed in differentiated lens fibers. As expected, no CREB-2 mRNA was detectable in eyes from the CREB-2-deficient animals {Fig. 5A). The Pax-6 transcription factor is required for the development of the eye, the olfactory system, and the brain (Cvekl and Piatigorsky, 1996; Grindley et al., 1995; Hanson and Van Heyningen, 1995). During early eye development, Pax-6 is essential for the formation of lens placode from surface ectoderm. Mice homozygous for mutations in the Pax-6 gene (Sey-/-j are unable to form a lens placode (Grindley et al., 1995). As described previously, we found that Pax-6 is expressed in the neural retina, in the anterior epithelium of the cornea, and in the lens epithelium at ED 14;5 IWalther and Gruss, 1991)/Fig. 5B}. No differences in Pax-6 expression were observed between wild-type and

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved,

119

CREB-2 and Lens D e v d o p m e n t

CREB-2-deficient eyes, suggesting that Pax-6 expression does not require CREB-2, Pax-6 is required for the expression of crystallin genes (including c~A-crystallin), which together constitute 90% of the soluble protein in the ocular lens (Cvekl and Piatigorsky, 1996). Previous studies have suggested that members of the CREB/ATF family synergize with Paxo6 to activate the murine ~A-crystallin promoter (Cvekl et al., 1995). However, we did not observe any alterations in c~Acrystallin expression in the CREB-2-deficient mutant mice (Fig. 5D), Thus, either CREB-2 is not involved in regulating c~A-crystallin expression in lens cells or other CREB proteins can compensate for the loss of CREB-2 in the CREB2 /~ mice. In either case, the defects in lens formation observed in the CREB-2-deficient mice are not the result of dysregulated c~A-crystallin expression. L-mar and cqnaf are members of the maf family of transcription factors that have been shown to be important for lens development in the chicken and mouse, respectively (Kawauchi et al., 1999; Kim et al, 1999; Ogino and Yasuda, 1998). Mice lacking c-mar display abnormal lens formation and like CREB-2-deficient mice are microphthalmic (Kawauehi et al., 1999; Kim et al., 1999). The similar phenotypes of the c-mar-deficient and CREB-2-deficient mice prompted us to examine the expression pattern of c - m a f in wild-type and CREB-2 -/- eyes. As shown in Fig. 5C, c-maf was detected at equivalent levels predominantly in the lens fiber cells of both wild-type and CREB-2 -llenses. Thus, the developmental defects caused by the loss of CREB-2 do not appear to result from defects in c-maf expression. Moreover there are important differences between the phenotypes of the c-mar and CREB-2 mutant mice. Mice homozygous for the c-maf mutation fail to elongate posterior lens fibers, resulting in a hollow lens at ED 14.5. In contrast posterior lens fiber elongation in the CREB-2-deficient mice appears to be normal at this stage of development (Fig. 3B). PDGF is an important regulator of lens cell differentiation and proliferation. PDGF is expressed in ocular tissues adjacent to the lens (Reneker and Overbeek, 1996). Expression of the PDGF receptor c~(PDGF-Rc~) is restricted to lens epithelial cells and this is the only form of the PDGF-Rs expressed in these cells. Accordingly, we examined the expression of PDGF-R~ in wild-type and CREB-2-deficient eyes (Fig. 5E). At ED 14.5, expression of PDGt?-RR in the anterior epithelial cells was comparable in the wild-type and CREB-2 - / mice (Fig. 5E). Thus, the defect in survival of the anterior epithelial cells seen in the absence of CREB-2 does not appear to result from defective expression of the PDGF-Rc~.

Apoptosis of the Anterior Epithelial Cells of the Lens in CREB-2-Deficient Mice Is p53-Dependent p53 has been shown to be an important regulator of the survival of differentiated lens fiber cells tFromm et al., 1994; Morgenbesser et aL, 1994). Thus, it was of interest to

ask if the apoptosis of anterior epithelial cells seen in the CREB-2~deficient mice was also p53-dependent, p53 is known to induce the expression of several downstream target genes, including the cdk inhibitor p21 (E1-Deiry et al, 1993). Therefore, in an initial series of experiments we used in situ hybridization to analyze the expression of p21 in anterior epithelial cells from wild-type and CREB-2deficient lenses (Fig. 6). Expression of p21 was markedly up-regulated in the anterior epithelial cells of ED 15.5 CREB-2-deficient mice compared to wild-type control mice. At ED 16.5, p21 transcripts were also detected in the CREB-2-deficient epithelial cells present in the equatorial regions of the lens in a pattern overlapping that of the TUNEL-positive cells in these regions (data not shown). These findings were consistent with the activation of a p53-dependent apoptotic pathway in these CREB-2deficient cells. To test this hypothesis directly, we bred CREB-2deficient mice with p53-deficient mice to produce doublehomozygous mutant mice. Ocular examination revealed normal eye and eyelid morphology and a marked suppression in microphthalmia in the p53 -/- × CREB-2 -/- mice (compare Figs. 7E and 7G). Histological analyses demonstrated that eyes from the double-homozygous mice contained lenses (Fig. 7H). These lenses contained viable epithelial cells positioned at the anterior pole and surrounded by a basement membrane (capsule) but displayed an in~ creased number of cells in the bow region {Fig. 7I). In addition, they contained decreased numbers of lens fiber cells. These data suggested that the programmed cell death observed in CREB-2-deficient anterior epithelial cells is p53-dependent. In addition, they suggested that there may be p53-independent effects of CREB-2-deficiency involving the differentiation and survival of elongated lens fiber cells.

DISCUSSION In the studies described in this report we have used gene targeting to assess the function of the CREB-2 transcription factor in mouse development. CREB-2-deficient mice displayed a partial perinatal mortality with approximately 70% of the animals dying between ED 17.5 and postnatal day 14. Thirty percent of the CREB-2-deficient mice survived until adulthood. The surviving CREB-2-deficient mice -were runted and displayed severe microphthalmia due to complete aphakia. Studies of embryonic eye development in the CREB-2-deficient mice demonstrated a specific defect in the survival of the anterior epithelial cells of the lens that give rise to the secondary lens fibers. Although these cells developed normally and began to proliferate between EDs 12.5 and 14.5, they subsequently underwent synchronous and complete apoptotic death. This abnormality in lens development was not associated with qualitative defects in the expression of known regulators of lens development, including c-maf, Pax-6, PDGF-Rc~, and c~Acrystallin. However, the death of the anterior lens epithelial

Copyright © 2000 by Academic Press. All rights of reproduction in any form reserved.

120

Hettmann, Barton, and Leiden

B wt

D p53-/-

F CREB.2 -/-

u ~••i/•,/ ,• •

CREB.2 -/p53-/-

FIG. 7. p53-depeudent aphakia in the CREB-2-deficient mice. Gross appearance (A, C, E, and G) and histological analysis (H and E staining) (B, D, F, H, and I) of eyes from 4oweek-old wild-type (wt), p53-deficient (p53-/-), CREB-2-deficient (CREB-2 /), and p53/CREB-2 doubly-deficient (CREB-2-/- p53-/-) mice. Note the rescue of the rnicrophthahnia and aphakia of the CREB-2-deficient mice by concomitant p53-deficiency {compare E and F to G and H, respectively). Original magnification ×2.5 for B and D, ×5 for F and H, and ×10 for I.

cells seen in these animals appeared to be dependent upon a p53-regulated death pathway, as the defect was rescued significantly by breeding the CREB-2-deficient mice to p53-deficient animals.

While this work was in progress, Tanaka et al. (1998) reported the phenotype of CREB-2-mutant animals produced by using a distinct gene targeting strategy that involved the deletion of the 3' portion of the third exon of

Copyright © 2000 by Academic Press. All rights of reproduction in any for m reserved,

121

CREB-2 a n d Lens D e v e l o p m e n t

FIG. 7~---Coz~tinued

the CREB~2 gene (which was erroneously labeled exon 2 in their studies). Tanaka et al. also observed severe microphthalmia and aphakia in the resultant CREB-2-deficient animals. However their reported phenotype differs significantly from that described in our report. Most importantly, Tanaka et al, did not observe apoptosis of the CREB-2deficient anterior epithelial cells at EDs 14.5M6.5. Instead, they described apoptosis of differentiated elongated secondary lens fiber cells at ED 18.5 as the cause of the aphakia in their CREB-2-deficient animals. There are several possible explanations for the divergent phenotypes observed in these two studies. First, in contrast to the complete deletion of the CREB-2 geue described in our studies, it is possible that the partial deletion of the CREB-2 gene introduced by Tanaka et al. resulted in the production of a truncated CREB-2 protein with some residual function. Alternatively, the different phenotypes n:ay reflect differences in the background strains used to produce the mutant mice. Finally, it is possible that Tanal
CREB-2 is a necessary component of a growth (survival 1 factor signaling pathway that is normally required to maintain the viability of the anterior epithelial cells during embryonic development. The fact that CREB-2 is expressed at high levels in these ceils between EDs 14.5 and 16.5 would suggest that it might control the expression of a specific growth factor receptor or alternatively might he a downstream transcriptional response element of such a signaling pathway. Relatively little is currently known about the growth factor pathways that regulate t h e later stages of lens development, Previous studies have suggested that PDGF produced by periocular cells may regulate lens fiber cell proliferation and elongation {Reneker and Overbeek, 1996 I. Indeed the PDGF-Ra chain is expressed on proliferating anterior epithelial ceils in the developing lens. However, our in situ hybridization studies failed to demonstrate a qualitative defect in PDGF-R~ expression in the CREB-2-deficient lens. CREB-2 can dimerize with several other CREB/ATF fan:fly members, including c~maf (Motohashi et aI., 1997). Recent gene-targeting experiments have demonstrated that c-mar is an important regulator of lens formation and eye development (Kawauchi et al., 19991 Kin: et aI., 1999). Thus, it seemed possible that CREB-2 might heterodimerize with c-mar to regulate anterior epithelial cell survival. However, several recent findings would seem to negate such a n:odel. First, c-mar and CREB-2 are expressed in distinct compartments of the developing lens. c-maf is expressed in the posterior primary lens fiber cells, whereas CREB-2 is expressed predominantly in the anterior epithelial cells and their descendants at the equatorial poles of the lens. Consistent with these spatially distinct patterns of expression, c-maf-deficient mice display defects in the elongation of posterior lens fibers cells resulting in a hollow lens, whereas the CREB-2 mice display normal posterior fiber cell elongation but subsequent defective survival of the anterior epithelial cells. Nevertheless, it is possible that c-mar-related genes are coexpressed with CREB-2 in the anterior epithelial cell compartment and that heterodimers between CREB-2 and these c-mar-related proteins are important for regulating the survival or differentiation of these cells. CREB-2 has also been shown to physically interact with a leucine zipper-containing serine/threonine kinase, called ZIP (Kawai et al., 1998). Interestingly, overexpression of ZIP led to apoptosis in N1H 3T3 cells. Thus, it is possible that homodimers of ZIP lead to apoptosis and that heterodimerization of ZIP with CREB-2 inhibits the apoptotic potential of ZIP. In such a model, deletion of CREB-2 in the anterior epithelial cells :night lead to increased levels of ZIP homodimers and subsequent apoptosis. Future studies designed to identify the target genes and partners of CREB-2 in the anterior epithelial cells of the lens should help to further elucidate the cell survival pathways regulated by CREB-2 during eye development.

Copyright 63 2000 by Academic Press. All rights of reproduction in any form reserved,

122

H e t t m a n n , Barton, and Leiden

ACKNOWLEDGMENTS This work was supported in part by a grant from the NIAID (R37 AI29673). We thank Mr~ Tim Lis for help with the preparation of illustrations and Mrs. Kirsten Sigrist for help with the production of the CREB-2/- mice.

REFERENCES Arnold, D. R., Moshayedi, P., Schoen, T. J., Jones, B. E,, Chader, G. J., and Waldbillig, R. J. (1993 ). Distribution of IGF-I and -II, IGF binding proteins (IGFBPs) and IGFBP mRNA in ocular fluids and tissues: Potential sites of synthesis of IGFBPs in aqueous and vitreous. Exp. Eye Res. 56, 555-565. Bartsch, D., Ghirardi, M., Skehel, P. A., Karl, K. A., Herder, S. P., Chen, M., Bailey, C. H., and Kandel, E. R. (1995). Aplysia CREB2 represses long-term facilitation: Relief of repression converts transient facilitation into long-term functional and structural change. Cell 83, 979-992. Beebe, D. C., Silver, M. H., Belcher, K. S., Van Wyk, J. J., Svoboda, M. E., and Zelenka, P. S. (1987}. Lentropin, a protein that controls lens fiber formation, is related functionally and immunologically to the insulin-like growth factors. Proc. NatI. Acad. Sci. USA 84, 2327-2330. Benbrook, D. M., and Jones, N. C. (1990). Heterodimer formation between CREB and JUN proteins. Oncogene 5, 295-302. Benbrook, D. M., and Jones, N. C. (1994). Different binding specificities and transactivation of variant CRE~s by CREB complexes. Nucleic Acids Res. 22, 1463-1469. Cvekl, A., Kashanchi, F., Sax, C. M., Brady, J. N., and Piatigorsky, J. (1995). Transcriptional regulation of the mouse alpha A-crystallin gene: Activation dependent on a cyclic AMPresponsive element (DE1/CRE) and a Pax-6-binding site. MoL Cell Biol. 15, 653-660. Cvekl, A., and Piatigorsky, J. (1996). Lens development and crystallin gene expression: Many roles for Pax-6. BioEssays 18~ 621-630. Donehower, L. A., Harvey, M., Slagle, B. L., McArthur, M. J., Montgomery, C. A., Jr., Butel, J. S., and Bradley, A. {1992). Mice deficient for p53 are developmentally normal but susceptible to spontaneous turnouts. Nature 356, 215-221. Dudley, A. T., Lyons, K. M., and Robertson, E. J. (1995). A requirement for bone morphogenetic protein-7 during development of the mammalian kidney and eye. Genes Dev. 9, 27952807. El-Deity, W~ S., Tokino, T., Velculescu, V. E., Levy, D. B., Parsons, R., Trent, J. M., Lin, D., Mercer, W. E., Kinzler, K. W., and Vogelstein, B. (1993). WAF1, a potential mediator of p53 tumor suppression. Cell 75r 817-825. Fini, M. E., Strissel, K. J., and West-Mays, J. A. (1997). Perspectives on eye development. Dev. Genet. 20, 175-185. Fromm, L., Shawlot, W., Gunning, K., Butel, J. S., and Overbeek, P. A. {1994). The retinoblastoma protein-binding region of simian virus 40 large T antigen alters cell cycle regulation in lenses of transgenic mice. Mol. Cell Biol. 14, 6743 6754. Gavrieli, Y., Sherman, Y., and Ben-Sasson, S. A. (1992). Identification of programmed cell death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol. 119, 493-501. Graw, J. (1996). Genetic aspects of embryonic eye development in vertebrates. Dev. Genet. 18, 181-197.

Grindley, J. C., Davidson, D. R., and Hill, R. E. (1995]. The role of Pax-6 in eye and nasa], development. Develqpment 121, 14331442. Hai, T., and Curran, T. (1991). Cross-family dimerization of transcription factors Fos/Jun and ATF/CREB alters DNA binding specificity. Proc. Natl. Acad. Sci. USA 88, 3720-3724. Hal, T. W., Liu, F., Coukos, W. J., and Green, M. R. {1989). Transcription factor ATF cDNA clones: An extensive family of teucine zipper proteins able to selectively form DNA-binding heterodimers. Genes Dev. 3, 2083-2090. Hanson, I., and Van Heyningen, V. 11995). Pax6: More than meets the eye. Trends Genet. 11, 268-272. Hettmann, T., DiDonato, J., Karin, M., and Leiden, J. M. [19991. An essential role for nuclear factor kappaB in promoting double positive thymocyte apoptosis. 1. Exp. Med. 189, 145-158. Hogan, B., Constantini, F., and Lacy, E. (1986). "Manipulating the Mouse Embryo: A Laboratory Manual." Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY. Ivashkiv, L. B., Liou, H. C., Kara, C. I., Lamph, W. W., Verma, I. M., and Glimcher, L. H. (1990). mXBP/CRE-BP2 and cqun form a complex which brads to the cyclic AMP, but not to the 12-OtetradecanoylphorboI-13-acetate, response element. Mol. Ceil Biol. 10, 1609-1621. Karpinski, B. A., Morle, G. D., Huggenvik, J., Uhler, M. D., and Leiden, J. M. (1992). Molecular cloning of human CREB-2: An ATF/CREB transcription factor that can negatively regulate transcription from the cAMP response element. Proc. Natl. Acad. Sci. USA 89, 4820-4824. Kato, Y., Koike, Y., Tomizawa, K, Ogawa, S., Hosaka, K., Tanaka, S., and Kato, T. (1999}. Presence of activating transcription factor 4 (ATF4) in the porcine anterior pituitary. Mol. Ceil Endocrinol. 154, 151-159, Kawai, T., Matsumoto, M., Takeda, K., Sanjo, H., and AkJra, S. 11998}. ZIP kinase, a novel serine/threonine kinase which mediates apoptosis. Mol. Ceil Biol. 18, 1642-1651. Kawauchi, S., Takahashi, S., Nakajima, O., Ogino, H., Morita, M., Nishizawa, M., Yasuda, K., and Yamamoto, M. (1999}. Regulation of lens fiber cell differentiation by transcription factor c-Maf. ]. BioL Chem. 274, 19254-19260. Kerppola, T. K., and Curran, T. (1994). Mar and Nrl can bind to APd sites and form heterodimers with Fos and Jun. Oncogene 9, 675-684. Kim, J. I., Li, T., Ho, I. C., Grusby, M. J., and Glilncher, L. H. (1999}. Requirement for the c-Mar transcription factor in crystallin gene regulation and lens development. Proc. NatL Acad Sci. USA 96, 3781-3785. Kuo, C. T., Veselits, M. L., Barton, K. P., Lu, M. M., Clendenin, C., and Leiden, J. M. [1997). The LKLF transcription factor is required for normal tunica media formation and blood vessel stabilization during murine embryogenesis. Genes Dev. 11, 2996-3006. Luo, G., Hofinann, C., Bronckers, A. L., Sohocki, M, Bradley, A, and Karsenty, G. [1995}. BMP-7 is an inducer of nephrogenesis, and is also required for eye development and skeletal patterning. Ga~es Dev. 9, 2808-~2820. Morgenbesser, S. D., Williams, B. O., Jacks, T., and DePinho, R. A. (1994), p53-dependent apoptosis produced by Rb-deficiency in the developing mouse lens. Nature 371, 72-74. Morrisey, E. E., Ip, H~ S, Lu, M. M., and Parmacek, M. S. {1996). GATA-6: A zinc finger transcription factor that is expressed in multiple cell lineages derived from lateral mesoderm. Dev. Biol. 177, 309--322.

Copyright © 2000 by Academic Press. All rights of reproduction m any form reserved.

CREB~2 and Lens Development

Motohashi, H., Shavit, J. A., Igarashi, K., Yamamoto, M., and Engel, J. D. (1997). Tire world according to Mar. Nucleic Acids Res, 25, 2953--2959. Nishiguchi, S., Wood, H., Kondoh, H., LovelloBadge~ R., and Epis~ kopou, V. {19981. Soxl directly regulates the gamma-crystallin genes and is essential for lens development in mice. Gel~es Dev. 12, 776~781. Ogino, H., and YasudG K~ (1998}. Induction of lens differentiation by activation of a bZIP transcription factor, L~Maf. Science 280, 115-118. Piatigorsky, J. (1981 }. Lens differentiation in vertebrates. A review of cellular and molecular features Differe~tiatioT? 19, 134-153. Reddy, T. R., Tang, H., Li, X., and WongGtaal, F. (1997). Functional interaction of the HTLV-I transactivator Tax wRh activating transcription factor-4 {ATF4). Oncogene 14, 2785-2792. Reneker, L. W., and Overbeek, P. A. (1996}. Lens-specific expression of PDGGA alters lens growth and development. Dev. Biol 180, 554-565. Robinson, M. L, Ohtaka-Maruyama, C., Chan, C. C., Jamieson, s., Dickson, C., Overbeek, P. A., and Chepelinsky, A. B. tt998). Disregulation of ocular morphogenesis by lens-specific expression of FGF-3/int~2 in transgenic mice. Dev. Biol. 198, 13-31. Tanaka, T., Tsujimura, T., Takeda, K., Sugihara, A., Maekawa, A., Terada, N., Yoshida, N., and Akira, S. (1998). Targeted disruption

123

of ATF4 discloses its essential role in the formation of eye lens fibres. Ge~es Cells 31 80t---810. Tsujimoto, A., Nyunoya, H:, Morita, T., Sato, T., and Shimotohno, K. (1991). Isolxion of cDNAs for DNA&inding proteins which specifically bind to a tax°responsive enhancer element in the long terminal repeat of hmnan T-cell leukemia virus type I. 1. ViroL 65, t420-1426. Tybutewicz, V. L., Crawford, C. E., Jackson, P. K., Bronson, R. T., and Mulligan, R. C. (199 t ). Neonatal lethality and lymphopenia in mice with a homozygous disruption of the c-abl proto~ oncogene. Cell 65, 1153~-1163. Valle}o, M., Ron, D , Miller, C. P., and Habener, J. F. 11993). C/ATF, a member of the activating transcription factor family of DNAbinding proteins, dimerizes with CAAT/enhancer-binding pro~ reins and directs their binding to cAMP response elements. Proc. Natl. Acad. Sci. USA 90, 4679-4683. Walther, C., and Gruss, P. (1991}. Paxo6, a routine paired box gene, is expressed in the developing CNS. Developmen~ 113, 14351449. Received for publication January 31, 2000 Revised March 7, 2000 Accepted March 9, 2000

Copyright©2000by AcademicPress.All rightsof reproductionin any formreserved.