p53 Deficiency Rescues Neuronal Apoptosis but Not Differentiation in DNA Polymerase {beta}-Deficient Mice

In mammalian cells, DNA polymerase ß (Polß)functions in base excision repair. We have previously shownthat Polß-deficient mice exhibit extensive neuronalcell death (apoptosis) in the developing nervous system andthat the mice die immediately after birth. Here, we studiedpotential roles in the phenotype for p53, which has been implicatedin DNA damage sensing, cell cycle arrest, and apoptosis. Wegenerated Polß^–/– p53^–/– double-mutantmice and found that p53 deficiency dramatically rescued neuronalapoptosis associated with Polß deficiency, indicatingthat p53 mediates the apoptotic process in the nervous system.Importantly, proliferation and early differentiation of neuronalprogenitors in Polß^–/– p53^–/–mice appeared normal, but their brains obviously displayed cytoarchitecturalabnormalities; moreover, the mice, like Polß^–/–p53^+/+ mice, failed to survive after birth. Thus, we stronglysuggest a crucial role for Polß in the differentiationof specific neuronal cell types.

INTRODUCTION

Top

Introduction

Repair of DNA damage is essential for maintaining the integrityof the genetic information necessary for normal developmentand physiological consequences (28). DNA polymerase ß(Polß) is a 39-kDa protein of a single polypeptide,consisting of two catalytic, functional domains. The N-terminal8-kDa domain carries a 5'-deoxyribose phosphate lyase activity,whereas the C-terminal 31-kDa domain carries a polymerase activitythat fills a short gap with a 5'-phosphate (26, 40). Polßis a critical component of the base excision repair (BER) pathway.The BER pathway repairs DNA damage, such as apurinic/apyrimidinic(AP) sites and base modifications, which spontaneously occuror are induced by a variety of endogenous and exogenous agents,including reactive oxygen species and DNA alkylating agents.Biochemical studies have identified two types of BER in mammaliancells: a short-patch pathway involving replacement of one nucleotideand a long-patch pathway involving gap-filling of several nucleotides(46). The BER pathway is generally initiated by a specific DNAglycosylase that recognizes and removes a damaged base to generatean AP site in DNA, followed by incision of the site by an APendonuclease. In the short-patch BER pathway, Polßremoves the 5'-deoxyribose phosphate and fills the single nucleotidegap, and finally DNA ligase I or a complex of XRCC1 and DNAligase III ligates the nick. On the other hand, the long-patchBER pathway requires proliferating cell nuclear antigen (PCNA),flap endonuclease 1 (FEN-1), and DNA ligase I to excise a flap-likestructure resulting from strand displacement by Polßand/or Pol ${delta}$ / ${varepsilon}$ and to ligate the nick (6).

We and others previously showed that Polß-deficientmice exhibit a reduced size and weight and die with a respiratorydefect immediately after birth (13, 42). In Polß-deficientmice, extensive cell death (apoptosis) occurs in postmitoticneurons in the developing central and peripheral nervous systems(42). This neuronal apoptosis is closely associated with theperiod between the onset and cessation of neurogenesis. Abnormalitiesin embryonic tissues other than the nervous system have notyet been reported (13, 19, 21). Therefore, we suggested thatPolß plays an essential role specifically in the developmentof the nervous system (42). However, the cause of this neuronalapoptosis remains entirely unknown. Mouse embryonic fibroblastcells in culture, derived from a Polß knockout mouse,are viable and show normal growth characteristics (41). Althoughthe mutant cells exhibit BER defects, as evidenced by increasedsensitivity to DNA alkylating agents, their cell extracts stillretain an activity to repair a damaged base residue in DNA substrate(6), indicating that there are both Polß-dependentand -independent BER pathways in vivo.

The tumor suppressor protein p53 plays a prominent role in themaintenance of genomic integrity (27). It is activated by differenttypes of DNA damage, including single-strand breaks (SSBs),double-strand breaks (DSBs), and adducts, which are generatedby endogenous or exogenous mechanisms. The activated p53 hasa choice of cell cycle arrest for repair or apoptosis, dependingon the level of damaged DNA; i.e., unless the damage is repaired,p53 leads to apoptosis. Recent studies show that p53 directlyinteracts with Polß, stimulating BER activity (37,50). In the nervous system, it has been shown that p53 regulatesneuronal apoptosis after neuronal injury induced by excitotoxins,hypoxia, and ischemia that cause oxidative damage (3, 29, 33,47). In p53-deficient mice, kainic acid excitotoxicity- or ischemia-inducedbrain damage is significantly reduced (10, 34). These observationssuggest the involvement of p53 in the control of neuronal apoptosis.

A close link between DNA damage and neurodegeneration appearsevident from many pathological data and observations with mouseknockouts (8, 38). In mice deficient in DNA ligase IV (Lig4)and XRCC4, the main components of nonhomologous-end-joiningapparatus for DSB repair, differentiating neurons undergo massivecell death (4, 14, 17, 22). However, this apoptosis is completelyrescued by p53 deficiency (15, 16). It would be important toexamine whether neuronal cell death found in Polß-deficientmice (42) is mediated by p53 activation. Here we study a potentialrole for p53 in the phenotypes, including neuronal cell deathassociated with Polß deficiency. We found that p53deficiency rescues the neuronal apoptosis in a Polß-deficientbackground. However, it should be noticeable that Polß^–/–p53^–/– mice still exhibit cytoarchitectural defectsin the development of the nervous system and die shortly afterbirth. These observations strongly suggest that Polßis crucial for the differentiation process of specific neuronalcell types.

MATERIALS AND METHODS

Top

Materials and Methods

Mice. Polß-deficient mice were described previously (42).p53-deficient mice (C57BL/6J-Trp53 tm1Tyj) were obtained fromThe Jackson Laboratory (24). PCR genotyping protocol for p53targeted allele are directed on the website of JAX MICE, TheJackson Laboratory (http://jaxmice.jax.org). Noon of the dayon which the vaginal plug was detected in the morning was designatedembryonic day 0.5 (E0.5). All mice were maintained in a pathogen-freeenvironment under the guidelines of Kihara Institute for BiologicalResearch, Yokohama City University, for laboratory animals.

Histology, immunohistochemistry, and TUNEL assay. Embryos were perfused with 4% paraformaldehyde and 7% picricacid in 0.1 M sodium phosphate buffer (pH 7.4); the brain wasremoved and postfixed in the same fixative for 2 h, equilibratedwith 25% sucrose-phosphate-buffered saline, frozen in OCT compound(Sakura Finetechnical Co.) and sectioned on a cryostat (10 µm).The sections were incubated with rabbit anti-p53 polyclonalantibody CM-5 (Novocastra Laboratories, 1:3,000), rabbit anti-cleavedcaspase-3 polyclonal antibody (Cell Signaling Technology; 1:100),mouse anti-PCNA monoclonal antibody PC10 (Sigma; 1:100), mouseanti-neuron specific type-III ß-tubulin monoclonalantibody Tuj1 (BabCO; 1:1,000), mouse anti-phosphorylated neurofilamentSMI31 (Sternberger Monoclonal Antibodies; 1:4,000), and rabbitanti-calbindin/spot 35 polyclonal antibody (a kind gift of T.Yamakuni) (1). Cy3 or horseradish peroxidase-conjugated antibodywas used for a secondary antibody to visualize primary antibody.TSA Biotin System (Perkin-Elmer Life Sciences) was applied toanti-p53, anti-calbindin immunohistochemistry. The TUNEL (terminaldeoxynucleotidyltransferase-mediated dUTP-biotin nick end labeling)assay was performed on cryosections by using 0.12 U of TdTase(Roche)/µl with 0.5 µM biotin-14-dATP (Invitrogen)in 1x TdT buffer (Roche) with 1.5 mM CoCl₂. Horseradish peroxidase-conjugatedbiotin (Jackson Immunoresearch Laboratories) was used for signaldetection. Cell nuclei were stained with DAPI (4',6'-diamidino-2-phenylindole)for immunofluorescence. Cresyl violet (Sigma) staining was performedto show neuronal architecture.

Western blot analysis. Cell extracts were prepared from developing telencephalons inE13.5 embryos, electrophoresed in an 8.0% sodium dodecyl sulfate-polyacrylamidegel, and transferred to an Immobilon membrane (Millipore) asdescribed previously (42). The membrane was probed with anti-humanphospho-p53 (Ser-15) antibody (Cell Signaling Technology; 1:1,000)and peroxidase-conjugated goat anti-rabbit immunoglobulin G(Chemicon) and detected with enhanced chemiluminescent detectionreagents (ECL Plus kit; Amersham).

RESULTS

Top

Results

Polß deficiency activates the p53-dependent apoptosis pathway. p53 is activated and stabilized after excessive DNA damage generatedby endogenous or exogenous mechanisms, and induces apoptosis(27). We examined whether neuronal cell death found in Polß-deficientmice (42) was induced by such p53 activation. To detect DNAfragmentation in cells undergoing apoptosis, we performed terminaldeoxynucleotidyltransferase-mediated dUTP biotin nick-end labeling(TUNEL) assay. In Polß^+/+ (wild-type) mice, TUNEL-positivecells were not detected in developing neocortex at E13.5 (Fig.1A). In contrast, in Polß^–/– developingneocortex, an extensive number of TUNEL-positive cells wereobserved in the primordial plexiform layer (PPL) and intermediatezone (IZ), where postmitotic neuronal cells were present; butthese cells were less detectable in the ventricular zone (VZ),where proliferating neuronal progenitor cells were present (Fig.1D).

View larger version (72K):
[in this window]
[in a new window]
FIG. 1. Polß deficiency induces apoptosis and activates p53 in the developing nervous system. Coronal sections of E13.5 developing neocortices in Polß^+/+ (wild-type) (A to C) and Polß^–/– (D to F) embryos were assayed by TUNEL staining (A and D) or immunohistochemistry with anti-cleaved caspase-3 antibody (B and E) and anti-p53 antibody (C and F). LV, lateral ventricle.

We next examined the expression of cleaved caspase-3 and p53by immunohistochemistry in E13.5 developing neocortex. Sincecaspase-3 is activated by proteolytic cleavage of its inactivezymogen by apoptosis (35), cleaved caspase-3 is a useful markerfor apoptotic cells. Anti-cleaved caspase-3 antibody could obviouslydetect apoptotic neuronal cells in Polß^–/–embryos (Fig. 1E), as observed by the above TUNEL assay, whereaswild-type embryos were not stained (Fig. 1B). Importantly, alarger number of cells with higher p53 levels were seen in Polß^–/–embryos than in wild-type (Fig. 1C versus 1F); in the VZ, suchp53-stained cells were more frequent than the TUNEL- or cleavedcaspase-3-positive cells. It is known that serine-15 of humanp53 (equivalent to serine-18 in mouse p53) is phosphorylatedin response to DNA damage (39). Therefore, we performed Westernblot analysis with anti-phosphorylated serine-15 of human p53specific antibody with cell extracts from E13.5 developing telencephalonsof Polß^+/+ p53^+/+ (wild-type), Polß^–/–p53^+/+, and Polß^–/– p53^–/–embryos (generation of the double-mutant mice will be citedbelow). We found that the serine-18 of p53 in Polß^–/–p53^+/+ extracts was strongly phosphorylated compared to thatin wild-type control, with no staining in Polß^–/–p53^–/– extracts (Fig. 2). These results suggestan intriguing possibility that neuronal apoptosis observed inPolß-deficient embryos is induced by p53 activationin response to DNA damage, immediately after final mitosis ofthe progenitor cells.

View larger version (49K):
[in this window]
[in a new window]
FIG. 2. Western blot analysis of phosphorylated serine-18 of p53 with cell extracts prepared from E13.5 developing telencephalons of Polß^+/+ p53^+/+ (wild-type), Polß^–/– p53^+/+, and Polß^–/– p53^–/– embryos. The same amounts of protein were analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, followed by immunoblotting with anti-human phospho-p53 (Ser-15) antibody.

To explore potential physiological interactions between Polßand p53, we produced mice null for both Polß and p53.We bred Polß^+/– and p53^+/– mice and generatedPolß^+/– p53^+/– double-heterozygous mice,which exhibited normal development and fertility. When Polß^+/–p53^+/– mice were intercrossed, developing embryos withnearly the expected Mendelian ratios of all genotypes were observedat E11.5 to E18.5 (Table 1). However, no offspring with Polß^–/–p53^+/+, Polß^–/– p53^+/–, and Polß^–/–p53^–/– genotypes was detected at weaning (ca. 4weeks after birth). It should be noted that, like Polß^–/–p53^+/+ neonates, all Polß^–/– p53^+/–and Polß^–/– p53^–/– neonatesdied at postnatal day 1 (data not shown). With anti-cleavedcaspase-3 antibody, we examined neuronal apoptotic phenotypesby staining sections of the developing telencephalon in E13.5embryos and the spinal cord and dorsal root ganglion in E11.5embryos of wild-type controls and littermates. In tissues ofPolß^–/– 53^+/+ mice, a large number ofcleaved caspase-3-positive neuronal cells were observed comparedto those of control mice (compare Fig. 3A and E with 3B andF). However, these stained cells were dramatically decreasedin Polß^–/– p53^+/– mice (Fig. 3C and G)and, more importantly, completely disappeared in Polß^–/–p53^–/– double-mutant mice (Fig. 3D and H). Theseobservations indicate that p53 deficiency rescues the neuronalapoptosis associated with Polß deficiency and thatp53 haploinsufficiency also substantially does so. These resultsindicate that the neuronal apoptosis in Polß-deficientmice is mediated by the p53-dependent apoptosis pathway.

View this table:
[in this window]
[in a new window]
TABLE 1. Genotypic analysis of Polß^+/– p53^+/– intercrosses

View larger version (112K):
[in this window]
[in a new window]
FIG. 3. Neuronal apoptosis in Polß-deficient mice is mediated by the p53-dependent pathway. Coronal sections of E13.5 telencephalons (A to D) and E11.5 spinal cords and dorsal root ganglions (E to H) in Polß^+/+ p53^+/+ (wild-type) (A and E), Polß^–/– p53^+/+ (B and F), Polß^–/– p53^+/– (C and G), and Polß^–/– p53^–/– (D and H) embryos were stained with anti-cleaved caspase-3 antibody. DRG, dorsal root ganglion; LGE, lateral ganglionic eminence; MGE, medial ganglionic eminence; NCX, neocortex; SC, spinal cord.

Proliferation and early differentiation of progenitors during neurogenesis appear normal in E13.5 Polß^–/– p53^–/– mice. Proliferation and differentiation of progenitors are temporallyand spatially controlled during neurogenesis (12). The generationof neurons from the progenitors involves successive steps incommitment and differentiation, which are progressively generatingmore restricted cell types. These steps include (i) cell typespecification, (ii) exit from the cell cycle, (iii) differentiationinto distinct cell types, (iv) migration into a correct destination,and (v) production of correct cell-cell contacts through dendriticand axonal processes. We examined effects of p53 deficiencyon histogenesis of the neocortex by cresyl violet staining (Fig.4A to F and 4P to R). In E13.5 wild-type lateral regions ofthe neocortex, generation of the PPL, IZ, and cortical plate(CP) was clearly observed (Fig. 4A), but the CP in the dorsalregion was not (Fig. 4D). The ventrolateral-to-dorsomedial,morphological gradients are evident in the development of theneocortex (5). In contrast, in Polß^–/–p53^+/+ neocortices, the CP was not clearly seen due to an extraordinarynumber of both pyknotic and cleaved caspase 3-positive cellsundergoing apoptosis (Fig. 4B, E, and H). However, in Polß^–/–p53^–/– neocortices, the CP was almost normally generated(Fig. 4C), parallel with a striking disappearance of apoptoticcells (Fig. 4F and I). With the use of antibody against eitherPCNA, a proliferating cell marker (Fig. 4J to L), or neuron-specifictype III ß-tubulin, an early neuron marker (Fig. 4Mto O), we observed the state of neuronal cells escaping fromapoptosis in E13.5 developing neocortex. In the PPL and IZ ofPolß^–/– p53^+/+ mice, expression of PCNAappeared in some apoptotic cells (Fig. 4K), whereas expressionof type III ß-tubulin was reduced with increasingapoptotic cells (compare Fig. 4N and M). Expression of PCNAis regulated by p53 in response to ionizing radiation in neuronalcells (45). Therefore, the PCNA activation in the apoptoticcells of Polß^–/– p53^+/+ mice (Fig. 4K)might be due to a response to certain DNA damage. On the otherhand, in Polß^–/– p53^–/– mice,we found intact expression of both PCNA and type III ß-tubulin,similar to wild-type controls (Fig. 4L and O, respectively).Neuronal progenitors expressing PCNA were restricted in theVZ (Fig. 4L); the differentiating neurons present in the PPLand IZ expressed type III ß-tubulin at a normal level(Fig. 4O). Finally, at E14.5, formation of the CP in Polß^–/–p53^–/– mice was more clearly observed than thatin Polß^–/– p53^+/+ mice (compare Fig. 4R and Q).Together, it appears that in Polß^–/–p53^–/– embryos early differentiation of neuronalcells escaping from apoptosis proceeds normally.

View larger version (73K):
[in this window]
[in a new window]
FIG. 4. Neuronal progenitors in E13.5 Polß^–/– p53^–/– embryos appear to normally proliferate and differentiate. Coronal sections of telencephalons in Polß^+/+ p53^+/+ (wild-type) (A, D, G, J, M, and P), Polß^–/–p53^+/+ (B, E, H, K, N, and Q), and Polß^–/– p53^–/– (C, F, I, L, O, and R) embryos at E13.5 (A to O) and E14.5 (P to R) were stained with cresyl violet (A to F and P to R). The sections were also stained with anti-cleaved caspase-3 antibody (G to I), anti-PCNA antibody (red) and DAPI (blue) (J to L), or anti-neuron specific type III ß-tubulin antibody TuJ1 (M to O). Arrows in panels A, C, P, Q, and R indicate the CP. Arrowheads in panel K indicate PCNA-activated cells. LV, lateral ventricle.

Formation of the nervous system is incomplete in E18.5 Polß^–/– p53^–/– mice. p53 deficiency dramatically rescued neuronal apoptosis in Polß-deficientmice (Fig. 3). As mentioned above, early neuronal differentiationappeared to proceed normally in E13.5 Polß^–/–p53^–/– embryos (Fig. 4). As shown in Table 1, Polß^–/–p53^–/– embryos could survive during gestation anddied at postnatal day 1. Therefore, we examined the developmentof their brains at E18.5 by immunohistochemical analysis (Fig.5). We observed slight but significant, neuronal defects inthe telencephalon of Polß^–/– p53^–/–embryos. Cresyl violet staining revealed that, compared to Polß^–/–p53^+/+ mice, the size of the telencephalon and its cytoarchitecturewere moderately recovered in Polß^–/– p53^–/–mice (compare Fig. 5B and C). We analyzed the axonal tract formationby staining phosphorylated neurofilaments. In Polß^–/–p53^+/+ brains, the major axonal tract, anterior commissure didnot cross at the midline (Fig. 5E); notably, this defect couldnot be rescued by p53 deficiency (Fig. 5F). We also observedmore aberrant axonal tracts in the striatum of Polß^–/–p53^+/+ brains (Fig. 5H) than in that of wild-type controls (Fig.5G), and similar aberrations were observed in Polß^–/–p53^–/– brains (Fig. 5I). These results indicatethat, at least in some areas, the brain development in Polß^–/–p53^–/– embryos was not complete. Recently identifiedis a cell type migrating from the lateral ganglionic eminenceand the medial ganglionic eminence of the basal ganglia to theneocortex of telencephalons (2, 32). The tangentially migratingcells become interneurons synthesizing inhibitory neurotransmitter ${gamma}$ -aminobutyric acid (GABA) in the CP of the neocortex. A greatnumber of neuronal apoptotic cells were observed in both thelateral ganglionic eminence and the medial ganglionic eminencein Polß^–/– p53^+/+ embryos at E13.5 (compareFig. 3A and B). We therefore examined whether GABAergic interneuronsexisted in the E18.5 neocortex. The interneurons in developingneocortex are known to express calbindin D28k (Calbindin), anintracellular calcium-binding protein (2, 32). In Polß^–/–p53^+/+ embryos (Fig. 5K), we found a significant decrease inthe number of calbindin-positive cells in the CP compared towild-type embryos (Fig. 5J). More importantly, this decreasewas not rescued by p53 deficiency in Polß^–/–p53^–/– embryos (Fig. 5L). Taken together, thesedata, shown in Fig. 5, indicate that p53 deficiency does notcompletely rescue developmental defects in the central nervoussystem, associated with Polß deficiency. These phenotypesare in sharp contrast to those of Lig4^–/– p53^–/–and XRCC4^–/– p53^–/– mice, which canbe alive for several weeks after birth. These results suggesta crucial role for Polß in the formation of the intactneuronal circuit and migration of certain neuronal cell types.

View larger version (82K):
[in this window]
[in a new window]
FIG. 5. Development of the brain is incomplete in E18.5 Polß^–/– p53^–/– embryos. Coronal sections of telencephalons in Polß^+/+ p53^+/+ (wild-type) (A, D, G, and J), Polß^–/– p53^+/+ (B, E, H, and K), and Polß^–/–p53^–/– (C, F, I, and L) embryos at E18.5 were stained with cresyl violet (A to C), anti-phosphorylated neurofilament antibody SMI31 (D to I), and anticalbindin antibody (J to L). Photographs of the sections in each genotype were taken with the same magnification. The arrowheads in panels A, B, and C indicate part of the cytoarchitecture recovered moderately in Polß^–/– p53^–/– mice compared to Polß^–/– p53^+/+. The arrows in panels D, E, and F indicate the anterior commissure (AC). The arrows in panels G, H, and I indicate aberrant axonal tracts in the striatum (St). The arrows in panel J indicate calbindin-positive cells. MZ, marginal zone.

DISCUSSION

Top

Discussion

We have shown here that p53 deficiency rescues neuronal apoptosisin Polß-deficient mice and that haploinsufficiencysubstantially does so as well (Fig. 3). We have also shown thatp53 is activated by its serine-18 phosphorylation in developingtelencephalons of Polß-deficient mice (Fig. 2). Theseresults indicate that the neuronal apoptosis associated withPolß deficiency is mediated by the p53-dependent pathway.In addition, it should be noted that, like Polß^–/–p53^+/+ neonates, Polß^–/– p53^–/–neonates die shortly after birth.

The onset of p53-dependent repair or apoptosis is determinedby the level of accumulated damaged DNA (36). There is amplebiochemical evidence for functioning of Polß mainlyin the short-patch BER pathway to repair SSBs, which are mostlygenerated as intermediates of damaged bases metabolized by DNAglycosylase and AP endonuclease (46). Mouse embryonic fibroblastsdefective in Polß are highly sensitive to DNA alkylatingagents but not to X-ray radiation (41, 42), supporting in vivothat Polß deficiency leads to defects in SSB repairbut not in DSB repair. Thus, Polß deficiency shouldresult in increased levels of SSBs, even if the Polß-independentlong-patch BER is able to partially substitute for the short-patchBER. The increased SSB levels would stabilize and activate p53,leading to apoptosis during neuronal differentiation in Polß^–/–mice. In mice defective in Lig4, XRCC4, Ku70/80, or XRCC2, whichall function in DSB repair, differentiating neurons undergomassive apoptosis (4, 11, 14, 17, 22). Therefore, in these mice,unrepaired DSBs are thought to be the cause of the apoptosis(4, 11, 14, 17, 22). The apoptosis in Lig4^–/– andXRCC4^–/– embryos is rescued by p53 deficiency (15,16). Lig4^–/– p53^–/– and XRCC4^–/–p53^–/– neonates can survive several weeks afterbirth without behavioral or neurological abnormalities. Thisis in sharp contrast with our observation that Polß^–/–p53^–/– neonates die shortly after birth. The degreeand quality of rescue by p53 deficiency in repair-deficientmice appear to vary depending on the type and level of DNA damage.As discussed above, in Polß^–/– mice, thedamage is most likely SSBs, but the possibility that these SSBsare subsequently converted into DSBs in the final DNA replicationof neuronal progenitor cells cannot be ruled out.

In Polß^–/– p53^–/– mice atE13.5, early steps of neuronal differentiation seem to proceednormally, as judged by immunohistochemical analysis (Fig. 4).However, at E18.5, these and Polß^–/– p53^+/+mice displayed serious cytoarchitectural defects in the majoraxonal tract (Fig. 5E and F) (with more aberrant axonal tractsin the striatum [Fig. 5H and I]) and the migration in GABAergicinterneurons (Fig. 5K and L). These results suggest that, althoughp53 deficiency indeed rescues neuronal apoptosis, these neuronsare still incomplete as mature ones, implying that the deficiencycannot fully restore the neuronal development of at least certaincell types. The brain is composed of remarkably complex neuronalcell types and networks. In the development of the brain, cellmigration, axon growth, and pathfinding are fundamental processes(12). Recent studies with knockout mice have identified a numberof molecules responsible for such processes (30, 32). Loss ofthese molecules severely affects the brain development and iscritical for survival. The abnormal development of the nervoussystem observed in both Polß^–/– p53^–/–and Polß^–/– p53^+/+ mice at E18.5 may beresponsible for death shortly after birth. In Lig4^–/–p53^–/– or XRCC4^–/– p53^–/–mice, severe defects in lymphogenesis are never recovered byp53 deficiency, implying that Lig4 or XRCC4 is a critical factorfor lymphogenesis (15, 16). Similarly, our finding that theneuronal differentiation in Polß^–/– miceis not completely rescued by p53 deficiency strongly suggeststhat Polß is a critical factor for neurogenesis; thatis, Polß may absolutely be required for neuronal differentiation.

The reason why Polß is required for neuronal differentiationremains obscure. One possibility is that in neuronal differentiation,a large amount of damaged bases and SSBs are generated by reactiveoxygen species, which might occur particularly in some neuronalcell types actively undergoing migration and/or axon pathfinding.Recently, the Polß-dependent pathway was shown tobe induced in response to oxidative base damage (7). Polßmight specifically be required to repair those damaged basesand SSBs. Thus, Polß deficiency would lead to increasedlevels of DNA damage and activation of p53, eventually resultingin apoptosis. A second possibility is that Polß isinvolved in chromatin remodeling and transcription in neuronaldifferentiation. When neuronal progenitor cells become postmitoticneurons, they exit cell cycle and drastically alter the patternof gene expression from immature to mature neurons (12). Transcriptionalactivation of a gene involves recruitment of not only a sequence-specificDNA-binding protein but also a coactivator complex, includingproteins with chromatin-modifying activity. For example, DNAtopoisomerase IIß alters DNA topology and forms complexeswith proteins involved in chromatin remodeling and transcription(25, 44). The enzyme-deficient mice show defects in the laminarorganization of the neocortex and motor axon growth, resultingin a breathing impairment and death of the pups shortly afterbirth (31, 49). This finding suggests that the control of chromatinreorganization is indispensable for neuronal differentiation.Interestingly, we note that transcriptional coactivator p300forms a physical and functional interaction with Polß(23). p300 integrates a diverse signaling pathway for a numberof sequence-specific transcription factors and activates transcriptionthrough chromatin remodeling via intrinsic histone acetyltransferaseactivity (20). Therefore, in association with p300 or relatedproteins, Polß might function to maintain the integrityof genes being, or to be, expressed in certain neuronal celltypes. A third possibility is that during neuronal differentiation,a genomic rearrangement factor(s) is expressed and generatesa certain type of DNA damage (repairable by Polß)to initiate a specific differentiation. In the immune system,the molecular mechanism of diversity by rearrangement of theimmunoglobulin or T-cell receptor gene clusters is well understood(43). In V(D)J recombination, the lymphocyte-specific endonucleasesRAG1 and RAG2 initially cleave specific recognition sequencesin immunoglobulin loci, followed by completion of rearrangementsthrough DSB repair by the action of nonhomologous-end-joiningfactors (18). Similarly, in the nervous system, neuronal diversitymight be created by such genomic rearrangement (9, 48). If thisis the case, DNA repair by Polß would be an essentialpart of the diversity mechanism.

In conclusion, our studies show that p53 deficiency dramaticallyrescues neuronal apoptosis associated with Polß deficiency,indicating that p53 mediates the apoptotic process in the nervoussystem. However, p53 deficiency cannot restore complete differentiationof neuronal progenitors and leads to lethality shortly afterbirth. These observations suggest a crucial role for Polßin differentiation of specific neuronal cell types. In addition,it is evident that in neuronal differentiation, p53 acts asa gatekeeper to maintain genomic stability against various typesof DNA damage (27). Further studies will be needed to elucidatethe precise role of Polß in neurogenesis.

ACKNOWLEDGMENTS

We thank T. Yamakuni (Tohoku University) for the gift of theanti-calbindin antibody, Y. Tanabe (Mitsubishi Kagaku Instituteof Life Sciences) for helpful discussion, and N. Adachi (YokohamaCity University) for critical reading of the manuscript. Wealso thank C. Nishigaki for technical support and F. Oonumafor animal care.

N.S. is a recipient of Research Fellowship of the Japan Societyfor the Promotion of Science for Young Scientists. This studywas supported in part by a Grant-in-Aid for Scientific Research(C) from the Ministry of Education, Culture, Sports, Science,and Technology of Japan.

FOOTNOTES

* Corresponding author. Mailing address for Hideki Koyama: Kihara Institute for Biological Research and Graduate School of Integrated Science, Yokohama City University, 641-12 Maioka-cho, Totsuka-ku, Yokohama 244-0813, Japan. Phone: 81-45-820-1907. Fax: 81-45-820-1901. E-mail: koyama@yokohama-cu.ac.jp. Mailing address for Keiko Takiguchi-Hayashi: Research Laboratory 1 (CNS), Pharmaceuticals Research Unit, Research and Development Division, Mitsubishi Pharma Corp., 1000, Kamoshida-cho, Aoba-ku, Yokohama 227-0033, Japan. Phone: 81-45-963-3413. Fax: 81-45-963-3967. E-mail: Hayashi.Keiko{at}mp.m-pharma.co.jp.

Present address: Graduate School of Frontier Biosciences, OsakaUniversity, Suita, Osaka 565-0871, Japan.

REFERENCES

Top