Introduction

Within the nervous system, appropriate responses to DNA damage are required to maintain homeostasis and prevent disease (McMurray, 2005; Lee and McKinnon, 2006). DNA double-strand breaks (DSBs) trigger a signalling cascade that leads to repair and resolution of the break, or as is frequent in the developing nervous system, apoptosis. The repair of DNA DSBs occurs via two mechanistically distinct pathways non-homologous end-joining (NHEJ) or homologous recombination (HR). Each pathway involves a distinct repertoire of repair enzymes and associated proteins. HR requires a group of RAD51-related proteins and a variety of other factors, including BRCA2, to ensure high-fidelity DNA repair using an undamaged homologous DNA template to replace an adjacent damaged one (Thompson and Schild, 2002; West, 2003). In contrast, NHEJ facilitates direct modification and ligation of the two DNA ends at the DSB. Efficient NHEJ requires among other factors, KU heterodimers (KU 70 and KU 80), DNA-PKcs, DNA ligase IV (LIG4) and XRCC4 (Lees-Miller and Meek, 2003; Lieber et al, 2003; Mills et al, 2003; O'Driscoll and Jeggo, 2006). Inactivation of either of these pathways in mice can lead to embryonic lethality or tumor development.

Broad insights illuminating DNA repair and nervous system development have been gained from mice in which gene targeting has disabled various DNA repair pathways (Friedberg and Meira, 2006). For example, inactivation of NHEJ in the mouse can lead to defective neurogenesis or brain tumors depending on the particular gene disruption and genetic background (Gao et al, 1998, 2000; Lee and McKinnon, 2002; Yan et al, 2006). Disruption of HR can also affect neural development (Deans et al, 2000; Orii et al, 2006), although when some key HR components such as Brca2 or Rad51 are inactivated, the result is early embryonic lethality during gastrulation (before neural development) (Lim and Hasty, 1996; Tsuzuki et al, 1996; Ludwig et al, 1997; Sharan et al, 1997; Suzuki et al, 1997). Moreover, DNA repair activity exhibits a clear tissue- and cell type specificity, as in the nervous system, HR only functions in proliferating cells, while NHEJ is particularly important in differentiated neural cells (Orii et al, 2006). Together, these data show that DNA DSB repair pathways operate in a complementary manner during neural development.

BRCA2 has a key role in HR and substantial DNA repair-associated defects result from its inactivation, including DNA damage hypersensitivity, chromosomal rearrangements and defective mammalian gametogenesis (Sharan et al, 1997; Tutt and Ashworth, 2002; Yang et al, 2002; Daniels et al, 2004). BRCA2 is a large protein of 3418 amino acids that physically interacts through its carboxyl terminus with RAD51, a protein essential for HR, and is responsible for the translocation of RAD51 to sites of DNA damage processing (Powell et al, 2002; Pellegrini and Venkitaraman, 2004; Shivji and Venkitaraman, 2004; Yang et al, 2005). BRCA2 has also been implicated in cell-cycle regulation via interaction with BRAF35 or Smad3 (Marmorstein et al, 2001; Preobrazhenska et al, 2002), and a role during an intra-S phase checkpoint has also been reported (Taniguchi et al, 2002). Brca2 functions as a tumor suppressor, as its loss confers susceptibility to breast, ovarian and brain tumors (Hughes-Davies et al, 2003; Offit et al, 2003; Shivji and Venkitaraman, 2004). The BRCA2-binding protein BALB2 is important for the tumor suppressor function of BRCA2 (Xia et al, 2006; Erkko et al, 2007; Rahman et al, 2007).

Biallelic hypomorphic mutations of BRCA2 are also responsible for some cases of Fanconi Anemia (FA), a rare autosomal recessive cancer susceptibility syndrome characterized by congenital abnormalities, progressive bone narrow failure and cellular hypersensitivity to DNA cross-linking agents (Kennedy and D'Andrea, 2005; Taniguchi and D'Andrea, 2006). FA results from inactivating mutations in any one of a multiprotein complex that functions during DNA repair (D'Andrea, 2003; Kennedy and D'Andrea, 2005). FA patients carrying BRCA2 mutations exhibit a more severe phenotype compared with other FA groups, including predisposition to medulloblastoma brain tumors (Offit et al, 2003; Shivji and Venkitaraman, 2004). Defects in the BRCA2 partner-protein PALB2 can also result in FA (Reid et al, 2007; Xia et al, 2007).

Given the importance of BRCA2 during HR and the lethality associated with germ-line inactivation, we used conditional gene inactivation via Cre/LoxP technology to determine the requirement for Brca2 during neural development. Inactivation of Brca2 resulted in microcephaly associated with defects in neurogenesis particularly during cerebellar development. Loss of Brca2 also led to medulloblastoma when p53 was disabled. Thus, Brca2 fulfils a critical role during nervous system development, highlighting the tissue-specific requirements for DNA repair during neurogenesis.

Results

Inactivation of Brca2 leads to microcephaly and cerebellar defects

To determine the requirement for Brca2 during development of the nervous system, we used a Brca2 conditional mutant allele in which exon 11 was flanked by LoxP sites (Jonkers et al, 2001). To inactivate Brca2, we used Cre that was driven by the Nestin promoter, resulting in expression throughout the central and peripheral nervous systems from embryonic day 10.5 (Dahlstrand et al, 1995; Frappart et al, 2005). Similar to other reports using this Nestin-cre transgenic line (Graus-Porta et al, 2001; Frappart et al, 2005), we observed efficient deletion of Brca2 throughout the nervous system, as determined using genomic DNA or RNA extracted from the mutant cerebrum or cerebellum (Supplementary Figure 1).

In contrast to early embryonic lethality after germ-line inactivation of Brca2 (Ludwig et al, 1997; Sharan et al, 1997; Suzuki et al, 1997), Brca2^LoxP/LoxP;Nestin-cre mice (hereafter referred to as Brca2^Nes-cre) were viable and were relatively normal in overall appearance, size and behavior. However, inactivation of Brca2 led to microcephaly as Brca2^Nes-cre brains were smaller and weighed significantly less (P<0.0001) compared with littermate controls (Brca2^LoxP/LoxP or Brca2^+/+;Nestin-cre littermates; hereafter referred to as Brca2^Ctrl) (Figure 1A–C). All Brca2^Nes-cre brain structures were proportionally affected, including the hippocampus, cortex and olfactory bulb (Figure 1A). However, despite the effect on brain size, overall neural development was relatively normal and normal cortical lamination was present (Figure 1A). Notably, the Brca2-deficient cerebellum exhibited stunted foliation and lobule morphology (Figure 1D), and ectopic localization of some Purkinje cells was observed (Figure 1E), suggesting that migration defects might occur during Brca2^Nes-cre cerebellar development. Although Cre expression in the brain can lead to microcephaly in some situations (Forni et al, 2006), on no occasion did we observe this with Nestin-cre mice that were wild type (WT) or heterozygous for the Brca2 mutant allele.

As the phenotypic severity of many DNA repair mutant mice are rescued by associated p53 inactivation (Gao et al, 2000; Lee and McKinnon, 2002), we generated Brca2^Nes-cre mice in which p53 was also inactivated. We found that p53 inactivation significantly rescued the microcephaly of Brca2-deficient mice (P<0.0001; Figure 1B–D) and contributed to restoration of cerebellar structure (Figure 1D and E);p53 heterozygosity also significantly improved cerebellar development, but had no significant effect on the occurrence of microcephaly (data not shown).

Loss of Brca2 activates DNA damage-induced apoptosis

As p53 functions to signal DNA damage, we questioned if the phenotypic rescue resulting from p53 loss resulted from a block in DNA damage signalling to activate apoptosis. Therefore, to determine if Brca2 loss in the nervous system resulted in DNA damage, we assessed phosphorylated H2AX (H2AX) levels in Brca2^Nes-cre brains; H2AX occurs in response to DNA DSBs and is a canonical marker for this type of damage (Rogakou et al, 1998; Fernandez-Capetillo et al, 2004). We found that the cells of the cerebellar external germinal layer (EGL) of mutant mice exhibited increased H2AX foci (Figure 2A). These foci were concentrated in the outer EGL of the cerebellum that corresponds to the proliferative granule cell progenitors, suggesting that DNA DSBs resulting from Brca2 loss likely occurred during replication. This is consistent with a role for BRCA2 in stabilizing DNA structures at stalled replication forks (Lomonosov et al, 2003).

DNA damage during development of the nervous system can result in apoptosis (Lee and McKinnon, 2006). To determine if the DNA DSBs identified by -H2AX resulted in apoptosis, we used TUNEL or apoptosis-associated single-stranded DNA (ssDNA) assays (Frankfurt et al, 1996; Kawarada et al, 1998) to assess the WT and mutant developing cerebellum. We found that Brca2 inactivation resulted in apoptosis in the cerebellar EGL, while co-inactivation of p53 led to a significant decrease of apoptosis (Figure 2B and C). Apoptosis was restricted to proliferative granule cell progenitors and some early differentiating post-mitotic cells (that may have incurred sublethal damage during proliferation) in the Brca2^Nes-cre EGL, as demarcated by immunostaining for Tag-1, a marker for premigratory granule cells (Supplementary Figure 2). However, some apoptosis still occurred in Brca2^Nes-cre;p53^-/- cerebella, indicating that while p53-dependent signalling accounts for most apoptosis in the Brca2-deficient cerebellum, there is some that is p53 independent. To confirm that the apoptosis assays specifically reflect apoptotic cells, we also quantified cells with pyknotic nuclei that are indicative of apoptosis. As shown in Figure 2D, the abundance of pyknotic cells reflects the number found using the TUNEL assay. Under the conditions used here, these distinct assay methods are accurate indicators of apoptosis and do not simply label damaged DNA. Both methods failed to show any positive apoptotic signal in various brain tissues up to 2 h after ionizing radiation treatment despite the presence of substantial levels of -H2AX in these tissues (data not shown). Finally, both ssDNA (Figure 2E) and TUNEL (not shown) positive signal were associated with pyknotic cells. Therefore, loss of Brca2 invokes a DNA damage response that can lead to apoptosis during development.

Brca2 inactivation promotes increased apoptosis but not proliferation defects

While apoptosis results from loss of Brca2, it is possible that proliferation defects also contribute to microcephaly. Therefore, we used bromodeoxyuridine (BrdU) incorporation to quantify proliferation, but found no statistically significant differences in BrdU incorporation between the mutant or WT genotypes at 90 min (Figure 3A) or 6 h (data not shown). Thus, Brca2 loss results in apoptosis, but does not appear to affect granule neuron progenitor proliferation.

Because Brca2-deficient cells in vitro can accumulate in G2 (Patel et al, 1998; Marmorstein et al, 2001), we examined cell-cycle progression using phosphohistone H3-specific immunostaining to identify cells in G2/M. We found a significant increase (P=0.0229) of cells that were immunopositive for phosphorylated histone H3 in Brca2^Nes-cre cerebellar EGL compared with control tissue (Figure 3B). We also found that the increase of G2/M cells after Brca2 loss was present in the Brca2^Nes-cre;p53^-/- EGL, indicating that this cell-cycle block is independent of p53 (Figure 3B). These data are consistent with DNA damaged cells activating a G2 arrest before either DNA repair or apoptosis (Kastan and Bartek, 2004; Sancar et al, 2004).

Apoptosis induced by Brca2 loss is present throughout neural development

To further assess the effects of Brca2 loss, we analyzed neural development at different developmental stages. We quantified BrdU-positive cells in the cerebellar primordia at E14.5 and similar to postnatal cerebellar development, we did not observe any differences in proliferation between control (Brca2^+/+;Nestin-cre) and Brca2^Nes-cre at this age (Figure 4A), although apoptosis was increased at this stage and remained elevated through postnatal development (Figure 4B).

We also analyzed other brain regions including the E14.5 neuroepithelium of the hindbrain, and again found no difference in proliferation, but there was an increase in apoptosis within this region of the central nervous system (CNS) (Figure 4C and D). Therefore, cell loss from apoptosis most likely accounts for the occurrence of microcephaly present in Brca2^Nes-cre animals.

Atm deficiency partially restores morphology of the Brca2^Nes-cre cerebellum

Atm is required for DNA DSB-induced apoptosis in select neural populations (Lee et al, 2000; Sekiguchi et al, 2001; Kruman et al, 2004; Orii et al, 2006), although DNA damage from defective HR does not signal Atm (Orii et al, 2006; Adam et al, 2007). However, in those studies, the HR mutation was germ-line inactivation of Xrcc2, and resulted in early embryonic lethality. This early embryonic lethality precluded analysis of Atm signalling during later neural development. Therefore, to further determine Atm function after disruption of HR, we crossed Brca2 mutant mice with Atm^+/- mice and obtained Brca2^Nes-cre;Atm^-/- animals. Notably, we found that Atm deficiency promoted partial recovery of cerebellar development (Figure 5A–C). However, in contrast to p53 deficiency, Atm loss did not rescue the microcephaly resulting from Brca2 inactivation (Figure 5A–C). While Atm deficiency contributed to a reduction of apoptosis in the EGL (Figure 5D), it did not alter proliferation of the granule cell progenitors (Figure 5D). However, when we measured TUNEL-positive cells in the EGL, the statistical significance (P=0.1219) between the Brca2^Nes-cre and the Brca2^Nes-cre; Atm^-/- mice suggested that Atm loss contributes to the rescue of only a small fraction of the apoptotic cells in the Brca2^Nes-cre EGL. This partial rescue of the Brca2^Nes-cre cerebellum is consistent with a role for Atm after granule precursors exit the cell cycle, rather than a primary function in proliferating cells (Lee et al, 2001). ATM may therefore act as a backup surveillance to ensure cells containing DNA damage don't become incorporated into mature neural tissue.

Brca2 deficiency leads to defects in neural progenitor cell self-renewal and proliferation

The previous analyses indicate apoptosis is increased after loss of Brca2, leading to defective neural development, and both proliferative progenitor cells and early post-mitotic neurons are affected. To further investigate potential targets of Brca2-deficiency, we examined neurosphere cultures.

Neurospheres were established from Brca2^Ctrl, Brca2^Nes-cre and Brca2^Nes-cre;p53^-/- brains at E14.5 and P0 (Figure 6A and B). PCR analysis showed that there was efficient gene deletion in E14.5 Brca2^Nes-cre neurospheres (data not shown). There was a substantial reduction in the number and size of neurospheres derived from the Brca2^Nes-cre brains compared with those from Brca2^Ctrl or Brca2^Nes-cre;p53^-/- brains (Figure 6A and B). When E14.5 neurospheres were cultured, the number of spheres was significantly reduced in Brca2-deficient animals after 7 days in culture (P<0.0001) (Figure 6B) compared with Brca2^Ctrl and Brca2^Nes-cre;p53^-/- animals, potentially indicating less CNS stem/progenitor cells in Brca2-deficient animals, or alternatively, decreased proliferation or survival in culture. The Brca2-deficient neurospheres propagated less readily, as indicated by a smaller number of cells per sphere (Figure 6C). Notably, at later stages, we were also unable to isolate P0 Brca2-deficient neurospheres after two independent attempts (data not shown). However, inactivation of p53 restored the number of Brca2^Nes-cre neurospheres and number of cells per neurosphere after 7 days in culture of either E14.5 or P0 neural stem cells (Figure 6B–D). To determine why Brca2-deficient neurospheres were compromised, we analyzed apoptosis and proliferation. Similar to Brca2 loss in vivo, we found that there was increased apoptosis in Brca2^Nes-cre neurospheres, but we found reduced BrdU incorporation in neural progenitor cells in vitro (Figure 6E). However, p53 inactivation restored normal proliferation and substantially reduced apoptosis in Brca2^Nes-cre;p53^-/- neurospheres (Figure 6E and F). The discrepancy between proliferation defects in vitro but not in vivo after Brca2 loss may reflect culture stress activating cell-cycle checkpoints. Alternatively, as stem cells are a minor population, our BrdU assays may have missed this compartment in the tissue we examined. Overall, these data suggest that Brca2^Nes-cre neural progenitors undergo increased rates of apoptosis and their loss may additionally contribute to the microcephaly observed in the Brca2^Nes-cre nervous system.

Brca2 is required to suppress medulloblastoma

In the nervous system, inactivation of DNA DSB repair can lead to medulloblastoma (Lee and McKinnon, 2002; Holcomb et al, 2006; Yan et al, 2006), and BRCA2 mutations in FANCD1 are also associated with this brain tumor (Offit et al, 2003). Therefore, to determine if Brca2 functions as a tumor suppressor in the brain, we monitored tumor formation in five experimental groups over a period of 32 weeks: Brca2^Ctrl, Brca2^Nes-cre, Brca2^Nes-cre;p53^+/-, p53^-/- and Brca2^Nes-cre;p53^-/-. Although Brca2^Nes-cre mice exhibit a significantly shorter lifespan compared with control mice (Figure 7A) (P<0.0001), analysis of these mice did not reveal tumors (Table I). To determine the cause of the premature death of Brca2^Nes-cre mice, we performed full necropsies and blood tests on 8-week-old Brca2^Nes-cre animals. Although these animals did not exhibit obvious histological defects, hematological analysis revealed abnormalities in numbers of red blood cells and platelets in Brca2-deficient animals (data not shown). These results are consistent with aplastic anemia found in FA and explained by the occurrence of gene deletion in bone narrow using Nestin-cre (Betz et al, 1996).

However, from 10 weeks of age onwards, most Brca2^Nes-cre;p53^-/- mice became moribund with medulloblastoma (n=19/23; Table I and Figure 7A). Although p53^-/- mice generally succumb to lymphoid tumors, in no case did we observe medulloblastoma in these mice (n=29; Table I and Figure 7A). Notably, medulloblastoma also occurred in p53 heterozygous mice that were Brca2^Nes-cre (n=34/47), although with a significantly increased tumor latency compared with Brca2^Nes-cre;p53^-/- mice (13 weeks versus 21 weeks; P<0.0001) (Figure 7A and B; Table I). Consistent with this, arrayCGH or spectral karyotyping (SKY) identified genomic rearrangements of chromosome 11 (on which p53 resides) in Brca2^Nes-cre;p53^+/- tumors (Figure 7C, Table I and Supplementary Figure 2). SKY analysis showed chromosome 11 translocations involved various other chromosomes (Figure 7C), while arrayCGH showed loss of chromosomal material spanning the region of chromosome 11 containing p53 (Supplementary Figure 2). In contrast, no rearrangements or loss of chromosome 11 was found in Brca2^Nes-cre;p53^-/- tumors, indicating that p53 loss underpinned tumorigenesis in the Brca2^Nes-cre;p53^+/- mice (Supplementary Figure 2). Finally, PCR analysis performed on the same Brca2^Nes-cre;p53^+/- samples showed a clear loss of heterozygosity of p53, even in samples which did not show major p53 genomic rearrangements (Figure 7D). Therefore, similar to FA, loss of Brca2 can lead to medulloblastoma.

Discussion

Repair of DNA via HR is critical for maintenance of genomic integrity, and BRCA2 is a central component of this pathway (West, 2003; Pellegrini and Venkitaraman, 2004). Cells lacking BRCA2 show pronounced genetic instability and susceptibility to DNA damaging agents, while BRCA2 mutations predispose to breast, ovarian and prostate cancer (Tutt and Ashworth, 2002; Shivji and Venkitaraman, 2004). Initial in vivo studies revealed that germ-line deletion of Brca2 in the mouse was lethal at E6 (Ludwig et al, 1997; Sharan et al, 1997; Suzuki et al, 1997; Patel et al, 1998; Jonkers et al, 2001). While this finding pointed to the critical importance of Brca2, the early lethality precluded spatiotemporal analysis of this protein's function during development. In this current report, we find that in contrast to germ-line inactivation, Brca2^Nes-cre mice are viable. Notwithstanding this, it is clear that Brca2 fulfils an important role during neurogenesis, as Brca2^Nes-cre mice are microcephalic and exhibit altered neural development arising from the effects of chronic genotoxic stress. Brca2 is also required to suppress the formation of medulloblastoma brain tumors. These data highlight the importance of Brca2 during normal neural development for maintaining genomic stability and uncover novel tissue-specific requirements for this DNA repair factor.

Inactivation of many important DNA repair genes including Brca2, Mre11, Nbs1 or Rad51 disrupts gastrulation and therefore tissue formation, making broader biological interpretation of gene loss problematic (Friedberg and Meira, 2006). In other cases, disruption of DNA repair genes such as Xrcc4, Lig4 or Xrcc2 while also resulting in lethality do allow more substantial embryonic development, leading to phenotypes associated with DNA damage effects in specific tissue regions during development (Barnes et al, 1998; Frank et al, 1998; Gao et al, 1998; Orii et al, 2006). Bypassing embryonic lethality, via tissue-specific gene inactivation, provides unique functional insights and is an essential counterpoint for comparative analyses of other conditional DNA repair mouse mutants and myriad cellular in vitro studies.

This point is illustrated by comparing our current study with Brca2 to neural inactivation of Nbs1, where an overlapping but different phenotype was observed (Frappart et al, 2005). NBS1 is important for DNA DSB repair because Nijmegen Breakage syndrome individuals are radiosensitive, predisposed to tumors and isolated cells are hypersensitive to DNA DSB-inducing agents (Shiloh, 1997; Digweed and Sperling, 2004). Two important differences distinguish our current study from the Nbs1^Nes-cre conditional mutant mice; these are the lack of overt proliferation defects and the occurrence of medulloblastoma in Brca2^Nes-cre animals. Neural inactivation of Nbs1 leads to substantial impairment of proliferation of granule neuron progenitor cells resulting in increased apoptosis in post-mitotic neurons primarily in the cerebellum (Frappart et al, 2005). However, the impressive phenotypic rescue of the Nbs1^Nes-cre cerebellum by p53 loss does not result in medulloblastoma. This contrasting phenotype between DNA DSB repair factors could reflect different cellular roles, whereby NBS1 may be critical for monitoring DNA breaks during DNA replication possibly in collaboration with ATR (Pichierri and Rosselli, 2004; Stiff et al, 2005), while BRCA2 functions primarily to repair DNA DSBs during HR. Notably however, NBS1 is reportedly essential for HR (Tauchi et al, 2002; Yang et al, 2006), so the different phenotypes between Nbs1^Nes-cre and Brca2^Nes-cre may relate to cell- or tissue-specific effects, that the damage induced by Nbs1 loss is more severe or could indicate that NBS1 does not participate in HR in a physiological setting.

Although Brca2 is clearly important for maintenance of DNA integrity in proliferating granule neuron precursor cells, partial rescue of the Brca2^Nes-cre cerebellar phenotype occurs in Brca2^Nes-cre;Atm^-/- animals. Because Atm is not required for DNA damage signalling after disruption of HR in vivo (Orii et al, 2006; Adam et al, 2007), it is most likely that Atm-dependent apoptosis eliminates post-mitotic granule cells with unrepaired DNA damage (resulting from Brca2 loss during proliferation) to prevent migration of these cells to the inner granule layer. The kinase responsible for signalling DNA damage in the proliferative neural populations may well be the ATM-related kinase, ATR (Shiloh, 2003). While Atm involvement during cerebellar development in Nbs1^Nes-cre mice has not been reported, the synthetic lethality between Nbs1^B/B and Atm^-/- mice (Williams et al, 2002) may make interpretation of this compound genetic background difficult, potentially further highlighting the different roles these respective DNA repair factors fulfil during development.

We also found that Brca2 was a potent tumor suppressor in the brain. When Brca2^Nes-cre was introduced onto a p53 deficient background (either p53^+/- or p53^-/-), most mice succumbed to brain tumors, resulting from enhanced DNA damage during proliferation in the EGL. Previous studies showed that medulloblastoma can occur when DNA DSB repair factors are disabled, as loss of NHEJ in collaboration with defective p53 signalling led to this type of tumor, although in those cases, no tumors were reported in p53 heterozygotes (Gao et al, 1998, 2000; Lee and McKinnon, 2002; Yan et al, 2006). The high incidence of medulloblastoma in Brca2^Nes-cre;p53^+/- mice probably reflects the more severe effects of Brca2 disruption toward genomic integrity. Importantly, the human FA syndrome, which results from Brca2 mutations (FANCD1), also develops medulloblastoma (Offit et al, 2003), and microcephaly is a prevalent feature of FA (Gennery et al, 2004), indicating that Brca2 loss in the mouse recapitulates the neural aspects of this disease. Similar to other mutations common in human medulloblastoma, we found that the Ptch1 was generally lost in the Brca2^Nes-cre;p53^-/- medulloblastoma, while N-Myc was often amplified (data not shown). Loss of p53, or the p53 pathway, also features strongly in human medulloblastoma (Woodburn et al, 2001; Giordana et al, 2002; Lee et al, 2003; Frank et al, 2004), underscoring the relevance of the Brca2^Nes-cre;p53^-/- model to human disease. Maintenance of genomic integrity by Brca2 is also likely to be critical for prevention of other brain tumor types in addition to medulloblastoma.

Our data help to clarify the relative developmental contributions of DNA damage response factors in a biological setting, and indicate that genomic stability is critical for proper neural development. As more tissue-restricted DNA repair mutant mice are generated, we will further determine the relationship between different DNA repair pathways during development, thereby providing a biological context for understanding the interplay of DNA repair factors and their role in preventing disease.

Materials and methods

Generation of mice with Brca2 deleted in the nervous system

The Brca2-floxed mice (Jonkers et al, 2001) were obtained from the MMHCC repository at the NIH: (http://mouse.ncifcrf.gov/available_
strains.asp) and Nestin-cre transgenic mice (B6.Cg-Tg(Nes-cre)1Kln/J; JAX #003771) were obtained from the Jackson Laboratory. These were interbred in order to obtain Brca2^LoxP/LoxP;Nes-cre mice. Mice used in this study resulted from backcrossing mixed C57BL/J 129Ola for between two and four generations. During breeding, the Nes-cre transgene was routinely carried by the female to avoid germ-line Brca2 disruption due to spurious Cre expression in the testis. The control group (Brca2^Ctrl) was consisted of the following genotypes: Brca2^LoxP/LoxP, Brca2^+/LoxP;Nes-cre or Brca2^+/+;Nes-cre. Genotyping for the mutant Brca2 allele was as described (Jonkers et al, 2001), and the primers used for Cre genotyping were Cre-1: 5'-CGGTCGATGCAACGAGTGATG-3' and Cre-2: 5'-CCAGAGACGGAAATCCATCGC-3'. Inactivation of p53 or Atm was achieved using p53^-/- or Atm^-/- mice (Herzog et al, 1998) and these were interbred with Brca2^LoxP/+;Nes-cre mice, and F1 mice were used to generate Brca2^Nes-cre;p53^-/- or Brca2^Nes-cre;Atm^-/-.

Histological and immunohistochemical analyses

Histological analysis was carried out on 5 m paraffin sections stained with hematoxylin and eosin (H&E), or by immunostaining on 10 m cryosections, as described previously (Orii et al, 2006). Antibodies used were anti-calbindin D-28K, (1/500; Sigma-Aldrich) and anti-BrdU (1/500; Oxford Biotechnology). For the in vivo proliferation assays, newborn mice or pregnant females were injected intraperitoneally with BrdU (50 g/g of body weight) (Sigma-Aldrich). Embryos or brains were removed 90 min or 6 h after injection and fixed in 4% PBS-buffered paraformaldehyde. Apoptosis was assessed by ssDNA immunoreactivity and TUNEL. Cryosections were incubated with ssDNA antibody (1:300; IBL Co.) overnight and visualized with indocarbocyanine (Cy3; Jackson ImmunoResearch) mounted with Vectashield containing DAPI (Vector Laboratories). TUNEL analysis was performed using cryosections with the ApopTag^® fluorescein in situ apoptosis detection kit (Chemicon) according the manufacturer's instruction. For each genotype, at least 1000 cells from four different sections and two mice were counted using a Zeiss axioskop with epiflourescence. Images from matched fields were captured in Adobe Photoshop and cell numbers were scored. The H2AX staining was performed on cryosections after antigen retrieval using polyclonal anti-H2AX (phospho S139; 1/250; Abcam).

Neurosphere cultures

Culturing of neurospheres was essentially as described (Frappart et al, 2005). Briefly, neurospheres were obtained from dissociated E14.5 brains and grown in Dulbecco's modified Eagle's media (NutMix/F12)/B27 media (Invitrogen) supplemented with EGF (Peprotech) and bFGF (Peprotech).

Spectral karyotyping

Medulloblastoma primary tumors were collected 4 h after an intraperitoneal injection of colcemid (1.5 g/g body weight) and single cell suspensions of medulloblastoma were subject to SKY analysis using a commercial SKY probe according to the SkyPaint hybridization and detection protocol (Applied Spectral Imaging). Pretreatment of samples used RNase A (100 g/ml) for 1 h at 37°C and pepsin for 2 min at 20°C (50 g/ml in 10 mM HCl), with counterstaining by 4',6-diamidino-2-phenylindole (DAPI).

Supplementary data

Supplementary data are available at The EMBO Journal Online (http://www.embojournal.org).

Acknowledgements

We thank the Hartwell Center and the Cancer Center Cytogenetics Core at SJCRH for their support of this work. These studies were supported by the NIH (NS-37956 and CA-21765) and the CCSG (P30 CA21765), and the American Lebanese and Syrian Associated Charities (ALSAC) of St Jude Children's Research Hospital.

References

Adam J, Deans B, Thacker J (2007) A role for Xrcc2 in the early stages of mouse development. DNA Repair (Amst) 6: 224–234 | PubMed | ChemPort |

Barnes DE, Stamp G, Rosewell I, Denzel A, Lindahl T (1998) Targeted disruption of the gene encoding DNA ligase IV leads to lethality in embryonic mice. Curr Biol 8: 1395–1398 | Article | PubMed | ISI | ChemPort |

Betz UA, Vosshenrich CA, Rajewsky K, Muller W (1996) Bypass of lethality with mosaic mice generated by Cre-loxP-mediated recombination. Curr Biol 6: 1307–1316 | Article | PubMed | ISI | ChemPort |

D'Andrea AD (2003) The Fanconi road to cancer. Genes Dev 17: 1933–1936 | Article |

Dahlstrand J, Lardelli M, Lendahl U (1995) Nestin mRNA expression correlates with the central nervous system progenitor cell state in many, but not all, regions of developing central nervous system. Brain Res Dev Brain Res 84: 109–129 | Article | PubMed | ChemPort |

Daniels MJ, Wang Y, Lee M, Venkitaraman AR (2004) Abnormal cytokinesis in cells deficient in the breast cancer susceptibility protein BRCA2. Science 306: 876–879 | Article | PubMed | ISI | ChemPort |

Deans B, Griffin CS, Maconochie M, Thacker J (2000) Xrcc2 is required for genetic stability, embryonic neurogenesis and viability in mice. EMBO J 19: 6675–6685 | Article | PubMed | ISI | ChemPort |

Digweed M, Sperling K (2004) Nijmegen breakage syndrome: clinical manifestation of defective response to DNA double-strand breaks. DNA Repair (Amst) 3: 1207–1217 | PubMed | ChemPort |

Erkko H, Xia B, Nikkila J, Schleutker J, Syrjakoski K, Mannermaa A, Kallioniemi A, Pylkas K, Karppinen SM, Rapakko K, Miron A, Sheng Q, Li G, Mattila H, Bell DW, Haber DA, Grip M, Reiman M, Jukkola-Vuorinen A, Mustonen A, Kere J, Aaltonen LA, Kosma VM, Kataja V, Soini Y, Drapkin RI, Livingston DM, Winqvist R (2007) A recurrent mutation in PALB2 in Finnish cancer families. Nature 446: 316–319 | Article | PubMed | ChemPort |

Fernandez-Capetillo O, Lee A, Nussenzweig M, Nussenzweig A (2004) H2AX: the histone guardian of the genome. DNA Repair (Amst) 3: 959–967 | Article | PubMed | ChemPort |

Forni PE, Scuoppo C, Imayoshi I, Taulli R, Dastru W, Sala V, Betz UA, Muzzi P, Martinuzzi D, Vercelli AE, Kageyama R, Ponzetto C (2006) High levels of Cre expression in neuronal progenitors cause defects in brain development leading to microencephaly and hydrocephaly. J Neurosci 26: 9593–9602 | Article | PubMed | ChemPort |

Frank AJ, Hernan R, Hollander A, Lindsey JC, Lusher ME, Fuller CE, Clifford SC, Gilbertson RJ (2004) The TP53-ARF tumor suppressor pathway is frequently disrupted in large/cell anaplastic medulloblastoma. Brain Res Mol Brain Res 121: 137–140 | PubMed | ChemPort |

Frank KM, Sekiguchi JM, Seidl KJ, Swat W, Rathbun GA, Cheng HL, Davidson L, Kangaloo L, Alt FW (1998) Late embryonic lethality and impaired V(D)J recombination in mice lacking DNA ligase IV. Nature 396: 173–177 | Article | PubMed | ISI | ChemPort |

Frankfurt OS, Robb JA, Sugarbaker EV, Villa L (1996) Monoclonal antibody to single-stranded DNA is a specific and sensitive cellular marker of apoptosis. Exp Cell Res 226: 387–397 | Article | PubMed | ISI | ChemPort |

Frappart PO, Tong WM, Demuth I, Radovanovic I, Herceg Z, Aguzzi A, Digweed M, Wang ZQ (2005) An essential function for NBS1 in the prevention of ataxia and cerebellar defects. Nat Med 11: 538–544 | Article | PubMed | ISI | ChemPort |

Friedberg EC, Meira LB (2006) Database of mouse strains carrying targeted mutations in genes affecting biological responses to DNA damage version 7. DNA Repair (Amst) 5: 189–209 | PubMed | ChemPort |

Gao Y, Ferguson DO, Xie W, Manis JP, Sekiguchi J, Frank KM, Chaudhuri J, Horner J, DePinho RA, Alt FW (2000) Interplay of p53 and DNA-repair protein XRCC4 in tumorigenesis, genomic stability and development. Nature 404: 897–900 | Article | PubMed | ISI | ChemPort |

Gao Y, Sun Y, Frank KM, Dikkes P, Fujiwara Y, Seidl KJ, Sekiguchi JM, Rathbun GA, Swat W, Wang J, Bronson RT, Malynn BA, Bryans M, Zhu C, Chaudhuri J, Davidson L, Ferrini R, Stamato T, Orkin SH, Greenberg ME, Alt FW (1998) A critical role for DNA end-joining proteins in both lymphogenesis and neurogenesis. Cell 95: 891–902 | Article | PubMed | ISI | ChemPort |

Gennery AR, Slatter MA, Bhattacharya A, Barge D, Haigh S, O'Driscoll M, Coleman R, Abinun M, Flood TJ, Cant AJ, Jeggo PA (2004) The clinical and biological overlap between Nijmegen Breakage syndrome and Fanconi Anemia. Clin Immunol 113: 214–219 | Article | PubMed | ChemPort |

Giordana MT, Duo D, Gasverde S, Trevisan E, Boghi A, Morra I, Pradotto L, Mauro A, Chio A (2002) MDM2 overexpression is associated with short survival in adults with medulloblastoma. Neuro-oncol 4: 115–122 | Article | PubMed | ChemPort |

Graus-Porta D, Blaess S, Senften M, Littlewood-Evans A, Damsky C, Huang Z, Orban P, Klein R, Schittny JC, Muller U (2001) Beta1-class integrins regulate the development of laminae and folia in the cerebral and cerebellar cortex. Neuron 31: 367–379 | Article | PubMed | ISI | ChemPort |

Herzog KH, Chong MJ, Kapsetaki M, Morgan JI, McKinnon PJ (1998) Requirement for Atm in ionizing radiation-induced cell death in the developing central nervous system. Science 280: 1089–1091 | Article | PubMed | ISI | ChemPort |

Holcomb VB, Vogel H, Marple T, Kornegay RW, Hasty P (2006) Ku80 and p53 suppress medulloblastoma that arise independent of Rag-1-induced DSBs. Oncogene 25: 7159–7165 | Article | PubMed | ChemPort |

Hughes-Davies L, Huntsman D, Ruas M, Fuks F, Bye J, Chin SF, Milner J, Brown LA, Hsu F, Gilks B, Nielsen T, Schulzer M, Chia S, Ragaz J, Cahn A, Linger L, Ozdag H, Cattaneo E, Jordanova ES, Schuuring E, Yu DS, Venkitaraman A, Ponder B, Doherty A, Aparicio S, Bentley D, Theillet C, Ponting CP, Caldas C, Kouzarides T (2003) EMSY links the BRCA2 pathway to sporadic breast and ovarian cancer. Cell 115: 523–535 | Article | PubMed | ISI | ChemPort |

Jonkers J, Meuwissen R, van der Gulden H, Peterse H, van der Valk M, Berns A (2001) Synergistic tumor suppressor activity of BRCA2 and p53 in a conditional mouse model for breast cancer. Nat Genet 29: 418–425 | Article | PubMed | ISI | ChemPort |

Kastan MB, Bartek J (2004) Cell-cycle checkpoints and cancer. Nature 432: 316–323 | Article | PubMed | ISI | ChemPort |

Kawarada Y, Miura N, Sugiyama T (1998) Antibody against single-stranded DNA useful for detecting apoptotic cells recognizes hexadeoxynucleotides with various base sequences. J Biochem (Tokyo) 123: 492–498 | PubMed | ChemPort |

Kennedy RD, D'Andrea AD (2005) The Fanconi Anemia/BRCA pathway: new faces in the crowd. Genes Dev 19: 2925–2940 | Article | PubMed | ISI | ChemPort |

Kruman II, Wersto RP, Cardozo-Pelaez F, Smilenov L, Chan SL, Chrest FJ, Emokpae Jr R, Gorospe M, Mattson MP (2004) Cell cycle activation linked to neuronal cell death initiated by DNA damage. Neuron 41: 549–561 | Article | PubMed | ISI | ChemPort |

Lee Y, Barnes DE, Lindahl T, McKinnon PJ (2000) Defective neurogenesis resulting from DNA ligase IV deficiency requires Atm. Genes Dev 14: 2576–2580 | Article | PubMed | ISI | ChemPort |

Lee Y, Chong MJ, McKinnon PJ (2001) Ataxia telangiectasia mutated-dependent apoptosis after genotoxic stress in the developing nervous system is determined by cellular differentiation status. J Neurosci 21: 6687–6693 | PubMed | ISI | ChemPort |

Lee Y, McKinnon PJ (2002) DNA ligase IV suppresses medulloblastoma formation. Cancer Res 62: 6395–6399 | PubMed | ISI | ChemPort |

Lee Y, McKinnon PJ (2006) Responding to DNA double strand breaks in the nervous system. Neuroscience doi: 10.1016/j.neurosciece.2006.07.026

Lee Y, Miller HL, Jensen P, Hernan R, Connelly M, Wetmore C, Zindy F, Roussel MF, Curran T, Gilbertson RJ, McKinnon PJ (2003) A molecular fingerprint for medulloblastoma. Cancer Res 63: 5428–5437 | PubMed | ISI | ChemPort |

Lees-Miller SP, Meek K (2003) Repair of DNA double strand breaks by non-homologous end joining. Biochimie 85: 1161–1173 | Article | PubMed | ISI | ChemPort |

Lieber MR, Ma Y, Pannicke U, Schwarz K (2003) Mechanism and regulation of human non-homologous DNA end-joining. Nat Rev Mol Cell Biol 4: 712–720 | Article | PubMed | ISI | ChemPort |

Lim DS, Hasty P (1996) A mutation in mouse rad51 results in an early embryonic lethal that is suppressed by a mutation in p53. Mol Cell Biol 16: 7133–7143 | PubMed | ISI | ChemPort |

Lomonosov M, Anand S, Sangrithi M, Davies R, Venkitaraman AR (2003) Stabilization of stalled DNA replication forks by the BRCA2 breast cancer susceptibility protein. Genes Dev 17: 3017–3022 | Article | PubMed | ISI | ChemPort |

Ludwig T, Chapman DL, Papaioannou VE, Efstratiadis A (1997) Targeted mutations of breast cancer susceptibility gene homologs in mice: lethal phenotypes of Brca1, Brca2, Brca1/Brca2, Brca1/p53, and Brca2/p53 nullizygous embryos. Genes Dev 11: 1226–1241 | PubMed | ISI | ChemPort |

Marmorstein LY, Kinev AV, Chan GK, Bochar DA, Beniya H, Epstein JA, Yen TJ, Shiekhattar R (2001) A human BRCA2 complex containing a structural DNA binding component influences cell cycle progression. Cell 104: 247–257 | Article |