Genetic Inactivation of Notch1 Synergizes with Loss of Trp53 to Induce Tumor Formation in the Adult Mouse Forebrain

Simple Summary Notch signaling plays a context-dependent role in multiple cancer types by either promoting or suppressing tumor development. The role of the Notch receptors in the formation of brain tumors remains controversial. By exploiting conditional genetics and lineage tracing approaches to study unperturbed solid tumor growth in vivo, we uncover a tumor suppressor function for the Notch1 receptor in the forebrain and show that p53 and Notch1 cooperate to inhibit tumor formation. Abstract Simultaneous genetic inactivation of the key Notch signaling mediator RBP-Jk and p53 leads to the formation of forebrain tumors in mice, suggesting a tumor suppressor role of the Notch pathway in this context. However, the contribution of individual Notch receptors to the tumor-suppressive activity of Notch signaling in the brain remains elusive. Here, we show that simultaneous Notch1 and Notch2 deletion, similar to complete ablation of canonical Notch signaling by Rbpj inactivation, cooperates with Trp53 deletion to promote tumor growth in the adult forebrain. We also demonstrate that inactivation of Notch1 and Trp53 in cells with active Notch signaling is sufficient to induce brain tumor or hyperplasia formation. Analysis of tumor location suggests a multifocal origin and shows that ventral forebrain regions and olfactory bulbs are the most affected sites. Hence, Notch1 cooperates with p53 to repress malignant transformation in the adult mouse forebrain.


Introduction
The Notch signaling pathway plays pleiotropic roles during tissue development and tumor formation [1]. Depending on the cellular context and tissue type, Notch signaling can function either as an oncogene or a tumor suppressor [1,2]. This dual function of Notch in the regulation of tumorigenesis has, in some instances, been observed even within the same organ, such as in the lung or the hematopoietic system [3][4][5].
Opposing oncogenic and tumor-suppressive roles of Notch have also been described for the central nervous system (CNS) in mammals [4,6]. Notch signaling is fundamental for the maintenance of neural stem cell (NSC) identity and for the regulation of cell fate decisions in the CNS [7,8]. The Notch1 and Notch2 receptors, the indispensable mediator of canonical Notch signaling RBP-Jk, and the Notch downstream targets of the Hes/Hey gene family are all essential in NSCs [9][10][11][12][13][14][15][16][17][18][19]. Despite the stem cell-promoting activity of Notch, which certainly plays roles in Notch oncogenic function in various forms of primary CNS cancers, Notch signaling can also have a tumor-suppressive function linked to the regulation of mitotic quiescence and immune evasion in the brain [4,6,20]. This dual role of Notch is likely related to the vast genomic and epigenomic heterogeneity of CNS tumors [21][22][23][24][25]. Decoding the context dependent functions of Notch signaling is fundamental in order to develop more effective strategies to treat primary brain malignancies.
Reduced Notch signaling activity associated with increased aggressiveness and worse prognosis has been observed in subtypes of glioma and of CNS primitive neuroectodermal tumors [21,23,[26][27][28][29][30][31]. Inactivating mutations in Notch pathway components, particularly in the NOTCH1 and NOTCH2 receptor and RBPJ genes, have been detected in isocitrate dehydrogenase (IDH) mutant gliomas [21,23,26,27,29,32]. In support of these observations in human brain tumor subtypes, simultaneous genetic inactivation of Notch1 and Notch2 or Rbpj accelerates the growth of PDGF-driven gliomas in mice [20,28]. TP53 mutations and elevated PDGF signaling are common in human proneural glioma, a transcriptional subtype where Notch signaling can act as a tumor suppressor [24,28]. Since deletion of Trp53 predisposes to hyperplasia formation but does not induce tumor development in the murine brain, it can be exploited in combination with inhibition of Notch signaling components to reveal their tumor suppressive activity [28]. Strikingly, genetic deletion of Rbpj together with Trp53 induces aggressive, de novo forebrain tumors even in the absence of PDGF signal activation [28]. However, it is not known whether inactivation of Notch1 and Notch2 with Trp53 can induce brain tumor formation. Moreover, the relative contribution of Notch1 and Notch2 receptors to the tumor-suppressive activity of Notch signaling in this context remains elusive. Here, by exploiting long-term genetic lineage tracing approaches to study unperturbed solid tumor growth in vivo, we show that the inactivation of Notch1 and Trp53 in Notch signaling-active cells is sufficient to induce tumor or hyperplasia formation in the ventral forebrain and olfactory bulbs (OBs).

Tamoxifen Administration
Stock solutions of Tamoxifen (TAM, Sigma-Aldrich, St Louis, MO, USA ) were prepared at a concentration of 20 mg/mL in corn oil (Sigma-Aldrich). Adult mice (8-10 weeks of age) were injected intraperitoneal with TAM once per day for 5 consecutive days at a dose of 2 mg/day. Animals were killed 10 months after the last injection, or when they developed symptoms of brain tumor formation (lethargy, poor grooming, weight loss, or macrocephaly), and the brains were prepared for immunohistochemistry as described below.

Tissue Preparation and Immunohistochemistry
For histology, mice were deeply anaesthetized by injection of a ketamine/xylazine/ acepromazine solution (130 mg, 26 mg and 4 mg/kg body weight, respectively) and perfused with ice-cold 0.9% saline solution followed by ice-cold 4% paraformaldehyde (PFA) solution in 0.1 M phosphate buffer (PB). Brains were post-fixed with 4% PFA overnight, washed in PB, cryoprotected in a 30% sucrose solution in 0.1 M PB for 48 h, embedded and frozen in OCT (TissueTEK). Free floating coronal sections (30 µm) were collected in multi-well dishes (Corning) and stored at −20 • C in anti-freeze solution until use.
For immunostaining, sections were incubated overnight at 4 • C with the primary antibody diluted in blocking solution of 2% normal donkey serum (Jackson ImmunoResearch), 0.5% Triton X-100 in phosphate-buffered saline (PBS). Sections were washed three times in PBS and incubated at room temperature for 1 h with the corresponding secondary antibodies in blocking solution. When necessary, sections were counter-stained with DAPI (1 µg/mL). For PCNA detection, antigen was recovered at 80 • C for 20 min in Sodium Citrate solution (10 mM, pH 6). Stained sections were mounted on Superfrost glass slides (Thermo Scientific, Waltham, MA, USA), embedded in mounting medium containing diazabicyclo-octane (DABCO, Sigma-Aldrich) as an anti-fading agent and visualized using a Zeiss Observer.Z1 equipped with Apotome. Tumors were defined as discernible masses (increased cellularity based on DAPI staining) of proliferating (PCNA + ) cells that were spanning at least 2 consecutive sections in a series of sections spaced 360 µm apart. Hyperplasias were defined as areas containing increased numbers of diffuse, highly proliferative PCNA + cells, with only a limited increase in cellularity (DAPI staining).

Deletion of Notch1 and Trp53 in Hes5 + Cells Leads to the Formation of Forebrain Tumors in Adult Mice
The relative contribution of Notch1 and Notch2 receptors to the tumor-suppressive activity of Notch signaling in this context is unknown. We addressed whether loss of p53 in combination with loss of either Notch1 or Notch2 is tumorigenic in the adult brain. We astrocytic (GFAP) and neuronal (DCX) markers, and were highly proliferative ( Figure 3E, Table 1). We concluded that Notch1 and Trp53 deletion in Hes5 + cells is sufficient to induce tumor or hyperplasia formation in the forebrains of adult mice.   Tumor n/a 89.6 ± 0.6 (n = 4) 85.4 ± 4.9 (n = 4) 76.7 ± 5.5 (n = 5) n/a

Hyperplasia and Tumors Are Preferentially Located in Ventral Forebrain Regions and OBs
Examination of the tumor border showed moderate invasion of Trp53 -/-Rbpj -/-, Trp53 -/-Notch1 -/-Notch2 -/-, and Trp53 -/-Notch1 -/tumor cells into the surrounding brain parenchyma (Figure 4). All tumors that we identified had a supratentorial location, in agreement with our previous results from Trp53 -/-Rbpj -/mutants [28]. However, the large size of Trp53 -/-Rbpj -/tumors often precluded a more precise characterization of their location. We took advantage of the relatively smaller size of Trp53 -/-Notch1 -/-Notch2 -/-, Trp53 -/-Notch1 -/-, and Trp53 -/-Notch2 -/tumors and hyperplasias compared to Trp53 -/-Rbpj -/tumors to better define the prevalent sites of tumor formation within the forebrain after genetic inhibition of Notch and p53 ( Figure 5, Table S1). We scored the number and location of tumors and hyperplasias in Trp53 -/-Notch1 -/-Notch2 -/-, Trp53 -/-Notch1 -/-, Trp53 -/-Notch2 -/-, and Trp53 -/mutants. Tumors and hyperplasias were predominantly located in ventral and anterior forebrain regions, including the OBs, the rostral migratory stream (RMS) and surrounding areas, and the subventricular zone (SVZ) and adjacent striatum ( Figure 5A-G, Table S1). The OB was the brain region where malignant cells were most frequently identified ( Figure 5A, Table S1). OB tumors and hyperplasias in Trp53 -/-Notch1 -/-Notch2 -/and Trp53 -/-Notch1 -/animals were either large and invaded all OB layers ( Figure 5D,F) or were superficial and preferentially occupied the glomerular and nerve layers ( Figure 5E). The OBs of most Trp53 -/-Notch2 -/animals did not show signs of hyperproliferation and looked normal ( Figure 5G). However, the rare, small hyperplasias that could be detected in Trp53 -/-Notch2 -/and Trp53 -/mutants were also preferentially located in the OBs (Table S1). In contrast to the anterior and ventral forebrain, tumor formation very rarely affected dorsal forebrain regions such as the cerebral cortex and the hippocampus ( Figure 5A, Table S1). Large tumors and hyperplasias were also relatively common in the amygdala region ( Figure 5H, Table S1). In addition, hyperproliferation was observed in the hypothalamus and thalamus of a proportion of the mice ( Figure 5I, Table S1). In some animals, multiple tumors and hyperplasias co-occurred at distant locations in the brain, for example OB and amygdala, OB and hypothalamus/thalamus, and OB and SVZ (Table S1).

Discussion
Our gene-ablation and long-term genetic lineage tracing experiments in mice indicate that Notch/RBP-Jk signaling, and particularly the Notch1 receptor, can inhibit brain tumor initiation and growth in the adult forebrain. Trp53 deletion alone in cells expressing the Notch target gene Hes5 predisposes to hyperplasia formation with very low penetrance. We have previously shown that deletion of Rbpj, Notch1, or Notch2 in the presence of an intact Trp53 gene alters NSC proliferation and differentiation, but does not induce formation of brain tumors [12,28]. In contrast, simultaneous deletion of Trp53 and Rbpj, Notch1/Notch2, or Notch1 induces tumor or hyperplasia formation in over 80% of the mice. Thus, Notch signaling cooperates with p53 to inhibit tumor formation in the adult murine brain. These data confirm and extend our previous findings from PDGF-driven gliomas and indicate that Notch signaling can behave as a tumor suppressor in the brain even in the absence of PDGF overexpression [28]. Although Trp53 -/-Rbpj -/tumors are clearly separable from PDGF + Trp53 -/-Rbpj -/tumors at the transcriptional level [28], it is conceivable that tumor cells in both models acquire oligodendrocyte progenitor cell (OPC)-like features, a cellular state that is associated with a proneural glioma transcriptional subtype and a tumor suppressor role of Notch signaling [20]. This scenario is in agreement with a previous study showing that the transcriptional profiles of murine glioma models driven by different oncogenic alterations can convergence to a proneural signature [38].
RBP-Jk can act as a transcriptional repressor and loss of the RBPJ gene promotes the growth of human breast cancer cells by de-repressing target gene promoters in a Notchindependent manner [39,40]. However, our data showing that not only Trp53 -/-Rbpj -/mutants, but also Trp53 -/-Notch1 -/-Notch2 -/and Trp53 -/-Notch1 -/mutants develop tumors suggest that the tumor suppressor function of RBP-Jk in the brain is linked to the regulation of Notch signaling. This hypothesis is supported by an in vivo CRISPR screen that identified Notch1 among potential tumor suppressors in glioma [41], and is further in line with the occurrence of NOTCH1, NOTCH2, and RBPJ inactivating mutations and/or reduced Notch signaling activity in human glioma subtypes [21,23,[26][27][28][29][30][31].
We found that Trp53 -/-Notch2 -/mutants rarely develop hyperplasia, in contrast to Trp53 -/-Notch1 -/and Trp53 -/-Rbpj -/mutants. We cannot exclude that differences in recombination efficiency of Notch1, Notch2, and Rbpj conditional alleles contribute to the different penetrance of the phenotypes. However, we have recently shown that glioma cells lacking Notch1 are more aggressive than glioma cells lacking Notch2, despite efficient deletion of both Notch1 and Notch2 floxed alleles [20]. Thus, our current finding confirms our observation that Notch1 is a stronger tumor suppressor than Notch2 in a PDGF-driven mouse model of glioma [20] and implies that Notch1 could partially compensate for Notch2 loss. Although Notch2 deletion does not accelerate hyperplasia formation in Trp53 -/mutants, it significantly accelerates the growth of PDGF + Trp53 -/gliomas [20], suggesting that a strong mitogenic stimulus, provided, for instance, by PDGF, is required to release the tumorigenic effects of Notch2 inhibition. Our data do not exclude the residual expression of Hes/Hey gene family members or other Notch downstream targets, particularly in cells deleted for one Notch receptor, as a result of compensatory mechanisms or direct induction of Hes genes expression by BMP signaling [42]. In agreement with this hypothesis, we have recently shown that Notch1 or Notch2 deletion in PDGF + Trp53 -/gliomas leads to changes in the expression of only partially overlapping sets of genes [20]. We observed tumor or hyperplasia formation in roughly 85% of Trp53 -/-Notch1 -/-Notch2 -/and Trp53 -/-Notch1 -/mutants, but complete penetrance in Trp53 -/-Rbpj -/mutants. This suggests that other receptors beyond Notch1 and Notch2 might be involved. It would be interesting to determine whether Notch3, which represses the proliferation of adult NSCs [43,44], contributes to the tumor suppressor activity of RBP-Jk in cooperation with Notch1 and Notch2.
We found that a proportion of the tumors and hyperplasias were located in the proximity of neurogenic regions including the RMS and SVZ and adjacent striatum, implying that adult neural stem and progenitor cells are probable cells of origin of some of the tumors [45]. However, a significant proportion of the tumors and hyperplasias were confined to anterior and ventral brain regions such as the surface of the OB, the amygdala, the hypothalamus and the thalamus, raising the possibility that some of the tumors originate locally. Parenchymal OPCs express Notch receptors [46][47][48] and can be an important cell type of origin of tumors, notably in the anterior and ventral brain [49][50][51][52]. Further work is needed to explore whether the resilience of ventral OPCs to pathological stimuli can make them exceptionally vulnerable to malignant transformation [49,53]. It is also possible that the glycosylation status of the Notch receptors, which can affect ligand binding, as well as varying expression levels of the Notch ligands contribute to regional heterogeneity in tumorigenesis [54,55]. In the future, it would be interesting to determine whether the frequency of tumor formation correlates with domain specific expression of Jagged and Delta ligands in different brain areas.
We show that genetic inactivation of Notch signaling and p53 in a subpopulation of brain cells induces CNS tumors with long latency but high penetrance. Notch-regulated competitive interactions, triggered by dishomogeneus Notch inactivating mutations in adjacent cells, contribute to the initiation of epithelial cancers by favoring the slow and progressive expansion of mutant clones at the expense of wild-type cells [56]. Intriguingly, data suggest that transformed NSCs can outcompete neighboring NSCs by repressing their proliferation through Notch signaling [57]. Whether Notch inhibition in malignant NSCs exacerbates cell competition and fitness selection is unknown. Likewise, OPC populations compete for space in the developing brain [58], but whether a similar phenomenon occurs during glioma formation from OPCs remains unclear. We speculate that Notch-regulated cell competition mechanisms could be in place during tumor initiation and progression in the brain.

Conclusions
In conclusion, we show that Notch1 deletion cooperates with Trp53 deletion to promote tumor growth in the adult forebrain. Our data also suggest that tumors induced by Notch signaling inhibition have a multifocal origin. These findings shed light on the pleiotropic roles of Notch signaling in the development of primary brain malignancies.

Informed Consent Statement: Not applicable.
Data Availability Statement: All data generated or analyzed during this study are included in this article and its Supplementary Material files. Further enquiries can be directed to the corresponding author.