Search

Pediatric high-grade glioma (pHGG) is a major contributor to cancer-related death in children. In vitro and in vivo disease models reflecting the intimate connection between developmental context and pathogenesis of pHGG are essential to advance understanding and identify therapeutic vulnerabilities. Here we report establishment of 21 patient-derived pHGG orthotopic xenograft (PDOX) models and eight matched cell lines from diverse groups of pHGG. These models recapitulate histopathology, DNA methylation signatures, mutations and gene expression patterns of the patient tumors from which they were derived, and include rare subgroups not well-represented by existing models. We deploy 16 new and existing cell lines for high-throughput screening (HTS). In vitro HTS results predict variable in vivo response to PI3K/mTOR and MEK pathway inhibitors. These unique new models and an online interactive data portal for exploration of associated detailed molecular characterization and HTS chemical sensitivity data provide a rich resource for pediatric brain tumor research., Patient-derived xenografts provide a resource for basic and translational cancer research. Here, the authors generate multiple pediatric high-grade glioma xenografts, use omics technologies to show that they are representative of primary tumours and use them to assess therapeutic response.

BackgroundMalignant peripheral nerve sheath tumors (MPNST) are aggressive sarcomas. Somatic inactivation of NF1 and cooperating tumor suppressors, including CDKN2A/B, PRC2, and p53, is found in most MPNST. Inactivation of the LATS1/2 kinases of the Hippo pathway was recently shown to cause tumors resembling MPNST histologically, although Hippo pathway mutations are rarely found in MPNST. Because existing genetically engineered mouse (GEM) models of MPNST do not recapitulate some of the key genetic features of human MPNST, we aimed to establish a mouse MPNST model that recapitulated the human disease genetically, histologically, and molecularly.MethodsWe combined two genetically modified alleles, an Nf1;Trp53 cis-conditional allele and an inducible Plp-CreER allele (NP-Plp), to model the somatic, possibly postnatal, mutational events in human MPNST. We also generated conditional Lats1;Lats2 knockout mice. We performed histopathologic analysis of mouse MPNST models and transcriptomic comparison of mouse models and human nerve sheath tumors.ResultsPostnatal Nf1;Trp53 cis-deletion resulted in GEM-MPNST that was histologically more similar to human MPNST than the widely used germline Nf1;Trp53 cis-heterozygous (NPcis) model and showed partial loss of H3K27me3. At the transcriptome level, Nf1;p53-driven GEM-MPNST were distinct from Lats-driven GEM-MPNST and resembled human MPNST more closely than do Lats-driven tumors.ConclusionsThe NP-Plp model recapitulates human MPNST genetically, histologically, and molecularly.Key PointsPostnatal Nf1;p53 cis-deletion in NP-Plp mice results in tumors similar to MPNST.The transcriptomes of Nf1;p53-driven and Lats-driven MPNST models are distinct.NP-Plp model resembles human MPNST genetically, histologically, and molecularly.Importance of the StudyMalignant peripheral nerve sheath tumors (MPNST) are aggressive sarcomas with a poor prognosis and limited treatment options. Existing genetically engineered mouse (GEM) models of MPNST do not recapitulate some of the key genetic features of human MPNST. To model the somatic, possibly postnatal, mutational events seen in MPNST patients, we generated a GEM-MPNST model by combining two genetically modified alleles, an Nf1;Trp53 cis-conditional allele and a Plp-CreER allele. Our histologic and transcriptomic analyses showed that this NP-Plp model resembles human MPNST genetically, histologically, and molecularly—more so than the widely used NPcis model and the recently published Lats-driven model. The NP-Plp model is genetically simple, making it easy to maintain and an ideal platform for preclinical studies. Given its tamoxifen-inducible nature, this model can be used to study the time/stage dependency of the tumorigenic potential of Schwann cells.

Pediatric high-grade glioma (pHGG) is a major contributor to cancer-related death in children.In vitroandin vivodisease models reflecting the intimate connection between developmental context and pathogenesis of pHGG are essential to advance understanding and identify therapeutic vulnerabilities. We established 21 patient-derived pHGG orthotopic xenograft (PDOX) models and eight matched cell lines from diverse groups of pHGG. These models recapitulated histopathology, DNA methylation signatures, mutations and gene expression patterns of the patient tumors from which they were derived, and included rare subgroups not well-represented by existing models. We deployed 16 new and existing cell lines for high-throughput screening (HTS).In vitroHTS results predicted variablein vivoresponse to inhibitors of PI3K/mTOR and MEK signaling pathways. These unique new models and an online interactive data portal to enable exploration of associated detailed molecular characterization and HTS chemical sensitivity data provide a rich resource for pediatric brain tumor research.

Cancer genomics has illuminated a wide spectrum of genes and core molecular processes contributing to human malignancy. Still, the genetic and molecular basis of many cancers remains only partially explained. Genetic predisposition accounts for 5-10% of cancer diagnoses(1,2) and genetic events cooperating with known somatic driver events are poorly understood. Analyzing established cancer predisposition genes in medulloblastoma (MB), a malignant childhood brain tumor, we recently identified pathogenic germline variants that account for 5% of all MB patients(3). Here, by extending our previous analysis to include all protein-coding genes, we discovered and replicated rare germline loss-of-function (LoF) variants across Elongator Complex Protein 1 (ELP1) on 9q31.3 in 15% of pediatric MB(SHH) cases, thus implicating ELP1 as the most common MB predisposition gene and increasing genetic predisposition to 40% for pediatric MB(SHH). Inheritance was verified based on parent-offspring and pedigree analysis, which identified two families with a history of pediatric MB. ELP1-associated MBs were restricted to the molecular SHHα subtype(4) and were characterized by universal biallelic inactivation of ELP1 due to somatic loss of chromosome 9q. The majority of ELP1-associated MBs exhibited co-occurring somatic PTCH1 (9q22.32) alterations, suggesting that ELP1-deficiency predisposes to tumor development in combination with constitutive activation of SHH signaling. ELP1 is an essential subunit of the evolutionary conserved Elongator complex, whose primary function is to enable efficient translational elongation through tRNA modifications at the wobble (U(34)) position(5,6). Biochemical, transcriptional, and proteomic analyses revealed that ELP1-associated MB(SHH) are characterized by a destabilized core Elongator complex, loss of Elongator-dependent tRNA wobble modifications, codon-dependent translational reprogramming, and induction of the unfolded protein response (UPR), consistent with deregulation of protein homeostasis due to Elongator-deficiency in model systems(7–9). Our findings suggest that genetic predisposition to proteome instability is a previously underappreciated determinant in the pathogenesis of pediatric brain cancer. These results provide a strong rationale for further investigating the role of protein homeostasis in other cancer types and potential opportunities for novel therapeutic interference.

To determine the role of microRNA regulation in brain tumor development, we incorporated a conditional allele of the microRNA processing enzyme Dicer to a previously characterized glioma mouse model based on inactivation of the tumor suppressors Nf1, Trp53, and Pten using the Nestin-creERT2 transgene. Loss of Dicer and tumor suppressors at adult ages led to glioma development; however, mutant mice tamoxifen induced at early postnatal ages developed medulloblastoma instead of glioma. The switch in tumor spectrum occurred with 100% penetrance and tumors were histologically indistinguishable from human medulloblastoma (MB). The minimum genetic mutations required for MB formation were Dicer and Trp53. Nf1 was dispensable, while additional loss of Pten produced more invasive tumors and leptomeningeal metastases. The time window for initiation of tumorigenesis was until the 2nd postnatal week, coinciding with the disappearance of the external granule layer (EGL), where cerebellar granule neuron precursors (CGNPs) undergo proliferation. Analysis of pre-symptomatic mutant mice showed proliferative defects and retained cells in the EGL, suggesting that the tumors may arise from CGNPs. However, targeting a subset of CGNPs using Math1-creERT2 did not lead to MB development, suggesting that an earlier EGL precursor may be required for tumorigenesis. Analysis of tumor transcriptome and MB subtype-specific genes and markers show that Dicer tumors most resemble extremely high risk p53-mutated SHH MB. Small RNA and mRNA sequencing analyses showed downregulation of microRNAs and dysregulation of its targets such as N-Myc. These studies demonstrate a role for microRNAs in MB development and show a fully penetrant genetic mouse model of highly metastatic MB.

Understanding the interplay between normal development and tumorigenesis, including the identification and characterization of lineage-specific origins of MB, is a fundamental challenge in the field. Recent studies have highlighted novel associations between biologically distinct MB subgroups and diverse murine cerebellar lineages via cross-species single-cell transcriptomics. Specifically, Group 4-MB correlated with the unipolar brush cell lineage and Group 3-MB resembled Nestin+ stem cells of the early cerebellum. However, these analyses were hampered by low resolution due to the sparsity of pertinent cerebellar cell types and the cross-species nature of the approach. Herein, we profoundly expand the depth of these rare developmental populations in the murine cerebellum using a combination of lineage tracing and integrative multi-omics. Isolation and enrichment of spatially and temporally unique developmental trajectories of key rhombic lip-derived glutamatergic lineages provided an enhanced reference for mapping MB subgroups based on molecular overlap, especially for poorly defined Group 3- and Group 4-MB. Further comparisons to a novel single-cell atlas of the human fetal cerebellum, companioned with laser-capture microdissected transcriptional and epigenetic datasets, reinforced developmental insights extracted from the mouse. Characterization of compartment-specific transcriptional programs and co-expression networks identified in the human upper rhombic lip implicated convergent cellular correlates of Group 3- and Group 4-MB, suggestive of a common developmental link. Together, our results strongly implicate developmental lineages of the upper rhombic lip as the probable origins of poorly defined Group 3- and Group 4-MB. These important findings will shape future efforts to accurately model the biological heterogeneity underlying these subgroups and provide unprecedented opportunities to explore their cellular and mechanistic basis.

Medulloblastoma (MB), a common malignant pediatric brain tumor, comprises at least four distinct molecular entities: WNT, SHH, Group 3, and Group 4. SHH-MB is driven by aberrant activation of the Sonic hedgehog (SHH) pathway in granule neuron progenitors (GNPs) and is associated with hereditary cancer predisposition syndromes including Li Fraumeni and Gorlin. We recently identified germline loss of function (LoF) mutations affecting IKBKAP/ELP1, the primary scaffolding subunit of the Elongator complex in a subset of SHH-MB patients. Germline ELP1 mutations account for ~15% of all pediatric SHH-MBs and position ELP1 as the most prevalent hereditary predisposition gene in MB. We genetically engineered Elp1 LoF in mouse GNPs to determine Elp1 function in cerebellar development and SHH-MB. Results of both mechanistic and phenotypic experiments demonstrate that GNPs harboring Elp1 loss exhibit ribosome pausing and protein aggregation, reinforcing the critical role of Elp1 in translational elongation and protein homeostasis. Further, we generated new transgenic mouse models mimicking germline ELP1 LoF mutations observed in SHH-MB patients. Elp1+/- transgenic mice exhibit purkinje cell degeneration and an increased DNA damage response. These mice are currently being evaluated for their capacity to recapitulate ELP1-associated SHH-MB. Additional analyses carried out on SHH-MB patient-derived xenografts showed that ELP1-mutant tumor cells specifically exhibit defects in tRNA biogenesis. Therefore, the function of ELP1 as a translational regulator is severely impaired in ELP1-mutant SHH-MBs. Our ongoing molecular and functional studies will provide important insights into the most common MB predisposition gene and will lay the foundation for future preclinical studies.

BACKGROUND Our previous analysis of established cancer predisposition genes in medulloblastoma (MB) identified pathogenic germline variants in ~5% of all patients. Here, we extended our analysis to include all protein-coding genes. METHODS Case-control analysis performed on 795 MB patients against >118,000 cancer-free children and adults was performed to identify an association between rare germline variants and MB. RESULTS Germline loss-of-function variants of Elongator Complex Protein 1 (ELP1; 9q31.3) were strongly associated with SHH subgroup (MBSHH). ELP1-associated-MBs accounted for ~15% (29/202) of pediatric MBSHH cases and were restricted to the SHHα subtype. ELP1-associated-MBs demonstrated biallelic inactivation of ELP1 due to somatic chromosome 9q loss and most tumors exhibited co-occurring somatic PTCH1 (9q22.32) alterations. Inheritance was verified by parent-offspring sequencing (n=3) and pedigree analysis identified two families with a history of pediatric MB. ELP1-associated-MBSHH were characterized by desmoplastic/nodular histology (76%; 13/17) and demonstrated a favorable clinical outcome when compared to TP53-associated-MBSHH (5-yr OS 92% vs 20%; p-value=1.3e-6) despite both belonging to the SHHα subtype. ELP1 is a subunit of the Elongator complex, that promotes efficient translational elongation through tRNA modifications at the wobble (U34) position. Biochemical, transcriptional, and proteomic analyses revealed ELP1-associated-MBs exhibit destabilization of the core Elongator complex, loss of tRNA wobble modifications, codon-dependent translational reprogramming, and induction of the unfolded protein response. CONCLUSIONS We identified ELP1 as the most common MB predisposition gene, increasing the total genetic predisposition for pediatric MBSHH to 40%. These results mark MBSHH as an overwhelmingly genetically-predisposed disease and implicate disruption of protein homeostasis in MBSHH development.

Medulloblastoma is a malignant childhood brain tumor arising from the developing cerebellum. In Sonic Hedgehog (SHH) subgroup medulloblastoma, aberrant activation of SHH signaling causes increased proliferation of granule neuron progenitors (GNPs), and predisposes these cells to tumorigenesis. A second, cooperating genetic hit is often required to push these hyperplastic cells to malignancy and confer mutation-specific characteristics associated with oncogenic signaling. Somatic loss-of-function mutations of the transcriptional corepressor BCOR are recurrent and enriched in SHH medulloblastoma. To investigate BCOR as a putative tumor suppressor, we used a genetically engineered mouse model to delete exons 9/10 of Bcor (BcorΔE9–10) in GNPs during development. This mutation leads to reduced expression of C-terminally truncated BCOR (BCORΔE9–10). While BcorΔE9–10 alone did not promote tumorigenesis or affect GNP differentiation, BcorΔE9–10 combined with loss of the SHH receptor gene Ptch1 resulted in fully penetrant medulloblastomas. In Ptch1+/−;BcorΔE9–10 tumors, the growth factor gene Igf2 was aberrantly up-regulated, and ectopic Igf2 overexpression was sufficient to drive tumorigenesis in Ptch1+/− GNPs. BCOR directly regulates Igf2, likely through the PRC1.1 complex; the repressive histone mark H2AK119Ub is decreased at the Igf2 promoter in Ptch1+/−;BcorΔE9–10 tumors. Overall, our data suggests that BCOR–PRC1.1 disruption leads to Igf2 overexpression, which transforms preneoplastic cells to malignant tumors.

Pediatric brain tumors comprise a distinct spectrum of diseases compared to adult brain tumors and are distinguished by their unique clinical and histopathological features, developmental context, mutation burden, and genomic, epigenomic, and transcriptomic alterations. Access to in vivo models that recapitulate pediatric brain tumors has been limited and inadequate to represent these heterogeneous diseases. Here we introduce the Pediatric Brain Tumor Portal (PBTP, pbtp.stjude.cloud), an open resource to access molecular characterization, including whole-genome sequencing, whole-exome sequencing, RNA-seq, and DNA methylome profiling, of patient-derived orthotopic xenograft (PDOX) models of pediatric brain tumors. The portal will host more than 70 models, which currently include pediatric High-Grade Glioma (pHGG), Medulloblastoma, Atypical Teratoid/Rhabdoid Tumors (AT/RT), Ependymoma, and Embryonal Tumors with Multilayered Rosettes (ETMR), and reflects close to ten years of effort to generate and extensively characterize in vivo models that recapitulate primary pediatric brain tumors. PBTP is a database-driven and user-friendly platform that enables multi-omics views of PDOX and matched patient tumor and germline samples at multiple levels to help identify appropriate lines for studies of tumorigenesis or preclinical testing. To evaluate the representation of different molecular features within a disease subgroup, we implemented various visualization tools for searching and comparing somatic mutations, gene expression profiles, and methylation groups of our models and matched patient tumors. Details for each tumor including de-identified clinical data, histology, growth characteristics, DNA methylation classification, mutation status, and copy number variations are also readily available for side-by-side comparison. Furthermore, in vitro chemical sensitivity profiling is presented for selected models with matched cell lines. Users can investigate genes of interest for single nucleotide variants, small insertions and deletions, copy number variations, gene fusions, and mRNA expression in PDOX and matched patient tumors. This portal has been integrated into the St. Jude Cloud platform, through which users can explore patient and PDOX sequencing data in the context of larger cohort data sets and download raw genomics data files. PBTP provides a platform to share PDOX models with in-depth genome/epigenome-wide characterization to support advances in basic and translational research in pediatric brain tumors. *co-first, #co-corresponding Citation Format: Paige S. Dunphy, Ke Xu, Darrell T. Gentry, Chen He, Kimberly Mercer, Xiaoyan Zhu, Kyle Smith, Brian Gudenas, Sarah Robinson, Junyuan Zhang, Lawryn H. Kasper, Chang-Hyuk Kwon, Laura D. Hover, Jon D. Larson, Nathaniel Twarog, Aksana Vasilyeva, Nedra Robison, Daniel Alford, Cynthia Williams, Anthony Woodard, Xin Zhou, Edgar Sioson, J. Robert Michael, Austyn Trull, Irina McGuire, Brandon McMahan, Swapnali Mohite, Ashok Boddu, Kirby Birch, Clay McLeod, Michael A. Dyer, Paul Klimo, Frederick A. Boop, Amar Gajjar, Christopher L. Tinkle, Giles Robinson, Brent A. Orr, Jason Chiang, Paul A. Northcott, Jinghui Zhang, Keith Perry, Gang Wu, Anang A. Shelat, Ed Suh, Martine F. Roussel, Suzanne J. Baker. St. Jude Pediatric Brain Tumor Portal: Cloud-based resource for patient-derived orthotopic xenograft (PDOX) models of pediatric high-grade glioma, ependymoma, and CNS embryonal tumors [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 6144.

Summary: Loss of heterozygosity (LOH) at 1p36 occurs in multiple cancers, including neuroblastoma (NBL). MYCN amplification and 1p36 deletions tightly correlate with markers of tumor aggressiveness in NBL. Although distal 1p36 losses associate with single-copy MYCN tumors, larger deletions correlate with MYCN amplification, indicating two tumor suppressor regions in 1p36, only one of which facilitates MYCN oncogenesis. To better define this region, we genome-edited the syntenic 1p36 locus in primary mouse neural crest cells (NCCs), a putative NBL cell of origin. In in vitro cell transformation assays, we show that Chd5 loss confers most of the MYCN-independent tumor suppressor effects of 1p36 LOH. In contrast, MYCN-driven tumorigenesis selects for NCCs with Arid1a deletions from a pool of NCCs with randomly sized 1p36 deletions, establishing Arid1a as the MYCN-associated tumor suppressor. Our findings reveal that Arid1a loss collaborates with oncogenic MYCN and better define the tumor suppressor functions of 1p36 LOH in NBL. : Garcia-Lopez et al. present a mouse model of high-risk neuroblastoma that includes 1p36 loss and Mycn overexpression. This study substantiates previous predictions from NBL genetic studies, which proposed that two tumor suppressor regions exist in 1p36. It further demonstrates that Mycn overexpression selects for loss of Arid1a during tumorigenesis. Keywords: chromatin remodelers, 1p36 LOH, ARID1A, CHD5, high-risk neuroblastoma

Genetic studies have identified many risk loci for autism spectrum disorder (ASD) although causal factors in the majority of cases are still unknown. Currently, known ASD risk genes are all protein-coding genes; however, the vast majority of transcripts in humans are non-coding RNAs (ncRNAs) which do not encode proteins. Recently, long non-coding RNAs (lncRNAs) were shown to be highly expressed in the human brain and crucial for normal brain development. We have constructed a computational pipeline for the integration of various genomic datasets to identify lncRNAs associated with ASD. This pipeline utilizes differential gene expression patterns in affected tissues in conjunction with gene co-expression networks in tissue-matched non-affected samples. We analyzed RNA-seq data from the cortical brain tissues from ASD cases and controls to identify lncRNAs differentially expressed in ASD. We derived a gene co-expression network from an independent human brain developmental transcriptome and detected a convergence of the differentially expressed lncRNAs and known ASD risk genes into specific co-expression modules. Co-expression network analysis facilitates the discovery of associations between previously uncharacterized lncRNAs with known ASD risk genes, affected molecular pathways and at-risk developmental time points. In addition, we show that some of these lncRNAs have a high degree of overlap with major CNVs detected in ASD genetic studies. By utilizing this integrative approach comprised of differential expression analysis in affected tissues and connectivity metrics from a developmental co-expression network, we have prioritized a set of candidate ASD-associated lncRNAs. The identification of lncRNAs as novel ASD susceptibility genes could help explain the genetic pathogenesis of ASD.