Year: 2018
Type: article
Abstract: The SWI/SNF chromatin regulator, ATRX, is frequently mutated in IDH-MUT gliomas with co-occurrence of TP53 mutations. Although ATRX loss is considered an oncogenic event, the exact role of ATRX in glioma progression remains unclear. Previous studies involving ATRX have been mostly carried out in the context of an intact TP53 but not in cells lacking this gene. Furthermore, the role of ATRX has been predominantly examined in murine embryonic stem cells (mESCs)1 and not neural progenitor cells (NPCs), the putative cell of origin for gliomas. To address these deficiencies, in a recent study Danussi et al2 used mouse NPCs (mNPCs) derived from mice harboring either intact or deleted Atrx (Atrx+ or Atrx-) with inactivated Tp53 (Tp53-). Although there was no difference in proliferation rates, Atrx-/Tp53- mNPCs showed dramatic alteration in cellular morphology characterized by increased attachment and a monolayer growth pattern instead of neurospheres. Transcriptome profiling demonstrated that Atrx-/Tp53- mNPCs showed signatures resembling astrocytic differentiation rather than neuronal or oligodendrocytic lineages. It is noteworthy that these findings correlate with observations in human gliomas where ATRX and TP53 mutations associate with astrocytomas but not oligodendrogliomas. In addition, the authors found increased expression of genes involved in motility. This was corroborated in cell migration as well as Rho GTPAse activation assays. Whole-genome chromatin immunoprecipitation revealed differences in occupancy of Atrx between mESCs and mNPCs: Atrx positioning was significantly higher on gene bodies and promoter regions in mNPCs, suggestive of a gene-regulatory role of Atrx in these cells. Using ATAC-seq, the authors found that altered chromatin accessibility largely overlapped with the Atrx-bound regions. Furthermore, open and closed ATAC regions correlated with genes upregulated and downregulated by Atrx deficiency, respectively, suggesting that in the absence of Atrx, chromatin accessibility and gene transcription is altered in regions bound by Atrx. Employing a combination of bio-informatic and siRNA-based approaches, the authors identified Gna13 as a downstream target of Atrx that promotes migratory phenotypes in the Atrx-deficient cells. Importantly, given that ATRX/DAXX/histone H3.3 complexes regulate heterochromatin silencing, the authors found differential H3.3 composition in vacant ATRX binding sites proximal to Gna13 and Gfap, suggesting that Atrx may regulate these genes by causing a shift in H3.3 incorporation. This study not only reveals a cell type-specific role for Atrx in the nervous system but also demonstrates that Atrx contributes to oncogenesis by regulating differentiation and migration, important hallmarks of diffusely infiltrating astrocytomas. Immortalized cell division in cancer requires telomere maintenance to escape cell cycle arrest and senescence. In most neoplasia, this is accomplished by activation of telomerase, which adds repeats to the telomeres of each chromosome, maintaining their respective lengths. However, small but significant minorities of tumors, including large subgroups of diffuse glioma, employ alternative lengthening of telomeres (ALT), a telomerase-independent mechanism of telomere maintenance based on homologous recombination. ALT has been strongly associated with inactivating mutations in the chromatin regulator gene ATRX (α-thalassemia mental retardation X-linked). However, studies to this point have been unable to demonstrate that ATRX deficiency is itself sufficient to induce the ALT phenotype. In adults, the vast majority of ATRX-deficient gliomas also harbor mutations in isocitrate dehydrogenase 1 or 2 (IDH1 or IDH2). Mutant IDH proteins produce the oncometabolite 2-hydroxyglutarate, which broadly impacts cellular transcriptional profiles by inducing hypermethylation across the genome. In a recent Cancer Research paper, Mukherjee et al. investigated whether the combined effects of IDH-mutant transcriptional shifts and ATRX inactivation were sufficient to induce ALT in glioma.1 Using a p53/pRb-deficient astrocyte model system, they demonstrated that combined IDH mutation and ATRX inactivation, but neither alone, effectively induced the phenotypic hallmarks of ALT along with anchorage-independent growth in soft agar. They went on to identify RAP1 and XRCC1 and two telomere-associated genes consistently downregulated in IDH-mutant astrocytes and IDH-mutant, ATRX-deficient gliomas. Restored expression of RAP1 and/or XRCC1 abrogated ALT in IDH-mutant ATRX-deficient astrocytes as well as ATRX-mutant, ALT-positive human glioma cells. In contrast, knockdown of RAP1 and/or XRCC1 in IDH-wild-type, ATRX-deficient astrocytes promoted ALT. Combined IDH mutation and ATRX loss induced γ-H2AX-positive DNA damage at telomeres, which was abrogated by RAP1 overexpression. By contrast, RAP1 knockdown in IDH-wild-type, ATRX-deficient astrocytes increased γ-H2AX positivity at telomeres. These findings indicat that RAP1, a component of the shelterin complex, serves to maintain telomeric integrity, and its downregulation in the setting of IDH mutation promotes telomeric dysfunction. Finally, the authors overexpressed XRCC1, which encodes a critical component of the alternative non-homologous end-joining (aNHEJ) pathway, in IDH-mutant, ATRX-deficient astrocytes, reducing ALT and increasing telomere fusion with associated cell death. These results suggest that XRCC1 silencing suppresses telomere fusion in response to DNA damage, enabling homologous recombination and ALT. Taken together, these findings are highly significant in that they establish for the first time a viable mechanism for the induction of ALT in ATRX-deficient glioma. According to the current international consensus there are four main molecular medulloblastoma (MB) subgroups (WNT, SHH, and Groups 3 and 4) with divergent histology, biology, and clinical outcomes.1 Medulloblastoma with extensive nodularity (MBEN) is a rare histological variant that usually occurs in the first 3 years of life and is associated with good prognosis. Molecular analyses of MBEN, mostly limited to single cases or small series, have shown that they are always classified as sonic hedgehog (SHH)-driven MB. In a recent issue of Acta Neuropathologica, Korshunov and co-workers2 analyzed 25 MBEN cases through genome-wide DNA methylation, copy number profiling, and targeted next-generation sequencing. These analyses were then compared with molecular profiles of other SHH MB histological variants. The vast majority of MBEN (23/25) disclosed SHH-associated epigenetic signatures and mutational landscapes, but, surprisingly, two MBEN were classified as Group 3/4 MB. Most MBEN classified as SHH MB displayed SHH-related and mutually exclusive mutations in either SUFU, or PTCH1, or SMO at similar frequencies. However, only SUFU mutations were also identified in the germ line. Most of SUFU-associated MBEN eventually recurred, but patients were treated successfully with second-line high-dose chemotherapy. Altogether, these data show that risk stratification even for well-recognizable histologies such as MBEN cannot rely on histology alone but should include additional molecular analyses such as methylation profiling and DNA sequencing. In conclusion, for all patients with “MBEN” histology, sequencing SUFU and PTCH1 in the tumor as well as in the germ line for further clinical stratification and choice of the optimal treatment strategy should be recommended. Further clarification of this provisional stratification scheme through detailed analysis of additional MBEN cohorts included in future clinical trials would be of significant value. Recent immunotherapies for glioblastoma (GBM) have shown great promise to kill tumor cells with long-term protective effects. One therapy that has garnered much interest is the engineering of T lymphocytes to express chimeric antigen receptors (CAR-Ts) that recognize tumor-specific antigens and deliver a powerful cytotoxic response. A major limitation of CAR T efficacy against GBM is the molecular heterogeneity of these tumors, which allows the escape of malignant cells lacking or downregulating the target antigen. In a recent study, Pellegata and colleagues describe a novel target (CSPG4) that is widely expressed in GBMs and may serve to enhance current CAR-T therapies.1 CSPG4 (also known as neuron/glia antigen-2 or NG2) is a transmembrane proteoglycan expressed in different cell lineages in the brain,2 predominantly in perivascular pericytes and glial precursors as well as in a subset of activated microglia during neural injury.3 Although the authors did not detect this antigen in normal neural tissue, they report elevated expression of CSPG4 in GBMs of all molecular subtypes, with widespread intratumoral expression that was also retained in cultures of GBM stem-like cells (GSCs). More importantly, GSCs expressing low or moderate levels of CSPG4 in vitro increased the expression of this antigen in vivo in response to TNF-α secreted by tumor-associated microglia. The sustained expression of CSPG4 in vivo provided a good target for T cells carrying an anti-CSPG4 CAR; accordingly, the treatment of intracranial GSC-derived tumors with local CAR-T inoculation resulted in a striking improvement of median survival and a large proportion of long-term survivors. Some tumors progressed after CAR-T inoculation due to the limited persistence of these lymphocytes in vivo. However, those tumors did not show diminished expression of CSPG4, suggesting that their escape was not due to antigen loss but to other factors that are currently known to limit the efficacy of CAR-Ts in solid malignancies. Although the functions of CSPG4 in GBM are largely unknown, prior work has shown that targeting this protein results in significant effects against these tumors.4 Therefore, CSPG4 is an attractive addition to the list of targets for CAR-T therapy of GBM. As advances in molecular profiling of medulloblastoma progress, subtype-specific associations regarding prognosis, patient demographics, histologic features, and treatment response continue to evolve. Among the current subgroup classification, group 3 medulloblastomas comprise about 25% of these tumors and are often associated with MYC amplification, albeit not exclusively, and possess the worst prognosis of all the groups, perhaps partly owing to the frequently metastatic presentation and young age of affected patients. As early as 2005, Boon and colleagues argued the role of OTX2 as a medulloblastoma oncogene.1 Then in 2010, Adamson’s group also identified OTX2 as a potential player in the development of medulloblastoma that transcriptionally up-regulates the oncogene Myc.2 If we fast-forward almost 13 years from the initial observation we find Garancher and colleagues highlighting the role of NRL and CRX, two essential transcription factors of the photoreceptor lineage controlled by OTX2 in the maintenance of group 3 medulloblastoma.3 Interestingly, NRL and CRX are expressed only in the retina and pineal gland with no reported expression in the cerebellum at any time during development. Moreover, despites the relationship of NRL to the large Musculo-Aponeurotic-Fibrosarcoma family and its oncogenic cousins c-MAF, MAFB, and MAFA, NRL has not been formally implicated in human cancer. Garancher et al investigate the role of NRL and CRX in group 3 medulloblastoma using a number of medulloblastoma cell lines and patient-derived xenografts.3 Within group 3 medulloblastomas, the authors identified heterogeneity with respect to expression of these two entities, consistent with prior reports of further subgrouping within group 3 medulloblastoma based on the expression of Myc. Functional assays on NRL revealed its role in modulating cell cycle progression and protecting against apoptosis by inducing CyclinD2 and BCL-XL expression. The existence of BCL inhibitors offers an exciting opportunity to target this group of patients for whom the current therapeutic standard of care, which includes cranial spinal irradiation, is often not an option given their young age and the corresponding risk of devastating neurocognitive outcomes. Preclinically, the authors were able to show the effects of two such inhibitors in medulloblastoma cell lines and a group 3 patient-derived xenograft. The treatment had no effect on normal cells but did decrease tumor cell viability and induce apoptosis in vitro. In vivo, decreased tumor mass and increased survival were observed in treated mice. These promising studies highlight the potential for future CNS-penetrant BCL-XL inhibitors with better pharmacokinetic and side effect profiles than currently available agents as a means to treat this aggressive subtype of medulloblastoma.
Source: Neuro-Oncology
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: bronze
OpenAlex ID: W4213038318