A greater understanding of the hypersensitivity and curability of testicular germ cell tumors (TGCTs) has the potential to inform strategies to sensitize other solid tumors to conventional chemotherapies. The mechanisms of cisplatin hypersensitivity and resistance in embryonal carcinoma (EC), the stem cells of TGCTs, remain largely undefined. To study the mechanisms of cisplatin resistance we generated a large panel of independently derived, acquired resistant clones from three distinct parental EC models employing a protocol designed to match standard of care regimens of TGCT patients. Transcriptomics revealed highly significant expression changes shared between resistant cells regardless of their parental origin. This was dominated by a highly significant enrichment of genes normally repressed by H3K27 methylation and the polycomb repressive complex 2 (PRC2) which correlated with a substantial decrease in global H3K27me3, H2AK119 ubiquitination, and expression of BMI1. Importantly, repression of H3K27 methylation with the EZH2 inhibitor GSK-126 conferred cisplatin resistance to parental cells while induction of H3K27 methylation with the histone lysine demethylase inhibitor GSK-J4 resulted in increased cisplatin sensitivity to resistant cells. A gene signature based on H3K27me gene enrichment was associated with an increased rate of recurrent/progressive disease in testicular cancer patients. Our data indicates that repression of H3K27 methylation is a mechanism of cisplatin acquired resistance in TGCTs and that restoration of PRC2 complex function is a viable approach to overcome treatment failure.

Ablation of microRNA synthesis by deletion of the microRNA-processing enzyme Dicer has demonstrated that microRNAs are necessary for normal hematopoietic differentiation and function. However, it is still unclear which specific microRNAs are required for hematopoiesis and at what developmental stages they are necessary. This is especially true for immune cell development. We previously observed that overexpression of the products of the mirn23a gene (microRNA-23a, -24-2, and 27a) in hematopoietic progenitors increased myelopoiesis with a reciprocal decrease in B lymphopoiesis, both in vivo and in vitro. In this study, we generated a microRNA-23a, -24-2, and 27a germline knockout mouse to determine whether microRNA-23a, -24-2, and 27a expression was essential for immune cell development. Characterization of hematopoiesis in microRNA-23a, -24-2, and 27a mice revealed a significant increase in B lymphocytes in both the bone marrow and the spleen, with a concomitant decrease in myeloid cells (monocytes/granulocytes). Analysis of the bone marrow progenitor populations revealed a significant increase in common lymphoid progenitors and a significant decrease in both bone marrow common myeloid progenitors and granulocyte monocyte progenitors. Gene-expression analysis of primary hematopoietic progenitors and multipotent erythroid myeloid lymphoid cells showed that microRNA-23a, -24-2, and 27a regulates essential B cell gene-expression networks. Overexpression of microRNA-24-2 target Tribbles homolog 3 can recapitulate the microRNA-23a, -24-2, and 27a phenotype in vitro, suggesting that increased B cell development in microRNA-23a, -24-2, and 27a null mice can be partially explained by a Tribbles homolog 3-dependent mechanism. Data from microRNA-23a, -24-2, and 27a mice support a critical role for this microRNA cluster in regulating immune cell populations through repression of B lymphopoiesis.

Testicular germ cell tumours (TGCTs) respond well to cisplatin-based therapy. However, cisplatin resistance and poor outcomes do occur. It has been suggested that a shift towards DNA hypermethylation mediates cisplatin resistance in TGCT cells, although there is little direct evidence to support this claim. Here we utilized a series of isogenic cisplatin-resistant cell models and observed a strong association between cisplatin resistance in TGCT cells and a net increase in global CpG and non-CpG DNA methylation spanning regulatory, intergenic, genic and repeat elements. Hypermethylated loci were significantly enriched for repressive DNA segments, CTCF and RAD21 sites and lamina associated domains, suggesting that global nuclear reorganization of chromatin structure occurred in resistant cells. Hypomethylated CpG loci were significantly enriched for EZH2 and SUZ12 binding and H3K27me3 sites. Integrative transcriptome and methylome analyses showed a strong negative correlation between gene promoter and CpG island methylation and gene expression in resistant cells and a weaker positive correlation between gene body methylation and gene expression. A bidirectional shift between gene promoter and gene body DNA methylation occurred within multiple genes that was associated with upregulation of polycomb targets and downregulation of tumour suppressor genes. These data support the hypothesis that global remodelling of DNA methylation is a key factor in mediating cisplatin hypersensitivity and chemoresistance of TGCTs and furthers the rationale for hypomethylation therapy for refractory TGCT patients.

Purpose Germ cell tumors (GCTs) are cured with therapy based on cisplatin, although a clinically significant number of patients are refractory and die of progressive disease. Based on preclinical studies indicating that refractory testicular GCTs are hypersensitive to hypomethylating agents (HMAs), we conducted a phase I trial combining the next‐generation HMA guadecitabine (SGI‐110) with cisplatin in recurrent, cisplatin‐resistant GCT patients. Methods Patients with metastatic GCTs were treated for five consecutive days with guadecitabine followed by cisplatin on day 8, for a 28‐day cycle for up to six cycles. The primary endpoint was safety and toxicity including dose‐limiting toxicity (DLT) and maximum tolerated dose (MTD). Results The number of patients enrolled was 14. The majority of patients were heavily pretreated. MTD was determined to be 30 mg/m2 guadecitabine followed by 100 mg/m2 cisplatin. The major DLTs were neutropenia and thrombocytopenia. Three patients had partial responses by RECIST criteria, two of these patients, including one with primary mediastinal disease, completed the study and qualified as complete responses by serum tumor marker criteria with sustained remissions of 5 and 13 months and survival of 16 and 26 months, respectively. The overall response rate was 23%. Three patients also had stable disease indicating a clinical benefit rate of 46%. Conclusions The combination of guadecitabine and cisplatin was tolerable and demonstrated activity in patients with platinum refractory germ cell cancer.