Search

Polycomb repressive complex 2 (PRC2) is an epigenetic regulator of gene expression that possesses histone methyltransferase activity. PRC2 trimethylates lysine 27 of histone H3 proteins (H3K27me3) as a chromatin modification associated with repressed transcription of genes frequently involved in cell proliferation or self-renewal. Loss-of-function mutations in the PRC2 core subunit SUZ12 have been identified in a variety of tumors, including malignant peripheral nerve sheath tumors (MPNSTs). To determine the consequences of SUZ12 loss in the pathogenesis of MPNST and other cancers, we used CRISPR-Cas9 to disrupt the open reading frame of each of two orthologous suz12 genes in zebrafish: suz12a and suz12b. We generated these knockout alleles in the germline of our previously described p53 (also known as tp53)- and nf1-deficient zebrafish model of MPNSTs. Loss of suz12 significantly accelerated the onset and increased the penetrance of MPNSTs compared to that in control zebrafish. Moreover, in suz12-deficient zebrafish, we detected additional types of tumors besides MPNSTs, including leukemia with histological characteristics of lymphoid malignancies, soft tissue sarcoma and pancreatic adenocarcinoma, which were not detected in p53/nf1-deficient control fish, and are also contained in the human spectrum of SUZ12-deficient malignancies identified in the AACR Genie database. The suz12-knockout tumors displayed reduced or abolished H3K27me3 epigenetic marks and upregulation of gene sets reported to be targeted by PRC2. Thus, these zebrafish lines with inactivation of suz12 in combination with loss of p53/nf1 provide a model of human MPNSTs and multiple other tumor types, which will be useful for mechanistic studies of molecular pathogenesis and targeted therapy with small molecule inhibitors.

SUMMARY Neurofibromatosis type 1 (NF1) is a common, dominantly inherited genetic disorder that results from mutations in the neurofibromin 1 (NF1) gene. Affected individuals demonstrate abnormalities in neural-crest-derived tissues that include hyperpigmented skin lesions and benign peripheral nerve sheath tumors. NF1 patients also have a predisposition to malignancies including juvenile myelomonocytic leukemia (JMML), optic glioma, glioblastoma, schwannoma and malignant peripheral nerve sheath tumors (MPNSTs). In an effort to better define the molecular and cellular determinants of NF1 disease pathogenesis in vivo, we employed targeted mutagenesis strategies to generate zebrafish harboring stable germline mutations in nf1a and nf1b, orthologues of NF1. Animals homozygous for loss-of-function alleles of nf1a or nf1b alone are phenotypically normal and viable. Homozygous loss of both alleles in combination generates larval phenotypes that resemble aspects of the human disease and results in larval lethality between 7 and 10 days post fertilization. nf1-null larvae demonstrate significant central and peripheral nervous system defects. These include aberrant proliferation and differentiation of oligodendrocyte progenitor cells (OPCs), dysmorphic myelin sheaths and hyperplasia of Schwann cells. Loss of nf1 contributes to tumorigenesis as demonstrated by an accelerated onset and increased penetrance of high-grade gliomas and MPNSTs in adult nf1a+/−; nf1b−/−; p53e7/e7 animals. nf1-null larvae also demonstrate significant motor and learning defects. Importantly, we identify and quantitatively analyze a novel melanophore phenotype in nf1-null larvae, providing the first animal model of the pathognomonic pigmentation lesions of NF1. Together, these findings support a role for nf1a and nf1b as potent tumor suppressor genes that also function in the development of both central and peripheral glial cells as well as melanophores in zebrafish.