Faculty Bibliography

Introduction: Gene rearrangements play a critical role in the development of brain tumors. RNA next-generation sequencing (NGS) panels cover a limited number of genes, are rarely successful in FFPE samples > 5 years old, and cannot detect rearrangements between genes and non-coding regulatory regions. We evaluated whole genome Hi-C NGS for detection of gene fusions and cryptic rearrangements.
Method(s): DNA was extracted from FFPE scrolls of 55 glial and non-glial brain tumors and processed using Arima-HiC+ FFPE Sample protocol, consisting of chromatin fragmentation, labeling, and re-ligation, followed by DNA purification and library preparation for paired-end Illumina sequencing with an average of 10X genome coverage (100M PE reads per sample). Data were analyzed using the Arima-SV pipeline using Juicer and HiCUP, SV detection using HiC-Breakfinder, loop calling using Juicer Tools, and integrative data visualization using Juicebox. Overexpression of putative driver genes was confirmed by immunohistochemistry.
Result(s): Hi-C libraries were prepared and sequenced from FFPE tissues including samples that failed RNA NGS. Hi-C successfully detected gene-gene fusions including actionable EML4-NTRK3, ETV6-NTRK3, fusions. We detected rearrangements missed by RNA NGS (i.e., complex MYBL1 rearrangement) or between non-coding regions and known cancer genes (i.e. PDL1, PAX5, NRAS, TERT, KAT6A, GATA6, and ARID1B). Since Hi-C data captures 3D genome structural features such as chromatin loops and topological domains, datasets were of high quality and capable of detecting up to 13,000 chromatin loops per tumor.
Conclusion(s): Genome-wide Hi-C NGS is successful in detecting gene fusions and cryptic rearrangements between coding and non-coding regions in archival FFPE tissue including degraded samples. Because Hi-C data captures 3D genome structures, these datasets simultaneously inform gene regulatory mechanisms that may play a role in oncogenesis or tumor progression. Whole-genome Hi-C NGS expands our ability to detect actionable and novel drivers, and potentially new therapeutic targets in a single NGS workflow

Image-based analysis as a method for mutation detection can be advantageous in settings when tumor tissue is limited or unavailable for direct testing. Here, we utilize two distinct and complementary machine learning methods of analyzing whole slide images (WSI) for predicting mutated BRAF. In the first method, WSI of melanomas from 256 patients were used to train a deep convolutional neural network (CNN) in order to develop a fully automated model that first selects for tumor-rich areas (Area Under the Curve AUC=0.96) then predicts for mutated BRAF (AUC=0.71). Saliency mapping was performed and revealed that pixels corresponding to nuclei were the most relevant to network learning. In the second method, WSI were analyzed using a pathomics pipeline that first annotates nuclei and then quantifies nuclear features, demonstrating that mutated BRAF nuclei were significantly larger and rounder nuclei compared to BRAF WT nuclei. Lastly, we developed a model that combines clinical information, deep learning, and pathomics that improves the predictive performance for mutated BRAF to AUC=0.89. Not only does this provide additional insights on how BRAF mutations affect tumor structural characteristics, machine learning-based analysis of WSI has the potential to be integrated into higher order models for understanding tumor biology.

Embryos are diagnosed as mosaic if their chromosomal copy number falls between euploid and aneuploid. The purpose of this study was to investigate the impact of mosaicism on global gene expression. This study included 42 blastocysts that underwent preimplantation genetic testing for aneuploidy (PGT-A) and were donated for IRB approved research. Fourteen blastocysts were diagnosed as mosaic with Next-generation Sequencing (NGS). Three NGS diagnosed euploid embryos, and 25 aneuploid embryos (9 NGS, 14 array Comparative Genomic Hybridization, 2 Single Nucleotide Polymorphism array) were used as comparisons. RNA-sequencing was performed on all of the blastocysts. Differentially expressed genes (DEGs) were calculated using DESeq2/3.5 (R Bioconductor Package) with p < 0.05 considered significantly differentially expressed. Pathway analysis was performed on mosaic embryos using EnrichR with p < 0.05 considered significant. With euploid embryo gene expression used as a control, 12 of 14 mosaic embryos had fewer DEGs compared to aneuploid embryos involving the same chromosome. On principal component analysis (PCA), mosaic embryos mapped separately from aneuploid embryos. Pathways involving cell proliferation, differentiation, and apoptosis were the most disrupted within mosaic embryos. Mosaic embryos have decreased disruption of global gene expression compared to aneuploid embryos. This study was limited by the small sample size, lack of replicate samples for each mosaic abnormality, and use of multiple different PGT-A platforms for the diagnosis of aneuploid embryos.

ARID2 is the most recurrently mutated SWI/SNF complex member in melanoma; however, its tumor-suppressive mechanisms in the context of the chromatin landscape remain to be elucidated. Here, we model ARID2 deficiency in melanoma cells, which results in defective PBAF complex assembly with a concomitant genomic redistribution of the BAF complex. Upon ARID2 depletion, a subset of PBAF and shared BAF-PBAF-occupied regions displays diminished chromatin accessibility and associated gene expression, while BAF-occupied enhancers gain chromatin accessibility and expression of genes linked to the process of invasion. As a function of altered accessibility, the genomic occupancy of melanoma-relevant transcription factors is affected and significantly correlates with the observed transcriptional changes. We further demonstrate that ARID2-deficient cells acquire the ability to colonize distal organs in multiple animal models. Taken together, our results reveal a role for ARID2 in mediating BAF and PBAF subcomplex chromatin dynamics with consequences for melanoma metastasis.

Resistance to targeted therapies is an important clinical problem in HER2-positive (HER2+) breast cancer. "Drug-tolerant persisters" (DTPs), a sub-population of cancer cells that survive via reversible, non-genetic mechanisms, are implicated in resistance to tyrosine kinase inhibitors (TKIs) in other malignancies, but DTPs following HER2 TKI exposure have not been well characterized. We found that HER2 TKIs evoke DTPs with a luminal-like or a mesenchymal-like transcriptome. Lentiviral barcoding/single cell RNA-sequencing reveal that HER2+ breast cancer cells cycle stochastically through a "pre-DTP" state, characterized by a G0-like expression signature and enriched for diapause and/or senescence genes. Trajectory analysis/cell sorting show that pre-DTPs preferentially yield DTPs upon HER2 TKI exposure. Cells with similar transcriptomes are present in HER2+ breast tumors and are associated with poor TKI response. Finally, biochemical experiments indicate that luminal-like DTPs survive via estrogen receptor-dependent induction of SGK3, leading to rewiring of the PI3K/AKT/mTORC1 pathway to enable AKT-independent mTORC1 activation.

Developmental origins of dendritic cells (DCs) including conventional DCs (cDCs, comprising cDC1 and cDC2 subsets) and plasmacytoid DCs (pDCs) remain unclear. We studied DC development in unmanipulated adult mice using inducible lineage tracing combined with clonal DNA "barcoding" and single-cell transcriptome and phenotype analysis (CITE-seq). Inducible tracing of Cx3cr1⁺ hematopoietic progenitors in the bone marrow showed that they simultaneously produce all DC subsets including pDCs, cDC1s, and cDC2s. Clonal tracing of hematopoietic stem cells (HSCs) and of Cx3cr1⁺ progenitors revealed clone sharing between cDC1s and pDCs, but not between the two cDC subsets or between pDCs and B cells. Accordingly, CITE-seq analyses of differentiating HSCs and Cx3cr1⁺ progenitors identified progressive stages of pDC development including Cx3cr1⁺ Ly-6D⁺ pro-pDCs that were distinct from lymphoid progenitors. These results reveal the shared origin of pDCs and cDCs and suggest a revised scheme of DC development whereby pDCs share clonal relationship with cDC1s.

Deregulation of the BCL-2 family interaction network ensures cancer resistance to apoptosis and is a major challenge to current treatments. Cancer cells commonly evade apoptosis through upregulation of the BCL-2 anti-apoptotic proteins; however, more resistant cancers also downregulate or inactivate pro-apoptotic proteins to suppress apoptosis. Here, we find that apoptosis resistance in a diverse panel of solid and hematological malignancies is mediated by both overexpression of BCL-XL and an unprimed apoptotic state, limiting direct and indirect activation mechanisms of pro-apoptotic BAX. Both survival mechanisms can be overcome by the combination of an orally bioavailable BAX activator, BTSA1.2 with Navitoclax. The combination demonstrates synergistic efficacy in apoptosis-resistant cancer cells, xenografts, and patient-derived tumors while sparing healthy tissues. Additionally, functional assays and genomic markers are identified to predict sensitive tumors to the combination treatment. These findings advance the understanding of apoptosis resistance mechanisms and demonstrate a novel therapeutic strategy for cancer treatment.

The contribution of epigenetic dysregulation to metastasis remains understudied. Through a meta-analysis of gene expression datasets followed by a mini-screen, we identified Plant Homeodomain Finger protein 8 (PHF8), a histone demethylase of the Jumonji C protein family, as a previously unidentified prometastatic gene in melanoma. Loss- and gain-of-function approaches demonstrate that PHF8 promotes cell invasion without affecting proliferation in vitro and increases dissemination but not subcutaneous tumor growth in vivo, thus supporting its specific contribution to the acquisition of metastatic potential. PHF8 requires its histone demethylase activity to enhance melanoma cell invasion. Transcriptomic and epigenomic analyses revealed that PHF8 orchestrates a molecular program that directly controls the TGFβ signaling pathway and, as a consequence, melanoma invasion and metastasis. Our findings bring a mechanistic understanding of epigenetic regulation of metastatic fitness in cancer, which may pave the way for improved therapeutic interventions.

CCCTC-binding factor (CTCF) is critical to three-dimensional genome organization. Upon differentiation, CTCF insulates active and repressed genes within Hox gene clusters. We conducted a genome-wide CRISPR knockout (KO) screen to identify genes required for CTCF-boundary activity at the HoxA cluster, complemented by biochemical approaches. Among the candidates, we identified Myc-associated zinc-finger protein (MAZ) as a cofactor in CTCF insulation. MAZ colocalizes with CTCF at chromatin borders and, similar to CTCF, interacts with the cohesin subunit RAD21. MAZ KO disrupts gene expression and local contacts within topologically associating domains. Similar to CTCF motif deletions, MAZ motif deletions lead to derepression of posterior Hox genes immediately after CTCF boundaries upon differentiation, giving rise to homeotic transformations in mouse. Thus, MAZ is a factor contributing to appropriate insulation, gene expression and genomic architecture during development.

The production of noncanonical mRNA transcripts is associated with cell transformation. Driven by our previous findings on the sensitivity of T cell acute lymphoblastic leukemia (T-ALL) cells to SF3B1 inhibitors, we identified that SF3B1 inhibition blocks T-ALL growth in vivo with no notable associated toxicity. We also revealed protein stabilization of the U2 complex component SF3B1 via deubiquitination. Our studies showed that SF3B1 inhibition perturbs exon skipping, leading to nonsense-mediated decay and diminished levels of DNA damage response-related transcripts, such as the serine/threonine kinase CHEK2, and impaired DNA damage response. We also identified that SF3B1 inhibition leads to a general decrease in R-loop formation. We further demonstrate that clinically used SF3B1 inhibitors synergize with CHEK2 inhibitors and chemotherapeutic drugs to block leukemia growth. Our study provides the proof of principle for posttranslational regulation of splicing components and associated roles and therapeutic implications for the U2 complex in T cell leukemia.