Jiamin Chen

An array-based analysis of microRNA expression comparing matched frozen and formalin-fixed paraffin-embedded human tissue samples.

MicroRNAs (miRNAs) are small, noncoding RNAs that suppress gene expression at the posttranscriptional level via an antisense RNA-RNA interaction. miRNAs used for array-based profiling are generally purified from either snap-frozen or fresh samples. Because tissues found in most pathology departments are available only in formalin-fixed and paraffin-embedded (FFPE) states, we sought to evaluate miRNA derived from FFPE samples for microarray analysis. In this study, miRNAs extracted from matched snap-frozen and FFPE samples were profiled using the Agilent miRNA array platform (Agilent, Santa Clara, CA). Each miRNA sample was hybridized to arrays containing probes interrogating 470 human miRNAs. Seven cases were compared in either duplicate or triplicate. Intrachip and interchip analyses demonstrated that the processes of miRNA extraction, labeling, and hybridization from both frozen and FFPE samples are highly reproducible and add little variation to the results; technical replicates showed high correlations with one another (Kendall tau, 0.722 to 0.853; Spearman rank correlation coefficient, 0.891 to 0.954). Our results showed consistent high correlations between matched frozen and FFPE samples (Kendall tau, 0.669 to 0.815; Spearman rank correlation coefficient, 0.847 to 0.948), supporting the use of FFPE-derived miRNAs for array-based, gene expression profiling.

Stathmin 1 is a potential novel oncogene in melanoma

In previous studies, we demonstrated that miR-193b expression is reduced in melanoma relative to benign nevi, and also that miR-193b represses cyclin D1 and Mcl-1 expression. We suggested that stathmin 1 (STMN1) might be a target of miR-193b. STMN1 normally regulates microtubule dynamics either by sequestering free tubulin heterodimers or by promoting microtubule catastrophe. Increased expression of STMN1 has been observed in a variety of human malignancies, but its association with melanoma is unknown. We now report that STMN1 is upregulated during the progression of melanoma relative to benign nevi, and that STMN1 is directly regulated by miR-193b. Using an experimental cell culture approach, overexpression of miR-193b using synthetic microRNAs repressed STMN1 expression, whereas inhibition of miR-193b with anti-miR oligos increased STMN1 expression in melanoma cells. The use of a luciferase reporter assay confirmed that miR-193b directly regulates STMN1 by targeting the 3′-untranslated region of STMN1 mRNA. We further demonstrated that STMN1 is overexpressed in malignant melanoma compared with nevi in two independent melanoma cohorts, and that its level is inversely correlated with miR-193b expression. However, STMN1 expression was not significantly associated with patient survival, Breslow depth, mitotic count or patient age. STMN1 knockdown by small-interfering RNA in melanoma cells drastically repressed cell proliferation and migration potential, whereas ectopic expression of STMN1 using lentivirus increased cell proliferation and migration rates. Subsequent gene expression analysis indicated that interconnected cytoskeletal networks are directly affected following STMN1 knockdown. In addition, we identified deregulated genes associated with proliferation and migration, and revealed that p21Cip1/Waf1 and p27Kip could be downstream effectors of STMN1 signaling. Taken together, our study suggests that downregulation of miR-193b may contribute to increased STMN1 expression in melanoma, which consequently promotes migration and proliferation of tumor cells.

The Oncogenic Roles of DICER1 RNase IIIb Domain Mutations in Ovarian Sertoli-Leydig Cell Tumors

DICER1, an endoribonuclease required for microRNA (miRNA) biogenesis, is essential for embryogenesis and the development of many organs including ovaries. We have recently identified somatic hotspot mutations in RNase IIIb domain of DICER1 in half of ovarian Sertoli-Leydig cell tumors, a rare class of sex-cord stromal cell tumors in young women. These hotspot mutations lost IIIb cleavage activity of DICER1 in vitro and failed to produce 5p-derived miRNAs in mouse Dicer1-null ES cells. However, the oncogenic potential of these hotspot DICER1 mutations has not been studied. Here, we further revealed that the global expression of 5p-derived miRNAs was dramatically reduced in ovarian Sertoli-Leydig cell tumors carrying DICER1 hotspot mutations compared with those without DICER1 hotspot mutation. The miRNA production defect was associated with the deregulation of genes controlling cell proliferation and the cell fate. Using an immortalized human granulosa cell line, SVOG3e, we determined that the D1709N-DICER1 hotspot mutation failed to produce 5p-derived miRNAs, deregulated the expression of several genes that control gonadal differentiation and cell proliferation, and promoted cell growth. Re-expression of let-7 significantly inhibited the growth of D1709N-DICER1 SVOG3e cells, accompanied by the suppression of key regulators of cell cycle control and ovarian gonad differentiation. Taken together, our data revealed that DICER1 hotspot mutations cause systemic loss of 5p-miRNAs that can both drive pseudodifferentiation of testicular elements and cause oncogenic transformation in the ovary.

Linked read sequencing resolves complex genomic rearrangements in gastric cancer metastases

BackgroundGenome rearrangements are critical oncogenic driver events in many malignancies. However, the identification and resolution of the structure of cancer genomic rearrangements remain challenging even with whole genome sequencing.MethodsTo identify oncogenic genomic rearrangements and resolve their structure, we analyzed linked read sequencing. This approach relies on a microfluidic droplet technology to produce libraries derived from single, high molecular weight DNA molecules, 50 kb in size or greater. After sequencing, the barcoded sequence reads provide long range genomic information, identify individual high molecular weight DNA molecules, determine the haplotype context of genetic variants that occur across contiguous megabase-length segments of the genome and delineate the structure of complex rearrangements. We applied linked read sequencing of whole genomes to the analysis of a set of synchronous metastatic diffuse gastric cancers that occurred in the same individual.ResultsWhen comparing metastatic sites, our analysis implicated a complex somatic rearrangement that was present in the metastatic tumor. The oncogenic event associated with the identified complex rearrangement resulted in an amplification of the known cancer driver gene FGFR2. With further investigation using these linked read data, the FGFR2 copy number alteration was determined to be a deletion-inversion motif that underwent tandem duplication, with unique breakpoints in each metastasis. Using a three-dimensional organoid tissue model, we functionally validated the metastatic potential of an FGFR2 amplification in gastric cancer.ConclusionsOur study demonstrates that linked read sequencing is useful in characterizing oncogenic rearrangements in cancer metastasis.

Identification of large rearrangements in cancer genomes with barcode linked reads

Abstract Large genomic rearrangements involve inversions, deletions and other structural changes that span Megabase segments of the human genome. This category of genetic aberration is the cause of many hereditary genetic disorders and contributes to pathogenesis of diseases like cancer. We developed a new algorithm called ZoomX for analysing barcode-linked sequence reads—these sequences can be traced to individual high molecular weight DNA molecules (>50 kb). To generate barcode linked sequence reads, we employ a library preparation technology (10X Genomics) that uses droplets to partition and barcode DNA molecules. Using linked read data from whole genome sequencing, we identify large genomic rearrangements, typically greater than 200kb, even when they are only present in low allelic fractions. Our algorithm uses a Poisson scan statistic to identify genomic rearrangement junctions, determine counts of junction-spanning molecules and calculate a Fishers exact test for determining statistical significance for somatic aberrations. Utilizing a well-characterized human genome, we benchmarked this approach to accurately identify large rearrangement. Subsequently, we demonstrated that our algorithm identifies somatic rearrangements when present in lower allelic fractions as occurs in tumors. We characterized a set of complex cancer rearrangements with multiple classes of structural aberrations and with possible roles in oncogenesis.

Joint single cell DNA-Seq and RNA-Seq of gastric cancer reveals subclonal signatures of genomic instability and gene expression

Sequencing the genomes of individual cancer cells provides the highest resolution of intratumoral heterogeneity. To enable high throughput single cell DNA-Seq across thousands of individual cells per sample, we developed a droplet-based, automated partitioning technology for whole genome sequencing. We applied this approach on a set of gastric cancer cell lines and a primary gastric tumor. In parallel, we conducted a separate single cell RNA-Seq analysis on these same cancers and used copy number to compare results. This joint study, covering thousands of single cell genomes and transcriptomes, revealed extensive cellular diversity based on distinct copy number changes, numerous subclonal populations and in the case of the primary tumor, subclonal gene expression signatures. We found genomic evidence of positive selection – where the percentage of replicating cells per clone is higher than expected – indicating ongoing tumor evolution. Our study demonstrates that joining single cell genomic DNA and transcriptomic features provides novel insights into cancer heterogeneity and biology. SIGNIFICANCE We conducted a massively parallel DNA sequencing analysis on a set of gastric cancer cell lines and a primary gastric tumor in combination with a joint single cell RNA-Seq analysis. This joint study, covering thousands of single cell genomes and transcriptomes, revealed extensive cellular diversity based on distinct copy number changes, numerous subclonal populations and in the case of the primary tumor, subclonal gene expression signatures. We found genomic evidence of positive selection where the percentage of replicating cells per clone is higher than expected indicating ongoing tumor evolution. Our study demonstrates that combining single cell genomic DNA and transcriptomic features provides novel insights into cancer heterogeneity and biology.

Single-cell transcriptome analysis identifies distinct cell types and intercellular niche signaling in a primary gastric organoid model

The diverse cellular milieu of the gastric tissue microenvironment plays a critical role in normal tissue homeostasis and tumor development. However, few cell culture model can recapitulate the tissue microenvironment and intercellular signaling in vitro. Here we applied an air-liquid interface method to culture primary gastric organoids that contains epithelium with endogenous stroma. To characterize the microenvironment and intercellular signaling in this model, we analyzed the transcriptomes of over 5,000 individual cells from primary gastric organoids cultured at different time points. We identified epithelial cells, fibroblasts and macrophages at the early stage of organoid formation, and revealed that macrophages were polarized towards wound healing and tumor promotion. The organoids maintained both epithelial and fibroblast lineages during the course of time, and a subset of cells in both lineages expressed the stem cell marker Lgr5. We identified that Rspo3 was specifically expressed in the fibroblast lineage, providing an endogenous source of the R-spondin to activate Wnt signaling. Our studies demonstrate that air-liquid-interface-derived organoids provide a novel platform to study intercellular signaling and immune response in vitro.

Abstract 2443: Massively parallel single-cell RNA-Seq identifies diverse subpopulations displaying EMT and stem-like features

Single-cell transcriptome analysis enables a new paradigm for studying complex systems in cancer. As opposed to bulk sequencing, which averages genomic signals across thousands or millions of cells and obscures the presence of rare subtypes, single cell sequencing enables the interrogation of individual cells. In cancer, intratumoral heterogeneity is observed at both genomic and epigenomic levels, and its analysis enables the discovery of new actionable targets and treatment modalities tailored to individual subpopulations. As an example, many cancer cell lines and those derived from patients contain subpopulations marked by distinct patterns of surface markers such as CD44, and are linked to drug resistant and tumor initiating phenotypes. A complete characterization of such cellular populations ideally requires marker-free sampling, followed by clustering into distinct subgroups. In this study, we demonstrate the significant advantages of such an approach; we utilize a high-throughput single-cell RNA-Seq method to characterize the transcriptomic profiles of cellular populations. We performed single-cell RNA-Seq on thousands of cells in the matched SW480 (primary) and SW620 (metastatic) colorectal cell lines using a microfluidic droplet barcoding technology that enables the tracking of single cells during library preparation. By focusing on genes with high inter-cell variability, we discovered a small subpopulation of cells that displayed a distinct gene expression signature from the major subpopulation. Differential gene expression analysis of this subpopulation yielded genes virtually all enriched in the epithelial-to-mesenchymal transition (EMT) pathway. These cells showed significant increases in canonical mesenchymal marker genes such as VIM, CD44, and SOX9. Gene expression profiles of these subpopulations also correlated with established EMT signatures. Remarkably, this subpopulation did not display mutual exclusivity in gene expression with the epithelial marker EPCAM, which possibly indicates an intermediate mesenchymal phenotype. We also observed in the major population cluster a small subset of cells totaling less than 1% of the population that were significantly enriched for LGR5 expression, a common stem-like marker in colorectal cancer. Overall, we demonstrate the use of single-cell RNA-Seq to discover and characterize a diversity of cellular states that would otherwise be impossible from bulk analysis. Citation Format: Billy Lau, Jiamin Chen, Hanlee P. Ji. Massively parallel single-cell RNA-Seq identifies diverse subpopulations displaying EMT and stem-like features [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 2443. doi:10.1158/1538-7445.AM2017-2443

Abstract 4339: Single cell RNA sequencing dissects cellular growth factor dependencies and oncogenic driver effects in an organoid model of gastric cancer

Gastric cancer is a lethal malignancy with few therapeutic options. Gastric tumors rely on complex intercellular signaling “crosstalk” that enables tumor development, metastasis, and therapeutic resistance. To recapitulate the intercellular communications among various cell populations that exist in vivo, we are using a three-dimensional culture system to grow and manipulate mouse gastric tissue in vitro, otherwise referred to as organoids. Importantly, this organoid model contains epithelium with its endogenous mesenchymal niche and does not require exogenous Wnt stimulation. To systematically analyze the distinct cellular lineages and their interactions, we applied a massively-scaled single cell RNA-Seq platform to sequence thousands of individual cells from organoid cultures. With PCA and t-SNE analysis of the high-dimensional data generated from single cell RNA-seq, we characterized two major cell types, i.e. epithelial and mesenchymal cells. Leverage the information from single cell transcriptome profiles, we identified specific niche factors of the Wnt signaling pathways that are activated in different stomach cell lineages. These results suggest that the mesenchymal cell populations provide a potential source of the R-spondin, a Wnt agonist, that sustains the growth of epithelium. Furthermore, we compared cell populations from Cdh1-/-/Trp53-/- and Trp53-/-organoids, and characterized changes on the transcriptome profiles due to the loss of Cdh1, an early oncogenic event in diffuse gastric cancer development. Overall, using organoid model and high-throughput single cell RNA-Seq provides a novel approach to study early tumor transformation and critical cancer-stroma interactions. Citation Format: Jiamin Chen, Noemi Andor, Susan M. Grimes, Billy Lau, Hanlee P. Ji. Single cell RNA sequencing dissects cellular growth factor dependencies and oncogenic driver effects in an organoid model of gastric cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 4339. doi:10.1158/1538-7445.AM2017-4339

Abstract 3693: Single cell RNA-Seq of primary lymphomas reveals the diverse transcriptional states of the cancer immunologic milieu

Follicular lymphoma (FL) is an incurable B-cell malignancy that has the potential to transform into highly aggressive, lethal lymphomas. To resolve the cellular characteristics of the underlying cancer immunologic milieu at single cell resolution, we leveraged droplet-based barcoding technology for highly parallel single cell RNA-Seq. We analyzed the transcriptomes across tens of thousands of indviidual cells derived from five primary FL tumors (average > 5,000 cells per sample). These tumors were obtained via direct surgical biopsies. In parallel, we conducted multi-dimensional flow cytometry that orthogonally confirmed our characterization of the various cell types including the various subsets of T cells that existed within each tumor. We identified multiple cellular subpopulations, matching known hematopoietic lineages that were present within each tumor. Despite some common features, such as MYC and BCL overexpression, transcriptional patterns and regulatory programs destinguished the different tumors from each other. We compared the transcriptional profiles of malignant cellular subpopulations to the matched normal B-cells within each sample. Malignant B-cells were characterized by expression of restricted immunoglobulin light chain type (either kappa or lambda), BCL2 expression, and CD20 expression. In each case a minor population of normal B-cells was characterized by expression of the opposite light chain type and a pattern of gene expression common to B cells in all tumor samples and normal peripheral blood. Tumor B cells consistently demonstrated downregulation of β-2-microglobulin and for those cases that harbored a founder CREBBP mutation, a downregulation of MHC II, that would cause an alteration in antigen presentation and evasion of the T cell immune system. Moreover, we characterized the transcriptional profiles of the infiltrating T-cell populations within each tumor, providing a high-resolution perspective of immunologic cellular interactions. Overall, we identified the diverse transcriptional states at single cell resolution among different tumors and discovered specific genes that were aberrantly expressed compared to normal B cells within the same microenvironment. Our findings provide an unprecedented resolution of distinct immune lineages as seen by transcriptionally characterized cell diversity. Studies using single cell genomics may have implications for considering immunotherapeutic efficacy. Citation Format: Noemi Andor, Erin Simonds, Jiamin Chen, Christina Wood, Susan Grimes, Debra Czerwinski, Grace Zheng, Ronald Levy, Hanlee P. Ji. Single cell RNA-Seq of primary lymphomas reveals the diverse transcriptional states of the cancer immunologic milieu [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3693. doi:10.1158/1538-7445.AM2017-3693