Ruli Gao

Punctuated copy number evolution and clonal stasis in triple-negative breast cancer

Aneuploidy is a hallmark of breast cancer; however, knowledge of how these complex genomic rearrangements evolve during tumorigenesis is limited. In this study, we developed a highly multiplexed single-nucleus sequencing method to investigate copy number evolution in patients with triple-negative breast cancer. We sequenced 1,000 single cells from tumors in 12 patients and identified 1–3 major clonal subpopulations in each tumor that shared a common evolutionary lineage. For each tumor, we also identified a minor subpopulation of non-clonal cells that were classified as metastable, pseudodiploid or chromazemic. Phylogenetic analysis and mathematical modeling suggest that these data are unlikely to be explained by the gradual accumulation of copy number events over time. In contrast, our data challenge the paradigm of gradual evolution, showing that the majority of copy number aberrations are acquired at the earliest stages of tumor evolution, in short punctuated bursts, followed by stable clonal expansions that form the tumor mass.

Tumor evolution: Linear, branching, neutral or punctuated? ☆

Intratumor heterogeneity has been widely reported in human cancers, but our knowledge of how this genetic diversity emerges over time remains limited. A central challenge in studying tumor evolution is the difficulty in collecting longitudinal samples from cancer patients. Consequently, most studies have inferred tumor evolution from single time-point samples, providing very indirect information. These data have led to several competing models of tumor evolution: linear, branching, neutral and punctuated. Each model makes different assumptions regarding the timing of mutations and selection of clones, and therefore has different implications for the diagnosis and therapeutic treatment of cancer patients. Furthermore, emerging evidence suggests that models may change during tumor progression or operate concurrently for different classes of mutations. Finally, we discuss data that supports the theory that most human tumors evolve from a single cell in the normal tissue. This article is part of a Special Issue entitled: Evolutionary principles - heterogeneity in cancer?, edited by Dr. Robert A. Gatenby.

Highly multiplexed targeted DNA sequencing from single nuclei

Single-cell DNA sequencing methods are challenged by poor physical coverage, high technical error rates and low throughput. To address these issues, we developed a single-cell DNA sequencing protocol that combines flow-sorting of single nuclei, time-limited multiple-displacement amplification (MDA), low-input library preparation, DNA barcoding, targeted capture and next-generation sequencing (NGS). This approach represents a major improvement over our previous single nucleus sequencing (SNS) Nature Protocols paper in terms of generating higher-coverage data (>90%), thereby enabling the detection of genome-wide variants in single mammalian cells at base-pair resolution. Furthermore, by pooling 48–96 single-cell libraries together for targeted capture, this approach can be used to sequence many single-cell libraries in parallel in a single reaction. This protocol greatly reduces the cost of single-cell DNA sequencing, and it can be completed in 5–6 d by advanced users. This single-cell DNA sequencing protocol has broad applications for studying rare cells and complex populations in diverse fields of biological research and medicine.

Nanogrid single-nucleus RNA sequencing reveals phenotypic diversity in breast cancer

Single cell RNA sequencing has emerged as a powerful tool for resolving transcriptional diversity in tumors, but is limited by throughput, cost and the ability to process archival frozen tissue samples. Here we develop a high-throughput 3′ single-nucleus RNA sequencing approach that combines nanogrid technology, automated imaging, and cell selection to sequence up to ~1800 single nuclei in parallel. We compare the transcriptomes of 485 single nuclei to 424 single cells in a breast cancer cell line, which shows a high concordance (93.34%) in gene levels and abundance. We also analyze 416 nuclei from a frozen breast tumor sample and 380 nuclei from normal breast tissue. These data reveal heterogeneity in cancer cell phenotypes, including angiogenesis, proliferation, and stemness, and a minor subpopulation (19%) with many overexpressed cancer genes. Our studies demonstrate the utility of nanogrid single-nucleus RNA sequencing for studying the transcriptional programs of tumor nuclei in frozen archival tissue samples.Single cell RNA sequencing is a powerful tool for understanding cellular diversity but is limited by cost, throughput and sample preparation. Here the authors use nanogrid technology with integrated imaging to sequence thousands of cancer nuclei in parallel from fresh or frozen tissue.

Chemoresistance Evolution in Triple-Negative Breast Cancer Delineated by Single-Cell Sequencing

Triple-negative breast cancer (TNBC) is an aggressive subtype that frequently develops resistance to chemotherapy. An unresolved question is whether resistance is caused by the selection of rare pre-existing clones or alternatively through the acquisition of new genomic aberrations. To investigate this question, we applied single-cell DNA and RNA sequencing in addition to bulk exome sequencing to profile longitudinal samples from 20 TNBC patients during neoadjuvant chemotherapy (NAC). Deep-exome sequencing identified 10 patients in which NAC led to clonal extinction and 10 patients in which clones persisted after treatment. In 8 patients, we performed a more detailed study using single-cell DNA sequencing to analyze 900 cells and single-cell RNA sequencing to analyze 6,862 cells. Our data showed that resistant genotypes were pre-existing and adaptively selected by NAC, while transcriptional profiles were acquired by reprogramming in response to chemotherapy in TNBC patients.

Abstract 418: Adaptive resistance to chemotherapy in triple-negative breast cancer revealed by single cell DNA and RNA sequencing

Triple-negative breast cancer (TNBC) is an aggressive subtype that displays extensive intratumor heterogeneity and frequently (46%) develops resistance to neoadjuvant chemotherapy (NAC). Currently, the genomic basis of chemoresistance remains poorly understood. An important question is whether resistance to chemotherapy is driven by the selection of rare pre-existing subclones with genomic mutations and transcriptional programs that confer resistance to chemotherapy (adaptive resistance) or by the spontaneous induction of new mutations and expression changes that confer a resistant phenotype (acquired resistance). To investigate this question we applied single cell DNA and RNA sequencing methods and deep-exome sequencing to longitudinal time-point samples collected from a cohort of 20 TNBC patients. Deep-exome sequencing of the cohort at three time points revealed a random death model, wherein multiple clones were targeted, as opposed to the selection of specific somatic mutations. In contrast, single cell copy number profiling of ~800 cells from 8 patients identified an adaptive resistance model, wherein minor subclones from the pre-treatment tumors were selected and expanded in response to NAC. Similarly, single cell RNA sequencing of ~8000 cells from 8 patients identified subclones with chemoresistant phenotypes that were selected in response to NAC, resulting in the expansion of the resistant tumor mass. These data suggest that chemoresistance evolves through the selection of copy number changes and expression changes in signaling pathways associated with chemoresistance, rather than point mutations. This adaptive resistance model has important translational implications in clinical diagnostics, by suggesting that resistant clones can be detected in TNBC patients prior to the administration of chemotherapy. Citation Format: Charissa Kim, Ruli Gao, Emi Sei, Rachel Brandt, Nicola Crosetto, Theodoros Foukakis, Nicholas Navin. Adaptive resistance to chemotherapy in triple-negative breast cancer revealed by single cell DNA and RNA sequencing [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 418. doi:10.1158/1538-7445.AM2017-418

Abstract 5399: Investigating phenotypic plasticity in breast cancer with high-throughput nanogrid single-nucleus RNA sequencing

Single-cell RNA sequencing (RNA-seq) is a powerful tool for investigating rare tumor subpopulations and resolving intra-tumor heterogeneity, but is low throughput, expensive, and requires fresh tissue samples. To address these limitations, we developed a 5’ high-throughput single-nucleus RNA sequencing (SNRS) approach that uses nanogrid technology to perform single-cell imaging and sequencing of 500-2500 nuclei in parallel. The automated image scanning procedure allowed us to exclude doublets and select live cells with DAPI/PI staining. This approach allows the transcriptomic profiling of frozen tissue samples, in which the cytoplasmic membrane is ruptured in cells, but leaves the nuclear membrane intact. We validated SNRS in a breast cancer cell line (SK-BR-3) and compared the transcriptomes of 500 nuclei to 500 whole cells, which revealed a high concordance in the number of genes expressed as well as their expression levels. We also performed bulk RNA-seq of isolated nuclear and cellular fractions from 5 breast cancer cell lines, which showed a high concordance in genes and expression levels. Differentially expressed genes in the nucleus mainly included lincRNAs, pseudogenes and mitochondria genes, but did not affect most cancer genes and pathway analysis. We further applied SNRS to sequence 500 nuclei from a triple-negative breast cancer patient and identified diverse phenotypes in tumor cells, including variation in cell proliferation, migration, invasion, and epithelial-to-mesenchymal transition. These studies demonstrated the technical feasibility of using a nanogrid platform to perform high-throughput single-cell RNA sequencing and showed that nuclei from cell lines and tumors can be used to study signaling pathways and gene networks that play an important role in tumor progression. Citation Format: Ruli Gao, Charissa Kim, Emi Sei, Jie Yang, Leo Chan, Maithreyan Srinivasan, Hong Zhang, Funda Meric-Bernstam, Nicholas E. Navin. Investigating phenotypic plasticity in breast cancer with high-throughput nanogrid single-nucleus RNA sequencing [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 5399. doi:10.1158/1538-7445.AM2017-5399

Abstract 2375: Multiclonal invasion in breast cancer identified by single cell DNA sequencing

Ductal carcinoma in situ (DCIS) is the most common form of early stage breast cancer and is frequently detected by mammography. Only 10% of low-grade and 30% of high-grade DCIS patients will later present with an invasive carcinoma, making it difficult to determine which patients to treat aggressively. Although DCIS is often considered a precursor to invasive breast carcinoma, a major question in the field is whether invasive subpopulations evolve directly from the in situ populations or independently. We hypothesize that early tumor cells evolve over an extended period of time in the ducts, generating multiple subpopulations that migrate out of the ducts together. To test this hypothesis, we developed an approach that combines laser-capture microdissection (LCM) with single cell sequencing to perform DNA copy number profiling while preserving the topographic location of cells. We applied this method to 10 high-grade breast cancer patients with defined in situ, invasive, and normal regions. Using these data, we delineated the clonal substructure and location of the in situ and invasive subpopulations. We found that most high-grade tumors are composed of 1-4 major clonal subpopulations that are intermixed in both the in situ and invasive regions. These subpopulations share a common evolutionary lineage or cell of origin. We also performed deep-exome sequencing (100X) of matched microdissected regions of in situ, invasive and adjacent normal tissue from these breast cancer patients. In each tumor we identified a large number of shared somatic mutations between the in situ and invasive regions, further supporting our direct genomic lineage hypothesis. We also identified a few mutations confined to the invasive tumor regions that may play an important role in the invasive phenotype. These data support a multiclonal invasion model in which distinct clones evolve in the ducts and subsequently migrate into the adjacent tissues to establish invasive tumors. This model has important implications for the diagnosis and therapeutic treatment of DCIS breast cancer patients since multiple clones may need to be targeted to inhibit invasion. Citation Format: Anna K. Casasent, Annalyssa N. Long, Aislyn Schalck, Emi Sei, Ruli Gao, Alexander Davis, Yong Wang, Mary E. Edgerton, Nicholas E. Navin. Multiclonal invasion in breast cancer identified by single cell DNA sequencing. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2375.

Abstract 2371: Multi-cell phylogenetic inference of copy number evolution in breast cancer

Aneuploidy is a hallmark of breast cancer, but little is known about how these complex genomic rearrangements evolve during tumor growth. To investigate this question, we developed a Highly-Multiplexed Single Nucleus Sequencing (HM-SMS) method to profile genome-wide copy number in individual tumor cells and compared multiple cells to infer evolutionary lineages. We applied this method to analyze thousands of single cells from 12 triple-negative breast cancer patients and 10 ductal-carcinoma-in-situ (DCIS) patients. Integer copy number profiles were calculated from a combination of sequencing data and ploidy estimation using flow cytometry. A multi-cell segmentation algorithm was used to identify common CNA events that are shared between tumor cells in each patient. Trinary event matrices were calculated for each patient to determine the presence or absence of individual copy number aberrations (CNAs) in individual cells. Using the maximum parsimony criterion, we inferred a phylogenetic tree for each tumor and estimated the temporal order of the CNAs.. Our data suggest that most tumors consisted of 1-4 major clonal subpopulations that shared a common evolutionary lineage, suggesting that they evolved from a single normal cell in the breast tissue. Furthermore our data support a punctuated model of copy number evolution, which which the majority of CNAs were acquired in a short evolutionary burst, followed by one or more stable clonal expansions to form the tumor mass. Citation Format: Alexander J. Davis, Ruli Gao, Emi Sei, Pei-Ching Tsai, Anna Casasent, Amy Zhang, Xiuqing Shi, Yong Wang, Jill Waters, Funda Meric-Bernstam, Mary Edgerton, Nicholas Navin. Multi-cell phylogenetic inference of copy number evolution in breast cancer. [abstract]. In: Proceedings of the 107th Annual Meeting of the American Association for Cancer Research; 2016 Apr 16-20; New Orleans, LA. Philadelphia (PA): AACR; Cancer Res 2016;76(14 Suppl):Abstract nr 2371.