The neuroendocrine transition in prostate cancer is dynamic and dependent on ASCL1

Lineage plasticity is a recognized hallmark of cancer progression that can shape therapy outcomes. The underlying cellular and molecular mechanisms mediating lineage plasticity remain poorly understood. Here, we describe a versatile in vivo platform to identify and interrogate the molecular determinants of neuroendocrine lineage transformation at different stages of prostate cancer progression. Adenocarcinomas reliably develop following orthotopic transplantation of primary mouse prostate organoids acutely engineered with human-relevant driver alterations (e.g., Rb1-/-; Trp53-/-; cMyc+ or Pten-/-; Trp53-/-; cMyc+), but only those with Rb1 deletion progress to ASCL1+ neuroendocrine prostate cancer (NEPC), a highly aggressive, androgen receptor signaling inhibitor (ARSI)-resistant tumor. Importantly, we show this lineage transition requires a native in vivo microenvironment not replicated by conventional organoid culture. By integrating multiplexed immunofluorescence, spatial transcriptomics and PrismSpot to identify cell type-specific spatial gene modules, we reveal that ASCL1+ cells arise from KRT8+ luminal epithelial cells that progressively acquire transcriptional heterogeneity, producing large ASCL1+;KRT8- NEPC clusters. Ascl1 loss in established NEPC results in transient tumor regression followed by recurrence; however, Ascl1 deletion prior to transplantation completely abrogates lineage plasticity, yielding adenocarcinomas with elevated AR expression and marked sensitivity to castration. The dynamic feature of this model reveals the importance of timing of therapies focused on lineage plasticity and offers a platform for identification of additional lineage plasticity drivers.

Metastatic site influences driver gene function in pancreatic cancer

Driver gene mutations can increase the metastatic potential of the primary tumor1-3, but their role in sustaining tumor growth at metastatic sites is poorly understood. A paradigm of such mutations is inactivation of SMAD4 - a transcriptional effector of TGFβ signaling - which is a hallmark of multiple gastrointestinal malignancies4,5. SMAD4 inactivation mediates TGFβ's remarkable anti- to pro-tumorigenic switch during cancer progression and can thus influence both tumor initiation and metastasis6-14. To determine whether metastatic tumors remain dependent on SMAD4 inactivation, we developed a mouse model of pancreatic ductal adenocarcinoma (PDAC) that enables Smad4 depletion in the pre-malignant pancreas and subsequent Smad4 reactivation in established metastases. As expected, Smad4 inactivation facilitated the formation of primary tumors that eventually colonized the liver and lungs. By contrast, Smad4 reactivation in metastatic disease had strikingly opposite effects depending on the tumor's organ of residence: suppression of liver metastases and promotion of lung metastases. Integrative multiomic analysis revealed organ-specific differences in the tumor cells' epigenomic state, whereby the liver and lungs harbored chromatin programs respectively dominated by the KLF and RUNX developmental transcription factors, with Klf4 depletion being sufficient to reverse Smad4's tumor-suppressive activity in liver metastases. Our results show how epigenetic states favored by the organ of residence can influence the function of driver genes in metastatic tumors. This organ-specific gene-chromatin interplay invites consideration of anatomical site in the interpretation of tumor genetics, with implications for the therapeutic targeting of metastatic disease.

scKINETICS: inference of regulatory velocity with single-cell transcriptomics data

MOTIVATION

Transcriptional dynamics are governed by the action of regulatory proteins and are fundamental to systems ranging from normal development to disease. RNA velocity methods for tracking phenotypic dynamics ignore information on the regulatory drivers of gene expression variability through time.

RESULTS

We introduce scKINETICS (Key regulatory Interaction NETwork for Inferring Cell Speed), a dynamical model of gene expression change which is fit with the simultaneous learning of per-cell transcriptional velocities and a governing gene regulatory network. Fitting is accomplished through an expectation-maximization approach designed to learn the impact of each regulator on its target genes, leveraging biologically motivated priors from epigenetic data, gene-gene coexpression, and constraints on cells' future states imposed by the phenotypic manifold. Applying this approach to an acute pancreatitis dataset recapitulates a well-studied axis of acinar-to-ductal transdifferentiation whilst proposing novel regulators of this process, including factors with previously appreciated roles in driving pancreatic tumorigenesis. In benchmarking experiments, we show that scKINETICS successfully extends and improves existing velocity approaches to generate interpretable, mechanistic models of gene regulatory dynamics.

AVAILABILITY AND IMPLEMENTATION

All python code and an accompanying Jupyter notebook with demonstrations are available at http://github.com/dpeerlab/scKINETICS.

Epigenetic plasticity cooperates with cell-cell interactions to direct pancreatic tumorigenesis

The response to tumor-initiating inflammatory and genetic insults can vary among morphologically indistinguishable cells, suggesting as yet uncharacterized roles for epigenetic plasticity during early neoplasia. To investigate the origins and impact of such plasticity, we performed single-cell analyses on normal, inflamed, premalignant, and malignant tissues in autochthonous models of pancreatic cancer. We reproducibly identified heterogeneous cell states that are primed for diverse, late-emerging neoplastic fates and linked these to chromatin remodeling at cell-cell communication loci. Using an inference approach, we revealed signaling gene modules and tissue-level cross-talk, including a neoplasia-driving feedback loop between discrete epithelial and immune cell populations that was functionally validated in mice. Our results uncover a neoplasia-specific tissue-remodeling program that may be exploited for pancreatic cancer interception.

The covariance environment defines cellular niches for spatial inference

The tsunami of new multiplexed spatial profiling technologies has opened a range of computational challenges focused on leveraging these powerful data for biological discovery. A key challenge underlying computation is a suitable representation for features of cellular niches. Here, we develop the covariance environment (COVET), a representation that can capture the rich, continuous multivariate nature of cellular niches by capturing the gene-gene covariate structure across cells in the niche, which can reflect the cell-cell communication between them. We define a principled optimal transport-based distance metric between COVET niches and develop a computationally efficient approximation to this metric that can scale to millions of cells. Using COVET to encode spatial context, we develop environmental variational inference (ENVI), a conditional variational autoencoder that jointly embeds spatial and single-cell RNA-seq data into a latent space. Two distinct decoders either impute gene expression across spatial modality, or project spatial information onto dissociated single-cell data. We show that ENVI is not only superior in the imputation of gene expression but is also able to infer spatial context to disassociated single-cell genomics data.

Abstract NG05: Multi-omic approaches to study the role of plasticity in therapy resistance and metastasis in lung cancer

Lineage plasticity, the ability of cells to transdifferentiate between committed developmental pathways, has been proposed as a source of intratumoral heterogeneity and as a mechanism for tumor adaptation to stringent environmental conditions. Lineage plasticity is increasingly recognized as a contributor to both drug resistance and metastasis, as recently highlighted by our team (Quintanal-Villalonga et al., Nat Rev Clin Oncol 2020), thus representing a biological phenomenon with high clinical relevance. Leveraging bulk and single-cell multi-omic approaches in clinical specimens, we aimed to perform a comprehensive molecular characterization of lung cancer cellular plasticity in the settings of (1) histological transdifferentiation as a mechanism of resistance to targeted therapy, and (2) disease progression. In lung adenocarcinomas (LUADs), lineage plasticity drives small cell (SCLC) and squamous cell (LUSC) transdifferentiation in the context of acquired resistance to targeted inhibitors. In the EGFR -mutant cancers, transformation to SCLC and to LUSC as a mechanism of acquired TKI resistance has been reported in 14% and 9% of cases, respectively. Transdifferentiated tumors portend poorer prognosis than non-transformed tumors. To date, no preventative therapies for transformation are available, although tumor subsets at high risk of transformation (concomitant TP53/RB1 inactivation in the case of SCLC transdifferentiation) have been identified. Defining molecular mechanisms of histological transformation in lung cancer has been challenging due in part to a paucity of well-annotated pre- and post-transformation clinical samples. We hypothesized that mixed histology tumors (LUAD/SCLC and LUAD/LUSC), containing different histological components and sharing common driver mutations, may represent an intermediate step of transdifferentiation, and an ideal substrate to study lineage plasticity, with both histological components sharing location, time, and treatment influence. By selecting mixed histology specimens amenable for clean microdissection of each histological component (10 LUAD/LUSC and 11 LUAD/SCLC), as well as pre- and post-transdifferentiated samples (N=17), we performed comprehensive genomic, epigenomic, transcriptomic, and protein characterization. Our data supports that histological transdifferentiation from LUAD to LUSC or SCLC tumors is driven by epigenetic remodeling rather than by mutational events, and indicates that transdifferentiated tumors retain epigenomic features of their previous LUAD state. Integrative epigenomic, transcriptomic, and protein analysis revealed divergent biological pathways dysregulated for each histological outcome, such as upregulation of genes involved in Hedgehog and Notch signaling and MYC targets in LUSC-transdifferentiated tumors. Most interestingly, these analyses identified commonly dysregulated pathways in both SCLC- and LUSC-transdifferentiating tumors, including remarkable downregulation of a variety of immune-related pathways and upregulation of genes involved in the PRC2 epigenetic remodeling complex and AKT signaling. To validate drivers of transdifferentiation from our multi-omic data, preclinical in vivo experiments indicated that concurrent activation of AKT and MYC overexpression induced a LUSC phenotype with increased P40 and CK5/6 expression in EGFR -mutant LUAD patient-derived models. LUSC features in these models were further accentuated by the EGFR inhibitor osimertinib, which enriched for transdifferentiated LUSC cells. With the aim to validate potential therapeutic targets to prevent transdifferentiation from our multi-omic data, we tested the efficacy of EZH2 (the catalytic subunit of PRC2) or AKT pharmacological inhibitors in combination with osimertinib in an EGFR -mutant patient-derived xenograft model of LUAD-to-LUSC transdifferentiation and observed that inhibition of either pathway dramatically delayed relapse and prevented emergence of LUSC phenotypic markers. Additionally, pharmacological targeting of AKT and EGFR delayed relapse in a mixed LUAD/SCLC patient-derived xenograft model representing an intermediate step of LUAD-to-SCLC transdifferentiation. Interestingly, AKT inhibition selectively targeted the SCLC compartment of the tumor and prevented full SCLC transformation.

This work was published in two manuscripts, of which I am the first and co-corresponding author (Quintanal-Villalonga et al., Cancer Discov 2021; Quintanal-Villalonga et al., J Hematol Oncol 2021) and defines a novel landscape of potential drivers and therapeutic vulnerabilities of histological transdifferentiation in lung cancer. Both transformed and de novo SCLCs are aggressive and rapidly metastatic lung tumors. Limited treatment options and transient responses translate to poor prognosis for patients with SCLC; 5-year survival rates are <1% for extensive disease, and SCLC accounts for >200,000 annual deaths worldwide. Metastasis is the main cause of mortality among patients with SCLC. To characterize SCLC metastasis, we combined single-cell RNA sequencing (scRNA-seq) and multiplexed ion beam imaging (MIBI) technologies to study intratumoral heterogeneity and the surrounding tumor microenvironment (TME). Efforts to apply these technologies to human SCLC tumors have been limited, as surgical resections of primary tumors are performed in <5% of patients with SCLC, and scRNA-seq processing of biopsied samples is extremely challenging. Additionally, since resection is only clinically indicated for exceptionally early stage de novo disease, these samples fail to capture the spectrum of disease progression. Through the optimization of protocols allowing single cell profiling of difficult samples such as small tissue biopsies, pleural effusions, and fine needle aspirations, along with larger volume resections, we constructed a single-cell atlas of SCLC patient tumors, with 155,098 transcriptomes, including 54,523 transcriptomes from 21 SCLC clinical specimens. Despite substantial heterogeneity among SCLC tumors, we detected a minor cell subpopulation that was shared among tumors across subtypes, treatments, and tissue locations, pointing to a potentially universal characteristic of this malignancy. This subpopulation demonstrated a pro-metastatic, highly plastic (stem-like) phenotype and exhibited profound PLCG2 overexpression. Direct genetic manipulation validated that PLCG2 expression promotes metastatic features and induced plasticity in vitro and in vivo. Consistently, we found that higher representation of this subpopulation in clinical samples, as well as PLCG2 expression itself, are strong predictors of shorter overall survival in patients with SCLC. Additionally, we found that SCLC is enriched for a profibrotic, immunosuppressive monocyte/macrophage population associated with the recurrent pro-metastatic PLCG2-high subpopulation, with potential implications in the metastatic process. I am the co-first author of this work published in Cancer Cell (Chan JM*, Quintanal-Villalonga* et al., Cancer Cell 2021) defining a novel mechanism of plasticity-mediated metastasis in SCLC. These works highlight the key role of plasticity in disease progression and therapy resistance in lung cancer, and describe molecular events occurring during fate reprogramming, thus nominating potential drivers and therapeutic vulnerabilities in plasticity-driven clinically relevant biological processes.

Citation Format: Alvaro Quintanal-Villalonga, Joseph M. Chan, Vianne R. Gao, Yubin Xie, Dana Pe’er, Charles M. Rudin. Multi-omic approaches to study the role of plasticity in therapy resistance and metastasis in lung cancer. [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr NG05.

Single-cell analysis of bidirectional reprogramming between early embryonic states reveals mechanisms of differential lineage plasticities

Two distinct fates, pluripotent epiblast (EPI) and primitive (extra-embryonic) endoderm (PrE), arise from common progenitor cells, the inner cell mass (ICM), in mammalian embryos. To study how these sister identities are forged, we leveraged embryonic (ES) and eXtraembryonic ENdoderm (XEN) stem cells - in vitro counterparts of the EPI and PrE. Bidirectional reprogramming between ES and XEN coupled with single-cell RNA and ATAC-seq analyses uncovered distinct rates, efficiencies and trajectories of state conversions, identifying drivers and roadblocks of reciprocal conversions. While GATA4-mediated ES-to-iXEN conversion was rapid and nearly deterministic, OCT4, KLF4 and SOX2-induced XEN-to-iPS reprogramming progressed with diminished efficiency and kinetics. The dominant PrE transcriptional program, safeguarded by Gata4, and globally elevated chromatin accessibility of EPI underscored the differential plasticities of the two states. Mapping in vitro trajectories to embryos revealed reprogramming in either direction tracked along, and toggled between, EPI and PrE in vivo states without transitioning through the ICM.

Systematic Comparison of Pancreatic Ductal Adenocarcinoma Models Identifies a Conserved Highly Plastic Basal Cell State

Intratumoral heterogeneity and cellular plasticity have emerged as hallmarks of cancer, including pancreatic ductal adenocarcinoma (PDAC). As PDAC portends a dire prognosis, a better understanding of the mechanisms underpinning cellular diversity in PDAC is crucial. Here, we investigated the cellular heterogeneity of PDAC cancer cells across a range of in vitro and in vivo growth conditions using single-cell genomics. Heterogeneity contracted significantly in two-dimensional and three-dimensional cell culture models but was restored upon orthotopic transplantation. Orthotopic transplants reproducibly acquired cell states identified in autochthonous PDAC tumors, including a basal state exhibiting coexpression and coaccessibility of epithelial and mesenchymal genes. Lineage tracing combined with single-cell transcriptomics revealed that basal cells display high plasticity in situ. This work defines the impact of cellular growth conditions on phenotypic diversity and uncovers a highly plastic cell state with the capacity to facilitate state transitions and promote intratumoral heterogeneity in PDAC.

SIGNIFICANCE

This work provides important insights into how different model systems of pancreatic ductal adenocarcinoma mold the phenotypic space of cancer cells, highlighting the power of in vivo models.

Lineage plasticity in prostate cancer depends on JAK/STAT inflammatory signaling

Drug resistance in cancer is often linked to changes in tumor cell state or lineage, but the molecular mechanisms driving this plasticity remain unclear. Using murine organoid and genetically engineered mouse models, we investigated the causes of lineage plasticity in prostate cancer and its relationship to antiandrogen resistance. We found that plasticity initiates in an epithelial population defined by mixed luminal-basal phenotype and that it depends on increased Janus kinase (JAK) and fibroblast growth factor receptor (FGFR) activity. Organoid cultures from patients with castration-resistant disease harboring mixed-lineage cells reproduce the dependency observed in mice by up-regulating luminal gene expression upon JAK and FGFR inhibitor treatment. Single-cell analysis confirms the presence of mixed-lineage cells with increased JAK/STAT (signal transducer and activator of transcription) and FGFR signaling in a subset of patients with metastatic disease, with implications for stratifying patients for clinical trials.

Abstract 279: Optimal tumor metastatic gene programs in pancreatic cancer

Metastatic pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of cancer-related deaths with very few treatment options and low-success rates. Generally speaking metastatic PDAC remains incurable. One reason behind PDAC morbidity and mortality is intratumor heterogeneity which allows metastatic dispersion and treatment resistance. We hypothesize that given multiple selective bottlenecks during metastatic progression, only a set of non-random gene programs are required for tumor cells to metastasize to different organs. To identify gene programs that become selected during PDAC metastatic progression we collected multiple (>20) metastatic tissue samples from different local and distant organs, from two clinically non-redundant human rapid-autopsies as part of the Human Tumor Atlas Network. We generated single-nuclei RNA-seq and Multiplexed Ion Beam Imaging data from these samples. Using trajectory analysis we inferred gene program dynamics between primary and metastatic samples and found an epithelial-to-mesenchymal-to-epithelial axis general to the metastases, suggesting that epithelial-mesenchymal plasticity is needed for cells to colonize other tissues. However, different organs have vastly different cell type compositions and may represent significantly different evolutionary bottlenecks. We employed archetype analysis as a tool to distinguish optimized phenotypes that may be shared or unique across colonized tissues. Archetype analysis revealed multiple gene programs, some of which are specific to a particular tissue and others ubiquitous. An epithelial-to-mesenchymal gene program was found to be shared across all samples at varying proportions, together with an extracellular matrix deposition/interaction program. Other gene programs identified include angiogenesis, hypoxia, cell cycle, immune interaction, lipid metabolism, autophagy/stress response, and cell migration. Some of these programs are present in various different metastases while others are unique to a specific site (e.g. Lipid metabolism in peritoneal metastasis). Together these programs shed light into organ tropism, metastatic modes of spread, adaptation to local tumor microenvironments, and cell-cell interactions with stromal cells. Further validation of these optimized phenotypes and integration of their spatial context will provide a deeper molecular understanding of metastatic PDAC and provide a source for data-driven therapeutic targets.

[A. J-S. and Y.X contributed equally to this work.]

Citation Format: Alejandro Jiménez-Sánchez, Yubin Xie, Roshan Sharma, Tin Yi Chu, Vincent Liu, Wungki Park, Akimasa Hayashi, Shigeaki Umeda, Linas Mazutis, Tal Nawy, Christine Iacobuzio-Donahue, Dana Pe’er. Optimal tumor metastatic gene programs in pancreatic cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 279.

Radiotherapy orchestrates natural killer cell dependent antitumor immune responses through CXCL8

Radiotherapy is a mainstay cancer therapy whose antitumor effects partially depend on T cell responses. However, the role of Natural Killer (NK) cells in radiotherapy remains unclear. Here, using a reverse translational approach, we show a central role of NK cells in the radiation-induced immune response involving a CXCL8/IL-8-dependent mechanism. In a randomized controlled pancreatic cancer trial, CXCL8 increased under radiotherapy, and NK cell positively correlated with prolonged overall survival. Accordingly, NK cells preferentially infiltrated irradiated pancreatic tumors and exhibited CD56^dim-like cytotoxic transcriptomic states. In experimental models, NF-κB and mTOR orchestrated radiation-induced CXCL8 secretion from tumor cells with senescence features causing directional migration of CD56^dim NK cells, thus linking senescence-associated CXCL8 release to innate immune surveillance of human tumors. Moreover, combined high-dose radiotherapy and adoptive NK cell transfer improved tumor control over monotherapies in xenografted mice, suggesting NK cells combined with radiotherapy as a rational cancer treatment strategy.

LMD-16. Choroid plexus orchestrates anti-cancer immunity in leptomeninges

Choroid plexus (CP) forms an anatomically functional barrier between the blood and cerebrospinal fluid (CSF) that dictates the cellular and humoral composition of the CSF. The immunological response of CP to inflammatory stimuli, such as cancer, remains unclear. Here, we find that CP orchestrates the immune composition of CSF in the steady state as well as in the presence of metastatic cancer. We show that the circulation-derived leptomeningeal monocyte-macrophages entering the CSF through CP promote the growth of leptomeningeal metastasis (LM) by perturbing the environment with a storm of dozens of pro- and anti-inflammatory cytokines. Functional manipulation of Type II Interferon pathway specifically within inflamed leptomeninges revealed that IFN-γ can serve as a dominant signal, further recruiting peripheral myeloid cells and activating their protective anti-tumoral response. This preclinical strategy was sufficient to controll the growth of syngeneic LM cancer cells and delay the onset of lethal LM.

Integrated Single-Cell Atlas of Endothelial Cells of the Human Lung

BACKGROUND

Cellular diversity of the lung endothelium has not been systematically characterized in humans. We provide a reference atlas of human lung endothelial cells (ECs) to facilitate a better understanding of the phenotypic diversity and composition of cells comprising the lung endothelium.

METHODS

We reprocessed human control single-cell RNA sequencing (scRNAseq) data from 6 datasets. EC populations were characterized through iterative clustering with subsequent differential expression analysis. Marker genes were validated by fluorescent microscopy and in situ hybridization. scRNAseq of primary lung ECs cultured in vitro was performed. The signaling network between different lung cell types was studied. For cross-species analysis or disease relevance, we applied the same methods to scRNAseq data obtained from mouse lungs or from human lungs with pulmonary hypertension.

RESULTS

Six lung scRNAseq datasets were reanalyzed and annotated to identify >15 000 vascular EC cells from 73 individuals. Differential expression analysis of EC revealed signatures corresponding to endothelial lineage, including panendothelial, panvascular, and subpopulation-specific marker gene sets. Beyond the broad cellular categories of lymphatic, capillary, arterial, and venous ECs, we found previously indistinguishable subpopulations; among venous EC, we identified 2 previously indistinguishable populations: pulmonary-venous ECs (COL15A1neg) localized to the lung parenchyma and systemic-venous ECs (COL15A1pos) localized to the airways and the visceral pleura; among capillary ECs, we confirmed their subclassification into recently discovered aerocytes characterized by EDNRB, SOSTDC1, and TBX2 and general capillary EC. We confirmed that all 6 endothelial cell types, including the systemic-venous ECs and aerocytes, are present in mice and identified endothelial marker genes conserved in humans and mice. Ligand-receptor connectome analysis revealed important homeostatic crosstalk of EC with other lung resident cell types. scRNAseq of commercially available primary lung ECs demonstrated a loss of their native lung phenotype in culture. scRNAseq revealed that endothelial diversity is maintained in pulmonary hypertension. Our article is accompanied by an online data mining tool (www.LungEndothelialCellAtlas.com).

CONCLUSIONS

Our integrated analysis provides a comprehensive and well-crafted reference atlas of ECs in the normal lung and confirms and describes in detail previously unrecognized endothelial populations across a large number of humans and mice.

ETMM-03. CANCER CELLS DEPLOY LIPOCALIN- 2 TO COLLECT LIMITING IRON IN LEPTOMENINGEAL METASTASIS

The tumor microenvironment plays a critical regulatory role in cancer progression, especially in central nervous system metastases. Cancer cells within the cerebrospinal fluid (CSF)-filled leptomeninges face substantial microenvironmental challenges, including inflammation and sparse micronutrients. To investigate the mechanism by which cancer cells in these leptomeningeal metastases (LM) overcome these constraints, we subjected CSF from five patients with LM to single-cell RNA sequencing. We found that cancer cells, but not macrophages, within the CSF express the iron-binding protein lipocalin-2 (LCN2) and its receptor SCL22A17. These macrophages generate inflammatory cytokines that induce cancer cell LCN2 expression but do not generate LCN2 themselves. In mouse models of LM, cancer cell growth is supported by the LCN2/SLC22A17 system and is inhibited by iron chelation therapy. A Phase Ia/1b clinical trial focused on this novel treatment approach is underway.

Fully defined human pluripotent stem cell-derived microglia and tri-culture system model C3 production in Alzheimer's disease

Aberrant inflammation in the CNS has been implicated as a major player in the pathogenesis of human neurodegenerative disease. We developed a new approach to derive microglia from human pluripotent stem cells (hPSCs) and built a defined hPSC-derived tri-culture system containing pure populations of hPSC-derived microglia, astrocytes, and neurons to dissect cellular cross-talk along the neuroinflammatory axis in vitro. We used the tri-culture system to model neuroinflammation in Alzheimer's disease with hPSCs harboring the APP^SWE+/+ mutation and their isogenic control. We found that complement C3, a protein that is increased under inflammatory conditions and implicated in synaptic loss, is potentiated in tri-culture and further enhanced in APP^SWE+/+ tri-cultures due to microglia initiating reciprocal signaling with astrocytes to produce excess C3. Our study defines the major cellular players contributing to increased C3 in Alzheimer's disease and presents a broadly applicable platform to study neuroinflammation in human disease.

NCMP-05. DECODING THE IMMUNE SYSTEM RESPONSE TO LEPTOMENINGEAL METASTASIS

For decades, the central nervous system was considered to be an immune privileged organ with limited access to systemic immunity. However, the leptomeninges, the cerebrospinal fluid (CSF)-filled anatomical structure that protects the brain and spinal cord, represent a relatively immune-rich environment. Despite the presence of immune cells, complications in the CSF, such as infectious meningitis and a neurological development of cancer known as leptomeningeal metastasis, are difficult to treat and are frequently fatal. We show that immune cells entering the CSF are held in an ‘idle’ state that limits their cytotoxic arsenal and antigen presentation machinery. To understand this underappreciated neuroanatomic niche, we used unique mouse models and rare patient samples to characterize its cellular composition and critical signaling events in health and disease at a single-cell resolution. Revealing the mediators of CSF immune response will allow us to re-evaluate current therapeutic protocols and employ rational combinations with immunotherapies, therefore turning the patient’s own immune system into an active weapon against pathogens and cancer.

Abstract PO-134: Identification of the cells of origin and tumor heterogeneity in neuroendocrine prostate cancer (NEPC) by single-cell analysis

Lineage plasticity has emerged as an important mechanism of therapeutic resistance in prostate cancer treated with a newer generation of AR signaling inhibitor (ARSi). As a consequence of lineage plasticity, treatment resistant prostate cancer loses epithelial cell identity, acquires stem cell-like properties and transforms into a neuroendocrine lineage. Loss of function mutations in Tp53, Rb1 and Pten are enriched in human neuroendocrine prostate cancer (NEPC). To understand the cell of origin and the molecular mechanism underlying lineage plasticity and NEPC development, we have used single-cell RNA-sequencing (scRNA-seq) technology to profile a genetically engineered mouse model with probasin-Cre driven Tp53, Rb1 and Pten deletions (referred to as TKO mouse) and Rb1 and Pten deletions (referred to as DKO mouse). We have profiled ~70,000 single cells from 16 mice of various ages and 6 additional mice that have undergone castration with or without testosterone addback. Our scRNA-seq analysis captures a developmental trajectory from luminal adenocarcinoma to neuroendocrine tumor, suggesting a newly discovered luminal cell type (L2) within the normal prostate may be the preferred cell of origin for NEPC development. Furthermore, combining scRNA-seq, flow cytometry and immunofluorescence analysis, we reveal tremendous heterogeneity within the neuroendocrine tumors with differential expression of several putative drivers, including SOX2, EZH2, AURKA, NMYC and POU2F3. Many of these genes have been previously implicated as drivers of NEPC development, while POU2F3 has been identified as a defining marker for a Tuft-variant small cell lung cancer (SCLC). Lastly, we have preliminary evidence that castration may have the potential to accelerate the transition from adenocarcinoma to NEPC, while adding back testosterone may delay the process in the TKO mice. Overall, in this study, we have identified a luminal L2 population as the putative preferred cell of origin for NEPC in the TKO model and revealed a previous under-appreciated heterogeneity within NEPC with differential expression of driver transcriptional, epigenetic and cell cycle regulators, which mirrors what has been described in the SCLC field.

Citation Format: Jimmy L. Zhao, Samir Zaidi, Joseph Chan, Kristine Wadosky, Wouter Karthaus, Danielle Choi, Aura Agudelo Rivera, Anuradha Gopalan, Dana Rathkopf, Brett Carver, Wassim Abida, Howard Scher, Yu Chen, David Goodrich, Dana Pe’er, Charles L. Sawyers. Identification of the cells of origin and tumor heterogeneity in neuroendocrine prostate cancer (NEPC) by single-cell analysis [abstract]. In: Proceedings of the AACR Virtual Special Conference on Tumor Heterogeneity: From Single Cells to Clinical Impact; 2020 Sep 17-18. Philadelphia (PA): AACR; Cancer Res 2020;80(21 Suppl):Abstract nr PO-134.

Learning time-varying information flow from single-cell epithelial to mesenchymal transition data

Cellular regulatory networks are not static, but continuously reconfigure in response to stimuli via alterations in protein abundance and confirmation. However, typical computational approaches treat them as static interaction networks derived from a single time point. Here, we provide methods for learning the dynamic modulation of relationships between proteins from static single-cell data. We demonstrate our approach using TGFß induced epithelial-to-mesenchymal transition (EMT) in murine breast cancer cell line, profiled with mass cytometry. We take advantage of the asynchronous rate of transition to EMT in the data and derive a pseudotime EMT trajectory. We propose methods for visualizing and quantifying time-varying edge behavior over the trajectory, and a metric of edge dynamism to predict the effect of drug perturbations on EMT.

Abstract IA14: An atlas of the tumor immune ecosystem

We created a comprehensive atlas of breast tumor-immune ecosystem by collecting a dataset of over 50,000 single immune cell transcriptomes from different types of human breast cancer, normal breast tissue, peripheral blood, and the lymph node using single-cell RNA-seq, and analyzed with novel computational clustering and normalization techniques. Our analyses revealed remarkably increased heterogeneity of intratumoral cells of both lymphoid and myeloid cell lineages, which occupy significantly expanded contiguous phenotypic space in comparison to normal breast tissue. The observed continuum of cell states is associated with their progressive cellular activation and differentiation and argues strongly against the notion of few discrete states of differentiation or activation of individual cell types shaping the tumor microenvironment. The substantial heterogeneity of cell states within major cell types and across patients is also driven by complex coexpression relationships between genes, and by extension their potential suitability as therapeutic cotargets.

Citation Format: Dana Pe’er. An atlas of the tumor immune ecosystem [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2017 Oct 1-4; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2018;6(9 Suppl):Abstract nr IA14.

PhenoGraph and viSNE facilitate the identification of abnormal T-cell populations in routine clinical flow cytometric data

BACKGROUND

Flow cytometric identification of neoplastic T-cell populations is complicated by the wide range of phenotypic abnormalities in T-cell neoplasia, and the diverse repertoire of reactive T-cell phenotypes. We evaluated whether a recently described clustering algorithm, PhenoGraph, and dimensionality-reduction algorithm, viSNE, might facilitate the identification of abnormal T-cell populations in routine clinical flow cytometric data.

METHODS

We applied PhenoGraph and viSNE to peripheral blood mononuclear cells labeled with a single 8-color T/NK-cell antibody combination. Individual peripheral blood samples containing either a T-cell neoplasm or reactive lymphocytosis were analyzed together with a cohort of 10 normal samples, which established the location and identity of normal mononuclear-cell subsets in viSNE displays.

RESULTS

PhenoGraph-derived subpopulations from the normal samples formed regions of phenotypic similarity in the viSNE display describing normal mononuclear-cell subsets, which correlated with those obtained by manual gating (r²  = 0.99, P < 0.0001). In 24 of 24 cases of T-cell neoplasia with an aberrant phenotype, compared with 4 of 17 cases of reactive lymphocytosis (P = 1.4 × 10^-7 , Fisher Exact test), PhenoGraph-derived subpopulations originating exclusively from the abnormal sample formed one or more distinct phenotypic regions in the viSNE display, which represented the neoplastic T cells, and reactive T-cell subpopulations not present in the normal cohort, respectively. The numbers of neoplastic T cells identified using PhenoGraph/viSNE correlated with those obtained by manual gating (r²  = 0.99; P < 0.0001).

CONCLUSIONS

PhenoGraph and viSNE may facilitate the identification of abnormal T-cell populations in routine clinical flow cytometric data. © 2017 Clinical Cytometry Society.

Abstract A54: A cellular and molecular atlas of the human breast for dissecting mechanisms of cell and tissue function

A fundamental question in biology, central to our understanding of cancer, is determining how cells coordinate and cooperate to form and maintain tissues. Developing systems-wide knowledge of the numerous cell interactions within a tissue requires robust identification of the cell types it comprises. It demands comprehensive information about each of these cell types, including the genes they express and proteins they make. Unfortunately, the cellular composition and arrangement of cells in many complex tissue systems has yet to be fully resolved, and this includes the breast. This has resulted in a crucial knowledge gap where we know very little about the elaborate biology of how different normal cell types interact, and the consequences these interactions have on each other and the tissue as a whole. Using immunofluorescence and advanced flow cytometry, we have recently developed methods to resolve and isolate every known cell type in the normal human breast. This includes several different luminal and epithelial fractions, myoepithelial cells, adipocytes, leukocytes, pericytes, erythrocytes, adipose-derived mesenchymal stem cells, vascular smooth muscle cells, and both lymphatic and vascular endothelial cells—12 different types in all. Successful modification of this multiparameter FACS procedure allowed purification of enough cells and RNA to perform next-generation sequencing (NGS) from even the rarest of populations at high depth. Here, we present global transcriptome analysis of these different cell types. These data have clarified cell-type differences and revealed new insights, including the identification of genes and gene families contributing to the unique phenotype of each cell type and the delineation of lineage-specific marker expression. Our analysis has also exposed how the cell types each contribute to their individual microenvironments through production of extracellular matrix proteins and other secreted factors. Moreover, identification of cognate ligand and receptor pairs expressed by these cells has unveiled an elaborate paracrine and autocrine signaling network that has implications towards interpreting the biologic interactome underlying tissue homeostasis. Future studies are aimed at establishing coculture models and determining the biochemical processes essential for cell-coordination and tissue maintenance, and developing knowledge of how these processes go awry in breast cancer.

Citation Format: Rosalyn W. Sayaman, Ambrose Carr, Kate Thi, Zhenmao Wan, Dana Pe’er, Mina J. Bissell, Curt Hines. A cellular and molecular atlas of the human breast for dissecting mechanisms of cell and tissue function [abstract]. In: Proceedings of the AACR Special Conference: Advances in Breast Cancer Research; 2017 Oct 7-10; Hollywood, CA. Philadelphia (PA): AACR; Mol Cancer Res 2018;16(8_Suppl):Abstract nr A54.

Negative costimulation constrains T cell differentiation

T cell costimulation is a principal mechanism by which T cell activation is regulated, but it remains unclear whether costimulatory pathways also control T cell differentiation. This unresolved question is germane to our understanding of how T cell function is controlled and has therapeutic implications in the context of cancer immunotherapy. Using a mass cytometry based systems approach and murine models we investigated whether negative costimulation regulates T cell differentiation as well as activation. Unsupervised population identification reveals 30 T cell subsets in Ctla-4−/− and littermate control mice including 6 CD8, 3 Treg, 14 CD4 effector T cell populations. Expectedly, loss of CTLA-4 led to dramatic changes in the relative frequency of CD8 and CD4 T cell subsets. Strikingly, however, loss of CTLA-4 led to the generation of multiple non-canonical T cell populations outside of the normal boundaries of T cell phenotypes, which interestingly, were all in the CD4 T compartment. To determine if the role of negative costimulation in differentiation may be a generalizable phenomenon, we performed similar analyses of Pdcd-1−/− mice. This profiling revealed that loss of PD-1 leads to a subtle expansion of CD8 T cell phenotypes in addition to changes in the relative frequencies of CD4 and CD8 T cell populations. These observations indicate that CTLA-4 restricts CD4 T cell phenotypes while PD-1 restricts CD8 T cell phenotypes. Consistent with this model, loss of CTLA-4 differentially affects the expression of individual markers in CD4 and CD8 T cells. Together these findings indicate that in addition to attenuating T cell activation, negative costimulation acts to constrain T cell differentiation.

Chromosomal instability drives metastasis through a cytosolic DNA response

Chromosomal instability is a hallmark of cancer that results from ongoing errors in chromosome segregation during mitosis. Although chromosomal instability is a major driver of tumour evolution, its role in metastasis has not been established. Here we show that chromosomal instability promotes metastasis by sustaining a tumour cell-autonomous response to cytosolic DNA. Errors in chromosome segregation create a preponderance of micronuclei whose rupture spills genomic DNA into the cytosol. This leads to the activation of the cGAS-STING (cyclic GMP-AMP synthase-stimulator of interferon genes) cytosolic DNA-sensing pathway and downstream noncanonical NF-κB signalling. Genetic suppression of chromosomal instability markedly delays metastasis even in highly aneuploid tumour models, whereas continuous chromosome segregation errors promote cellular invasion and metastasis in a STING-dependent manner. By subverting lethal epithelial responses to cytosolic DNA, chromosomally unstable tumour cells co-opt chronic activation of innate immune pathways to spread to distant organs.

Multiparameter mass cytometry of T cell development and selection

T cell development and selection in the thymus play key roles in shaping the adaptive immune system and maintaining self-tolerance. Previous studies characterizing thymic development have largely relied on genetic perturbations and subsequent cell sorting that invariably perturb specific developmental compartments. Using mass-cytometry, allow us to capture the entire developmental progression of T cells, from the double-negative (DN) stages, through double-positive (DP) and the CD4 and CD8 branches, without perturbation. With 42 channels simultaneously measured, we collected a mass cytometry dataset profiling the mouse thymus with T-cell surface markers and transcription factors, chosen based on their broad functionality in T cell development. Using the newly developed “Wishbone” algorithm, we recovered the known stages in T cell development with high accuracy and developmental resolution. This allowed us to place DN, DP, CD4+ and CD8+ cells from a single thymus along a unified bifurcating trajectory. Our data allows for precise ordering of multiple events along the trajectory using un-sorted cells from mice that were not perturbed genetically. To characterize the selection process along the development trajectory we used Nur77-GFP mice in which GFP is induced by TCR signaling that occurs during positive and negative selection. By following GFP strength together with cell trajectory we were able to recognize and characterize the behavior of proteins that participate in selection and that are involved in the transition from the DP to SP stage. This approach enables us for the first time to place and follow coordinated changes in multiple markers that take part in T cell development and selection in the thymus using single snapshots

Detection of minimal residual disease in B lymphoblastic leukemia using viSNE

BACKGROUND

Minimal residual disease (MRD) following treatment is a robust prognostic marker in B lymphoblastic leukemia. However, the detection of MRD by flow cytometric immunophenotyping is technically challenging, and an automated method to detect MRD is therefore desirable. viSNE, a recently developed computational tool based on the t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm, has been shown to be capable of detecting synthetic "MRD-like" populations of leukemic cells created in vitro, but whether viSNE can facilitate the immunophenotypic detection of MRD in clinical samples has not been evaluated.

METHODS

We applied viSNE retrospectively to 8-color flow cytometric immunophenotyping data from normal bone marrow samples, and samples from B lymphoblastic leukemia patients with or without suspected MRD on the basis of conventional manual gating.

RESULTS

In each of 14 bone marrow specimens containing MRD or suspected MRD, viSNE identified a putative MRD population; an abnormal composite immunophenotype was confirmed for the putative MRD in each case. MRD populations were not identified by viSNE in control bone marrow samples from patients with increased normal B-cell precursors, or in post-treatment samples from B lymphoblastic leukemia patients who did not have detectable MRD by manual gating.

CONCLUSION

viSNE shows promise as an automated method to facilitate immunophenotypic MRD detection in patients treated for B lymphoblastic leukemia.

Context Sensitive Modeling of Cancer Drug Sensitivity

Recent screening of drug sensitivity in large panels of cancer cell lines provides a valuable resource towards developing algorithms that predict drug response. Since more samples provide increased statistical power, most approaches to prediction of drug sensitivity pool multiple cancer types together without distinction. However, pan-cancer results can be misleading due to the confounding effects of tissues or cancer subtypes. On the other hand, independent analysis for each cancer-type is hampered by small sample size. To balance this trade-off, we present CHER (Contextual Heterogeneity Enabled Regression), an algorithm that builds predictive models for drug sensitivity by selecting predictive genomic features and deciding which ones should-and should not-be shared across different cancers, tissues and drugs. CHER provides significantly more accurate models of drug sensitivity than comparable elastic-net-based models. Moreover, CHER provides better insight into the underlying biological processes by finding a sparse set of shared and type-specific genomic features.

Broadening horizons: holistic viewpoints from the Biology of Genomes

A report on the Cold Spring Harbor Laboratory 27th annual meeting on the Biology of Genomes, held in Cold Spring Harbor, New York, USA, 6-10 May 2014.