Deciphering intratumoral heterogeneity using integrated clonal tracking and single-cell transcriptome analyses

Abstract

Cellular heterogeneity is a major cause of treatment resistance in cancer. Despite recent advances in single-cell genomic and transcriptomic sequencing, it remains difficult to relate measured molecular profiles to the cellular activities underlying cancer. Here, we present an integrated experimental system that connects single cell gene expression to heterogeneous cancer cell growth, metastasis, and treatment response. Our system integrates single cell transcriptome profiling with DNA barcode based clonal tracking in patient-derived xenograft models. We show that leukemia cells exhibiting unique gene expression respond to different chemotherapies in distinct but consistent manners across multiple mice. In addition, we uncover a form of leukemia expansion that is spatially confined to the bone marrow of single anatomical sites and driven by cells with distinct gene expression. Our integrated experimental system can interrogate the molecular and cellular basis of the intratumoral heterogeneity underlying disease progression and treatment resistance.

Introduction

Cancer is a dynamic disease driven by continuous genetic and epigenetic changes^1,2,3,4. The accumulation of these molecular alterations generates tremendous intratumoral heterogeneity^1,2,3,5,6. Consequently, individual cancer cells differentially proliferate, selectively metastasize, and sporadically escape therapeutic treatment^1,2,3,5,6. Cellular heterogeneity has arisen as a major hurdle in cancer treatment^7,8,9. Identifying the genes underlying the heterogeneous behaviors of individual cancer cells is critically important to improving the efficacy of cancer treatment.

Recent advances in single-cell genomic and transcriptomic sequencing have greatly improved the detection of intratumoral heterogeneity at the molecular level^10,11,12. However, it remains difficult to relate the molecular profiles generated by these technologies to the cellular behaviors underlying disease progression and relapse. Some studies have managed to link a few genes using massive in-depth sequencing to identify naturally occurring genetic mutations that can be used to trace cell clones^13,14,15,16. Although clonal tracking through natural mutations is a powerful technique that does not require any invasive maniputation^{13,14,15,17,18,19,20}, it suffers from a prohibitive cost, low efficiency, and complications in clonal comparison that arise from the rarity and randomness of natural mutations. As an alternative to tracking natural mutations, some studies have employed synthetic DNA barcodes to simultaneously mark individual cancer cells^21,22,23,24. Two recent studies have also demonstrated the feasibility of integrating synthetic DNA barcode tracking with single-cell mRNA sequencing analyses^25,26. In conjunction with patient-derived xenograft (PDX) models, the use of synthetic DNA barcoding has greatly improved our understanding of the heterogeneous growth and metastasis of cancer cells^22,23,27,28.

In this study, we present an integrated experimental system that directly connects gene expression with cellular behavior at the single-cell level by combining synthetic DNA barcode tracking and single-cell mRNA sequencing in a PDX model. We have previously demonstrated the high sensitivity and precise quantification of our DNA barcode tracking using hematopoietic stem cells in vivo^{29,30,31,32,33}. Here, we adapt the barcode tracking to assay the activities of cancer cells and their gene expression profiles simultaneously in a PDX model xenografted by human B-cell acute lymphoblastic leukemia (B-ALL) samples. In the PDX model, B-ALL cells can autonomously home to their native micro-environment in the bone marrow and can be easily and repeatedly sampled over time to monitor cancer progression and therapeutic response. Using this integrated system, we show that primary B-ALL clones exhibit heterogeneous dynamics during expansion, circulation, and response to chemotherapy. Furthermore, their distinct temporal and spatial clonal dynamics are associated with unique gene expression.

Results

Integrating single-cell transcriptome and clonal tracking

Primary B-ALL cells were genetically barcoded using a GFP-encoding lentiviral vector (Fig. 1a and Supplementary Table 1)^29,30. After barcode labeling, they were transplanted into sub-lethally irradiated NSG or NSG-SGM3 mice. Subsequent clonal tracking assays were performed as previously described^29,30. Our data show that the barcoded leukemia cells expanded proportionally to non-barcoded leukemia cells in recipient mice (Fig. 1b and Supplementary Fig. 1), suggesting that barcoded cells are representative of the engrafted leukemic cell population in terms of their progression in mice. As DNA barcodes are inserted into cellular genomes, they are inherited by all descendants of barcoded cells, allowing us to track cellular proliferation and elimination. In this study, a “clone” refers to cells carrying identical barcodes.