Clonal tracking in cancer and metastasis

Abstract

The eradication of many cancers has proven challenging due to the presence of functionally and genetically heterogeneous clones maintained by rare cancer stem cells (CSCs), which contribute to disease progression, treatment refractoriness, and late relapse. The characterization of functional CSC activity has necessitated the development of modern clonal tracking strategies. This review describes viral-based and CRISPR-Cas9-based cellular barcoding, lineage tracing, and imaging-based approaches. DNA-based cellular barcoding technology is emerging as a powerful and robust strategy that has been widely applied to in vitro and in vivo model systems, including patient-derived xenograft models. This review also highlights the potential of these methods for use in the clinical and drug discovery contexts and discusses the important insights gained from such approaches.

Keywords: Cancer stem cells; Cellular barcoding; Clonal dynamics; Clonal tracking.

Fig. 1

Depiction of a functionally defined clone versus a genomic clone. Heterogeneous tumors are depicted at the top, with each color representing a different clone either defined based on cell-surface phenotypes (left) or constellation of genomic mutations (right). Clones that are defined functionally can only be assessed retrospectively based on their ability to propagate in a given assay. These clones are generally thought to be composed of two cell types, CSCs (indicated with a curved arrow), and non-CSCs (at the bottom of the hierarchy), where CSCs possess self-renewal activity and non-CSCs have limited or no proliferative activity. At the bottom right is a depiction of genomic clonal evolution over time, resulting in multiple genomic sub-clones through the acquisition of distinct mutations, all having originated from a common ancestor. Created with BioRender.com

Fig. 2

Schematic of a general preclinical platform to characterize CSC heterogeneity and response to therapy. A lentiviral DNA barcode library is constructed from plasmids each encoding distinct semi-random oligonucleotide sequences. Single cells from patient-derived tumor xenografts or cell lines are uniquely labeled with the lentiviral DNA barcode library (with each color representing a uniquely barcoded cell). These cells are then transplanted into an appropriate animal model and receive experimental therapies. Genomic DNA is then isolated from the tumors and metastatic deposits and sequenced with NGS. The resulting data is then analyzed for barcode composition, to interpret the clonal representation in primary tumors, relapsed tumors, and metastatic lesions. Comparison of barcode clones detected at different sites and at different time points will reveal how various CSC clones contribute to disease progression and metastatic spread. Machine learning can be employed to predict the dissemination pattern of an individual clone [62], Created with BioRender.com