Associate Professor University of Texas at Austin, United States
Introduction: : Triple negative breast cancer (TBNC) is marked by fewer standard-of-care treatment options and poorer treatment outcomes than other breast cancer subtypes, with approximately 40% of TNBC patients developing treatment resistance to standard-of-care cytotoxic chemotherapy. Recently, small molecule inhibitors of canonical cell survival pathways, known as targeted therapies, have demonstrated promising improvements in patient survival and reduction of disease progression in HR+ and Her2+ breast cancer. However, these treatments often show little to no improvements over standard-of-care therapy in TNBC. High intratumoral heterogeneity, a characteristic of TNBC, leads to its difficulty in treatment and rapid acquisition of resistance. In this study, the use of DNA barcoding technologies and single-cell RNA sequencing (scRNA-seq) reveals how transcriptomic heterogeneity in TNBC influences treatment response and resistance to targeted therapy.
Materials and
Methods: : ClonMapper is a DNA barcoding technology developed previously by our lab that utilizes integrated and heritable unique 20-nucleotide DNA barcodes to follow clonal cell populations across treatment. ClonMapper barcodes are expressed as polyadenylated RNA transcripts and are identifiable in scRNA-seq, which enables the assignment of cells to a clonal identity and allows for an analysis of the differences in transcriptomic state and trajectory between clonal subpopulations. To investigate heterogeneity in targeted therapy treatment, three patient-derived TNBC cell populations were integrated with ClonMapper barcodes such that each cell received a unique genomically-integrated barcode, which facilitates the downstream identification of that cell and its progeny. Barcoded populations were treated with three different clinically-relevant targeted therapies at a bottleneck dose and allowed to recover to their original cell number before being collected for analysis. Barcode regions were amplified from genomic DNA and sequenced using next-generation sequencing to approximate the number of cells that contain each unique barcode within each sample. By measuring differences in barcode abundance in clonal subpopulations between control and treated samples, the number of cells that belong to each clone before and after treatment can be estimated to calculate changes in abundance during treatment. Alongside clonal abundance measurements, scRNA-seq of treated and control cell populations was used to reveal differences in RNA expression between subpopulations and their unique transcriptomic trajectories across treatment.
Results, Conclusions, and Discussions:: The probability of survival for a cell in a clonal subpopulation was calculated as the difference between the treated and untreated barcode counts of that clone, giving a quantitative metric for treatment resistance, unique to each clone. Clonal measures of sensitivity or resistance were found to be highly diverse but consistent, with some clonal subpopulations demonstrating very high survivorship after treatment, and some nearly dying off consistently across replicates. There were no clonal subpopulations identified in any of the three cell lines that showed universal resistance to all three targeted inhibitors, with all subpopulations displaying sensitivity to at least one treatment. This suggests that heterogenous treatment response is mechanistically related to the molecular pathways targeted by the treatment itself, rather than related to the presence of a general, non-specific “persister” phenotype. A sample of treatment response metrics is shown in the bar plot in Figure 1. To find genes coorelated with survival during treatment, a predictive model was constructed that estimates changes in abundance for each clonal subpopulation, dependent on the gene expression data from those subpopulations. Changes in abundance were coorelated with clonal gene expression to parameterize the model, calculating weights for genes based on their ability to significantly and reliably predict cell survival during treatment. Based on their predictive capacity, genes were identified that are directly related to increased survival and persistence for subpopulations, and in turn directly influence intratumoral heterogeneity in treatment response with targeted inhibitors. The transcriptomic changes that occur during treatment were revealed by differential gene analyses (DGE) between control and treated populations. Because ClonMapper barcodes allow for the identification of clonal identities in scRNA-seq data, DGE analysis can identify specific transcriptomic trajectories unique to high-survivorship and low-survivorship clones. In this way, heterogenous changes in gene expression patterns during treatment are associated with the clonal identity of a cell subpopulation and that subpopulation’s specific treatment phenotype, providing insight into how a post-treatment tumor landscape is shaped by the pre-existing transcriptomic and phenotypic state of a tumor.
Acknowledgements and/or References (Optional):: ClonMapper - Gardner A, Morgan D, Al'Khafaji A, Brock A. Functionalized Lineage Tracing for the Study and Manipulation of Heterogeneous Cell Populations. Methods Mol Biol. 2022;2394:109-131. doi: 10.1007/978-1-0716-1811-0_8. PMID: 35094325.
