Unifying genomics and transcriptomics in single cells with ResolveOME amplification chemistry to illuminate oncogenic and drug resistance mechanisms

Abstract

ABSTRACTDiscovering genomic variation in the absence of information about transcriptional consequence of that variation or, conversely, a transcriptional signature without understanding underlying genomic contributions, hinders understanding of molecular mechanisms of disease. To assess this genomic and transcriptomic coordination, we developed a new chemistry and method, ResolveOME, to extract this information out of the individual cell. The workflow unifies template-switching full-transcript RNA-Seq chemistry and whole genome amplification (WGA), followed by affinity purification of first-strand cDNA and subsequent separation of the RNA/DNA fractions for sequencing library preparation. In the ResolveOME methodology we leverage the attributes of primary template-directed amplification (PTA)1 to enable accurate assessment of single-nucleotide variation as a DNA feature—not achieved with existing workflows to assess DNA + RNA information in the same cell.We demonstrated the validity of the technique in the context of two major phenomena in oncology: tumor heterogeneity (leading to cancer progression) and treatment resistance. Material from a primary patient breast cancer and an acute myeloid leukemia (AML) cell line, MOLM-13, was used to highlight multiomic biomarker paradigms enabled by this chemistry. Performance of the PTA-enabled genome amplification was largely unaffected by addition of RNA enrichment, with control WGS results showing > 95% genome coverage, precision > 0.99 and allele drop out < 15%. In the RNA fraction of the chemistry, we were able to routinely retrieve full-length transcripts that demonstrate a ratio of 1 for 5’/3’ bias, with increased coverage of intronic regions and 5’ regions that are indicative of novel transcripts, showing strength of the template switching mechanism to capture isoform information with sparsity rates < 75%. We find remarkable cellular variability of revealed biomarkers at both in the genome and transcriptome despite employing a relatively small number of individual cells. In our primary patient sample of ductal carcinoma in situ (DCIS)/invasive ductal carcinoma (IDC) we observed oncogenic PIK3CA driver mutations and prototypical DCIS copy number alterations binned into heterogenous single-cell classes of genomic lesions. Within our quizartinib-treated MOLM-13 cells, we identified multiple potential mechanisms of resistance within seemingly sporadic changes and were able to associate specific mutation, copy number and expression significantly correlated to treatment. In this latter scenario, the DNA arm of our combined workflow uncovered a secondary FLT3 (non-internal tandem duplication (ITD)) mutation as a candidate primary driver of resistance to drug while the RNA arm showed matched transcript upregulation of AXL signal transduction as well as enhancer factor modulation. Importantly, proximal candidate regulatory SNVs, outside of the CDS, were identified and associated to upregulated transcripts in cis. The study highlights that both the genome and transcriptome are dynamic, leading to a set of combinatorial alterations that affect cellular evolution and that fate can be identified through ResolveOME application to individual cells.