Clone-based full-length RNA-seq for early diagnosis of cancer

Introduction

Intra-tumor genetic heterogeneity imposes a great challenge on cancer therapy. Resistance to molecularly targeted therapies and chemotherapy can arise from selective growth of pre-existing sub-clones that carry drug-resistance mutations, altered metabolic and/or epigenomic states, providing a survival advantage. Early detection of these subclonal states can thus significantly aid cancer therapy.

Limitations of Current Techniques

However, attempts to profile various types of primary cancer cells using single-cell techniques are relatively poor. One of the major limitations is the significant dropout rate (i.e., loss of alleles) observed in single-cell RNA-seq.

Impact on Mutation Profiling

This dropout rate severely affects our ability to leverage single-cell RNA-seq to accurately profile somatic mutations, reveal cancer driver mutations, and even extract low/mid-level expressed genes and splicing.

Current Approaches and Their Limitations

For that reason, most of the efforts to expose mutations that are critical for cancer growth and can subsequently lead to more effective treatment are based on the sequencing of bulk populations. However, due to the noise introduced by PCR, sequencing, and alignment processes, bulk sequencing is limited to identifying mutations with a frequency higher than 5%.

Proposed Solution

Here we propose to develop a novel 3D clone-based full-length RNA-seq profiling technology. A preliminary version of this technology for digital profiling of mRNA has already allowed us to significantly improve sensitivity compared to gold-standard single-cell RNA-seq methods.

Preliminary Findings

Using this preliminary version on clones of lung adenocarcinoma, we revealed a novel cancer stem-like subpopulation that could not be detected using regular single-cell RNA-seq maps.

Conclusion

Altogether, improving the ability to detect rare mutations (<5%), splicing events, and transcriptional variation between cancer cells will be an extremely powerful tool for early diagnosis of cancer and an effective means to improve personalized drug treatment decision-making.