Novel single-cell genomics analysis approach provides direct insights into cell cycle and proliferation

SPRINTER is a novel method that infers proliferation rates from single-cell genomics data, shedding light on tumor evolution

In a recent study published in Nature Genetics, a large team of researchers, including members of the TRACERx and PEACE consortia, designed an algorithm called Single-cell Proliferation Rate Inference in Non-homogeneous Tumors through Evolutionary Routes or SPRINTER to analyze single-cell genomic data, focusing on the proliferative behavior of cancer cells. The study also investigated the variations in cell proliferation across genetically distinct tumor clones.

About the study

In the present study, the researchers employed the SPRINTER algorithm to investigate cancer cell proliferation at a single-cell level. The research also combined advanced sequencing techniques to explore the interplay between genetic mutations, cell cycle dynamics, and cancer evolution, providing insights into metastatic potential.

SPRINTER utilizes single-cell whole-genome sequencing data to classify tumor cells into distinct cell cycle phases, such as S-phase and G2-phase, and assigns them to specific genetic clones. The approach involved several methodological innovations to overcome the limitations of existing techniques.

The study relied on single-cell deoxyribonucleic acid (DNA) sequencing (scDNA-seq) data and focused on replication timing, which is an indicator of when specific genomic regions replicate during the cell cycle. SPRINTER uses a specialized method to adjust for errors caused by DNA replication, allowing it to measure cell activity accurately. It examines parts of the DNA that replicate early or late and uses this information to sort and assign active (S-phase) cells.

The researchers explained that the process involves six steps — identifying replication patterns, analyzing changes in DNA structure, spotting active cells, grouping similar cells into clones, matching cells to clones after correcting for replication effects, and identifying other active cells (G2 phase). This helps map how quickly different groups of cancer cells grow.

The study focused on non-small cell lung cancer and confirmed SPRINTER’s accuracy by comparing its findings with other tests such as imaging and Ki-67 staining. SPRINTER was also tested on breast and ovarian cancer to determine whether it would perform well with different cancers. The study combined statistical analyses and evolutionary mapping to explore links between cell growth, genetic changes, and metastatic ability.

Major findings

The study found that cancer proliferation rates vary significantly among tumor clones, and SPRINTER identified the clones with high proliferation as having greater metastatic potential. These findings were consistent across primary and metastatic tumor samples in the non-small cell lung cancer dataset. The algorithm also revealed that high-proliferation clones tend to shed more circulating tumor DNA (ctDNA), which is a marker linked to cancer progression.

Furthermore, SPRINTER’s ability to resolve proliferation heterogeneity within tumors showed that distinct clones in both primary and metastatic sites have unique growth patterns. For instance, metastasis-associated clones often had elevated proliferation rates compared to others. This heterogeneity was overlooked in the bulk estimation methods, underscoring SPRINTER’s precision in distinguishing proliferative behaviors.

In the breast and ovarian cancer datasets, SPRINTER demonstrated that high-proliferation clones contained elevated genomic mutation rates, including single-nucleotide variants, structural variants, and copy-number alterations. These findings supported the hypothesis that rapid cell division contributes to the accumulation of genomic changes.

Furthermore, SPRINTER also linked alterations in replication timing with changes in gene expression, especially in genes involved in proliferation and metastasis. Such alterations were more pronounced in high-proliferation clones, indicating a mechanistic link between non-genetic factors and aggressive cancer behaviors.

Conclusions

To summarize, the study showed that tumor proliferation is highly heterogeneous and is driven by genetic and non-genetic factors. The detailed analysis using the algorithm SPRINTER revealed that high-proliferation clones are critical to understanding cancer metastasis and progression.

Furthermore, the study showed that these clones exhibit unique genomic alterations and increased ctDNA shedding, providing potential biomarkers for clinical applications. Overall, the study showed that SPRINTER offers a robust framework for studying cancer evolution, paving the way for targeted therapeutic strategies based on clone-specific proliferation dynamics.