Cancer outcomes have rapidly improved over the first quarter of the 21st century, and in the ten years from 2013 to 2022, the overall cancer death rate in the United States dropped by 1.7 percent each year. These gains have been driven in large part by new insights from cancer research. One such insight that has enhanced treatment is the recognition that every tumor is unique.
Inside the tumor microenvironment
Developing a precision medicine approach to cancer care is a major goal of modern oncology. It requires recognizing the variability in tumors at multiple scales: between patients, between tumors in the same patient, and within individual tumors. This heterogeneity can be temporal and spatial; cancer cells’ unstable genomes mean that tumors evolve over time, while genetically varied cell populations can be found in different tumor regions. The tumor microenvironment (TME) is a chaotic mix of immune, stromal, and cancer cells that changes with disease progression.
Some cancers’ progression depends on the presence or absence of specific genome markers known as “driver genes.” Improved identification of these gene features could open new treatment opportunities. A recent analysis of over 10,000 cancer patients through the UK 100,000 Genomes Project found that 55 percent of patient genomes included at least one clinically actionable mutation that predicted cancer resistance or vulnerability.
Demystifying the TME is critical to precision medicine, but traditional genomic sequencing tools are limited in their ability to resolve this complex environment. These tools rely on bulk analysis of a heterogeneous mix of cells. While such approaches have helped researchers gather a global picture of tumor function, their averaged perspectives can obscure the signals from specialized or rare tumor cell populations.
The rise of single-cell sequencing technologies, such as single-cell whole-genome sequencing and single-cell RNA sequencing, has dramatically advanced our understanding of disease progression. Scientists are now gaining a detailed view of cellular and molecular diversity within the TME, bringing new targets into focus and advancing precise treatments.
Using single-cell DNA sequencing to identify rogue genomes
Available single-cell analysis techniques explore different types of omics data within tumor cells. Abnormal genomic replication is a hallmark of cancer cell growth that often leads to copy number variations (CNV) in certain genes or even abnormal copy numbers of entire genomes. These CNV mutations can drive cancer progression and predict treatment response. CNV mutations in cancer can be divided into two categories based on scale: focal and chromosome-arm. Focal variants are CNVs affecting a few genes over small regions, while chromosome-arm level variants affect larger chromosomal regions.
Traditional tools for identifying mutations of interest, like fluorescence in situ hybridization, have limited genomic coverage, which may lead to false positives or negatives in CNV detection. Whole-genome sequencing provides a full view of the genome but requires more input DNA than that contained within a single cell. That means effective and unbiased amplification methods are critical for single-cell whole-genome analysis. When working with such small starting amounts of DNA—a human cell contains just six to 10 picograms of DNA—amplification can involve risks including amplification bias, slippage in repetitive regions like microsatellites, and the emergence of chimeric amplicons.
Lab Management Certificate
The Lab Management certificate is more than training—it’s a professional advantage.
Gain critical skills and IACET-approved CEUs that make a measurable difference.
If these requirements can be met and the associated risks overcome, single-cell DNA sequencing provides valuable insights. A recent study of CNVs in high-grade serous ovarian carcinoma (HGSOC) cells found five disease-linked variants. The analysis connected a higher copy number of the frequently amplified MYC gene to improved in vitro and clinical responses to paclitaxel, a chemotherapy drug.
These studies also benefit from high-throughput sequencing approaches that facilitate single-cell genome sequencing at the cell-population level. A recent analysis that sequenced over 22,000 genomes demonstrated the effectiveness of this scaled sequencing at determining large-scale patterns of genome structural variations like CNVs within tumors in triple-negative breast cancer and HGSOC.
Hunting heterogeneity with single-cell RNA sequencing
Single-cell global RNA analysis can also provide valuable insights into disease progression. These techniques can be subdivided into 3’ or 5’ end-counting and full-length methods. The former is more cost-effective, but may miss critical events like alternative splicing that full-length methods can detect.
Interested in life sciences?
Single-cell sequencing of DNA and RNA content in cancer cells can be complementary and additive. Single-cell DNA analysis of cancer cells in acute myeloid leukemia (AML) revealed DNA clonal evolution during disease progression. However, they also showed that DNA mutations are sporadic in AML. This discovery informed further transcriptome analysis, revealing additional RNA clonal evolution. These data included significant heterogeneity, highlighting the importance of global coverage and sufficient scale. The analysis nonetheless identified that pathways linked to metabolism and apoptosis are useful signatures of cancer relapse.
Choosing the correct analysis pipeline
Single-cell sequencing is most powerful when paired with appropriate bioinformatics tools for processing and analysis steps. For example, variant callers used in bulk sequencing are ill-suited to single-cell data, so it is important to use callers that fully compensate for amplification artifacts. Further downstream analysis may include CNV calling, gene or transcript quantification, or immune profiling. Choosing the correct analysis method depends on the depth and quality of sequence data, ease of implementation, and the types of analysis desired.
Boosting biomarker discovery with single-cell techniques
Identifying genetic or transcriptomic signals that predict disease progression is essential to the broader move toward precision oncology. Single-cell techniques are fueling improvements in this area of biomarker discovery.
A low successful clinical translation rate—estimated at just 0.1 percent—has plagued biomarker research. TME heterogeneity has made identifying reliable biomarkers more challenging. Further, biomarkers should be robust, easy to quantify, and hold across gender and ethnic groups—qualities that are hard to attain when biomarker discovery research is done at a small scale. To meet these criteria and separate true biomarkers from other mutations inconsequential to progression, analysis should be large-scale and, ideally, longitudinal to account for spatial and temporal heterogeneity in biopsied tumors.
This need is partially met through initiatives like the UK Biobank, which has started integrating multiomic data to enhance understanding of human disease. Still, biobanks face issues regarding sample availability and quality. Automated and high-throughput sample processing and analysis tools can address these shortcomings by standardizing workflows and facilitating discovery of relevant data. The advent of highly sensitive single-cell techniques has enabled the identification of biomarkers such as circulating microRNAs and circulating tumor cell-derived markers from body fluid samples, including plasma, serum, and urine using liquid biopsy techniques. These additional sources further enhance the power of large-scale biomarker datasets.
Single-cell methods face challenges from unstandardized protocols, limited sample sizes, and high biological variability. During library preparation, small biases in adaptor ligation or amplification could hamper the clinical reproducibility of any identified biomarkers. High-throughput methods using higher sample cell counts, paired with standardized library preparation techniques, can minimize these risks.
Computational analysis of single-cell data can also introduce potential bias. The increased complexity of high-dimensional omics data requires the use of classification algorithms to detect molecular signatures. These machine learning-based techniques, like support vector machines, can overfit data to the cohort if an underpowered sample size is used, making it less relevant to populations not heavily featured in the sample. While this can be mitigated with additional analytical techniques, larger sample sizes can reduce the risk of overfitting occurring in the first place.
Cancer researchers have used high-throughput single-cell techniques to overcome these challenges in biomarker discovery. Ho and colleagues used single-cell RNA sequencing to investigate the transcriptional TME of hepatocellular carcinoma—a particularly heterogeneous solid tumor—and identify rare subclone populations with unique carcinogenic characteristics. Similarly, Wu and colleagues analyzed the TME in advanced non-small cell lung cancer using single-cell RNA sequencing. They identified the infiltration of unusual immune cell subsets, such as follicular dendritic cells and Th17 helper cells.
Single-cell analysis techniques are meeting the challenge of complex TMEs. New technologies have unveiled mutations and expression shifts linked to disease progression, offering valuable new strategies for biomarker discovery. By embracing higher throughput technologies and standardized, automated protocols, these pipelines can overcome hurdles in translation and lead to improved, precision treatments.
