Advances in Single-Cell Multiomics for Cancer Immunotherapy

Each of us consists of approximately 30 trillion human cells (Sender et al., 2016). The function of each cell—and thus the human body—depends on physical, signaling, and microenvironment interactions with neighboring cells (Regev et al. 2018). Single-cell multiomics (SCMO) focuses on understanding these myriad interactions by integrating various hierarchies of biomolecules, such as the transcriptome and epigenome (Dimitriu et al., 2022). For example, SCMO has revealed that the profile of immune cells in the tumor microenvironment can be a deciding factor in clinical outcomes (Danenberg et al., 2022).

Consequently, it’s no surprise that SCMO has a large, expanding market for cancer immunotherapy and other medical conditions. Allied Market Research estimates that the global market for SCMO was over $2 billion in 2020, and predicts that the market will reach over $15 billion by 2030 (Shraddha et al., 2022). Ongoing SCMO bench work and computational tools are necessary for advancing their utility for cancer immunotherapy (Vandereyken et al., 2023). Here, we present representative state-of-the-art developments.

Immunotherapy: Putting the CAR T before the tumor

Chimeric antigen receptor T-cell (CAR T) immunotherapy uses a patient’s genetically engineered immune cells to target cancer (Chen et al., 2023). CAR T therapy has been the focus of hundreds of clinical trials, although the risk of short-term relapse and deadly side effects can be high (Wang et al., 2023). SCMO might help advance the efficacy and safety of CAR T therapy (Yang et al., 2023). For example, acute lymphoblastic leukemia is a blood and bone marrow cancer that’s commonly fatal within a few weeks or months if untreated (Cancer Research UK, 2021). Although survival has dramatically increased over the past decades, treatment usually requires years of chemotherapy (Jabbour et al., 2023).

To identify a biological cause of patient relapse from acute lymphoblastic leukemia after CAR T therapy, Bai et al. (2022) conducted a 12-patient pediatric study of >10⁵ single-cell transcriptomes and cell surface proteomes. Specifically, the CAR T-cells of relapsed patients were more likely to be deficient in inducing a T-helper type 2 functional response and in retaining stem-cell-like memory, compared with non-relapsed patients. By complementary assays, they validated their findings in a group of 49 patients. Optimizing CAR T therapy might also be useful for treating other cancers.

Population-scale studies: Cancer in the biobank

Biobanks are large collections of cells, organoids, and other biological samples that are invaluable for conducting large-scale medical research (Xie et al., 2023). Because SCMO studies are frequently incompatible with frozen samples (i.e., analysis of biobanks), it’s necessary to optimize methodologies accordingly. For example, it’s difficult to obtain single, intact cells from certain frozen brain cancer samples (Habib et al., 2017).

To address this limitation, Yu et al. (2023) developed scONE-seq: a single-cell, whole-genome sequencing method. scONE-seq co-amplifies the genome (DNA) and transcriptome (RNA), is compatible with automation (liquid-handling robots) and standard single-cell analyses (fluorescence-activated cell sorting), eliminates the sample loss that’s common in multistep processes, and does not require intact cells. They applied scONE-seq to astrocytoma samples that had been frozen for two years. There were unique cell subpopulations in the sample that are likely pertinent to regulating immune repression and forming the tumor–neuron synapse. Using these results to tailor glioma treatment might help increase survival from particularly aggressive brain cancers.

Data analysis: Order from chaos

One can imagine that SCMO experiments output an enormous quantity of data that isn’t useful without appropriate procedures for interpreting it. Evaluating disparate omics data by computational data analysis methods remains an ongoing challenge (Stanojevic et al., 2022).

In a preprint, Huizing et al. (2023) developed a computational tool termed Mowgli that they seamlessly integrated as a Python package. They compared the results from immune and bone marrow SCMO profiles as analyzed with Mowgli, CITE-seq, 10X Genomics Multiome, and TEA-seq tools. In general, Mowgli was comparable with the state-of-the-art in identifying cell subpopulations, and was especially useful for data interpretation. For example, Mowgli identified several possible transcription factors for a type of immune cell and the factors’ corresponding target genes, which could be useful for optimizing cancer immunotherapy.

In another preprint, Song et al. (2023) focused on developing positive and negative benchmarks for computational tools, especially for cases in which clear experimental data is unavailable. They report on scDesign3, a statistical simulator for various cell states and other omics data. This tool simulated various omic features (resembled the output of a CITE-seq dataset) well and can be used to test specific hypotheses (such as gene up-regulation in immune cells).

Industry perspectives

What do industry leaders say about the potential and challenges of SCMO? According to Rachid El Morabiti, Senior Regional Account Manager at Nanostring, “For many years, valuable clinical samples preserved in formalin-fixed paraffin-embedded (FFPE) were not fully utilized. However, with cutting-edge multiomics tools, valuable information can now be extracted effortlessly. This sheds light on how complex diseases, such as cancer, operate and leads to the discovery of biomarkers for detecting diseases earlier and identifying new therapeutic targets, thus accelerating the era of immuno-therapies.”

How can you contribute?

Single-cell multiomics work is inherently cross-disciplinary and helps answer questions that are obscured by simply focusing on the epigenome or transcriptome of large cell populations. Cell biologists, computer scientists, mathematicians, and other researchers all have valuable knowledge and perspectives that can help advance the field. By capitalizing on advances in single-cell multiomics, you can provide insight for improved cancer immunotherapy and other modern medical therapies.

References

Bai Z, et al. (2022). Single-cell antigen-specific landscape of CAR T infusion product identifies determinants of CD19-positive relapse in patients with ALL. Sci. Adv. 8(23):eabj2820.

Cancer Research UK (2021). Acute lymphoblastic leukemia (ALL). Jul. 29, 2021. (last accessed Apr. 4, 2021)

Chen Y-J, et al. (2023). CAR-T: What is next? Cancers 15(3):663.Danenberg E, et al. (2022). Breast tumor microenvironment structures are associated with genomic features and clinical outcome. Nat. Genet. 54:660–669.

Dimitriu MA, et al. (2022). Single-cell multiomics techniques: From concepts to applications. Front. Cell Dev. Biol. 10:954317.

Habib N, et al. (2017). Massively parallel single-nucleus RNA-seq with DroNc-seq. Nat. Meth. 14(10):955–958.

Huizing G-J, et al. (2023). Paired single-cell multi-omics data integration with Mowgli. bioRxiv preprint 2023.02.02.526825. (last accessed Apr. 4, 2023)

Jabbour E, et al. (2023). The evolution of acute lymphoblastic leukemia research and therapy at MD Anderson over four decades. J. Hematol. Oncol. 16:22.

Regev A, et al. (2018). The Human Cell Atlas white paper. arXiv preprint 1810.05192. (last accessed Apr. 4, 2023)

Sender R, et al. (2016). Revised estimates for the number of human and bacteria cells in the body. PLoS Biol. 14(8):e1002533.

Shraddha M, et al. (2022). Single cell multiomics market by type, (single cell genomics, single cell proteomics, single cell transcriptomics, and single cell metabolomics), application (oncology, cell biology, neurology, immunology and stem cell research), technique (single-cell isolation & dispensing and single-cell analysis), and end user (academic institutes, contract research organizations, and pharmaceutical & biotech companies): Global opportunity analysis and industry forecast, 2021-2030UP. Report prepared for Allied Market Research, Feb. 2022 (last accessed Apr. 4, 2023)

Song D, et al. (2023). A unified framework of realistic in silico data generation and statistical model inference for single-cell and spatial omics. bioRxiv preprint 2022.09.20.508796. (last accessed Apr. 4, 2023)

Stanojevic S, et al. (2022). Computational methods for single-cell multi-omics integration and alignment. Genom. Proteom. Bioinform. 20(5):836–849.

Vandereyken K, et al. (2023). Methods and applications for single-cell and spatial multi-omics. Nat. Rev. Genet. 1–22. Published ahead of print (last accessed Apr. 4, 2023)

Wang V, et al. (2023). Systematic review of CAR-T cell clinical trials up to 2022: Academic center input. Cancers 15(4):1003.

Xie X, et al. (2023). Tumor organoid biobank-new platform for medical research. Sci. Rep. 13:1819.

Yang J, et al. (2023). Advancing CAR T cell therapy through the use of multidimensional omics data. Nat. Rev. Clin. Oncol. 20:211–228.

Yu L, et al. (2023). scONE-seq: A single-cell multi-omics method enables simultaneous dissection of phenotype and genotype heterogeneity from frozen tumors. Sci. Adv. 9(1):abp8901.