Blog
Jul 9, 2025 / Oncology

Advancing cancer drug discovery with large-scale, single cell CRISPR screens

Olivia Habern

Multiomic CRISPR screening innovations, single cell resolution, increased scale—these technologies and capabilities are coming together at a turning point in functional genomics and other crucial areas of research, including cancer drug discovery. 

Review the following publication summaries to learn about new approaches to high-throughput single cell multiomic CRISPR screening and how researchers are applying them to cancer drug target discovery and functional evaluation of cancer gene mutations. And learn how scientists can further expand the scale and affordability of these applications by taking advantage of GEM-X Flex Gene Expression for single cell CRISPR screens.  

Use the following links to read publication summaries demonstrating powerful CRISPR innovations built on 10x Genomics Chromium single cell technology and their value for cancer research and drug discovery:

Jump ahead to our overview of single cell CRISPR screens using GEM-X Flex Gene Expression, or explore these additional resources:

undefined

CRISPR screening with GEM-X Flex Gene Expression

Ready to get started? Dive into the details of how to perform CRISPR screening with GEM-X Flex Gene Expression. Download this technical note featuring an overview of CRISPR guide probe design, the CRISPR + GEM-X Flex workflow using fixed samples, example datasets, and more.

Download now

High-throughput single cell CRISPR screen reveals a potential gene target for rare pediatric brain cancer

What is MultiPerturb-seq?

Innovations in CRISPR screening are enabling deeper functional insights from every perturbation by offering single cell resolution and multiomic readouts. Adding to a growing mix of single cell CRISPR screening methods, such as Perturb-seq, ECCITE-seq, and others, the MultiPerturb-seq method (developed by New York Genome Center researchers Yan et al.) offers another unique advantage—scale (1). 

MultiPerturb-seq allows scientists to link pooled CRISPR perturbations to single cell readouts of chromatin accessibility and gene expression, providing a view of a cell’s epigenetic state as well as its overall cell state and developmental stage (1). Relating specific perturbations to linked changes in chromatin accessibility and gene expression can also reveal the functions of individual genes or groups of genes. 

The method is adapted from the Epi ATAC assay (previously known as Chromium Single Cell ATAC, or Assay for Transposase Accessible Chromatin). It leverages a unique step-wise workflow involving combinatorial indexing and Chromium instrument–enabled droplet microfluidics to scale the throughput of single cell CRISPR screens to 100,000 cells on a single 10x Chromium ATAC lane (1). 

You can explore a detailed breakdown of the MultiPerturb-seq method in this paper.

ZNHIT1 knockdown drives terminal cancer cell differentiation 

Yan et al. tested the MultiPerturb-seq method in a biological case where understanding the gene regulators of cellular differentiation is absolutely critical: cancer

Their model cancer system was a rare pediatric brain cancer called atypical teratoid/rhabdoid tumor (AT/RT), which is typically caused by a change in the gene SMARCB1 (2). This gene encodes a subunit of an important chromatin remodeling complex that, when dysfunctional because of a SMARCB1 mutation, prevents differentiation of neural progenitor cells and, therefore, promotes uncontrolled tumor growth (1,2). 

If this chromatin remodeling complex was broken, the authors reasoned that other genes may play a role in driving terminal differentiation. But which genes? If they could identify those alternative epigenetic regulators, there might be different ways to reprogram cancer cells to differentiate, even in the absence of functional SMARCB1, by knocking down those alternative genes. This hypothesis led Yan et al. to perform a large-scale single cell CRISPR screen, targeting 100 epigenetic remodelers in human AT/RT cells. 

Their strategy stems from emerging research into cancer reprogramming therapy, also known as differentiation therapy, which involves manipulating tumor epigenetic regulation to enhance therapeutic efficacy. For example, researchers could deliver transcription factors into tumor cells that can change them into antigen-presenting cells to recruit T cells to the tumor microenvironment (3); or, in this case, they could identify genes that can be regulated to stop uncontrolled tumor growth. But the future impact and feasibility of this approach relies on high-throughput screens that can help identify more possible reprogramming targets (1). 

With these goals in mind, the authors leveraged the multiomic nature of the MultiPerturb-seq readout to identify genes that, when knocked down, reprogram tumor cells into a state that resembles mature neurons—that is, a more terminally differentiated state. They discovered that perturbation of ZNHIT1, a gene known to be involved in maintaining the characteristics of stem cells in other biological systems (1), led to a genome-wide chromatin profile that was more similar to postnatal brain tissue than fetal brain, and transcriptomic changes that were more similar to adult brain tissue. 

They also found that ZNHIT1 perturbation led to a 19% decrease in the expression of S-phase genes, which guide the process of DNA replication between cell division during the cell cycle. They also saw a 43% decrease in cells progressing through the S phase compared to control cells that did not receive the ZNHIT1 perturbation (1). These findings suggest ZNHIT1 is a possible target for reprogramming therapy in AT/RT because of its role in progressing cellular division and, conversely, in driving terminal differentiation when knocked down.

Defining gene function and possible tumor suppressors in T-cell acute lymphoblastic leukemia

The power of single cell CRISPR screening is also evident in research that seeks to explore the function of specific gene mutations in cancer development and progression. In their 2024 publication in Haematologica, Meyers et al. explored a subset of recurrently mutated genes in individuals with T-cell acute lymphoblastic leukemia (T-ALL). This cancer type is of particular interest given its genetic complexity: some tumors can carry up to 20 different mutations (4), but the normal functions of many genes found to be mutated in T-ALL remain unclear. 

The authors leveraged single cell CRISPR screening built on the Universal 3’ Gene Expression assay (previously known as Chromium Single Cell 3’ Gene Expression) to perform two separate screens of, first, 17 genes and then 42 genes in cultured human T-cell precursors. Their smaller screen perturbed 17 T-cell transcriptional regulators known to play a role in leukemia, with the goal of validating their CRISPR methodology by observing expected differential expression for known downstream target genes and determining the core transcriptional signatures driving T-cell growth in T-ALL. T cells were transduced with the guide RNA (gRNA) library then subsequently cultured and harvested after 3, 7, and 14 days to show the impact of the perturbations on cell proliferation over time. This screen pointed to Spi1—a gene expressed in early T-cell progenitors that, when active, delays cellular commitment—as crucial for proliferation of immature T cells. When Spi1 was knocked down, the team observed a significant proliferation disadvantage by day 14, and also noted that Spi1 inactivation downregulated MYC target genes, which are known to be essential for cell growth and proliferation (4).  

Taking this approach to a larger screen, the team then perturbed 42 genes that are recurrently mutated in individuals with T-ALL, yet whose functions are not clear. Which genes might be tumor activators and which might be tumor suppressors? In particular, the screen pointed to the gene Bcl11b as a strong tumor suppressor, indicated by the significant cellular proliferation the team observed on day 14 for cells where Bcl11b was knocked down. In essence, when Bcl11b was functional, it served as the brakes on T-cell proliferation. Knocking it down, however, unleashed cellular proliferation. 

With access to the complete transcriptome readout of individual cells, the authors could explore if genetic perturbations had common downstream affected pathways, revealing deeper knowledge of the mechanisms behind cancer growth when these genes are mutated or dysfunctional. Combining the data from both of their single cell CRISPR screens, they observed a significant cluster of perturbations that upregulated the STAT and NOTCH signaling pathways, which are known pro-cancer pathways (4). This, in turn, suggested those perturbed genes served as tumor suppressors in T-ALL.

The power of scalable single cell CRISPR screens using GEM-X Flex

Why scale is important for single cell CRISPR screens

These studies demonstrate the critical insights to be gained from single cell CRISPR screens. The ability to further scale single cell CRISPR screens is important for screening efficiency and more confident biological insights in disease research and drug discovery applications. Here are a few reasons why:

  1. Enabling greater breadth of tested perturbations: Scaling allows researchers to test thousands of genetic perturbations across many cells in a single experiment. This means a greater breadth of genes can also be tested, revealing novel gene functions or regulatory networks. 
  2. Decoding complex genotype–phenotype relationships: When you have a large and diverse dataset, you can more confidently infer relationships between genes and phenotypes, including mapping how genes may interact in pathways and networks. Whole transcriptome, single cell–resolution screens can also deeply profile cellular phenotypes, and enable identification of rare cell subtypes, to better understand the effects of perturbation.
  3. Providing a deeper knowledge of disease and greater potential for target discovery: Scaling up means researchers can screen for many disease-associated mutations that could be underlying diseases like cancer or neurodegeneration. Researchers can efficiently screen for new drug targets or biomarkers and discern meaningful biological information about how specific perturbations drive phenotypic outcomes. 
  4. Enabling greater statistical confidence: When researchers can survey perturbation effects across more cells, they can be confident in determining phenotypic outcomes, distinguishing real effects from random variation. More cells per condition also leads to better statistical power and reproducibility.

Beyond these key considerations, visionary researchers are looking to build a perturbation cell atlas with AI and machine learning models using the data gleaned from high-throughput, high-content single cell CRISPR screens—which could untangle the complex molecular circuits governing cell function, tissue organization, and more (5). A recent study from Xaira Therapeutics demonstrated the feasibility and utility of increasingly large single cell CRISPR screens for these atlasing efforts. The team developed a CRISPR screening method called Fix-Cryopreserve-ScRNAseq (FiCS) Perturb-seq powered by the GEM X Universal 5’ assay, then perturbed all human protein-coding genes in approximately 8 million cells with the goal of training foundational models and supporting therapeutic discovery (6).

How GEM-X Flex enables greater scalability for CRISPR screens

We released the n-Plex kit for our Flex Gene Expression assay to enable higher throughput and cost efficiency across all types of single cell studies, but it also benefits single cell CRISPR screens. The kit allows researchers to study a maximum of 256 samples and 5.12 million cells at less than $0.01 USD per single cell, and leverages custom probes directed against guide RNAs in fixed samples. 

Our scientists developed the Flex protocol for large-scale CRISPR screens specifically to overcome traditional challenges in this area. You can view their work in this poster: 

undefined

Poster

Scaling high-throughput, multimodal single cell CRISPR screens using 10x Genomics Single Cell Gene Expression Flex

You can also perform powerful single cell CRISPR screens using GEM-X Universal 5’ Gene Expression with Feature Barcode technology. However, there may be some specific experiments where it makes more sense to use Flex, including: 

  • Screens that require cell numbers into the millions
  • Screens of 10s of thousands of perturbations
  • Screens that have a high number of observations for each perturbation (such as when the expected phenotypes are more subtle) 
  • Repeated screening of the same perturbations across a number of conditions, cell types, and timepoints 
  • Screens where fixation may be required or desirable for your samples

You can explore some of the differences between these CRISPR screening approaches in our knowledge-based article: “Which 10x Genomics Single Cell CRISPR approach should I use?”

And if you’re looking for a detailed breakdown of how to perform single cell CRISPR screening with GEM-X Flex Gene Expression, download our Technical Note:

undefined

CRISPR screening with GEM-X Flex Gene Expression

This Technical Note features an overview of CRISPR guide probe design, the CRISPR + GEM-X Flex workflow using fixed samples, example datasets, and more.

References: 

  1. Yan R, et al. Pooled CRISPR screens with joint single-nucleus chromatin accessibility and transcriptome profiling. Nat Biotechnol (2024). doi: 10.1038/s41587-024-02475-x
  2. Atypical Teratoid/Rhabdoid Tumor (AT/RT): Diagnosis and Treatment. https://www.cancer.gov/rare-brain-spine-tumor/tumors/atrt
  3. Guetter S, Fan K, and Poeck H. Cancer cell reprogramming: Turning the enemy into an ally. Signal Transduct Target Ther 10: 13 (2025). doi: 10.1038/s41392-024-02102-w
  4. Meyers S, et al. Single-cell CRISPR screening characterizes transcriptional deregulation in T-cell acute lymphoblastic leukemia. Haematologica 109: 3167–3181 (2024). doi: 10.3324/haematol.2023.284901
  5. Rood J, Hupalowska A, and Regev A. Toward a foundation model of causal cell and tissue biology with a Perturbation Cell and Tissue Atlas. Cell 187: 4520–4545 (2024). doi: 10.1016/j.cell.2024.07.035
  6. Huang A, et al. X-Atlas/Orion: Genome-wide Perturb-seq datasets via a scalable fix-cryopreserve platform for training dose-dependent biological foundation models. bioRxiv (2025). doi: 10.1101/2025.06.11.659105