Due to their high sensitivity, it is no surprise that CUT&Tag4, CUT&RUN6,7,10, and ATAC-seq9,11–13 have been adapted for single-cell analyses. Here we will focus on the single-cell applications of CUT&Tag, which have attracted many researchers due to the use of Tn5 tagmentation technology.
Perhaps the most powerful feature of CUT&Tag is that the entire reaction – from antibody binding to tagmentation of sequencing adapters – occurs within intact nuclei (or cells). Thus, creation of a single cell suspension following a bulk CUT&Tag experiment facilitates barcoding of individual cells via indexing PCR. The resulting libraries can then be pooled for multiplex sequencing, making CUT&Tag a highly attractive strategy for single-cell chromatin profiling.
In the original CUT&Tag paper, Kaya-Okur et al. used a Takara iCell8 nano-dispensing system to isolate over 1,000 individual cells for high-throughput single-cell CUT&Tag (scCUT&Tag) experiments4. Importantly, they demonstrated that scCUT&Tag data for H3K4me2 and H3K27me3 correlate with profiles from corresponding bulk cell populations. The resulting single cell histone PTM maps could also be used to distinguish between H1 and K652 cell types in silico with high accuracy.
Bartosovic et al. adapted the scCUT&Tag protocol to investigate histone PTMs and transcription factor occupancy in single cells isolated from mouse brain tissue14. Using scCUT&Tag they were able to distinguish unique cell types and characterize changes in histone PTMs (H3K4me3, H3K27ac, H3K36me3, and H3K27me3) throughout cell differentiation. They also applied scCUT&Tag to map the transcription factor OLIG2 and the cohesin complex component RAD51, low-abundant targets that have been challenging to profile at single-cell resolution. These results reveal the epigenetic heterogeneity within tissue samples and establish the benefit of single-cell epigenomics, despite technical challenges and sparse data compared to bulk methods.
Zhu et al. combine scCUT&Tag with RNA-seq to simultaneously profile five histone PTMs (H3K4me1/3, H3K27ac, H3K27me3, and H3K9me3) and gene expression in single cells isolated from mouse brain tissue15. The authors demonstrate the power of “multiomics” approaches where the integrated analysis of transcriptional activity, histone modifications, and chromatin accessibility (via Paired-seq16) can uncover the intricate epigenetic regulatory mechanisms governing different cell types.
Clinical Applications of CUT&Tag
Single-cell analysis has the potential to uncover epigenetic programs driving development and human disease at unprecedented resolution. Janssens et al. implemented a CUT&Tag strategy to profile heterogeneity in mixed-lineage leukemia (MLL) patient samples, healthy patient samples, and cell lines 17. Specifically, the authors investigated the H3K4 methyltransferase KMT2A, which is translocated in approximately 10% of acute leukemias and has been shown to fuse with transcriptional elongation factors (e.g. AF9, AF4). Although it is known that different KMT2A rearranged oncofusion proteins (referred to as KMT2Ar for brevity) have distinct genomic localization, it has been difficult to accurately map and characterize their function due to limited patient samples and the poor resolution of existing approaches.
To enable efficient mapping from small patient samples, Janssens et al. developed automated CUT&Tag assays, and used them to profile various KMT2Ar fusion proteins and a collection of histone PTMs denoting active and silenced chromatin. They identified a subset of KMT2Ar binding sites with bivalent (H3K4me3 + H3K27me3) chromatin signatures. They then used scCUT&Tag to determine that this bivalent signature in the bulk population was caused by heterogeneity among individual cells within the tumor.
Finally, the authors used automated CUT&Tag to characterize cellular responses and predict tumor sensitivity to epigenetic inhibitors. DOT1L is highly indicated in cancer development and has been shown to interact with KMT2Ar fusion proteins in some leukemias. CUT&Tag provides the sensitivity needed to detect differential recruitment of DOT1L by KMT2Ar and monitor the impact of DOT1L inhibitors in various leukemia cell lines. Although not yet at single-cell resolution, this approach demonstrates CUT&Tag as a robust platform to map patient-specific fusion proteins, compare with profiles of other important factors or histone PTMs, and subsequently use this information to rationally determine best courses of treatment.