Skip to main page content
U.S. flag

An official website of the United States government

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Mar 10;117(10):5442-5452.
doi: 10.1073/pnas.1919259117. Epub 2020 Feb 24.

Joint profiling of chromatin accessibility and CAR-T integration site analysis at population and single-cell levels

Wenliang Wang  1   2   3   4 , Maria Fasolino  1   2   3   4 , Benjamin Cattau  1   2   3   4 , Naomi Goldman  1   2   3   4 , Weimin Kong  5   6   7 , Megan A Frederick  1   2   3   4 , Sam J McCright  1   2   3   4 , Karun Kiani  1   2   3   4 , Joseph A Fraietta  5   6   7   8 , Golnaz Vahedi  9   2   3   4   6
Affiliations

Joint profiling of chromatin accessibility and CAR-T integration site analysis at population and single-cell levels

Wenliang Wang et al. Proc Natl Acad Sci U S A. .

Abstract

Chimeric antigen receptor (CAR)-T immunotherapy has yielded impressive results in several B cell malignancies, establishing itself as a powerful means to redirect the natural properties of T lymphocytes. In this strategy, the T cell genome is modified by the integration of lentiviral vectors encoding CAR that direct tumor cell killing. However, this therapeutic approach is often limited by the extent of CAR-T cell expansion in vivo. A major outstanding question is whether or not CAR-T integration itself enhances the proliferative competence of individual T cells by rewiring their regulatory landscape. To address this question, it is critical to define the identity of an individual CAR-T cell and simultaneously chart where the CAR-T vector integrates into the genome. Here, we report the development of a method called EpiVIA (https://github.com/VahediLab/epiVIA) for the joint profiling of the chromatin accessibility and lentiviral integration site analysis at the population and single-cell levels. We validate our technique in clonal cells with previously defined integration sites and further demonstrate the ability to measure lentiviral integration sites and chromatin accessibility of host and viral genomes at the single-cell resolution in CAR-T cells. We anticipate that EpiVIA will enable the single-cell deconstruction of gene regulation during CAR-T therapy, leading to the discovery of cellular factors associated with durable treatment.

Keywords: CAR-T cell; epigenetics; lentiviral integration site; single-cell genomics.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing interest.

Figures

Fig. 1.
Fig. 1.
Workflow of EpiVIA. (A) Schematic illustration of EpiVIA. By mapping the ATAC-seq data to the combined reference, EpiVIA is able to identify integration sites and provirus accessibility, in addition to host chromatin state. (B) Five different categories of ATAC-seq fragments. EpiVIA is able to identify five different categories of fragments based on the results of alignment to the combined reference genome.
Fig. 2.
Fig. 2.
Validation of EpiVIA with clonal cells. (A) Chimeric and provirus fragments in HEK293 clone #1 identified by EpiVIA. (B) Genome browser view of reads and identified chimeric reads from HEK293 clone #1 mapped to the context of predefined integration site measured by LM-PCR. (C) Chimeric and provirus fragments in HEK293 clone #2 identified by EpiVIA. (D) Genome browser view of reads and identified chimeric reads from HEK293 clone #2 mapped to the context of predefined integration site measured by LM-PCR. (E) Genome browser view of reads and identified chimeric reads from two replicates of CAR+ and CAR− CD8+ cells from a CLL patient at the predefined integration site measured by LM-PCR. (F) Chimeric and provirus fragments in CAR+ and CAR− CD8+ T cells from a CLL patient identified by EpiVIA.
Fig. 3.
Fig. 3.
Genomic features of EpiVIA identified integration sites in single cells. (A) Circos plot visualization of the integration sites across the genome and local genomic features from inner to outer circle: 1) Gray dots: the density of HIV-1 integration sites in RID; 2) purple circle: the distribution of integration sites, with the color indicating different classes of transposable elements of the host sequence; 3) genes that harbor these integration sites (gene location red bar), color of the gene names suggest the frequency of integration into these genes: black indicates the gene is a RIG; red indicates the RIG were integrated more than once in our study; gray indicates the integration into this gene only present once in our study. (B) Plot demonstrates the odds-ratio of EpiVIA’s single integration sites falling in various classes of TEs, exons, introns, and intergenic regions using permutation tests. The comparison was done between fragments used to identify integration sites in CAR-T cells and equal number of randomly selected fragments with similar GC content from the scATAC-seq data. The test was repeated for 100 times and P values were calculated by Fisher’s exact test, followed by FDR correction. (C) Bar plot demonstrates the number of integration sites located in different classes of TEs. (D) Accumulative distribution demonstrates distance of EpiVIA identified integration sites to HIV-1 integration sites reported in RID (17).
Fig. 4.
Fig. 4.
Chromatin state of provirus sequence and host genome at integration sites. (A) Heatmap demonstrates the accessibility of different regulatory elements in the provirus genome at single-cell level. The average profile at the viral genome is depicted in black. (B) Accumulative distribution demonstrates distance between identified chimeric fragments to the peaks in aggregated scATAC-seq (red) in comparison with the distance between permuted open chromatin fragments to peaks in the aggregated scATAC-seq (gray), with the P value calculated with Wilcoxon rank-sum test. (C) Host chromatin accessibility of a gene with two independent integration events in two individual CAR-T cells. Genome browser view depicts MINK1 locus with CAR-T integration sites identified by EpiVIA in two single CAR-T cells together with scATAC-seq data across all single cells. Red arrows indicate the CAR-T integration sites and navy blue bars represent genomic locations of scATAC-seq fragments in two single CAR-T cells. Heatmap depicts the scATAC-seq fragments across all CAR-T cells. (D) Heatmap demonstrates mapping depth of aggregated scATAC-seq data in the 10-kb host genome centered on EpiVIA’s single-cell integration sites. The integration sites are sorted by the average mapping depth in this region. (E) Bar plot demonstrates the number of host fragments within 5 and 10 kb of integration sites in the cell that harbor the integration site. The integration sites are sorted in the same order as in D. (F) Bar plot demonstrates the total number of host fragments within 5 and 10 kb of integration sites across all single cells. The integration sites are sorted in the same order as in D. (G) Bar plot demonstrates the number of scATAC-seq fragments in the proviral genome from the cell that have the integration site. The integration sites are sorted in the same order as in D.
Fig. 5.
Fig. 5.
Single-cell chromatin state of the host genome. (A) UMAP projection of scATAC-seq profiles of CAR+ CD8 T cells. Dots represent individual cells and the colors indicate the cluster of the cell identified by Snap-ATAC (35). (B) UMAP projection of scATAC-seq profiles of CAR+ CD8 T cells. The colors indicate integration site or provirus fragments present in individual cells, and the size of the dots suggests the number of fragments sequenced in the cell. (C) Heatmap visualization of distance of peaks in each cluster to the closest TSS. (D) Heatmap representation of enrichment level of transcription factor (TF) motifs in the unique peaks of each cluster. CTCF, CCCTC-binding factor; ETS, ETS proto-oncogene 1, transcription factor; IRF, interferon regulatory factor; RUNX, runt-related transcription factor (RUNX); and TCF, T cell factor.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Niederer H. A., Bangham C. R., Integration site and clonal expansion in human chronic retroviral infection and gene therapy. Viruses 6, 4140–4164 (2014). - PMC - PubMed
    1. Schröder A. R., et al. , HIV-1 integration in the human genome favors active genes and local hotspots. Cell 110, 521–529 (2002). - PubMed
    1. Schmidt M., et al. , High-resolution insertion-site analysis by linear amplification-mediated PCR (LAM-PCR). Nat. Methods 4, 1051–1057 (2007). - PubMed
    1. Mueller P. R., Wold B., In vivo footprinting of a muscle specific enhancer by ligation mediated PCR. Science 246, 780–786 (1989). - PubMed
    1. Cohn L. B., et al. , HIV-1 integration landscape during latent and active infection. Cell 160, 420–432 (2015). - PMC - PubMed

Publication types