High Throughput Epigenomic Mapping of Hematopoietic Stem and Progenitor Cells Using Low Cell Input Acoustic Enhanced Immunoprecipitation

Acute myeloid leukemia (AML) is a complex disease with several etiologies characterized by incomplete cellular differentiation and clonal proliferation of hematopoietic stem and progenitor cells (HSPCs) in place of normal hematopoiesis. Non-mutational epigenetic reprogramming is an emerging hallmark and enabling characteristic of cancer pathogenesis. Epigenetic regulation of gene activity is an essential component of chromatin stability and maintenance of proper cellular function and differentiation. Modifications of gene and histone landscape methylation and acetylation, three-dimensional chromatin structure, and protein-controlled activating and silencing switches are all regulated by epigenetic components, and thus are integral to understanding gene regulatory mechanisms during cell fate changes, disease onset, progression, as well as for optimizing therapeutic regime. In addition, this information holds immense potential for accurate identification of disease biomarkers, cell of origin, and targets for epigenetic therapy. However, existing chromatin immunoprecipitation sequencing (ChIP-Seq) workflows require significant cellular material and are very time-consuming, requiring cumbersome optimization. Moreover, cellular heterogeneity can influence accuracy of results from the perspective of sample preparation efficiency and also population-biased effects. Both normal and malignant bone marrow are heterogeneous in cell type and stage of differentiation that leads to imprecise results with regard to epigenetic determinates.

In our study, we implemented Covaris’s Adaptive Focused Acoustics® (AFA®) technology to miniaturize sample volumes for both chromatin shearing and immunoprecipitation in a 96-well format. This approach enables a standardized, robust, automatable, half-day workflow for both ChIP and Hi-ChIP applications. AFA ChIPseq- was applied to flow sorted hematopoietic stem and progenitor cells (HSPC) from individual mice. We utilized a range of epitopes including major histone modifications (H3K9me1/2/3, H3K27ac) and factors (CTCF, cohesin). Interestingly, we found epitope-dependent differences in AFA-enhanced binding with CTCF saturating at 30 minutes, whereas H3K27ac required 4 hours of AFA pulse to facilitate chromatin binding. Additionally, we evaluated low input in vitro and in vivo samples, with limited dilution of lineage negative HSPCs ranging from 500,000 to 5,000 cells with comparable sheared chromatin yield and immunoprecipitated material. Overlap of peaks called by MACS2 was consistent among the various cell inputs as well as when compared with gold standard CTCF ChIPseq- data previously performed on pooled samples from our previously published work. The low cell input capability permitted deconvolution of dynamic epigenetic changes in protein binding and chromatin structure during normal lineage specification of hematopoietic stem cells. Epigenetic remodeling of these sites was consistent with established gene programs essential for myeloid differentiation. Data were found to be in alignment with standard ChIPseq- and Hi-C datasets generated from similar cell populations.

The simplified and significantly shortened (Hi)ChIP-seq processes showcased in this study will be highly beneficial for both research and clinical diagnostic applications owing to their requirements for using 96-well plate format and their ability to become integrated into automated workflows. Taken together, this technology and workflow represents a high throughput epigenetic system for small cell numbers of cells with a multitude of applications enabling laboratories focused in cell biology and translational research.