ALBERT

Description: Pediatric AML is a heterogeneous disease in which treatment resistance remains an unsolved problem that is responsible for most deaths (Yeung and Radich 2017). Recently we have come to learn that resistance may be driven by mechanisms that extend beyond somatic mutations and DNA methylation changes (Ghasemi et al. 2020; van Galen et al. 2019; Bell et al. 2019). Transcriptional changes within specific primitive and committed cell types in AML tumours, which may be accompanied by alterations in chromatin structure and topology, can also contribute to disease progression (Ghasemi et al. 2020). To study such changes at the single-cell level, we analyzed single-cell RNA-seq (scRNA-seq) and matched scATAC-seq data from primary, remission and/or relapse samples obtained from three pediatric AML patients enrolled in the AAML1031 clinical trial (Alpenc et al. 2016) (Figure 1). Using the 10X Genomics single-cell platforms, we profiled a total of 39,738 cells using scRNA-seq (~4,826 cells per sample, 1,571 genes per cell), and 46,580 cells and 197,128 peaks using scATAC-seq (~6,718 cells per sample, 5,628 unique reads per cell). We then integrated these data types to determine the extent to which these two modalities corroborated and/or complemented each other in analyses of these longitudinally-obtained samples. Cell subpopulations detected in scRNA-seq through Leiden clustering on a k-nearest neighbor graph were generally consistent with recent observations of malignant and normal cell types detected in the bone marrow and peripheral blood compartments (van Galen et al. 2019; Hay et al. 2018). Malignant-like subpopulations at primary and relapse stages exhibited similar levels of cell type diversity along the myeloid lineage. These included hematopoietic stem-like cells, progenitors, granulocyte-monocyte progenitors, monocytes and dendritic cell-like subpopulations. Remission samples appeared to contain normal blood cell types including natural killers (NK), B and T cells, platelets and erythrocytes, consistent with the clearance of blasts. However, we also observed putative malignant-like conventional dendritic cell subpopulations at remission (50% and 16% in the respective samples), noting that these cells displayed increased expression of genes involved in antigen presentation and lysosomal protein processing. To integrate scATAC-seq with scRNA-seq data we performed clustering of transformed and reduced scATAC-seq data through iterative latent semantic indexing (Granja et al. 2020), and aligned cells in scATAC-seq to cells from scRNA-seq data using canonical correlation analysis (Stuart et al. 2019). We observed similar patterns of T cell expansion, presence of monocyte-like populations and NK cells at remission in the scATAC-seq data. However, scRNA-seq subpopulations dominated by malignant-like cells showed variability in mapping to distinctive chromatin states, with a few notable exceptions (Figures 2 and 3). One such exception is a subpopulation in scRNA-seq, found mostly at relapse, marked by high expression of genes involved in proliferation and growth factor-mediated cellular processes such as YBX3 (binds to GM-CSF promoter), CYTL1, and EGFL7 (regulator of vasculogenesis) (Figures 3 and 4). Cells within this subpopulation mapped to two scATAC-seq clusters whose significantly more highly accessible regions were enriched for functional processes such as blood vessel remodeling and neutrophil/granulocyte activation (Figure 4). These observations are consistent with recent evidence that AML tumour cells can activate the immune system to acquire resistance (Melgar et al. 2020). The scRNA-seq subpopulation, however, did not display high expression of myeloid/granulocyte factors such as CD15, ELANE, and MPO (Figure 4), perhaps consistent with the notion that such transcriptional programs may be primed but not yet activated within these malignant cells. We thus evaluated the potential of scATAC-seq to complement scRNA-seq in understanding transcriptional changes within cell types in AML tumours. We observed that normal cell types and specific malignant cell states could occupy distinctive chromatin states. Through integrative analyses, we conclude that scATAC-seq results can add additional information to complement scRNA-seq data, including identifying nascent transcriptional programs that may be poised for activation within malignant cells. Disclosures No relevant conflicts of interest to declare.

Description: Population sequencing often requires collaboration across a distributed network of sequencing centers for the timely processing of thousands of samples. In such massive efforts, it is important that participating scientists can be confident that the accuracy of the sequence data produced is not affected by which center generates the data. A study was conducted across three established sequencing centers, located in Montreal, Toronto, and Vancouver, constituting Canada’s Genomics Enterprise (www.cgen.ca). Whole genome sequencing was performed at each center, on three genomic DNA replicates from three well-characterized cell lines. Secondary analysis pipelines employed by each site were applied to sequence data from each of the sites, resulting in three datasets for each of four variables (cell line, replicate, sequencing center, and analysis pipeline), for a total of 81 datasets. These datasets were each assessed according to multiple quality metrics including concordance with benchmark variant truth sets to assess consistent quality across all three conditions for each variable. Three-way concordance analysis of variants across conditions for each variable was performed. Our results showed that the variant concordance between datasets differing only by sequencing center was similar to the concordance for datasets differing only by replicate, using the same analysis pipeline. We also showed that the statistically significant differences between datasets result from the analysis pipeline used, which can be unified and updated as new approaches become available. We conclude that genome sequencing projects can rely on the quality and reproducibility of aggregate data generated across a network of distributed sites.