Abstract
Tumor formation is the result of a stepwise process through which cells acquire genetic alterations that may confer growth advantages to enable clonal expansion. One model in leukemia is that mutations accumulate in normal hematopoietic stem cells (HSC) leading to pre-leukemic cells which carry some, but not all, leukemia-specific mutations. Chronic lymphocytic leukemia (CLL) has been traditionally viewed as a mature B cell malignancy, however, recent studies have identified leukemia-specific mutations in CD34+ cells from either peripheral blood or bone marrow. These data raise the question of whether marrow progenitor cells in CLL are dysregulated even before commitment to the B cell lineage and, if so, what are the characteristics of such cells in CLL.
To this end, we first quantified the absolute number of CD34+ cells from marrow collected from CLL patients (n=9) vs healthy adult volunteers (n=8). Using a florescence bead-based absolute quantification assay (BD Trucount), we detected a significant reduction of CD34+ progenitor cells in CLL compared to normal donors (Average: 55.6 vs 391/ml, p=0.002, Welch's t -test). To determine if this quantitative change was accompanied by changes in the cellular state of these cells, we performed low-input RNA sequencing (RNA-seq) from 7 marrow progenitor subpopulations derived from 5 CLL patients and 4 healthy donors, that were isolated on the basis of flow cytometric sorting. These subpopulations included HSC (Lin-CD34+CD38-CD90+CD45RA-), MPP (Lin-CD34+CD38+CD90-CD45RA-), MLP (Lin-CD34+CD38-CD90-CD45RA+), CMP (Lin-CD34+CD38+CD90-CD135+CD45RA-), MEP (Lin-CD34+CD38+CD90-CD135-CD45RA+), GMP (Lin-CD34+CD38+CD90-CD135+/-CD45RA+), and CLP (Lin-CD34+CD38+CD10+CD45RA+). This systematic analysis revealed both a CLL-specific transcriptome profile as well as unique expression signatures for each progenitor subpopulation. In total, 119 genes were identified as CLL progenitor cell-specific and were highly enriched for the regulation of mRNA catabolic processes and protein translation.
To determine if leukemia-specific mutations are present in these progenitor cells and may contribute to dysregulation, we performed targeted deep-sequencing with low-input DNA derived from these subpopulations. We focused on marrow cells collected from 3 CLL patients, whose matched peripheral blood leukemia cells had previously undergone whole-exome sequencing (WES). Of the 30 somatic mutations identified by WES in CLL cells from Patient 1, we detected presence of 4 and 26 mutations in HSC and Pro-B/Pre-B cells, respectively. Likewise, we detected 4 of 34 CLL-specific mutations in HSC cells for Patient 2. In contrast, none of 10 mutations detected in Patient 3's CLL cells by WES were evident in the HSC subpopulation, but 9 of 10 mutations could be identified at the Pro-B/Pre-B stage. None of the 74 mutations from CLL cells from Patients 1-3 were detected in DNA extracted from marrow cells collected from 3 normal healthy donors.
To confirm the presence of these somatic mutations in CLL progenitor cells, we utilized a microfluidics-based approach to perform multiplexed simultaneous targeted detection of up to 48 mutations and expression of up to 96 genes from cDNA of individual cells. We constructed panels of 96 genes based on the aforementioned RNA-seq data to confirm both identity of cellular differentiation stage and cellular source (normal vs CLL). We thus tested whether the single-cell marrow progenitor cells from CLL patients harbored leukemia-specific mutation while expressing progenitor cell genes. With this method, we validated the expression of 2 mutations in bona fide HSC at the single-cell level in Patient 1. Further validation studies of somatic mutation detection using low-input DNA and single-cell approach are ongoing.
In summary, we present evidence of pre-leukemic cells at the HSC stage and these cells have increasing similarity to the cognate CLL at the pro/preB stage. Our DNA and single-cell analysis highlight the existence of CLL progenitor cells harboring leukemia-specific mutations while transcriptome profiling identified a CLL-specific gene signature associated with these pre-leukemic cells. We anticipate that further systematic analysis linking mutation, epigenetic changes, and dysregulated gene expression will enhance our understanding of the stepwise transformation and pathogenesis of this disease.
Neuberg: Synta Pharmaceuticals: Other: Stock shares. Wiestner: Pharmacyclics: Research Funding; Acerta Pharma: Research Funding. Brown: Janssen: Consultancy; Janssen Oncology: Honoraria; Redx: Consultancy; Gilead: Consultancy, Research Funding; Celgene: Consultancy; Infinity Pharmaceuticals: Consultancy; Roche/Genentech: Consultancy; Astellas Pharma: Consultancy; Pfizer: Consultancy; Sun BioPharma: Consultancy, Research Funding; AbbVie: Consultancy, Honoraria; Pharmacyclics: Consultancy; AstraZeneca: Consultancy. Wu: Neon Therapeutics: Consultancy.
Author notes
Asterisk with author names denotes non-ASH members.
This feature is available to Subscribers Only
Sign In or Create an Account Close Modal