Abstract
Introduction: Childhood acute lymphoblastic leukaemia (ALL) is a disease that originates before birth and results from the proliferation of neoplastic lymphoid precursor cells in the bone marrow. In the majority of cases of B-ALL, “pre-leukaemic” cells, harbouring B- ALL genetic mutations, can be retrospectively detected at birth, suggesting the prenatal origin of the disease. Previously, dense biallelic DNA methylation of the TES promoter was described in over 90% of B-ALL and 70% of T-ALL cases. By exploring publicly available datasets, hundreds of genes were found to be differentially methylated in ALL. The large number and consistency of the epigenetically modified genes raises the possibility that methylation-induced gene silencing in B-ALL may not be an acquired cancer-related phenomenon. I propose that B-ALL arises from normal “ALL-like” fetal cell lineage, comparable to murine fetal B-1 lymphocytes.
Hypothesis: that there is a distinct population of normal fetal lymphocytes that have an epigenetic profile similar to that reported in ALL and that this population contains the fetal cell of origin for ALL. To test this hypothesis, this project has three aims: 1) identify a panel of methylation biomarkers by exploring publicly available datasets; 2) evaluate the panel of biomarkers by deep targeted methylation sequencing using leukaemia and normal DNA samples; and 3) use the panel of biomarkers to identify the ALL-like fetal cells in healthy newborns.
Methods: The ABC.RAP package was developed to identify a panel of methylation biomarkers for ALL from publicly available 450k DNA methylation array datasets. The identified biomarker panel was validated by deep targeted methylation sequencing (MiSeq; Illumina) using blood samples from healthy adults, newborns, B-ALL and T-ALL xenograft samples and cell lines. Cord blood samples were either enriched into CD19+ B cells using MicroBeads or sorted into CD34+ progenitor cells, CD19+ B cells, CD3+ T cells and CD3-CD19- lymphocytes using fluorescence-activated cell sorting.
Results: Analysis of nine leukaemia datasets revealed 36 genes that were consistently methylated across T- and B-ALL datasets. A panel of 11 genes was selected for validation by targeted sequencing and results showed dense DNA methylation in leukaemia cases compared to low level methylation in adult blood samples. The methylation biomarker panel was used to identify the ALL-like fetal cells in enriched cell population from 34 cord blood samples. Remarkably, the ALL-like methylation profile was detected in 39% of the study population. The number of methylated epialleles was low in the sequenced samples (methylation level ranged between 0.1% to 13%). Methylated TES sequences were detected in CD19+ B cells of a preterm neonate (4-week-old, 28 week-gestation). In addition, methylated NEFM sequences were found at low levels in CD34+ progenitor cells and CD19+ B cells from healthy newborns (median methylation of 1.6% (n=4) and 1.4% (n=6), respectively). Furthermore, methylated TES and NEFM epialleles were detected at low level in CD3+ T cells and CD3-CD19- lymphocytes from newborns. Methylated KITLG and PAK7 sequences were also detected in neonatal CD19- mononuclear cells and CD3-CD19- lymphocytes, respectively.
Conclusion: An ALL-like epigenotype is detectable in normal neonatal blood, suggesting the presence of a distinct fetal blood lineage from which ALL can arise. The identification of these cells will contribute hugely to our understanding of the development and pathogenesis of ALL.