Somatic Mutations in TP53 Identified in T-Cells of Patients with TP53-mutated AML

Abstract

Introduction: Genomic characterization of AML has delineated the frequent occurrence of somatic mutations in pre-malignant cell populations characterized as clonal hematopoiesis (CH). AML with mutant TP53 (TP53mt) is associated with an adverse prognosis and resistance to all available therapies. The cell-of-origin in AML with TP53mt is not well established. Delineating the clonal architecture of hematopoiesis in AML with TP53mt could improve understanding of resistance. Recurrent somatic DNMT3A mutations, commonly detected in CH, have been identified in purified hematopoietic stem cells (HSCs) and T-cells of patients (pts) with AML (Shlush, Nature 2014). We hypothesized that somatic TP53mt affect early cells in the hematopoietic hierarchy of AML leading to a dysfunction of multiple hematopoietic compartments including T-cells. Methods: Using whole exome sequencing (WES), we analyzed FACS-sorted bone marrow sub-populations from 39 pts with adverse risk, including 16 (41%) with TP53mt AML. Mononuclear cells were flow-sorted for bulk (CD45dim), and leukemia stem cells (LSCs) (Lin-CD34+CD38-CD123+) from diagnostic samples, whereas post-treatment samples were sorted for therapy residual disease based on patient-specific aberrant phenotype. CD3+ sorting was used to identify T-cells in available samples. Mutations were called using MuTect and Pindel against virtual normal. We used patient-specific primers to validate mutations by digital-droplet PCR (ddPCR). DAb-seq (joint profiling of DNA and proteins in single cells) was used to examine mutations in subpopulations. Results: We found that CD3+ sorted cells harbored, in addition to previously described mutations known to occur in CH, mutations in TP53 at relatively high variant allelic frequencies (VAF) (46%, 14% - 88%), in 5/7 (71%) samples from different pts. These were the exact mutations detected in leukemia sub-populations. These pts had no personal or family history indicative of a Li-Fraumeni Syndrome (LFS). To address whether the observed mutations in sorted CD3+ cells are a result of an imperfect selection and contamination by leukemia cells or whether these are mutations in T-cells, we extracted T-cell receptor (TCR) sequences from WES to use as barcodes for quantifying T-cell fractions in each sample (Bentham, Nature 2021). We found that T-cell diversity, a metric quantifying unique TCR sequences, was significantly higher in sorted CD3+ cells indicating a high fraction of T-cells in these samples (median TCRβ normalized diversity of 23.6 vs 1.1, P=0.005). In order to further corroborate these findings, we resorted to sorting of CD3+, CD33-, CD34- cells from an additional cohort yielding purity >97% (Fig. 1A). Subsequently, we used ddPCR with probes designed for corresponding mutations in TP53 detected in clinical testing for each patient. In highly pure sorted samples with these markers (CD3+, CD33-, CD34-), TP53mt were detected in more than 20% of droplets (range of droplets with TP53mt: 23% - 44%) and were detected in 3/9 (33%) samples from different pts (Figure 1A). Finally, we used DAb-seq for simultaneous capture of DNA genotype and cell surface phenotype in bone marrow cells from 5 pts with TP53mt AML. We identified the TP53 (R280G) mutation in the CD34-, CD3+, CD4+, or CD8+ cell cluster, the same mutation detected by the clinical sequencing assay in 1 patient (Fig. 1B). We had previously shown that T-cells from TP53mut AML have aberrant cytokine secretion with immune dysfunction. T cells from TP53mut AML (N=5) have decreased function vs TP53wt (N=5) (Polyfunctional Strength Index of 6.6 vs 73.5, P=0.01). This result was driven by lower secretion of effector cytokines such as granzyme B, interferon-g and TNF-α in TP53mut. All together, these findings identify multiple mechanisms for immune dysfunction in TP53mt AML. Conclusion: We found TP53mt in a fraction of CD3+, CD34- cells from pts with TP53mt AML, likely T-cells. These mutations are likely somatic given the range of VAFs and lack of family history indicative of LFS. This possibility suggests that TP53mt are early events in AML, occurring in HSCs and expanding in multiple cellular compartments including T-cells. Further evaluations of these findings are ongoing. Given the important role of TP53 in various cellular functions including immunity, alterations of TP53 could be a challenge to therapeutic strategies aimed at reinvigorating T-cells in TP53mt AML. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal