Loss of 5q Drives Evolution to Aneuploidy in an iPSC Model of Complex Karyotype AML

Aneuploidy is common in solid tumors but only comprises ~10% of acute myeloid leukemias (AML), defined by complex karyotype (AML-CK) and associated with TP53 mutations and 5q loss. The drivers and trajectories of AML-CK evolution have not been experimentally established. Here, we develop an experimental model of complex karyotype evolution. We utilize single cell copy number inference and karyotyping to track clonal evolution of TP53-mutant hematopoietic stem and progenitor cells (HSPCs) to aneuploidy.

Deletion of chromosome 5q (del5q) is the most common chromosome abnormality found in 80% of AML-CK patients. Loss of critical genes encoded within the 5q region may cooperate with TP53 mutation to promote genome instability. We established induced pluripotent stem cell (iPSC) lines from TP53^+/- and TP53^+/- with del5q (TP53/del5q) preleukemic clones reprogrammed from an AML-CK patient. Since mitotic errors are a common cause of aneuploidy in cancers, we treated TP53^+/- and TP53/del5q iPSC-derived hematopoietic stem and progenitor cells (HSPCs) with mitotic checkpoint inhibitors. Chromosomal evolution was monitored over extended culture using standard karyotyping and single cell “karyotyping” by RNAseq copy number inference. TP53/del5q but not the ancestral TP53^+/- clones progressively evolved to diverse, complex (>3 abnormalities) aneuploid states, with a greater number of distinct clones and abnormalities compared to TP53^+/- single mutant cells (95.1% vs 5.8% aneuploid clones, n=7, p < 0.0001). Frequently observed copy number alterations included trisomy 8, trisomy 10, and unbalanced translocations on chr. 21, also found in AML-CK. Evolution was marked by genetic drift and clonal competition between diverse clones, with often both numerical and structural chromosomal changes. We observed multiple subclones arising, further diversifying, and typically resulting in a single “winning” aneuploid clone with high relative fitness achieving dominance.

Clones with chr. 10q gain displayed a fitness advantage across multiple experiments. These clones harbored 10q amplified material on chr. 21p (add21p), a cytogenetic change seen in AML-CK patients. To correlate distinct aneuploid states with changes in gene expression, we used single cell RNAseq to compare cell states of aneuploid clones detected by copy number inference. Differential expression analysis between TP53/del5q +add(21p) and parental single cells revealed a ~116 gene signature (ADD-10 signature) marked by upregulation of 10q genes including PTEN, as well as BCL2. Single cell transcriptomics of TP53/del5q-mutant AML-CK primary patient samples revealed that aneuploid cells upregulated the ADD-10 signature compared to normal cells within each patient, even in the absence of chr. 10q abnormalities, suggesting that the signature is a common feature of aneuploidy in AML-CK. Venetoclax (VEN) is a clinically approved BCL2 inhibitor for AML, however TP53 mutations confer poor response. Treatment with VEN dramatically altered the clonal landscape, resulting in the elimination of the BCL2-high add(21p) clone and dominance of a del(1p) minor clone that conferred resistance to VEN. The minor clone displayed a 3-log increased IC50 of VEN, decreased BCL2 expression, and decreased BCL2/BAX ratio. These data identify a gene signature associated with aneuploidy in AML-CK and that its rational targeting by VEN triggers outgrowth of resistant subclones.

In summary, we describe the trajectories of chromosome evolution to aneuploidy in a novel model of preleukemic HSPCs derived from a patient with AML-CK. We demonstrate that 5q loss cooperates with TP53 mutations to drive evolution to complex karyotype. Evolution is marked by clonal competition between diverse aneuploid clones with distinct fitness, where winning clones display a shared gene signature with patient samples. Treatment with VEN alters the fitness landscape by selecting for minor clones with reduced BCL2 expression, providing an explanation for poor VEN response in TP53-mutant AML. Our experimental system can be harnessed to gain insights into the mechanisms and therapeutic vulnerabilities conferred by aneuploid cell states.