Acute lymphoblastic leukemia (ALL), a hematopoietic malignancy characterized by the accumulation of abnormal white blood cells in the bone marrow, is the most common cancer in children. B-ALL often exhibits chromosomal gains and losses, with near-tetraploidy and near-haploidy representing extreme cases. Near-haploid karyotypes, while also found in chronic myelogenous leukemia (CML), are less frequent. Ploidy is a strong predictor of clinical outcome in ALL, with hypoploidy associated with shorter event-free survival (EFS). Near-haploid B-ALL patients have a poor prognosis (EFS < 40% at 6-8 years). Hypodiploid leukemia cells can undergo whole genome duplication, leading to diploidized populations ('masked hypodiploid'), which don't show improved patient outcomes. Therapeutic resistance often originates from hypodiploid cells, highlighting the need to understand how leukemia cells thrive with significant chromosome loss and why this is linked to poor survival. A genomic study identified recurrent mutations in receptor tyrosine kinase, Ras signaling pathways, and the *IKZF3* gene in near-haploid ALL, suggesting PI3K inhibition as a potential treatment. However, it remains unclear if these alterations drive pathological differences or are later mutations. Another study demonstrated a p53-dependent response limiting the viability of near-haploid cells, with p53 deletion increasing viability and reducing diploidization. The molecular link between p53 and haploid-specific disadvantages, however, remained unaddressed. This study aimed to leverage single-cell RNA sequencing and computational modeling to infer cell cycle states in near-haploid and diploidized leukemia cells, perform comparative genome-wide knockout screens, and investigate the role of the homologous recombination (HR) pathway in near-haploid leukemia.

Previous studies have linked near-haploid karyotypes in acute lymphoblastic leukemia (ALL) to poor patient prognosis. While genomic analyses have revealed recurrent mutations in various signaling pathways, the specific mechanisms driving the unique biology of near-haploid leukemia cells and their resistance to therapy remained unclear. Research suggested a potential role for p53 in regulating the viability and diploidization of near-haploid cells, but the precise molecular mechanisms were not fully elucidated. These studies laid the groundwork for this investigation, which aimed to utilize multi-omics approaches to systematically identify the vulnerabilities of near-haploid leukemia cells and explore potential therapeutic targets.

This study employed a multi-faceted approach combining various experimental techniques and computational analyses. The KBM7 cell line, capable of spontaneous diploidization, served as the primary model system. Cell culture involved culturing near-haploid and diploidized KBM7 cells, as well as K562 cells (a near-triploid CML cell line), under standard conditions. Flow cytometry was used for cell cycle analysis, employing Hoechst 33342 staining for single-parameter DNA content analysis and BrdU pulse-labeling with 7-AAD staining for two-parameter analysis. Bulk RNA sequencing was performed on sorted near-haploid and diploid KBM7 cells to assess global gene expression differences. Genome-wide CRISPR-Cas9-mediated knockout screens were conducted in both haploid and diploid KBM7 cells to identify genes essential for near-haploid cell survival. shRNA-mediated knockdown of RAD51B was performed using a doxycycline-inducible system. Immunofluorescence staining was used to detect γ-H2AX and RAD51 foci, indicators of DNA damage and repair. Single-cell RNA sequencing (scRNA-seq) was performed on near-haploid and diploidized KBM7 cells and on cells from a patient-derived xenograft (PDX) model of near-haploid B-ALL treated with combination chemotherapy. Computational analyses, including cell cycle stage inference, pseudo-time trajectory analysis, and differential expression analysis, were used to analyze the generated datasets. Finally, the analysis incorporated microarray expression data from a large panel of pediatric hypodiploid B-ALL patients to validate the findings in a clinical setting.

This study revealed several key findings regarding the unique biology and therapeutic vulnerabilities of near-haploid leukemia cells: 1. **Slower Proliferation and Cell Cycle Alterations:** Near-haploid KBM7 cells exhibited slower proliferation rates compared to diploid cells, primarily due to prolonged G2/M phase. Single-cell RNA sequencing revealed altered cell cycle expression programs in near-haploid cells, with key cyclins showing atypical expression patterns. 2. **Homologous Recombination Pathway Dependency:** Genome-wide CRISPR-Cas9 knockout screens identified RAD51B as uniquely essential for near-haploid KBM7 cells, while RAD51 was essential in both haploid and diploid cells. Further analyses revealed that RAD51B and other genes were significantly depleted only in near-haploid cells, suggesting a unique dependency on this pathway. 3. **Impaired DNA Damage Repair:** Near-haploid cells showed increased γ-H2AX foci (indicating DNA damage) and reduced efficiency of RAD51-mediated DNA repair, particularly in the G2/M phase. Loss of RAD51B further exacerbated this defect, significantly impairing RAD51 recruitment to DNA damage sites in the G2/M phase of near-haploid cells, but not in diploid cells. 4. **RAD51B Signature in B-ALL:** A RAD51B expression signature was identified in a patient-derived xenograft (PDX) model of near-haploid B-ALL and was associated with responses to combination chemotherapy. Near-haploid cells showed elevated G2/M checkpoint signaling genes as part of the RAD51B co-expression program, and this program was significantly enriched in near-haploid B-ALL patients compared to their diploid counterparts. 5. **Clinical Relevance:** Analysis of microarray data from a large cohort of pediatric hypodiploid B-ALL patients confirmed the enrichment of the RAD51B signature and higher RAD51B expression in near-haploid B-ALL compared to other subtypes. This suggests a clinical relevance to the observed effects.

This study demonstrates a unique dependency on the homologous recombination (HR) pathway, specifically RAD51B, in near-haploid leukemia cells. The findings suggest that near-haploid cells have impaired DNA damage repair mechanisms, particularly in the G2/M phase, making them particularly sensitive to RAD51B loss. The observation that RAD51B depletion selectively affects near-haploid cells highlights the importance of understanding the context-specific roles of DNA repair pathways in cancer. The identification of a RAD51B-associated gene signature in both preclinical and clinical samples underscores the translational potential of these findings. The study also highlights the importance of integrating multi-omics data to uncover context-specific vulnerabilities in cancer cells. Future studies should explore the interaction between RAD51B and mitochondrial processes, further characterize the RAD51B signature, and evaluate RAD51B as a potential therapeutic target in near-haploid leukemia.

This study integrated multi-omics analyses to identify RAD51B as a unique dependency in near-haploid leukemia. The findings reveal impaired DNA damage repair, particularly in the G2/M phase, leading to increased sensitivity to RAD51B loss in near-haploid cells. A RAD51B expression signature is observed in both preclinical and clinical settings, highlighting the potential of RAD51B as a promising therapeutic target. Future research should focus on exploring the interactions between RAD51B and mitochondrial function and further evaluating RAD51B-targeted therapies.

The study primarily utilizes the KBM7 cell line and a single PDX model, limiting the generalizability of the findings. Genetic manipulation of human B-ALL samples was challenging, hindering more extensive functional studies in this context. Further research involving additional near-haploid leukemia models and clinical samples is necessary to confirm and expand upon these findings.