Epigenetic Silencing of BRCA1 Is Linked to Homologous Recombination Repair Defects and Elevated Mir-155 Expression in Myeloid Neoplasms

Myeloid neoplasms are a heterogeneous group of clonal hematological disorders ranging from chronic myeloproliferative neoplasms (MPNs) to more aggressive mixed myelodysplastic syndrome/myeloproliferative neoplasms (MDS/MPNs) and secondary acute myeloid leukemias (sAMLs). With the exception of BCR-ABL positive chronic myelogenous leukemia (CMLs), these diseases are clinically challenging to diagnose. Due to the clinical heterogeneity and lack of new biologic insights, current therapies fail to eradicate the malignant clone and alter the natural history of the disease. However, these diagnoses share overlapping features in morphologies, cytogenetics and genetic alterations suggesting common alterations in underlying disease pathways.

We have previously reported genomic loss of HR genes in MPN patients linked to defective RAD51 foci induction and sensitivity to poly (ADP-ribose) polymerase (PARP) inhibition (McDevitt et al., ASH 2011). In this study, we directly examined homologous recombination (HR) in primary human samples by measuring induction of RAD51 foci after ionizing irradiation. Freshly isolated mononuclear leukocytes from patients diagnosed with MPNs, mixed MDS/MPNs and sAMLs were collected at the Johns Hopkins Hospital from April 2012 to March 2014. 11 of the 22 samples (50%) exhibited impaired RAD51 foci induction. With little evidence for biallelic genetic inactivation of HR genes in previous work, we evaluated promoter CpG island methylation of BRCA1, which we had previously found in a pilot study of myeloid neoplasm samples (McDevitt et al., ASH 2011). Using quantitative methylation-specific PCR (qMSP) with melt curve analysis, we reported BRCA1 promoter hypermethylation in 22 of 104 samples (21.2%), and observed statistically significant down-regulation of BRCA1 transcript in samples with BRCA1 CpG methylation (p<0.05). To determine the relationship between BRCA1 epigenetic silencing and HR status, we categorized BRCA1 gene expression of the samples according to the promoter methylation and HR status. Strikingly, we noted that all BRCA1 methylated samples were defective for HR and 5/12 (41.7%) of HR defective samples had down-regulated BRCA1 levels.

We next validated the role of BRCA1 repression in vitro using the AML cell line OCI-AML3 using two independent shRNAs targeting BRCA1. We found that BRCA1-silenced cells have reduced induction of RAD51 foci, recapitulating our observations in primary samples with epigenetic silencing of BRCA1. Given the selective toxicity observed with HR-deficient cancers to PARP inhibition, we treated the BRCA1 knockdown cells with the PARP inhibitor ABT-888 and observed increased drug sensitivity.

Finally, we examined additional consequences of BRCA1 loss in myeloid malignancies by investigating its role in repressing miR-155. Frequent up-regulation of miR-155 in leukemia is linked to poor prognosis, and its overexpression in murine models results in myeloproliferative disorders. Our results showed a statistically significant inverse correlation between BRCA1 and miR-155 expression in patient samples (Pearson r:-0.36; p-value<0.005). We also found elevated miR-155 expression in the BRCA1-silenced OCI-AML3 cells that increased during long-term BRCA1 repression.

Our results demonstrate that defective HR repair occurs in a significant subset of myeloid malignancies. This provides a rationale for PARP inhibitors in this patient subpopulation who currently lack curative therapy, a strategy currently explored in two clinical studies (J0783, PI: Keith Pratz, MD; J1051, PI: Ivana Gojo, MD). Our finding showing BRCA1 promoter methylation in samples with defective HR highlights a novel mechanism underlying HR defects in myeloid malignancies, and offers a biomarker to identify potential responders to PARP inhibition. We also provide an alternative mechanism in which BRCA1 loss contributes to disease progression, via de-repressing miR-155 linked to aberrant expansion of myeloid cells.