BRCA1/2 Containing Complex 3 (BRCC36) Is Recurrently Mutated in AML with t(8;21) and Associated with Increased Sensitivity to Chemotherapy through Impairment of the DNA Damage Repair Pathway

Introduction

BRCA1/BRCA2-containing complex 3 (BRCC36) is a Lys63-specific deubiquitinating enzyme (DUB) involved in DNA damage repair. Mutations in BRCC36 have been identified in 2-3% of patients with myelodysplastic syndromes (MDS) and secondary AML (sAML). The role of BRCC36 mutations in de novo AML and their impact on DNA damage-inducing cytotoxic chemotherapy sensitivity is not clear.

Aim

We aimed to determine the incidence of BRCC36 mutations in AML and their impact on outcome and drug sensitivity in vitro.

Methods

We analyzed the entire coding region of BRCC36 for mutations in 191 AML cases with t(8;21) (q22;q22.1) and 95 cases with inv(16) (p13.1q22) using a customized targeted sequencing panel. Data for de novo AML was derived from The Cancer Genome Atlas Research Network (TCGA) data set (NEJM 2013). Lentiviral CRISPR/Cas9 was used to inactivate BRCC36 in t(8;21)-positive AML cell lines - Kasumi-1 and SKNO-1 - and murine hematopoietic stem and progenitor cells (LSKs). Knockout was confirmed by a cleavage assay as well as Western blot. AML1-ETO-9a was expressed by a retroviral vector. Cell lines and LSK cells were treated with different concentrations of doxorubicin or cytarabine and their viability was assessed seven days post treatment. DNA damage was assessed through phospho-γH2AX staining using flow-cytometry.

Results

BRCC36 mutations were identified in 7 out of 191 patients (3.7%) with t(8;21) AML and none of 95 patients with inv(16). In the TCGA data set one out of 200 patients (0.5%) with de novo AML had a BRCC36 mutation. This patient had a complex karyotype and would be considered as secondary AML with myelodysplastic-associated changes according to the 2016 WHO classification. Six of the 7 mutations were missense or nonsense mutations that were predicted to be deleterious to BRCC36 function. One mutation affected a splice site at exon 6, resulting in an impaired splicing capability. With intensive standard chemotherapy all patients with BRCC36 mutations achieved a complete remission and had an estimated relapse-free and overall survival of 100% after a median follow up of 4.2 years. Given its role in DNA damage repair, we hypothesized that BRCC36 inactivation sensitizes AML cells to DNA-damage inducing drugs. In order to test this, we generated BRCC36 knockout Kasumi-1 and SKNO-1 cell lines using CRISPR-Cas9. BRCC36 inactivation had no impact on cell growth on either of the cell lines. However, we found that BRCC36 knockout cells were significantly more sensitive to doxorubicin as compared to the parental cells with normal BRCC36. This was accompanied by a significant increase in DNA damage as assessed by phospho-γH2AX in BRCC36 knockout vs control cells after doxorubicin treatment. In contrast, BRCC36 inactivation had no impact on cytarabine sensitivity. We next assessed drug sensitivity in primary murine leukemic cells expressing AML1-ETO-9a. Again, inactivation of BRCC36 resulted in a significant higher sensitivity to doxorubicin but not cytarabine.

Conclusion

We found BRCC36 to be recurrently mutated in t(8;21)-positive AML Inactivation of BRCC36 was associated with impairment of the DNA damage repair pathway and thus higher sensitivity to DNA damage-inducing chemotherapy. This might be also reflected by the favorable clinical outcome of patients with BRCC36 mutated t(8;21)-positive AML, a finding which has to be confirmed in a large patient cohort.