19S Proteasome Subunits As Oncogenes and Prognostic Biomarkers in FLT3-Mutated Acute Myeloid Leukemia (AML)

Introduction. Acute myeloid leukemia (AML) is a heterogeneous leukemic disease that results from abnormal stem cells in the bone marrow. Despite initial high response rates to standard induction chemotherapy, most patients eventually relapse due to common driver mutations, such as the FMS-like tyrosine kinase-3 (FLT3) gene. Patients with FLT3 mutations present with a poor prognosis and reduced overall survival (OS). FLT3 tyrosine kinase inhibitors (TKIs) like Midostaurin or Gilteritinib improve remission rates of AML patients, but 60% of patients experience drug toxicity or acquired resistance mechanisms. One approach to circumvent drug resistance is to target the ubiquitin-proteasome system (UPS). The UPS is a multi-protein complex that regulates protein homeostasis. It is comprised of two main subcomplexes, the 19S regulatory complex and the 20S core complex, which together recognize ubiquitylated proteins to facilitate UPS-dependent degradation. The 19S regulatory complex serves dual roles by i) recognizing and binding ubiquitylated proteins and ii) unfolding and translocating peptides into the 20S core complex for degradation. Proteasome inhibitors like Bortezomib reversibly target the 20S proteasome, and FLT3⁺ AML cells were reportedly more sensitive to proteasome inhibition than AML cells harboring wild-type FLT3. Thus, the UPS represents a promising target for the development of therapeutic drugs with efficacy in FLT3⁺ AML. We reported a role for two members of the 19S regulatory complex, proteasome 26S non-ATPase subunits 1 (PSMD1) and 3 (PSMD3), in drug resistance of chronic myeloid leukemia (CML) and several solid tumors. In the present study, we hypothesized that non-ATPase members of the 19S proteasome could serve as novel prognostic biomarkers and potential therapeutic targets in FLT3-mutated AML.

Methods. Using data from The Cancer Genome Atlas (TCGA) available at UALCAN (http://ualcan/path.uab.edu/) and GEPIA2 (http://gepia2.cancer-pku.cn/), we analyzed expression of 19S proteasome subunits in patients with FLT3-mutated AML. We phenotypically assessed the potential for PSMD3 to act as an oncogene in FLT3⁺ AML in vitro, using shRNAs targeting PSMD3 (shPSMD3) combined with apoptosis, colony formation, and luciferase reporter assays. AML cell line xenografts were performed to assess the role of PSMD3 in vivo. Finally, we analyzed other 19S proteasome subunits that could serve as novel prognostic biomarkers.

Results. Data from GEPIA2 revealed no significant difference in PSMD1 expression comparing AML versus normal cells. Consistent with these results, data from UALCAN revealed no correlation in OS comparing AML patients with high versus low PSMD1 expression (p=0.1), including patients with FLT3 mutations (p=0.3). In contrast, data from GEPIA2 revealed that PSMD3 mRNA expression was surprisingly downregulated in AML versus normal controls. However, high levels of PSMD3 mRNA expression correlated with a significant reduction of OS when all AML subtypes were considered (p=0.0029), and this was especially true in patients with FLT3 mutations (p=0.0015, Figure 1A). These findings implicate PSMD3 as a potential oncogene in FLT3-mutated AML. shRNA-mediated knockdown of PSMD3 (shPSMD3) impaired survival of the FLT3⁺ AML cell lines, MOLM-13 and MOLM-14, in colony formation assays ± the FLT3 TKI, quizartinib. Mice receiving shPSMD3-expressing cells had a significant increase in OS compared with mice receiving the shNT-expressing control cells (p=0.0027, Figure 1B). While PSMD3 regulated nuclear factor-kappa B (NF-κB) activity in CML, we did not observe similar effects in FLT3⁺ AML cells. Rather, proteomics analyses suggested a role for PSMD3 in neutrophil degranulation and energy metabolism. We next assessed whether expression of other PSMD subunits is altered in AML patients with mutant versus wild-type FLT3, and whether or not they correlate with OS. TCGA data demonstrated that the genes encoding PSMD2 (p=0.046), PSMD4 (0.044), PSMD6 (p=0.03), PSMD7 (p=0.046), PSMD8 (p=0.0032), PSMD9 (p=0.048), and PSMD13 (p=0.00022) were found to be altered in FLT3⁺ AML and/or high expression levels correlated with worse outcomes.

Conclusions. Collectively, these findings suggest that different components of the 19S regulatory complex of the 26S proteasome have indications for OS and may serve as prognostic biomarkers in AML and other types of cancers.