Abstract 4733: Human functional brain imaging data support preclinical and clinical evidence that marizomib crosses the blood-brain barrier (BBB) to inhibit proteasome activity in the brain

Abstract

Abstract Background: Here we summarize key data suggesting, unlike bortezomib, marizomib (MRZ), an irreversible inhibitor of all proteasome subunits, crosses the BBB, supporting its evaluation in brain malignancies. Methods: Male Swiss Webster mice with microdialysis probes implanted in the cerebellum (CER) and prefrontal cortex (PFC) were administered MRZ (0.3 mg/kg IV, n = 15), amphetamine (1 mg/kg IP, positive control, n = 4), or vehicle (n = 6). Neurotransmitter levels were measured up to 180 minutes post-dose. Brain samples were collected from MRZ-treated mice at 30, 60, 120 (n = 3 each) and 180 (n = 6) minutes post-dose to determine proteasome activity (chymotrypsin-like [CT-L], trypsin-like [T-L], and caspase-like [C-L] subunits). Proteasome activity was assessed in normal human brain (CER and frontal lobe; n = 6) and glioblastoma tumor samples (GBM; n = 30), and in peripheral blood mononuclear cells (PBMCs) from patients with recurrent GBM (rGBM) prior to and after MRZ administration (MRZ-108 study, NCT02330562). Longitudinal resting-state functional magnetic resonance imaging (fMRI) of whole-brain volumes was acquired before and after MRZ administration to patients with rGBM (n = 6). Results: In mice, MRZ significantly increased 3,4-Dihydroxyphenylacetic acid (DOPAC) and dopamine levels in the CER and decreased levels of serotonin (CER and PFC), 5-Hydroxyindoleacetic acid (CER and PFC), Homovanillic acid (PFC), and DOPAC (PFC). Proteasome activity of all 3 proteasome subunits (CT-L, C-L, T-L) was significantly reduced in mouse CER and PFC after MRZ administration; CT-L activity was most potently inhibited. In humans, similar CT-L and C-L activity levels were observed in GBM tumor tissue compared with normal brain CER, while activity levels were 3- to 4-fold higher in normal brain frontal lobe. In patients with rGBM, CT-L activity in PBMCs was inhibited 80-100% 1 hour post-MRZ, however, activity levels had returned to baseline prior to the next MRZ infusion 7 days later. Resting-state fMRI data showed, after MRZ exposure, hallucination severity was associated with decreased functional connectivity between left lingual gyrus and both left CER (T[4] = −12.78; P < 0.03 FDR) and left temporal cortex (T[4] = −9.56; P < 0.04 FDR). Conclusions: MRZ demonstrated rapid, yet transient proteasome inhibition in PBMCs. This activity pattern may be related to proteasome turnover and occur in other nucleated cells, such as glioma cells. Lower proteasome activity levels were observed in CER and GBM tumor tissue compared with frontal lobe tissue. Proteasome inhibition and neurotransmitter level alterations in the mouse brain and functional connectivity changes in the human brain suggest that MRZ crosses the BBB and specifically affects the CER. Taken together, these data suggest GBM tumor tissue, like the CER, may be sensitive to MRZ treatment. Citation Format: Daniela A. Bota, Kaijun Di, David B. Keator, Robert G. Bota, Matthew Hoffmann, C. Dan Dumitru, Nancy Levin. Human functional brain imaging data support preclinical and clinical evidence that marizomib crosses the blood-brain barrier (BBB) to inhibit proteasome activity in the brain [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 4733.