Development of orthotopic glioblastoma (GBM) patient-derived xenograft (PDX) models for efficacy evaluation of sorafenib combined with a synthetic flavonoid dimer

Abstract

Breast Cancer Resistance Protein (BCRP, ABCG2) and P-glycoprotein (P-gp) at the blood-brain barrier (BBB) have been proposed to efflux anticancer agents out of the brain and thereby restrict their usage on glioblastoma (GBM) clinically. Sorafenib, a tyrosine kinase inhibitor (TKI), showed promising cytotoxicity on GBM cells in vitro, but Phase I/II clinical trials were unsuccessful. This project mainly investigated if a novel synthetic flavonoid dimer, Ac12Az9, can inhibit the BCRP/ABCG2- and P-gp­-mediated sorafenib efflux at the BBB to improve GBM treatment. From an in-house library of 300-member of triazole-bridged flavonoid dimer, a potent and nontoxic dual inhibitor, Ac12Az9, was previously discovered with an EC50 of 0.9­-1.4 nM and 285 nM for reversing BCRP/ABCG2- and P-gp-mediated drug resistance in vitro, respectively. Ac12Az9 not only inhibited BCRP/ABCG2-mediated efflux of sorafenib, increasing intracellular sorafenib from 54% to 108% (p<0.001) in MDCKII­-GFP-BCRP cells, but it also blocked P-gp-mediated efflux of sorafenib in MDCKII-P­-gp cells, increasing intracellular sorafenib from 70% to 78% (p < 0.01). The transepithelial transport assay further demonstrated that Ac12Az9 can decrease the basolateral (B) side to apical (A) side (Papp B-to-A) transport of sorafenib by 81% or 8% in MDCKII-GFP-BCRP or MDCKII-P-gp cell monolayers, respectively. In animal experiments, Ac12Az9 can enhance brain penetration of sorafenib, with a 1.8-fold increase in the brain-to-plasma ratio, to a therapeutic level without causing toxicity. In an orthotopic patient-derived xenograft (PDX) model called G22-FLuc, co-administration of Ac12Az9 and sorafenib could suppress tumor growth by 56.7% (p = 0.13) and prolong the life span of tumor-bearing nude mice by 47.4% (p = 0.01) when compared to the solvent control group. To improve the half-life (t1/2) of Ac12Az9, a chemical modification was performed. Metabolism of Ac12Az9 resulted in a metabolite M1 (Ac12Az9-COOH) that was inactive in reversing BCRP/ABCG2- (EC50 = 470 nM) and P-gp- (EC50 > 1000 nM) mediated drug resistance in vitro. Esterase was proposed to be responsible for the cleavage of the ester bond of Ac12Az9 to generate M1. Therefore, the chemical modification was performed by focusing on the ester linkage in Ac12Az9. Ac12Az9 was subjected to an ester-based modification to produce the derivatives D1-D11. Compound D6 was found to have increased plasma stability (t1/2 = 270 min versus < 60 min for Ac12Az9) and good BCRP/ABCG2 modulating activity, with EC50s of 1.2-1.3 nM for reversing BCRP/ABCG2-mediated topotecan (TPT) resistance in vitro. In a subcutaneous HEK293/R2 xenograft model, co-administration of D6 and TPT could increase the tumor accumulation of TPT by 2-fold (p = 0.0028) compared to TPT alone. Ac12Az9 was also found to increase the brain permeability for AZD1775, a Wee1 inhibitor. AZD1775 was not only found to be effective against GBM cells (IC50 = 0.59­-1.14 μM) but also to be a good substrate for both BCRP/ABCG2 and P-gp. Ac12Az9 was found to be able to reduce the BCRP/ABCG2- or P-gp-mediated efflux of AZD1775 by 143% or 63% in MDCKII-GFP-BCRP or MDCKII-P-gp cells, respectively. In PK studies, Ac12Az9 can enhance the brain permeability of AZD1775, resulting in an increase in the brain-to-plasma ratios of 16.5 or 29 times. In summary, Ac12Az9, a potent dual inhibitor of BCRP/ABCG2 and P-gp, can be used in combination with sorafenib for treating GBM. Ac12Az9 derivative - D6, a specific inhibitor of BCRP/ABCG2, could potentially be used in another combination therapy for the treatment of BCRP/ABCG2-overexpressing tumors.