Discovery and development of imaging-capable bacterial nitroreductases and their cytotoxic prodrugs for Clostridium sporogenes based gene therapy

Abstract

Clostridia Directed Enzyme Prodrug Therapy (CDEPT) has demonstrated anti-tumour efficacy in preclinical studies but has yet to be developed for human clinical trials. Challenges associated with stable genetic modification of the bacterial genome have recently been overcome, making CDEPT clinically feasible for the first time. The main goal of this project was to accelerate the development of nitroreductase-based CDEPT, initially by further improving nitroreductase activity with respect to PR- 104, a clinical-stage 3,5-dinitro-2-mustard prodrug. Random and site-directed mutagenesis techniques were used to optimise the E. coli NfsA polypeptide (UniProt accession P17117) for PR- 104A activation. The preferred multi residue variant #22GP (S41Y, E99G, L103M, R225P, F227S) was successfully expressed in three stable cell line models (HCT116, H1299 and C33A) and demonstrated significantly improved in vitro anti-proliferative activity with PR-104A compared to native NfsA (up to 2.3-fold), whilst retaining the ability to co-metabolise 2-nitroimidazole PET imaging substrates. In parallel, a series of PR-104A analogues were evaluated to eliminate off-mechanism aerobic metabolism by aldo-keto reductase 1C3 (AKR1C3) and to improve therapeutic activity by maximising bystander potency using 2D anti-proliferative and 3D multi-cellular layer assays for E. coli NfsA and/or AKR1C3 activity. In vivo efficacy studies on promising candidates resulted in the selection of a lead compound from each of three chemical classes; SN 34507 (an alcohol bearing 5- nitro-2-mustard phenylcarboxamide), SN 36001 (a piperazine bearing 2-nitro-4-alkylsulfone-5- mustard phenylcarboxamide) and SN 35288 (an alcohol bearing 2-nitro-4-alkylsulfone-5-mustard phenylcarboxamide). Finally, an in vivo model of CDEPT was identified. 1000mm3 H1299 xenografts treated with vascular disrupting agent DMXAA (5,6-dimethylxanthenone-4-acetic acid) produced widespread tumour necrosis and promoted colonisation of nitroreductase expressing Clostridia spores. The combination of DMXAA and Clostridia spores alone produced significant anti-tumour activity compared to untreated controls (TGD = 67%, P=0.004), but the inclusion of prodrug (PR-104) provided significant additional activity (TGD = 167%, P<0.001), validating the CDEPT concept. Successful preclinical evaluation of a transferable gene that metabolises both clinical stage PET imaging agents (for whole body vector visualisation) as well as chemotherapy prodrugs (for conditional enhancement of efficacy) is a valuable early step towards the prospect of nitroreductasearmed C. sporogenes entering clinical evaluation.