Abstract

Evolution and Production of a Clinical-Grade Adenovirus Expressing the Transgene Tissue Inhibitor of Metalloproteinase-3 (TIMP-3) Through Genome Engineering

Author(s): Stuart A. Nicklin, Simon D. Brown, Angelique Lemckert, Virginia Ammendola, Katie White, Rachel Dakin, Nicola Britton, Raul Alba Fernández, Oscar Lopez Franco, Colin Berry, Stefano Colloca, Menzo Havenga, Samantha Carmichael and Andrew H. Baker

Background: First-generation human adenoviral serotype 5 (HAdV5)-based vectors remain one of the most widely utilised gene transfer vectors for both experimental gene transfer and translational development for vaccine delivery and clinical gene therapy. We previously performed non-Good Manufacturing Practice (GMP), laboratory-grade production of a first-generation, replication-deficient vector expressing Tissue Inhibitor of Metalloproteinase-3 (TIMP-3) as a gene therapy approach to prevent vein graft disease following coronary artery bypass grafting. We have shown in pre-clinical in vitro, ex vivo and in vivo studies that adenoviral gene transfer of TIMP-3 reduces vascular smooth muscle cell (VSMC) proliferation, promotes VSMC apoptosis and reduces neointima formation in human saphenous vein ex vivo and in a pig vein graft model in vivo. Our next step was to translate HAdV5-TIMP-3 to the clinic; however, adenoviral-mediated TIMP-3 overexpression can be cytotoxic in multiple cell types. In this study, we describe the vector which led to successful production of a GMP batch of vector for a Phase 1 first-in-human clinical trial. Methods and Results: Initially, a low passage seed stock of a plaque-purified clone of HAdV5-TIMP-3 produced in either the pAdEASY system or based on the pJM17 system were compared and assessed in GMP production in a standard clinical grade 293 cell batch. Both vector configurations were assessed by PCR, Sanger DNA sequencing, and western blotting and immunofluorescence for TIMP-3 expression following transduction of HeLa cells. Transfer of the vector from the laboratory to standard clinical grade manufacturing protocols led to loss of TIMP-3 expression and the emergence of contaminating HAdV5 variants with different transgene configurations, including reversion of E1 DNA sequence. Loss of TIMP-3 expression, vector mutations and sequence rearrangements were observed in multiple clones from each configuration by vector passage 5-6. Next, three strategies were undertaken to facilitate vector production at clinical grade: (i) use of a less-active cytomegalovirus immediate early promoter (CMVIEP), (ii) reverse orientation of the expression cassette to reduce TIMP-3 expression levels and (iii) incorporation of tetracycline (Tet) operator sequences to silence TIMP-3 expression in 293 Tet Repressor (TetR) helper cell lines. Strategy iii resulted in successful completion of ten stable passages of vector amplification in commercial 293 T-Rex cells as evidenced by stable AdTIMP-3 batches (HAdV5.CMV.TO.TIMP-3) with expected DNA sequence and equivalent expression levels of TIMP-3 relative to the original non-GMP HAdV5-TIMP-3. AdTIMP-3 production over 10 passages in GMP-grade 293TetR cell lines was sufficient to maintain expected genome configuration assessed by PCR and sequencing, low particle: infectious unit ratios and stable TIMP-3 expression by western blotting. Transfer of the pre-master viral seed stock of HAdV5-CMV.TO.TIMP-3 to clinical grade manufacture produced a high titre batch of HAdV5.CMV.TO.TIMP-3 which passed QP (qualified person) release and was approved for use in a first-in-human clinical trial to assess its utility in preventing saphenous vein graft disease. Conclusions: In summary, transfer of experimental batches of first-generation HAdV5 vectors expressing TIMP-3 into clinical grade manufacture required silencing of TIMP-3 expression during the multi-passage scale-up required. Tetracycline regulation is sufficient to achieve this only where expression is completely silenced and can be achieved in GMP-certified 293TetR cell lines.