Despite the powerful impact in recent years of gene expression markers like the green fluorescent protein (GFP) to link the expression of recombinant protein for selection of high producers, there is a strong incentive to develop rapid and efficient methods for isolating mammalian cell clones secreting high levels of marker-free recombinant proteins. Recently, a method combining cell colony growth in methylcellulose-based medium with detection by a fluorescently labeled secondary antibody or antigen has shown promise for the selection of Chinese Hamster Ovary (CHO) cell lines secreting recombinant antibodies. Here we report an extension of this method referred to as fluorescent labeling in semi-solid medium (FLSSM) to detect recombinant proteins significantly smaller than antibodies, such as IGF-E5, a 25 kDa insulin-like growth factor derivative. CHO cell clones, expressing 300 μg/ml IGF-E5 in batch culture, were isolated more easily and quickly compared to the classic limiting dilution method. The intensity of the detected fluorescent signal was found to be proportional to the amount of IGF-E5 secreted, thus allowing the highest producers in the population to be identified and picked. CHO clones producing up to 9.5 μg/ml of Tissue-Plasminogen Activator (tPA, 67 kDa) were also generated using FLSSM. In addition, IGF-E5 high-producers were isolated from 293SF transfectants, showing that cell selection in semi-solid medium is not limited to CHO and lymphoid cells. The best positive clones were collected with a micromanipulator as well as with an automated colony picker, thus demonstrating the method's high throughput potential. FLSSM allows rapid visualization of the high secretors from transfected pools prior to picking, thus eliminating the tedious task of screening a high number of cell isolates. Because of its rapidity and its simplicity, FLSSM is a versatile method for the screening of high producers for research and industry.
Packaging or producer cell lines for scalable recombinant adeno-associated virus (rAAV) production have been notoriously difficult to create due in part to the cytostatic nature of the Rep proteins required for AAV production. The most difficult challenge being creating AAV packaging cell lines using HEK293 parental cells, currently the best mammalian platform for rAAV production due to the constitutive expression of E1A in HEK293 cells, a key REP transcription activator. Using suspension and serum-free media adapted HEK293SF carrying a gene expression regulation system induced by addition of cumate and coumermycin, we were able to create REP-expressing AAV packaging cells. This was achieved by carefully choosing two of the AAV Rep proteins (Rep 40 and 68), using two inducible promoters with different expression levels and integrating into the cells through lentiviral vector transduction. Three of our best clones produced rAAV titers comparable to titers obtained by standard triple plasmid transfection of their parental cells. These clones were stable for up to 7 weeks under continuous cultures condition. rAAV production from one clone was also validated at scale of 1 L in a wave bioreactor using serum-free suspension culture.
Lipoprotein lipase deficiency (LPLD) results from mutations within the lipoprotein lipase (LPL) gene that lead to a complete lack of catalytically active LPL protein. Glybera was one of the first adeno-associated virus (AAV) gene replacement therapy to receive European Medicines Agency regulatory approval for the treatment of LPLD. However, Glybera is no longer marketed potentially due to a combination of economical, manufacturing and vector related issues. The aim of this study was to develop a more efficacious AAV gene therapy vector for LPLD. Following preclinical biodistribution, efficacy and non-GLP toxicity studies with novel AAV1 and AAV8-based vectors in mice, we identified AAV8 pVR59. AAV8 pVR59 delivered a codon optimized, human gain-of-function hLPLS447X transgene driven by a CAG promoter in an AAV8 capsid. AAV8 pVR59 was significantly more efficacious, at 10- to 100-fold lower doses, compared to an AAV1 vector based on Glybera, when delivered intramuscularly or intravenously, respectively, in mice with LPLD. Efficient gene transfer was observed within the injected skeletal muscle and liver following delivery of AAV8 pVR59, with long-term correction of LPLD phenotypes, including normalization of plasma triglycerides and lipid tolerance, for up to 6 months post-treatment. While intramuscular delivery of AAV8 pVR59 was well tolerated, intravenous administration augmented liver pathology. These results highlight the feasibility of developing a superior AAV vector for the treatment of LPLD and provide critical insight for initiating studies in larger animal models. The identification of an AAV gene therapy vector that is more efficacious at lower doses, when paired with recent advances in production and manufacturing technologies, will ultimately translate to increased safety and accessibility for patients.