Development of in vitro 3D model systems for screening non-viral neurotrophic factor gene therapies for the retina

Abstract

Glaucoma is a neurodegenerative disease that can lead to a complete loss of vision due to retinal ganglion cell (RGC) death. Therapies that have the capacity to protect and rescue stressed RGCs remain a critical unmet need in glaucoma management. Neurotrophic factor (NF) gene therapy is a promising therapeutic approach that can address this current clinical deficiency by providing damaged RGCs with extrinsic neurotrophic support as a means of protection and repair. Moreover, a non-viral approach to the delivery of NF-encoding plasmid DNA (pDNA) confers many advantages over a viral approach for its improved immunogenicity and mutagenesis risks, patient compliance and large-scale manufacturing cost and feasibility. In this research, the main objective was to address the challenges facing non-viral NF gene therapy field for the retina, through development of three in vitro model systems that aim to facilitate the preclinical screening and identification of promising NF gene delivery systems. The first model system developed was a versatile co-culture model that simulates cellular interactions between “healthy” and “stressed” cells in the retina. Furthermore, through incorporation of techniques including enzyme-linked immunosorbent assay (ELISA), immunofluorescent imaging, and neurite tracing into the co-culture setup, the model system enables a systematic evaluation of the therapeutic potential of gene delivery systems through assessment of bioavailability and bioactivity of therapeutic proteins produced from transfected cells. The second model system was a new potential RGC cell line, termed XFC series of cells, that express key RGC characteristics and suitable for the evaluation of RGC-aimed gene therapies. Derived from multipotent retinal stem cells (RSCs), XFC cells express multiple RGC markers including Map-2, Rbpms, and Tubb3, and exhibit RGC-like neurite extension capacity in response to brain-derived neurotrophic factor (BDNF), ciliary neurotrophic factor (CNTF), and rho-kinase inhibitor (RKI) Y-27632 activation. The feasibility of the cell model was further validated in the described co-culture setup as XFC cells were able to validate the bioactivity of the BDNF proteins released by transfected cells. The third model system developed was a stem cell-derived 3D “mini-retina” culture model (termed MiEye series of retinal neurospheres) that contains multiple retinal cell types and enables in vivo-like gene delivery assessment. Derived from differentiating multipotent RSCs in 3D culture, MiEye retinal neurospheres with different retinal biomarker expressions can be generated using different protocols. Moreover, by harnessing the tissue-like arrangement of retinal cells in MiEye retinal neurospheres, it enables the assessment of infiltration and transfection capacity of gene delivery systems in tissue-like structure, towards the establishment of a more representative in vitro-in vivo correlation and prediction of in vivo gene delivery feasibility. Concurrent to model system development, aspects that focus on the development of non-viral gene delivery systems for the retina were also explored. The first aspect involved the optimization of gemini surfactant (GS) lipid nanoparticle systems (GL-NPs) physicochemical properties by evaluating the roles of minicircle plasmid (MC), sonication processing, and total NP component concentration (TNPC). Through dynamic light scattering and fluorescent correlation spectroscopy physicochemical characterizations, it was found that the size, particle size distribution, and zeta potential could be effectively optimized through sonication processing and TNPC. Moreover, the number of pDNA per particle homogeneity can be improved by formulating GL-NPs with MCs. The second aspect involved an investigation on the application of carbon nanotubes (CNTs) as a gene delivery vehicle to retinal cell types. More specifically, GS-functionalized SWNT gene delivery system (ƒ-ptSWNT) was developed and demonstrated the ability to deliver pDNA to a retinal astrocyte cell line. The results demonstrate the feasibility of utilizing ƒ-ptSWNT to deliver pDNA to retinal cells and serves as a starting point for future ƒ-ptSWNT retinal gene delivery system development. The development of the three in vitro model systems in this thesis collectively aims to facilitate the preclinical screening and development of non-viral NF gene therapies in a synergistic manner, covering key areas of assessments that are critical to in vivo therapeutic success. Furthermore, concurrent developments in non-viral gene delivery systems through GL-NP optimization and CNT exploration also advance the knowledge towards the development of better non-viral gene delivery systems.