Shedding light on the immune response to AAV gene therapy

Abstract

AAV vectors are one of the most promising tools in retinal gene therapy. However, accumulating evidence points to the relevance of AAV's immunogenicity in the clinical setting. In light of this, it is critical to have a better knowledge of the mechanisms underlying immunological responses to AAV vectors. In this PhD project we focused on the innate immune responses to AAV vector- and AAV production system-dependent factors that modulate vector immunogenicity in different in vitro models and its impact on AAV-mediated retinal gene therapy. In order to do that, a wide range of basic-research techniques, including biochemistry, molecular biology, microscopy, cell culture and organotypic culturing methodologies as well as preclinical-research analyses and interpretation of SD-OCT and FAF were performed. The main results achieved in this dissertation may be summarised as follows: 1. Experimental AAV8 and AAV2 vectors induced lot-specific innate immune responses in human pDCs which were neither specific to the capsid/vg ratio nor the production platform nor the manufacturer. Also, innate immune responses in pDCs were dependent on TLR9 signalling pathway, which could be reduced by pre-treatment with DNase. Furthermore, DNase treatment increased the transduction rate across all AAV8 vector lots in HEK293T cells. This suggests that both HEK293- and Sf9-cell derived preparations of experimental AAV vectors may contain a variety of extra-viral DNA impurities. Some of them cause lot-specific innate immune responses in human pDCs and impair transduction in HEK293T cells. 2. cgAAV lots without any known extra-viral DNA impurities but carrying different amounts of HCP were not able to elicit any innate immune response in neither pDCs nor in PMA-differentiated THP-1 cells. This may suggest that either the amount of HCP content found in our cgAAV vectors lots may not be high enough to induce immune responses in our in vitro models, or that HCP impurities are less immunogenic than extra-viral DNA impurities. 3. PRR ligands induced innate immune responses in ROs at early timepoints suggesting the presence of at least TLR2, 4 and 9. On the other hand, immunogenic AAV vectors that were sensed by pDCs were not able to elicit innate immune responses in retinal models. However, AAV vectors were able to transduce ROs. 4. SD-OCT analysis of NHPs after clinical grade AAV-mediated subretinal gene therapy revealed that, although the surgical procedure by itself induced the appearance of HRF in the ONL, the number of HRF increased over time in the high dose AAV-treated eyes. In addition, immune cell infiltration was detected in retinal sections from AAV treated eyes. Interestingly, the cgAAV lot used in this preclinical study contained lower HCP levels than the cgAAV lots used in THP-1 and pDCs in which no immune response was detected. This implies that an immune response to AAV and not to extra-viral impurities, appears to be the most plausible explanation for the group of animals injected with the high dose, and the HRF observed on SD-OCT could represent activated microglia activation into outer retinal layers. Finally, we showed that extra-viral DNA contaminants can affect the immunogenicity and potency of AAV vector lots. This has significant implications for the study of innate immune mechanisms involved in AAV vector recognition (basic research), the safety and efficacy of AAV-mediated gene therapy in animal models or human patients (translational and clinical research), and it also provides new information on how to adapt the AAV production process to generate safer and more effective gene therapy vectors (vector production methods).