MeiraGTx Holdings plc announced the Company will exhibit eight posters and deliver one oral presentation at the European Society of Gene and Cell Therapy (ESGCT) 2023 Annual Congress, which is being held from October 24-27, 2023, in Brussels, Belgium. Poster #122: Development of alternatives to Triton X-100 cell lysis for AAV2, AAV5 and AAV8 primary recovery; Category: AAV & non integrative vectors; Triton X-100 is an effective detergent for recovering biological products from intracellular compartments, but its use is now prohibited by Registration, Evaluation, Authorisation and Restriction of Chemicals (REACH) regulations as its degradation generates a compound that negatively impacts aquatic life. This work aimed to develop alternative methods to Triton X-100 for Adeno-Associated Virus (AAV) recovery from Human Embryonic Kidney 293 cells which were REACH-compliant, Good Manufacturing Practice-compatible, did not affect product quality and showed comparable product recovery relative to lysis with Triton X-100.

Two alternative AAV release methods were evaluated for AAV2, AAV5 and AAV8 serotypes: 1) A detergent-free hyperosmotic release method was initially tested at microplate scale using a Central Composite Design of Experiment approach, before scaling up to 250 mL stirred tank reactor (STR) scale. 2) When not suitable, a lysis method using Deviron C16 detergent was assessed at 250 mL STR scale with concentrations ranging from 0.1 to 0.5%. Finally, successful conditions for either method were scaled up to 10 L STR scale.

For AAV8 serotype, a hyperosmotic AAV release method employed the addition of 400 mM NaCl and incubation for 2 hours, leading to a ~70% VG recovery relative to the legacy Triton X-100 method. For AAV2 and AAV5 serotypes, a 0.5% Deviron C16 lysis method proved most suitable, resulting in VG recoveries of 87% and 97% relative to AAV release with Triton X-100, respectively. No downstream process changes were needed to accommodate these alternative methods.

Furthermore, both methods were scalable to 10 L STR scale, and process-related impurities including residual DNA and Host Cell Proteins, as well as product potency, were not impacted by these methods. In conclusion, this study demonstrates two optimized alternative AAV release methods to substitute the use of Triton X-100 that showed comparable product recovery and product quality profiles. Poster #167: Potency assay cell line development for ocular gene therapy vectors; Category: AAV & non integrative vectors; When producing GMP AAV drug products, the ability to assess their potency is an essential part of release and stability.

Cell-based potency assays are a regulatory requirement for commercialisation of AAV gene therapies but poor AAV in vitro transducibility as well as inability to test expression due to the use of tissue-specific promoters is hindering the development of efficient and robust assays. Similarly, basic pre-clinical research, where easy-to-implement tests for faster paced experiments and iterations are desirable, suffers from the same limitations. Company therefore set out to examine ways with which to increase vector transducibility as well as establishing ocular promoter trans-activation in the most commonly used cell lines (HEK293 and HeLa).

Company assessed different methods to increase the transducibility of AAV5 and AAV8 vectors as well as trans-activate photoreceptor-specific (Rhodopsin kinase) and RPE-specific (RPE65) promoters. In dose-ranging experiments, identified potent dCas9-mediated trans-activators for both promoters and established exogenous factor supplementation that increased the transducibility of both capsids multiple folds over baseline. Further studies will enable to combine these in an in vitro cell-based potency assay platform for GMP batch release and stability testing.

Poster #324: Use of mechanistic modeling to design a platform process for the separation of full and empty AAV capsids; Category: Manufacturing Adeno-associated viruses are a relative newcomer to the field of biopharmaceutical modalities and are used to deliver a therapeutic gene to a patient. During their upstream production, separate cellular processes are used to produce the viral capsid and the therapeutic transgene, and package the transgene within the capsid. This leads to the expression of empty capsids which must be removed during the downstream process as they may stimulate an immune response in the patient.

The separation of empty capsids presents a challenge due to the similarity in properties between the empty and full capsids. The proportion of empty capsids can vary widely and be as high as 90% depending on the maturity of the upstream process which can further increase the challenge in achieving this separation. Most attempts have focused on using anion exchange chromatography to exploit the difference in charge to achieve this separation.

However, there is a lot of diversity in the published approaches with some groups using different matrix types (resins, membranes, and monoliths), process conditions, and additives. Company has previously demonstrated that weak partitioning can be used to maximise the enrichment of full capsids which further increases the available options for this separation. Establishing a platform process typically involves screening a number of options while using heuristics to narrow the design space and reduce the experimental burden.

This typically leads to process options being compared at sub-optimal conditions and there is a risk that the optimal platform may not be identified. This presentation will focus on work that have done to demonstrate that mechanistic models can be used to identify an optimal platform process for the enrichment of AAV2 full capsids. In this study, developed mechanistic models for the separation of full and empty AAV2 capsids for three AEX matrices that had previously shown promise during experimental evaluation.

These models were then used to examine a range of process conditions (salt concentrations, load ratios) and the modes of operation (bind and elute, flowthrough, weak partitioning) and identify the optimum conditions for each matrix. Finally, the identified optimum conditions were verified experimentally.