Archived

This content is available here strictly for research, reference, and/or recordkeeping and as such it may not be fully accessible. If you work or study at University of Kentucky and would like to request an accessible version, please use the SensusAccess Document Converter.

Author ORCID Identifier

https://orcid.org/0009-0005-5655-7646

Date Available

1-11-2027

Year of Publication

2026

Document Type

Doctoral Dissertation

Degree Name

Doctor of Philosophy (PhD)

College

Engineering

Department/School/Program

Chemical and Materials Engineering

Faculty

Dibakar Bhattacharryya

Faculty

Gosia Chwatko

Abstract

The production of viral vectors, such as adeno-associated virus (AAV), poses a central challenge in translating gene therapies from research to clinical application. A persistent bottleneck in manufacturing is the downstream purification process, where maintaining high product yield, impurity reduction, and vector integrity at industrially relevant scales is crucial. Membrane-based clarification and purification offer significant advantages over ultracentrifugation, including scalability, reduced processing times, and lower capital costs. However, their broader adoption for nanoscale biologics has been constrained by limitations in membrane–particle selectivity, fouling control, and predictable recovery. This dissertation, supported by the NSF and NIEHS, addresses these limitations by investigating membrane morphology, functional chemistry, and responsive behavior for viral vector purification. In Phase 1, the role of microfiltration membrane morphology in nanoparticle separation was evaluated by varying porosity, thickness, pore structure, and hydrophilicity. Hydrophilicity and porosity correlated positively with model AAV2 recovery, while asymmetric membranes exhibited higher porosity and flux, reducing operating time by up to 100× compared to overnight centrifugation. An optimal hydrophilic asymmetric membrane achieved 100% model AAV2 recovery, while reducing model cell debris by ~44% and protein by ~35%. Phase 2 examined the pore-to-particle size ratio (λ = dₘ/dₚ) on recovery, retention, and flux stability, combining literature analysis with controlled experiments. Across both micro- and ultrafiltration, high retention (≥95%) occurred consistently for λ < 0.5, while λ << 1 yielded minimal flux decline. Amine-functionalized hydrogel-coated membranes extended retention beyond size-exclusion limits, capturing >70% of 20 nm negatively charged particles at λ > 40 and retaining ~80% of DNA and protein at λ > 160, demonstrating selective, scalable purification via electrostatic control. Phase 3 developed a nanoscale model of crude AAV2 harvest (lysed HEK293 cells + 20 nm silica particles) to evaluate primary (depth filtration) and secondary (membrane-based) clarification. At high flux (600 LMH), depth filters showed reduced DNA reduction due to weakened electrostatic/adsorptive interactions, whereas low flux (150 LMH) enabled >90% DNA removal without DNase. Secondary clarification revealed material-specific trade-offs: polyether sulfone (PES) membranes achieved >90% AAV2 recovery, while regenerated cellulose (RC) membranes reduced DNA by >80%.

Incorporating optimized primary clarification increased PES flux from ~60 LMH to ~600 LMH. Phase 4 assessed a pH-responsive carboxybetaine methacrylate (CBMA)-functionalized deconstructed depth filter made of sequential PES layers (1.2–0.1 μm), allowing electrostatic control over impurity removal and product recovery. With AAV2 lysate, pH adjustments tuned outcomes: pH 8 maximized recovery (>88%); pH 4 maximized impurity removal (>89% DNA, >70% protein) while maintaining moderate recovery, potentially saving ~$100,000 per batch by avoiding DNase; and pH 6 yielded higher flux but low recovery and separation, suggesting a cleaning mode to extend membrane life. Optimal conditions were validated in a two-stage CBMA-functionalized hollow fiber module (0.65 μm followed by 0.1 μm), with flat-sheet results correlating strongly, supporting rapid screening-to-scale-up translation. Collectively, these studies integrate morphological optimization, tunable electrostatics, and responsive control to expand the performance of membrane-based purification. The findings bridge mechanistic insight with process-scale application, advancing frameworks for high-yield, impurity-selective, and industrially scalable AAV2 manufacturing for gene therapy.

Digital Object Identifier (DOI)

https://doi.org/10.13023/etd.2026.10

Archival?

Archival

Funding Information

This study was supported by National Science Foundation (NSF) grant number 2218054 and the National Institute of Environmental Health Sciences of the National Institute of Health under Award Number P42ES007380.

Download

Available for download on Monday, January 11, 2027

Contact Author

Share

COinS