In vitro reconstituted virus-like particles (VLPs) are promising gene delivery platforms for mammalian cell gene therapy. However, transfection of VLPs must be conducted with lipofectamine, a transfection agent that is known to be toxic to mammalian cells. This study aims to enhance cellular intake of the particles by the addition of lipid envelopes, creating enveloped virus-like particles (EVLPs) that are spontaneously taken up by mammalian cells without lipofectamine. This encapsulation of VLPs in lipids is optimized by varying the mass ratio of VLPs to lipids, pH and composition of the buffer, and wrapping method. Electron microscopy was used to verify the formation of EVLPs, and their uptake efficiency into baby hamster kidney cells at varying concentrations was compared to that of VLPs. The EVLPs without lipofectamine were found to have a slightly higher uptake efficiency than that of VLPs without lipofectamine. RNA extracted from EVLPs produced by drying-sonication methods show degradation, suggesting that RNA has increased susceptibility to degradation during sonication and prompting the need to explore alternative VLP-wrapping methods, such as freeze-thawing. Additionally, the use of shorter reporter genes for more consistent particle size and stability are explored, as shorter RNA sequences demonstrated better success in EVLP formation. These findings can help develop safer and more effective gene delivery systems for downstream therapeutic applications, treating diseases like cancer and genetic disorders.

Primary Hyperoxaluria Type 1 (PH1) is a rare autosomal recessive metabolic disorder leading to end-stage renal disease (ESRD). Pathologically, point mutations in alanine-glyoxylate aminotransferase (AGT) cause a mislocalized protein—typically from peroxisomes to mitochondria—or aggregate elsewhere. Physiologically, this disrupts AGT’s detoxification role and leads to oxalate buildup and recurrent kidney stones. Current therapies such as kidney stone removal and liver transplantation are insufficient, with high mortality rates especially in young patients. Our project aims to attenuate PH1 by restoring AGT function pharmacologically, using small molecules to redirect a mislocalized mutant from mitochondria to peroxisomes. Compounds like MitoBloCKs inhibit mitochondrial import and have shown promise in rescuing the P11L-G170R AGT mutant. We engineered a human liver model (HepG2 FlpIn) expressing AGT P11L-G170R fused to GFP11, with GFP1-10 targeted to peroxisomes. Upon small molecule treatment, successful protein relocation to peroxisomes is confirmed by green fluorescence upon GFP reconstitution. A high-throughput screen of ~55,000 compounds identified one that restored peroxisomal AGT localization and function comparable to wildtype. Some also appeared to increase peroxisome numbers, suggesting enhanced peroxisomal biogenesis. These findings underscore the therapeutic potential of targeting protein mislocalization and highlight our screening platform as a valuable tool for identifying treatments for PH1 and related disorders.

Data-independent acquisition mass spectrometry (DIA-MS) offers a powerful strategy for in-depth, robust proteomic profiling, but achieving comprehensive coverage of the human proteome remains a challenge. While current DIA-MS workflows in our laboratory routinely identify 8,000-9,000 proteins in whole cell samples, this falls short of the estimated 20,000 unique proteins encoded by the human genome. This project investigates the use of high pH reversed-phase fractionation of whole cell samples to increase protein identification depth in DIA-MS experiments. HEK293 cells were used as a model system. Initial single-shot (SS) DIA-MS search of an in-silico tryptic digest library identified 8,861-8,898 unique proteins at a 1% false discovery rate. To improve coverage of SS results, a comprehensive spectral library was generated from fractionated samples using DIA-NN software. This library was used to re-analyze the SS data, and a notable increase in protein identifications was observed: 900-1000 additional proteins per sample. Majority (42.6%) of these newly identified proteins appear to be categorized as low abundance proteins–which are often involved in uncharacterized pathways and avoid detection due to limitations in mass spectrometry acquisition pipelines. We are now investigating the use of fractionation on enhancing low abundance protein identification and designing experiment-specific spectral libraries for deeper proteome coverage, with hopes of advancing our lab’s exploration of tissue specific iron metabolism pathways.

The directed evolution of proteins towards a particular task is resource intensive, requiring large libraries, efficient screens, and time-consuming iterative rounds. Computational evolution offers an alternative to experimental methods for exploring sequence space and guiding functional optimization with relatively little effort. The project aim is to develop and validate a computational workflow for improving small-molecule-binding protein affinity using the active metabolite of the therapeutically-relevant drug irinotecan, SN38. Our workflow starts with a seed sequence as input. Protein language models generate plausible mutations, followed by structure prediction tools and their subsequent docking with the small molecule target. Top candidates undergo molecular dynamics and MM/GBSA calculations for refined binding energy estimates, which produce candidate seeds for iterative runs for further optimization. Starting with a known low-micromolar SN38 binder sequence, we’ve preliminarily generated low-nanomolar binders after two rounds of the workflow. Experimental validation can assess the workflow’s accuracy. If successful, this approach could generalize to other small-molecule-binding proteins and targets, accelerating binder development for various applications.

Redox-switchable polymerization offers a powerful way for controlled synthesis of block copolymers, enabling the sequential incorporation of distinct monomer units through external stimuli. This study explores the use of ferrocene-based catalysts as redox-active centers for mediating polymerization processes. The electrochemical properties of ferrocene allow for reversible oxidation and reduction, which can be used to modulate catalytic activity and selectively initiate polymerization of specific monomers. By toggling between oxidation states, the catalyst system enables sequential polymerization of monomers with orthogonal reactivities, leading to the formation of block copolymers. This approach allows for high control over molecular weight and block sequence, which can offer a versatile platform for the design of advanced polymers through electrochemical control.