The Department of Chemical and Biological Engineering is pleased to present student seminars by Yash Chainani and Beth Dibiase.

Yash Chainani will present a seminar titled, "Enhancing computer-aided synthesis planning by merging enzymatic and synthetic organic chemistry."

ABSTRACT: Engineering microbial cell factories to produce valuable commodity chemicals at scale has the potential to reshape industries ranging from pharmaceuticals to transportation fuels. Yet designing novel bio-synthetic and semi-synthetic pathways de novo remains a slow and arduous process, often requiring exhaustive literature searches and numerous trial-and-error experiments in the lab. To accelerate this process, we extract reaction templates from publicly available databases of experimentally validated reactions, which we use to generate large-scale reaction networks (or knowledge graphs) in silico. We then develop machine learning models to sift through these resulting networks and eventually identify only the most feasible pathways. A key challenge in training such models is the absence of negative enzyme-reaction data in the literature. To address this gap, we strategically infer infeasible reactions from reported feasible ones and synthetically generate negative datapoints. These data allow us to train supervised classifiers that can generalize beyond known biochemistry, predicting enzyme specificity and reaction feasibility at scale. Building on this foundation, we then incorporated multifunctional polyketide synthases (PKSs) to create a new pathway design software, BioPKS Pipeline. PKSs are large modular enzymes that function as molecular assembly lines and are responsible for the biosynthesis of a wide range of valuable natural products, including antibiotics, immunosuppressants, and anticancer agents. While sometimes difficult to engineer, they are invaluable in catalyzing carbon-carbon bond formation reactions and consequently, creating elongated carbon backbones. By modeling PKS domain architectures and their catalytic logic, we extend pathway design from single-enzyme transformations to complex, multi-domain enzymatic systems capable of generating novel scaffolds. Most recently, we have expanded this framework even further by coupling enzymatic reactions with synthetic organic chemistry. Biology and chemistry offer complementary advantages: biology provides selectivity, sustainability, and modular enzymatic logic, while chemistry brings efficiency, robustness, and the ability to access transformations not easily achieved enzymatically. We finally demonstrate that by merging biology and chemistry, we can access a wider space of chemicals than would be possible using either route alone.

Beth Dibiase will present a seminar titled, "Engineering Extracellular Vesicles as Modular Synthetic Biology Tools for Scalable Delivery of Biotherapeutics."

ABSTRACT: Engineered extracellular vesicles (EVs) are emerging treatment modalities for many diseases. EVs are nano-sized lipid-based nanoparticles derived from nearly all cell types. As native vessels of intracellular communication, EVs possess natural advantages for use in biotherapeutic delivery, such as the ability to cross biological barriers, encapsulate unstable cargo, and exhibit low immunogenicity and toxicity. One of their most prominent advantages for use as modular synthetic biology tools, however, is the ease of engineering their properties and functions, including cargo loading, surface display, biophysical characteristics, and vesiculation rates, among others. These benefits make engineered EVs uniquely suited to address several key challenges in biotherapeutic delivery. Despite their promise, there remain some limitations to the widespread use of engineered EVs. These limitations include difficulties in quantifying surface functionalization of EVs in absolute terms, reducing nonspecific uptake and clearance, and low functional delivery efficiency.

While several strategies for reducing nonspecific uptake of biotherapeutic nanocarriers have been translated into EVs, fast clearance times and off-target effects remain a challenge relative to synthetic delivery vehicles such as lipid nanoparticles. Modulating key membrane biophysical properties of EVs may allow us to enhance their on-target cell association/uptake while reducing their nonspecific uptake, which may also have implications for in vivo clearance times and biodistribution. We hypothesized that by varying the loading amounts of a polyethylene glycol-like biopolymer called mucin (Muc1) on our EVs, we could tune their biophysical properties. To investigate this hypothesis, I engineered a tunable, inducible genetic circuit for mucin expression in parental cell lines and loading into EVs. Using this platform, I characterized key biophysical properties of the engineered EVs, such as morphology, size, zeta potential, membrane fluidity, and protein loading, developing both bulk and single vesicle quantification methods. Next, I optimized an assay to explore the effects of these biophysical changes on targeted delivery of a protein. The ability to modulate and measure the biophysical properties and cargo delivery of a cell-derived nanocarrier is a critical step towards enabling enhanced targeted biotherapeutic treatment.

Bagels served at 9:30am, seminar to start at 9:40am.