Research Talks

Title: Using genetics to improve disease screening: an application to prostate cancer

Bio

John Witte is a Professor of Epidemiology & Population Health, Biomedical Data Sciences, and Genetics at Stanford University. He is the Vice Chair of Epidemiology & Population Health and co-Leader of the Population Sciences program in the Stanford Cancer Institute. Before Stanford he was on the faculty at UC San Francisco and at Case Western Reserve University. He is an internationally recognized expert in genetic epidemiology, focused on developing and applying machine learning methods to decipher the genetic basis of cancer and other complex traits. His work has used comprehensive genome-wide studies of germline genetics, transcriptomics, and somatic genomics toward determining who is most likely to be diagnosed with clinically relevant disease and require additional screening or specific treatment. John Witte has mentored over 50 graduate students and postdoctoral fellows, serves on the executive committees of multiple graduate programs, and has directed National Institutes of Health funded pre- and post-doctoral training programs in genetic epidemiology for over 20 years.

Abstract

Biomarkers are commonly used to screen for numerous different diseases. For instance, prostate-specific antigen (PSA) levels are commonly used to screen men for early signs of prostate cancer. However, many of these biomarkers are highly heritable, making it difficult to distinguish between high levels due to genetic predisposition versus high levels due to early-stage disease. In addition to diluting useful prognostic signals, the use of heritable biomarkers puts individuals with naturally elevated biomarker levels at risk of overdiagnosis and overtreatment, unnecessarily exposing them to side effects and imposing avoidable financial costs. To address this, we show how one can use machine learning to perform robust genetic personalization of biomarkers with an application to PSA and prostate cancer. In particular, we demonstrate that the accuracy of prostate cancer screening can be significantly improved by removing “genetic noise” from PSA levels, and that this can further allow for developing a polygenic risk score predictive of aggressive disease. This personalization framework holds promise for improving any non-causal yet heritable disease biomarker.

Title: Leaps of Faith - My Biotech Journey

Bio

Patrice Gilooly is a senior quality executive with over 30 years of experience in biopharmaceutical manufacturing, spanning early clinical through commercial launch and post-marketing phases. Patrice was most recently Senior Vice President of Quality Assurance and Operations for Regeneron Pharmaceuticals, Inc., a multi-national, fully integrated biotechnology company that develops highly specialized parenteral therapies for serious unmet medical needs such as age-related macular degeneration, atopic dermatitis, asthma, cancers and chronic obstructive pulmonary disease. During Patrice’s 27-year career with Regeneron, she contributed to the company growth from a $100 million market capitalization to over $100 billion and was primarily responsible for the company’s GMP Quality Assurance program and personnel globally. This responsibility spanned multiple functions and geographies, including a global team that grew from approximately 20 US-based to over 1000 QA professionals across North America, Europe and Asia as well as oversight of dozens of contract manufacturing organizations (CMO). She was instrumental in growing the quality culture and strong compliance mindset for a clinical phase and small-scale manufacturing organization to a global leader in large scale commercial biologics manufacturing. She was also involved in the approval of over 10 biologic license approvals (BLA) for fusion proteins and monoclonal antibody therapies as well as an emergency use authorization for a COVID-19 therapy.

Abstract

This presentation will reflect on a transformative 30-year career in the biotechnology industry, highlighting pivotal moments, technological advancements, and evolving trends. From early-stage clinical development to the commercialization of groundbreaking therapies, the journey spans critical innovations in drug discovery, facility design, quality system maturity, operational excellence and organizational development. Key lessons learned in navigating regulatory landscapes, fostering cross-disciplinary collaborations, and driving innovation will be shared, along with insights into the future trajectory of biotech. The presentation aims to provide a comprehensive view of the industry’s progress and the personal and professional growth that has accompanied a dynamic and rewarding career.

Title: Reconstructing the History of Genomic Island Insertions in Clades of Microbes

Bio

Eliot Bush is a professor at Harvey Mudd College where he’s been teaching courses in computational biology, bioinformatics and evolution since 2007. He received an A.B. from Harvard University and a Ph.D. from the California Institute of Technology. In college he discovered his interest in evolution, and in graduate school realized he could have fun using computers to study it. Since then, he's worked on a wide range of research projects. Among other things he's used CT scans to study a 35-million-year-old primate fossil, modeled the evolution of metabolism, and built computational tools to identify horizontal transfer events in bacteria. He's co-author of two textbooks in computational biology and bioinformatics (the second one forthcoming). He also writes for a popular audience. Outside of work he enjoys playing ultimate frisbee, riding his mountain bike around the local hills, and baking things for his family.

Abstract

Genomic islands play an important role in microbial genome evolution, providing a mechanism for strains to adapt to new ecological conditions. A variety of computational methods have been developed to identify them. Some of these methods are explicitly designed to work in single strains, while others make use of multiple strains. In general, previous methods do not identify islands in the context of the phylogeny in which they evolved. We have developed a software package, xenoGI, that identifies genomic islands and maps their origin within a clade of closely related bacteria, determining which branch they inserted on. It takes as input a set of sequenced genomes and a tree specifying their phylogenetic relationships. Making heavy use of synteny information, the package builds gene trees and then reconciles them with the species tree in order to identify gene families. It then attempts to combine into islands those families whose members are adjacent and whose most recent common ancestor is shared. We demonstrate the capabilities of the package with several examples from enteric bacteria.

Title: Backstage Passes to the Brain: Delivering Therapeutics Past the Blood Brain Barrier

Bio

Barbara Bailus joined KGI in 2020 with an expertise in genetics and genetic engineering. Born and raised in Northern California, she completed both her bachelors and doctoral studies at University of California, Davis. Her doctoral work in Dr. David Segal’s lab focused on genetic engineering for Angelman Syndrome, a rare neurodevelopmental disorder. Her postdoctoral work in the laboratory of Dr. Lisa Ellerby focused on potential therapies for Huntington’s disease including small molecules and gene editing proteins. Her current laboratory at the Keck Graduate Institute focuses on delivery of therapeutics past the blood brain barrier for neurological disorders, with a special emphasis on Angelman syndrome and Huntington’s disease. As a recognized expert on Angelman Syndrome, Bailus serves as the Chair of the Scientific Advisory Board for The Foundation for Angelman Syndrome Therapeutics, the largest funder for Angelman research in the world.

Abstract

Angelman Syndrome (AS) is a rare neurodevelopmental disorder caused by the loss of UBE3A expression in the central nervous system (CNS). In the last fifteen years much, progress has been made in both understanding the basic genetics of AS and developing tools for treating AS. However, one area that still needs further optimization is the method of delivery to the brain for these novel therapies. Multiple promising technologies, including ATFs, cDNA, CRISPR, and ASOs, have been successful in the rodent models of AS, but achieving widespread delivery of these therapeutics in a human brain remains elusive and is a major limiting factor in the effectiveness of these therapeutic approaches. Our study has focused on optimizing a novel delivery technology for delivering therapeutic proteins to the brain, based on various cell penetrating peptides. This system has previously achieved widespread delivery of several proteins from an intravenous injection, in a murine model. We are adapting this system for use with UBE3A. Our presentation will highlight the progress we have made in optimizing and testing these proteins in cellular models, organoids and mouse models.

Title: Interrogating the Fitness Landscape of SARS-CoV-2 Receptor Binding Domain Mutants’ Interactions with ACE2

Abstract

Since the first cases of COVID-19, the SARS-CoV-2 coronavirus has accrued many mutations that have stabilized in the strains profiled today, settling into a range of Omicron subvariants. Mutations in the SARS-CoV-2 Receptor Binding Domain (RBD) are known to have consequences on binding affinity with the Human Angiotensin Converting Enzyme 2 (hACE2) receptor. To study the nature of the protein-protein interactions at play, our experiments were designed to investigate the fitness landscape of mutations observed between the earliest strains of SARS-CoV-2 and the current strains. To investigate which combinations of mutations conferred the greatest increase in binding affinity to hACE2, we generated a combinatorial library of SARS-CoV-2 RBD structures, expressed them utilizing a yeast display platform, and performed an affinity enrichment assay to see how the population of all possible mutations compete for binding.

Title: Hepatic Amyloid Precursor Protein (APP) Is Localized in Mitochondria and May Be a Source of Peripheral Amyloid-Beta (aB): Implications in Alcohol-Dependent Alzheimer’s Disease (AD)

Abstract

Alcohol-induced liver injury occurs in the pericentral region of the liver and can induce mitochondrial remodeling and exacerbate Alzheimer’s Disease (AD) progression. Subpopulations of mitochondria such as the general mitochondria (GM), peridroplet mitochondria (PDM) and endoplasmic reticulum (ER)-bound mitochondria (ERM) maintain cellular homeostasis and damage in these regions can contribute to AD pathogenesis. Hepatic amyloid precursor protein (APP) is a source of peripheral amyloid beta (aB) and can confer risk in AD pathology. Understanding the localization and expression of hepatic APP in mitochondrial subpopulations are therefore critical to understanding aB metabolism and may play a role in identifying a potential mechanism for metabolic dysfunction. Here, site-specific expression of hepatic APP and aB processing proteins are measured in the subpopulations of the mitochondria of ethanol-fed AD mice to identify the regions where aB metabolism occurs. Double transgenic (APP/PS1) AD mice were intra-gastric fed with ethanol or control diet for 5 weeks (n =7-11/group). Liver tissue was harvested for digital spatial profiling (DSP) analysis to measure pathogenic AD biomarkers (APP, PSEN1, BACE1) in the periportal and perivenous regions using sequence-specific fluorescent probes. Mitochondrial fractions were isolated using differential centrifugation and homogenized for protein measurement using western blotting and co-immunoprecipitation to identify whether aB is localized to the mitochondria. Hepatic APP expression increased 2-fold and PSEN1 increased over 5-fold in the mitochondria compared to the cytosol of both ethanol-fed and non-ethanol fed AD mice (p≤0.05). APP expression increased in the GM and ERM but not in the PDM of ethanol-fed mice compared to control (p≤0.05). APP transcript increased in the pericentral region of ethanol-fed mice (p≤0.05) and both APP and aB showed an increased trend at the protein level (p=0.06 and p=0.07, respectively). Differential expression of APP and aB across mitochondrial subpopulations suggests that hepatic mitochondria may be a source of peripheral aB and affect specific mitochondrial processes in the GM and ERM. Increased expression of APP in the pericentral region of AD mice correlates to the same site affected in alcohol-induced liver damage and may assist in identifying potential pathways involved in AD progression due to alcohol-induced liver impairment. Localization of aB in hepatic mitochondria suggests that aB processing in the liver may be a source of aB in the brain. Future studies modulating hepatic APP in vivo will provide future insights on the contribution of hepatic aB on AD pathogenesis.

Title: How to Manufacture a Better Autologous CAR-T Cell Therapy Product

Abstract

There are patients alive today because chimeric antigen receptor (CAR) T cell therapy exists. CAR-T cell therapy is a unique treatment that utilizes a patient’s own cells as the starting material for their disease. CAR-T cell therapy manufacturing technology has advanced considerably in the last eight years. Since the first approval by the Food and Drug Administration and European Medicines Agency in 2017, CAR-T cell therapy has become an important treatment modality against heme malignancies including lymphoma, leukemia, and multiple myeloma. Standard CAR-T cell therapy manufacturing methods take one to two weeks while accelerated methods take one week from receipt of leukapheresis (aph) to final drug product. The standard manufacturing method starts with T cell enrichment, activation, genetic engineering, expansion, harvest, formulation, and cryopreservation. Magnetic-activated cell selection has become the preferred tool to isolate T cell populations from aph while static T cell culture methods have given way to dynamic culture methods. Taken together, new manufacturing technologies demonstrate promise in shortening production time, raising manufacturing success rates, improving product quality, decreasing overall cost, and increasing patient access. This presentation will discuss existing and emerging methods for the manufacture of CAR-T cell therapy products.

Title: Regulatory Challenges and Future Directions for the Use of Artificial Intelligence in Drug Development

Melissa Mendez, Larry J. Davis, PharmD

From drug target identification and optimization to clinical trial design and post-market safety surveillance, Artificial Intelligence (AI) is quickly transforming pharmaceutical development. The challenges of drug development have been difficult to ignore and continue to hamper industry ability to address important unmet medical needs and drug shortages in a timely manner. AI is offering significant potential for addressing these challenges by increasing efficiency, reducing costs, and raising success rates. However, this wave of innovation also presents regulatory and ethical challenges that must be addressed to ensure safe and responsible use of AI in drug development. To make sure safe and effective treatments are reaching patients at a faster rate, it is crucial that regulatory bodies stay ahead of these emerging technologies while creating a clear framework for their implementation.

Title: Content Analysis of Price Communication in Direct-to-Consumer Pharmaceutical Television Advertising

Andrew Parker, Maxim Polonsky

Our study examines how different types of financial information convey value to consumers in a comprehensive sample of 127 prescription drug video advertisements for 101 unique products aired in the United States from January 2020 to March 2024. We analyze the quantity and nature of price communication messages, focusing on their modality, placement, and duration within the ads. Google Trends data was also used to investigate the link between aggregate price communication in DTC pharmaceutical advertising and consumer interest in advertised products. The results indicate that current price communication strategies are not well-aligned with best practices for advertising effectiveness, highlighting a need for improvement. These findings have significant implications for the pharmaceutical industry, suggesting a need for improved price communication strategies in ads. To our knowledge, this is the first study to systematically investigate price and financial communication in pharmaceutical video advertising.

Title: Evolutionary Implications of Cross-Species Transmission: Computational Analysis of SARS-CoV-2 Spike Protein Binding

Christopher Perucho, Jonathan Felix, Ilya Tolstorukov, Animesh Ray

The recent COVID-19 pandemic featured a wide variety of novel SARS-CoV-2 variants that all differ in their prevalence but the force behind their evolution is still being understood. This prevalence is related to the variant's fitness which includes many factors but when simplified to one factor such as binding affinity to native receptor vs the variants in question a landscape known as a fitness landscape can be generated. Investigation into the rise of SARS-CoV-2 suggests that cross-species transmission from a natural reservoir across a variety of hosts led to the human-specific strain seen in 2019. This strain is likely to have evolved through the use of an altered fitness landscape derived from these cross-species transmissions. To understand this altered fitness landscape and its effect on novel variant generation, computational modelling featuring SARS-CoV-2 spike protein variants complexed to mouse or human ACE2 receptors was performed. From this, affinity estimations served as a basis to generate fitness landscapes for each receptor-variant complex and utilized to determine the effects of multiple cross-species transmissions.