Poster Abstracts
Deep dives into the science are at the heart of what we do at BioCentury, and posters take the BioCentury Grand Rounds discussions further by putting data, methods and mechanism front and center. Posters will be displayed in the conference foyer, and featured during the interactive Networking Session "Discover hidden gems, the Old Fashioned Way: Poster Spotlight and Networking Reception" on Thursday, June 5 at 6:15 PM Central Time.
Get the conversation started by requesting 1x1 meetings with poster presenters via the BioCentury Grand Rounds 2025 digital platform.
PLATFORMS
- “Inside-Out” Proteins: Novel Biomarkers for Targeted Immunotherapy
- From synapse to solution: A scalable platform for longitudinal and high-throughput neuroplasticity profiling
- MSR-seq™: Comprehensive profiling of non-coding RNA
- Illuminating the Structural Surfaceome: Discovering a Novel Class of Tumor-specific Targets for First-in-Class Therapeutic Opportunities
- Lipid-protein interaction inhibitors as a new drug discovery platform
- Developing a High-Throughput Technique for Protein Stability Measurements: cDNA Display Pulse Proteolysis
BIOMARKERS & DIAGNOSTICS
CANCER THERAPIES
NEUROLOGY THERAPIES
THERAPIES FOR OTHER INDICATIONS
- “Inside-Out” Proteins: Novel Biomarkers for Targeted Immunotherapy
Associated Company/University: KossMek Biosciences | University of Chicago
Author: Anthony Kossiakoff
Abstract:
A major challenge in immunotherapeutic targeting is identifying cell surface protein biomarkers that are solely and abundantly expressed on malignant cells, while absent from healthy tissues. Despite extensive efforts, progress toward this goal remains static, hindering advancement of targeted biotherapeutic strategies in oncology. Using comparative cell surface proteomics, we identified a novel class of biomarkers, termed “Inside-Out” (I-O) proteins. In healthy cells, I-O proteins are sequestered within intracellular compartments such as the cytoplasm or nucleus. In malignant cells, these proteins are aberrantly translocated to the plasma membrane, conferring a tumor-specific surface phenotype. Importantly, I-O proteins exhibit resistance to antigen-loss escape, a common tumor immune evasion mechanism. The absence of I-O proteins on normal tissues, combined with their robust expression on cancer cells, positions them as optimal candidates for immunotherapeutic intervention. Using a high-throughput phage display platform, we generated and validated over 500 synthetic antibodies against 50 distinct I-O protein targets. These proteins displayed consistent, elevated surface expression across six diverse cancer cell lines, while remaining undetectable on healthy peripheral blood mononuclear cells (PBMCs), indicating regulation by oncogenic stress pathways. This tumor-restricted expression was further substantiated by analyses of patient-derived tumor specimens, which recapitulated the in vitro findings. Antibodies targeting several I-O proteins were efficiently internalized, making them attractive for antibody-drug conjugate (ADC) development. Whole animal imaging confirmed selective accumulation of I-O antibodies in tumor xenografts. These antibodies were engineered into bispecific T cell engagers (BiTEs), chimeric antigen receptor T cells (CAR-Ts), and ADCs, each demonstrating potent cytotoxicity across diverse cancer cell lines. In vivo, a BiTE-based approach resulted in significant regression of established tumors in a murine xenograft model. In vivo studies also established a good safety profile for targeting I-O proteins by showing virtually no observed toxicity in normal mice treated with CART constructs targeting them.
- From synapse to solution: A scalable platform for longitudinal and high-throughput neuroplasticity profiling
Associated University: Northwestern University
Authors: Pushpa Kumari and Yevgenia Kozorovitskiy
Abstract:
Neuroplasticity—the brain’s ability to rewire synaptic connections—is the foundation of learning and memory. Disrupted plasticity underlies a wide spectrum of brain disorders, from rare neurodevelopmental diseases to mood disorders, and neurodegeneration. Despite the promise of recently FDA-approved neuroplastogens like ketamine, drug discovery in this space remains paralyzed by outdated tools that lack scalability, precision, and physiological relevance. We have developed a next-generation, high-throughput screening platform that directly measures structural plasticity with unprecedented speed, specificity, and quantitative readout. Our platform uses genetically encoded nanoluciferase-based biosensors, transcriptionally driven by activity-dependent promoters, to quantitatively report dendritic spine potentiation—the core underlying phenomenon of neuroplasticity. Unlike traditional pathway-based or imaging-heavy screens, our luminescence-based readout is fast, scalable, and directly linked to synaptic strengthening. Optimized first in mature primary cortical neurons, this platform reflects true neuronal biology and allows cell-type-specific interrogation. Thus far, we have screened ~2,500 compounds—including FDA-approved drugs and natural product libraries—and have identified 20 potent neuroplasticity enhancers after rigorous statistical filtering. These hits show robust dose-response profiles and are being further validated using complementary assays and chemoinformatic prioritization based on bioavailability, toxicity, molecular targets, and opportunities for medicinal chemistry innovations. Critically, our platform also enables longitudinal, non-invasive tracking of synaptic plasticity in mouse models, opening the door to in vivo applications and personalized medicine. It breaks two major bottlenecks in CNS drug development: lack of high-throughput functional assays and absence of direct plasticity metrics. Our technology is more than a screening tool—it is a disruptive shift in how we identify and evaluate neuroplasticity-targeting therapeutics. By combining the precision of molecular neuroscience with the scale of modern drug discovery, we offer a powerful engine to accelerate the development of next-generation treatments for brain disorders rooted in dysfunctional plasticity.
- MSR-seq™: Comprehensive profiling of non-coding RNA
Associated Company: MesoRNA
Author: Douglas Liu
Abstract:
Non-coding RNAs ranging from 30-300 nucleotides, collectively termed “mesoRNAs,” represent 85% of nucleic acid molecules in human cells, orchestrate translation, and are markers of pathology ranging from cancer to neurodegenerative diseases. While next generation RNA technologies have transformed biological research and the practice of medicine, enabling new classes of therapies, vaccines, and diagnostics, most research has focused on only 15% of cellular RNA molecules, predominantly mRNA. mesoRNA’s rigid secondary structures and dense chemical modifications reduce conventional library preparation to fragmented snapshots. No current method offers the accurate and comprehensive view needed to study these abundant molecules. MSR-seq™ resolves these challenges with a patented hairpin adapter. The adapter ligates efficiently to any 3′ RNA end, and its 3′ ribonucleotide phosphate cap prevents self-ligation and adapter dimers. The adapter contains a barcode and avidin to allow for multiplexing and magnetic capture, so up to 96 samples can be processed in parallel without columns. A single run yields four data layers: mesoRNA abundance, nucleotide-level modification maps, amino-acyl charging ratios, and detailed fragmentation profiles. MSR-seq works equally well with cellular RNA or circulating RNA from plasma and urine. These innovations convert the previously inaccessible mesoRNA landscape into a high-throughput, quantitative, clinically practical assay. MSR-seq opens the door to a new generation of RNA based diagnostics and a richer systems level understanding of human biology. mesoRNAs have already shown diagnostic value in human studies in both tissue and blood. For example, tRNA fragments stratify metastatic risk, Y RNA fragments identify aggressive melanoma and exosomal snoRNAs flag malignant non-small cell lung cancer and colorectal lesions. MSR-seq can merge these isolated signals into multidimensional liquid biopsy assays with potentially greater sensitivity and longitudinal monitoring power.
- **Illuminating the Structural Surfaceome: Discovering a Novel Class of
Tumor-specific Targets for First-in-Class Therapeutic Opportunities
Associated Company: Immuto Scientific
Author: Faraz Choudhury
Abstract:
The discovery of truly tumor-specific targets remains one of the greatest challenges in oncology, as most conventional targets are shared with healthy tissues, leading to dose-limiting toxicities and therapeutic failures. Here, we explore a new class of targets—Tumor-Specific Surface Protein Conformations (SPC Targets)—that arise due to structural alterations in surface proteins unique to diseased cells.These previously undetectable targets provide exceptional specificity, addressing the critical challenge of “on-target, off-tumor” toxicity and enabling more effective therapeutic strategies. To unlock the discovery and identification of SPC targets, we have developed a target discovery platform that combines live-cell structural proteomics, ultra-sensitive mass spectrometry, and AI-powered computation. This SPC Target ID platform enables the identification of conformationally distinct epitopes, providing a transformative approach to target discovery. Further, the development of therapeutics against SPC targets requires a paradigm shift in antibody engineering. Our EPIC (Epitope-targeted In-Cell) Antibody Discovery Platform leverages structure-based antigen design, AI-constrained modeling, and precision counter-screening to generate antibodies that selectively bind SPC targets with high specificity and internalization potential. This strategy not only expands the druggable target space but also enhances the efficacy and safety of next-generation biologics. This poster will introduce SPC targets as a novel class of actionable oncology targets, detail the Immuto Scientific SPC Target ID and EPIC Antibody Discovery platforms, and discuss how this integrated approach is advancing our pipeline, including lead ADC programs in AML. By illuminating the structural surfaceome, we are unlocking a new frontier in precision oncology.
- Lipid-protein interaction inhibitors as a new drug discovery platform
Associated University: University of Illinois Chicago
Author: Wonhwa Cho
Abstract:
Lipids control numerous cellular processes via lipid-protein interaction (LPI) and dysregulated LPI caused by the reprogrammed lipid metabolism leads to diverse human diseases, including cancer. Taking advantage of the highly specific and variable nature of LPI sites, we recently pioneered the LPI inhibitor-based drug discovery platform (1,2). In a proof-of-principle study, we developed a first-in-class LPI inhibitor (WC36) for acute myeloid leukemia (AML), which potently and specifically blocked LPI of Syk kinase (1). Because WC36 only blocked aberrant LPI of Syk in AML cells, which is crucial for cancer cell survival and resistance to conventional Syk inhibitors, WC36 potently suppresses viability and drug resistance of AML cells without causing detectable site effects to normal cells. Further, AML cells could not develop resistance to WC36, demonstrating the potential of LPI inhibitors as potent and resistance-proof anti-cancer agents. In colorectal cancer (CRC) cells with commonly observed APC mutation, locally elevated cholesterol constitutively activated WNT-β-catenin signaling via a scaffold protein Dvl(2). Our LPI inhibitor for Dvl (WC522) potently and specifically blocked WNT-β-catenin signaling in CRC cells. Since WC522 blocked LPI of Dvl in CRC cells, which is essential for CRC cells but dispensable in normal cells, it suppressed CRC tumors in vivo without causing cytotoxic effects on somatic stem cells (2), unprecedented for a WNT-targeting inhibitor. Additional examples of LPI inhibitors showcasing their high potency, specificity, and safety as anti-cancer agents against various human cancers, including breast cancer, will be presented. References: 1) I. Singaram et al., Targeting lipid-protein interaction to treat Syk-mediated acute myeloid leukemia. Nat. Chem. Biol. 19, 239-250 (2023).
2) A. Sharma et al., Cholesterol-targeting Wnt/b-catenin signaling inhibitors for colorectal cancer Nature Chem Biol DOI: 10.1038/s41589-025-01870-y (2025).
- Developing a High-Throughput Technique for Protein Stability Measurements: cDNA Display Pulse Proteolysis
Associated University: Northwestern University
Author: Ted Litberg
Abstract:
Accurate prediction of protein stability is a critical challenge in protein engineering, with direct implications for drug development and industrial biotechnology. Proteins with low stability are prone to aggregation, degradation, and immunogenicity, making them poor therapeutic candidates. In contrast, highly stable proteins are better suited for clinical and industrial applications. Despite recent breakthroughs in machine learning (ML) and large language models (LLMs) for protein structure prediction and design, stability prediction has lagged due to datasets limited in both the amount of data and mutation types (overwhelmingly to alanine). To address this gap, we developed cDNA display pulse proteolysis, a high-throughputmethod that enables large-scale, quantitative measurements of protein stability. This approach builds on the cDNA display proteolysis platform pioneered by the Rocklin lab, which coupled display technologies with next-generation sequencing (NGS) to measure the stability of over 800,000 protein sequences. By incorporating elements of pulse proteolysis, our enhanced method overcomes key limitations of the original assay, including a limited dynamic range (0–5 kcal/mol) and an inability to distinguish local from global unfolding. Using this improved platform, we quantified the stability of a 12,000-protein test library containing five deep mutationally scanned proteins with hundreds of sequences that had been measured previously in low throughput by other labs. This library included sequences up to 108 amino acids long and with stabilities exceeding 5 kcal/mol, both beyond the range of our previous cDNA display proteolysis method. We have now scaled this to a library of over 500,000 sequences up to 120 amino acids, encompassing natural protein domains and de novo designs, with the original 12,000 serving as internal controls. This expansive dataset will help enhance the training and development of current and future protein stability prediction tools.
