Poster Judging and Prizes

All Early Career Researchers who submit a poster will be considered for a poster prize. There are three prizes available: Best Abstract, Best Lightning Talk,, Best Visual Presentation, and Best Overall.

Abstracts will be evaluated after submission closes, but in advance of the Annual Meeting, and will be evaluated on the following criteria:

• Scientific Background & Approaches: Is the background and/or purpose for the science well-described and clear? Are the approaches and methods included and well-defined?
• Scientific Conclusion & Results: Are the results (actual or anticipated) clearly presented? Are the conclusions or next steps reasonable and in line with the presented information? (Consider in relationship tothe stage of research.)
• Is the patient impact well understood?

The top 10 posters based on abstract performance will advance to the second round of voting. These top 10 posters will be evaluated on the lightning talk and a visual inspection of the poster. There will be no interaction between the poster presenters and the poster judges. This evaluation will use the following rubric.

Poster Abstracts

1-Predicting anti-cancer drug combination responses with a temporal cell state network model

Marc Birtwistle, Clemson University

Cancer chemotherapy combines multiple drugs, but predicting the effects of drug combinations on cancer cell proliferation remains challenging, even for simple in vitro systems. We hypothesized that by combining knowledge of single drug dose responses and cell state transition network dynamics, we could predict how a population of cancer cells will respond to drug combinations. We tested this hypothesis here using three targeted inhibitors of different cell cycle states in two different cell lines in vitro. We formulated a Markov model to capture temporal cell state transitions between different cell cycle phases, with single drug data constraining how drug doses affect transition rates. This model was able to predict the landscape of all three different pairwise drug combinations across all dose ranges for both cell lines with no additional data. While further application to different cell lines, more drugs, additional cell state networks, and more complex co-culture or in vivo systems remain, this work demonstrates how currently available or attainable information could be sufficient for prediction of drug combination response for single cell lines in vitro.

Other Authors: Deepraj Sarmah, Wesley Meredith, Ian Weber, Madison Price

2-  Development of a Glioblastoma Cell State Network Model to Implement Cell State-Directed Therapy

Brandon Bumbaca, SUNY Buffalo

The complete treatment armamentarium of glioblastoma multiforme (GBM) - drug, surgery, and radiation - has yielded meager results with a median survival of 15 months. Treatment failures are multi-faceted and may be related to the high degree of intratumoral heterogeneity and the epigenetic plasticity of GBM cells which allow them to transition to more resistant states. To study these phenomena, publicly available multi-omics datasets from GBM patient were analyzed to categorize patients into four major cell states: (i) neural progenitor-like, (ii) oligodendrocyte progenitor-like, (iii) astrocyte-like, and (iv) mesenchymal-like. Next, a protein-protein interaction network (PPIN) was constructed for each of the four patient subpopulations. These PPINs were combined to form one Boolean network that served as the basis to conduct in silico knockout simulations of protein nodes. Our analysis identified multiple potential drivers of cell state transitions including TP53, E2F1/E2F4, ATM, TFAP2A, and PTEN. To further explore these predictions, a second large GBM database was utilized with a boosted tree machine learning model. The model could accurately predict cell state after being trained on the results of the in-silico knockout simulations. We then applied Shapley Additive Explanations to understand the predictions of each patient individually. We found that hepatocyte growth factor is activated upon transition to the MES-like state, potentially by TFAP2A, while CDK2 is consistently activated upon transition to the NPC-like state. These hypotheses – consistent with cell-state directed therapy- can be used to explore how drugs can be used to affect cell state transitions and mitigate plasticity and treatment resistance.

Authors: James M. Gallo, Marc R. Birtwistle, Jonah R. Huggins

3- Extreme wrinkling of the nuclear lamina is a morphological marker of cancer

Christina Dubell, Texas A&M University

Nuclear atypia is a common characteristic of human cancers, but how nuclear shapes become abnormal during cancer development is not well understood. A recently proposed model for nuclear shaping posits that folds/wrinkles in the nuclear lamina allow nuclei to assume a range of shapes with little mechanical resistance. In extreme deformations such as flattened nuclei in 2D cell culture, the folds are smoothed geometrically producing a nucleus that is “stiff” to further flattening. However, whether nuclei conform to such a model in human tissues generally, and during cancer development specifically, is not known. Here, we imaged nuclear lamins in patient tissues and found that a) laminar wrinkles were present in both control and cancer tissue, but were generally masked in hematoxylin and eosin (H&E) images, b) nuclei rarely had a smooth lamina, and c) wrinkled nuclei assumed diverse shapes. Deep learning analysis demonstrated higher rates of extreme wrinkling in cancer tissues than in control tissues. Fourier analysis of nuclear contours revealed systematic differences between cancer tissues and control tissues, consistent with the increased prevalence of high-frequency contour waviness. These data support a model in which lamina wrinkles allow varied nuclear shapes in vivo, and they indicate that extreme wrinkling is a marker of diverse cancers. Further, sensitive imaging of the nuclear lamina may improve the diagnosis of cancer pathology.

Authors: Ting-Ching Wang*, Christina R. Dubell*, Sneha Mishra, Hailee Patel, Samere Abolghasemzade, Ishita Singh, Vilmos Thomazy, Daniel G. Rosen, Vlad C. Sandulache, Saptarshi Chakraborty, Tanmay P. Lele (*co-first authors)

4- The Mechano-Metabolic Crosstalk Driving 3D Breast Cancer Invasion is Regulated by YAP/TAZ Activity

Jacopo Ferruzzi, University of Texas at Dallas

Cancer progression is driven by both cell autonomous and micro-environmental factors. The role played by a collagen-rich extracellular matrix (ECM) is particularly acute in breast cancer, a disease associated which collagen densification and alignment. YAP and TAZ are established oncoproteins that regulate cell behavior in response to a variety of stimuli, including ECM mechanics and cell metabolism. However, it is yet unknown how enhanced nuclear activity of YAP/TAZ drive ECM mechanics and cell metabolism during breast cancer invasion, and how such mechano-metabolic crosstalk impacts the modality of invasion. Therefore, in this study we used tumor spheroids to investigate how nuclear activation of YAP/TAZ regulate a mechano-metabolic crosstalk between ECM mechanics and cell metabolism, thereby influencing invasion in complex 3D environments. Doxycycline-inducible nuclear mutants MCF-10A TAZ(4SA) and YAP(5SA) were used to form spheroids which were embedded in collagen I. By using a combination of biophysical measurements, metabolic assays, and imaging methods, we show that constitutive activation of nuclear YAP/TAZ impacts breast cancer spheroid proliferation, traction force generation, collagen remodeling, and matrix stiffening. We also show that activation of nuclear YAP/TAZ alters the metabolic profiles of tumor spheroids in a spatially and temporally heterogeneous manner. With regards to invasion, metabolic inhibition experiments show that collectively invading cells display a tendency for oxidative phosphorylation while single cell invasion appears to be fueled by a glycolytic metabolism. In conclusion, this work provides novel insights into the relationship between matrix mechanics and YAP/TAZ activity, specifically as it pertains to regulating cancer cell metabolism and invasion.

Other Authors: Adil Khan, Haider Ali, Bishant Karki, Xaralabos Varelas, Jacopo Ferruzzi

5- State-transition Modeling of Blood Transcriptome Predicts Disease Evolution and Treatment Response in Chronic Myeloid Leukemia

David Frankhouser, City of Hope

David E. Frankhouser1, Jihyun Irizarry2, Anupam Dey3, Sergio Branciamore1, Lisa Uechi1, Dandan Zhao2, Denis O’Meally4, Lucy Ghoda2, Haris Ali2, Jeffery M. Trent5, Stephen Forman2, Yu-Hsuan Fu2, Ya-Huei Kuo2, Adam MacClean3, Bin Zhang2, Guido Marcucci2, Russell C. Rockne1

1Department of Computational and Quantitative Medicine, Beckman Research Institute, City of Hope National Medical Center, Duarte, California, 91010, USA 2Department of Hematologic Malignancies Translational Science, Beckman Research Institute and Division of Leukemia, City of Hope National Medical Center, Duarte, California, 91010, USA

3Department of Quantitative and Computational Biology, University of Southern California, Los Angeles, California, 90089, USA

4Department of Diabetes and & Cancer Discovery Science, Beckman Research Institute, City of Hope National Medical Center, Duarte, California, 91010, USA

5Translational Genomics Institute, Phoenix, Arizona, 85004, USA

Chronic myeloid leukemia (CML) is initiated and maintained by BCR::ABL expression which can be induced in transgenic mouse models. Thus, CML is an ideal system to test our hypothesis that transcriptome-based state-transition math models can accurately predict cancer evolution and treatment response. In previous studies we have used our approach to model both the bulk transcriptome1 and the micro-RNA transcriptome2 in acute myeloid leukemia and the bulk transcriptome in CML3. Here, to investigate CML at the single cell level, we preformed scRNA-seq on time-sequential blood samples from tetracycline-off BCR::ABL-inducible transgenic mice. By creating pseudobulk (PsB) samples from the scRNA-seq data, we constructed a CML state-space that mapped into our previously generated bulk CML state-space. We then showed that the sc data could be fit using the same mechanistic model to define a three-well leukemogenic potential landscape used for the bulk data. Because the variance in the data at the level of single cells identifies and is dominated by the difference between cell types, we could not detect a state-transition at the sc-level. To determine how leukemia is encoded by individual cells, we created pseudobulk samples for each cell type and showed that the cell type PsB data does encode the leukemia state-transition. Therefore, we extended our mechanistic model to incorporate the different cell types as linear combinations of their PsB transcriptomes and found that the linear combination of the cell type PsB also created a three-well potential landscape. Importantly, the construction of the state-space produces a loading value (eigengene) for each gene which can be used to quantify how each gene expression change at different states of CML contributes to leukemia. Using our model to guide our bioinformatic analysis, we are currently investigating the level of individual cells to identify which gene expression dynamics and altered biological processes contribute most to CML progression. Our approach for scRNA-seq data integrates information from all genes and all cells to provide a theory-based approach to both investigate and model the dynamics of complex processes at the system-level.

References
[1] RC Rockne, S Branciamore, J Qi, DE Frankhouser, D O'Meally, WK Hua, G Coo, E Carnahan, L Zhang, A Marom, H Wu. State-transition analysis of time-sequential gene expression identifies critical points that predict development of acute myeloid leukemia. Cancer research. 80(15):3157-69. 2020

[2] DE Frankhouser, D O’Meally, S Branciamore, L Uechi, L Zhang, YC Chen, M Li, H Qin, X Wu, N Carlesso, G Marcucci. RC Rockne, Y-H Kuo. Dynamic patterns of microRNA expression during acute myeloid leukemia state-transition. Science advances. 8(16):eabj1664. 2022

[3] DE Frankhouser, RC Rockne, L Uechi, D Zhao, S Branciamore, D O’Meally, J Irizarry, L Ghoda, H Ali, JM Trent, S Forman, Y-H Fu, Y-H Kuo, B Zhang, G Marcucci. State-transition modeling of blood transcriptome predicts disease evolution and treatment response in chronic myeloid leukemia. Leukemia. Feb 2:1-2. 2024

Other Authors: Jihyun Irizarry2, Anupam Dey3, Sergio Branciamore1, Lisa Uechi1, Dandan Zhao2, Denis O’Meally4, Lucy Ghoda2, Haris Ali2, Jeffery M. Trent5, Stephen Forman2, Yu-Hsuan Fu2, Ya-Huei Kuo2, Adam MacClean3, Bin Zhang2, Guido Marcucci2, Russell C. Rockne1

6- Single-cell analysis of state-transition from health to leukemia and the leukemic cell heterogeneities in inv(16) acute myeloid leukemia (AML)

Yu-Hsuan Fu, City of hope

Acute myeloid leukemia (AML) is an aggressive malignancy characterized by genetic heterogeneity. Using a conditional knock-in mouse model (Cbfb 56M/+ /Mx1-Cre) that mimics the acquisition of the CBFB-MYH11 (CM) fusion gene, this study explores transcriptomic changes and cellular heterogeneity during leukemia progression through single-cell RNA sequencing (scRNA-seq). CM expression was induced in the mouse model, and scRNA-seq analysis was conducted on peripheral blood mononuclear cells and bone marrow from leukemic and wild-type controls, analyzing over 105,000 cells. Principal component analysis (PCA) revealed a Y-shaped distribution, with a distinct branch corresponding to leukemic cells. Key genes and pathways driving leukemia progression were identified at a single-cell level, highlighting upregulated inflammatory responses and downregulated pathways such as Myc targets, oxidative phosphorylation, and ribosomal biogenesis. Nine distinct leukemic clusters were identified among cKit+ bone marrow cells, displaying unique features such as high heme metabolism, high proliferation, and dysregulated oxidative phosphorylation. Specific clusters showed elevated Cd9 and Egfl7 expression, indicating potential targets for further study. This research uncovers the transition from health to leukemia at single-cell resolution, revealing significant biological heterogeneity in the inv(16) AML model. These findings provide a framework for future studies on leukemia evolution and transcriptomic changes at the single-cell level.

Other Authors: Yu-Hsuan Fu, Lianjun Zhang, Ying-Chieh Chen, David E. Frankhouser, Denis O'Meally, Lisa Uechi, Jihyun Irizarry, Sergio Branciamore, Guido Marcucci, Russell Rockne, Ya-Huei Kuo

7- Enhancing Reusability and Adoption of Cancer Research Tools: Development of the Cancer Complexity Toolkit

Aditi Gopalan, Sage Bionetworks

The Cancer Complexity Knowledge Portal (CCKP) is an open-access data repository that stores and shares resources from multiple cancer research programs. Maintained by Sage Bionetworks, in collaboration with The Multi-Consortia Coordinating (MC2) Center, which supports six NCI Division of Cancer Biology (DCB) funded programs made up of interdisciplinary communities of scientists, CCKP centralizes access to cancer research tools and data, improving their discoverability and reuse.

Reusing tools has been challenging, so we are developing the Cancer Complexity Toolkit to examine tool reusability and increase adoption. The toolkit aims to: (1) enhance discovery and reuse of cancer research tools through rich metadata and standardized evaluation frameworks, (2) create a catalog of cancer research tools to improve accessibility and confidence, and (3) enable standardized reporting and validation of tool functionality.

To achieve these goals, we are (1) refining a standardized schema to capture metadata on tool parameters and usage, (2) building infrastructure for automated checks to assess documentation, test coverage, and community metrics, and (3) developing tool cards to display relevant metrics and metadata, helping researchers select the right tools and allowing developers to highlight their tools' impact.

In Phase 1, we refined the tool metadata schema and developed a Python-based testing infrastructure to check for documentation, testing files, and deployment on non-native Linux environments. We selected ten tools from the CCKP for initial testing, generating logs that assess their reusability.

Other Authors: Aditi Gopalan, Brad MacDonald, James Eddy, Thomas Yu, Susheel Varma, Jineta Banerjee

8- Photoimmunotherapy to target cancer cells, spare T cells, and engage anti-tumor immunity

Rebecca Harman, Spring Lab Northeastern University

Photoimmunotherapy (PIT) uses a cancer cell-targeted antibody conjugated with photosensitizers used for photodynamic therapy (PDT) to deplete cancer cells while sparing neighboring off-target cells, including intra-tumoral T cells. Preserving and increasing the number of T cells in and around the tumor is a crucial treatment goal towards priming the immune system, as intra-tumoral T cell presence is an important predictor of outcome in many cancers and can be an indicator of response to immunotherapy. In an immunocompetent pancreatic cancer mouse model, we analyzed live-animal in vivo hyperspectral microendoscope images and ex vivo immunohistochemistry-stained slices pre- and post-PDT to show that the number of CD3+CD4+ and CD3+CD8+ T cells at the tumor site increases significantly post-PDT. In a 2D cancer cell—peripheral blood mononuclear cell (PBMC) coculture, PIT spares lymphocytes, including T cells, while significantly reducing cancer cells. In a 3D Matrigel dome model combining ovarian cancer spheroids and PBMCs, a synergistic effect of cancer cell-PIT combined with antibody-mediated immune cell killing of cancer cells is observed in both overall cancer cell viability and the depletion and disruption of tumor spheroids. Collectively, these experiments are being used to parameterize a mathematical model of tumor-immune dynamics that will help to design dosimetry for optimal local and distal anti-tumor immune stimulation.

Other Authors: Mohammad U. Zahid, Eric M. Kercher, Ryan T. Lang, Mohammad Ahsan Saad, Sudip Timilsina, Heiko Enderling, Tayyaba Hasan, Bryan Spring

9- Precision biomaterials sustain durable and massive expansion of human CAR-T cells in vitro

Xiao Huang, Drexel University

Enhancing cell persistence is essential for achieving durable outcomes in chimeric antigen receptor (CAR) T cell therapy. Unlike genetic modifications, which carry safety risks, our approach focuses on externally programming T cells using functional biomaterials. We developed a DNA scaffold-based technology on PLGA microparticles, enabling precise control over the density and ratio of various immune modulatory signals. Using this platform, CAR-engaging particles (CAREp) were created by coating CAR-antigens and CD28 agonist antibodies (a-CD28 Ab) on surfaces to stimulate human CD8+ CAR T cells in vitro. Remarkably, with an optimal ratio of the two signals, CAREp sustained T cell proliferation for over 100 days, achieving cumulative expansions of ~10¹²-10¹⁸ folds. This unprecedented expansion was consistent across multiple donors and CAR constructs, significantly surpassing results from U87 cancer cells or standard CD3/28-Dynabeads.

Critically, the massively expanded cells displayed enhanced mitochondrial fitness and effector function. Single-cell RNA sequencing revealed clonal enrichment of memory progenitor and durable effector cells with excellent gene set scores compared to clinical standards. Mechanistically, CAREp induced a gene expression program that restricted differentiation, supported cell proliferation, preserved telomeres, and enhanced mitochondrial respiration. This program contrasted sharply with the differentiation-focused pathways activated by U87 cells. Additionally, telomerase activity was temporarily elevated during initial CAREp stimulations, stabilizing telomere length and extending expansion capacity. These findings highlight the potential for CAREp to improve CAR-T cell persistence in vivo and offer valuable insights into T cell signaling, guiding ongoing research.

Other Authors: Landon Flemming, Wendell Lim, Qizhi Tang, Tejal Desai

10- Sex-distinct Transcriptomic Signatures Underlie MRI-defined Edema Patterns in Human Gliomas

Pamela Jackson, Mayo Clinic

Magnetic resonance imaging (MRI) is key to clinically managing brain tumor patients, however connecting the biology to imaging remains challenging. We developed a two-compartment model of MRI signal intensity to quantitatively estimate relative edema abundance from T2-weighted MRIs. The goal of our project was to delineate sex-distinct markers associated with MRI-estimated brain tumor edema abundance. We analyzed 179 bulk RNA-Seq multiregional samples (Female: 75; Male: 104) from 55 high grade glioma patients (Female: 21; Male: 34). Patients’ segmented pre-surgical multiparametric MRIs were utilized in the edema mathematical model to create edema scores. We performed differential expression for high and low edema, gene set enrichment analysis (GSEA) using MSigDB hallmarks, and leading edge interpretation. We found that the fatty acid metabolism (FAM) and oxidative phosphorylation (OxPhos) pathways were sex-distinct, with both pathways amplified for high edema in males and low edema in females. Of the OxPhos pathway leading edge genes, 36 were unique to females, 66 were unique to males and 45 were common. Of the FAM pathway leading edge genes,39 were unique to females, 29 were unique to males, and 8 were common. Notably, expression of both IDH1 and IDH2 were increased for males in regions of high edema in the OxPhos pathway. IDH3a was decreased for females in regions of low edema in the OxPhos pathway. These data suggest that there may be sex-distinct metabolism underlying MRI measurable edema formation.

Other Authors: Lee Curtin, Kyle W. Singleton, Maciej M. Mrugala, Richard S. Zimmerman, Bernard R. Bendok, Peter Canoll, Kristin R. Swanson

11- Integrating signaling and transcription to study c-Myc induced stress response in cancerous cells

Reshma Kalyan Sundaram, University of Pennsylvania

The transcription factor c-Myc (Myc) is known to regulate a multitude of genes and cellular processes. Myc is deregulated in 70% of human cancers, making it a potent oncogene. Myc deregulation through increased Myc concentration has been suggested to cause cells to experience stoichiometric stress which may result in stress-induced cellular reprogramming. We hypothesize that extracellular signals transduced through intracellular signaling pathways can contribute to increased Myc concentration by modulating Myc protein stability via phosphorylation. We also hypothesize that increased Myc concentration can alter Myc’s transcriptional function by changing Myc-DNA interactions and hence Myc regulated gene expression. To test the first hypothesis, we built an ODE-based systems model consisting of extracellular growth and mechanical signals (effected by EGFR and integrins) and intracellular signaling pathways that modulate Myc phosphorylation. Our modeling results show that in normal cellular phenotypes, Myc is primarily regulated by EGFR. However, in the Myc-upregulated cancerous phenotype, Myc stabilization is dependent on both EGFR and integrins. Therefore, we conclude that extracellular growth and mechanical signals synergistically regulate Myc protein stabilization and accumulation in cells. To test the second hypothesis, we performed bioinformatics analysis on Myc ChIP-seq and RNA-seq datasets with varied Myc expression levels. From this analysis, we identified a new DNA-binding site of Myc, and demonstrated that Myc binding to this new site varies as a function of Myc concentration. We also showed that Myc utilizes this binding site to reprogram cells by modulating several cancer hallmarks such as proliferation, apoptosis, and cell adhesion.

Other Authors: Ravi Radhakrishnan, Bomyi Lim

12- Towards Cancer Digital Twins: Simulating Tumor Heterogeneity and Patient Response using Tissue Level Modeling Protocol

Sharvari Kemkar, University of Pennsylvania

Cancer's heterogeneity undermines the efficacy of conventional treatments. While advances in multiomics have provided valuable insights, the complexity of the data requires robust modeling for improved interpretation. We propose a multiscale framework that integrates agent-based modeling (ABM) with cellular systems biology models (CSBM), utilizing patient-specific data to predict tissue-level responses. We use this framework to study the effect of genetic and spatial heterogeneity on treatments in Prostate cancer and Wilms tumor.

We have established CSBMs that predict cancer cell states with and without treatment, customized to patient data, and inform cell growth dynamics in the ABM. The ABM is seeded with a mixed population (sensitive and resistant types) to capture genetic heterogeneity. Local sensitivity analysis on extrinsic factors like drug concentrations and cell-intrinsic properties like genetic composition, drug uptake rates, cell adhesion and motility reveal significant effects on tumor proliferation and tissue-level heterogeneity.

We have built ML surrogates for the CSBM to reduce ABM computational costs. For global sensitivity on CSBM parameters, we performed Shapley value ranking on the ML surrogate to identify key features influencing cell fates. We established the significance of these species by conducting Kaplan-Meier survival analysis. Additionally, we are also exploring generative AI for global sensitivity and uncertainty quantification in the multiscale framework.

Our results demonstrate the potential of multiscale ABMs and coupled data-driven methods in explaining complex tumor behaviors. We aim to create cancer digital twin models that can inform patient-specific therapies and enhance clinical decision-making.

We acknowledge support from NCI PSON.

Other Authors: Sharvari Kemkar, Mengdi Tao, Alokendra Ghosh, Ravi Radhakrishnan

13- Mechanosensitivity of PD1 Localization on the CD8 T-cell Membrane as Interrogated Using STORM Microscopy and Monte Carlo Simulations

Seung-Hyun Ko, University of Pennsylvania

Solid tumors establish and develop within microenvironments capable of potently suppressing host immune responses against the tumor. Recent studies demonstrate that cancerous cell-secreted extracellular vesicles (EVs) contribute to tumor immunosuppression by carrying upregulated levels of the immune checkpoint molecule PD-L1. However, given the small size of EVs, it is unclear what contributes to their seemingly disproportionately potent immunosuppression observed experimentally. Our collaborators demonstrated that the adhesion molecule ICAM-1 is necessary for EVs to successfully suppress cytotoxic T-cells. Furthermore, our lab previously demonstrated that the potency of nanoparticle/cell interactions is dependent on mechanosensitive multivalent binding and clustering. Here, we utilize biophysical Monte Carlo simulations in conjunction with STORM superresolution microscopy experiments to provide novel insight into the mechanosensitivity of EV-induced PD-1 localization on the T-cell membrane. We find that various clustering patterns arise at the interaction interface in a manner dependent on cortical tension. In a stochastic model of T-cell signaling, we find that these patterns lead to significantly different levels of phosphorylated AKT, a critical transcription factor in cytotoxic T-cell activation and development. These results bolster the growing evidence supporting the importance of mechanosensitivity in cytotoxic T-cell activation, specifically providing novel insight into how PD-1 clustering may be a mechanism by which TAM EVs so potently suppress recipient T-cells.

Other Authors: Seung-Hyun Ko, Zizhao Li, Wenqun Zhong, Wei Guo, Jina Ko, Su Chin Heo, Ravi Radhakrishnan

14- Platelet roles in determining biophysical flow in pre-metastatic niche

Fransisca Leonard, Houston Methodist Hospital

The role of platelets in interaction with CTCs and other bone marrow-derived cells to increase tumor cell survival, invasion, and growth have been established. However, the biophysical roles of accumulated and aggregated platelets in the pre-metastatic soil have been elusive.

We hypothesized that primary tumors may alter biophysical transport phenomena in distant organs, promoting the arrest of CTCs on vessel walls via platelet contribution. To test this hypothesis, treated tumor-bearing animals with platelet inhibitors: clopidogrel, eptifibatide, anti-GPIba (CD42b) antibody or aspirin. Series of studies were conducted using flow cytometry and intravital microscopy (IVM) imaging of tumor imaging of liver capillaries of tumor bearing and non-tumor bearing animals followed by phyton-based computational analysis to understand the flow fundamentals in vessels as a function of development/evolution of the pre-metastatic niche. IVM data showed the irregularities in platelet flow dynamics only in the tumor-bearing animals. Further segmentation of the capillaries revealed distinct platelet population speeds, which may be caused by different populations of platelet that were involved in rolling and adhesion process to the activated endothelial cells.

The study provides new insights into the biophysical roles of platelets in the pre-metastatic niche. We plan to expand to study their effects on CTC circulation, immobilization, interaction with platelets, and the changes of flow biomechanical characteristics in the pre-metastatic niche. By understanding both the biological and physical roles of metastatic initiation, we can develop a new and more effective treatment to reduce the risk of metastasis for cancer patients.

Other Authors: Xuewu Liu

15- Sex-Specific Cellular Heterogeneity in Glioblastoma: Implications for Personalized Treatment and MRI-Based Characterization

Jing Li, Georgia Tech

BACKGROUND:
Glioblastoma (GBM) remains challenging to treat due to its complex intra-tumoral heterogeneity. Understanding cellular composition differences, particularly sex-specific variations, is crucial for developing personalized treatment strategies. This study aims to characterize these differences and their implications for GBM management, with a focus on the correlations between cell typesand MRI sequences

METHODS:
We analyzed 191 image-localized biopsies from 56 GBM patients (76 female vs. 115 male samples) using the CIBERSORTx dataset to quantify cellular compositions. We employed a mixed-effects logistic regression model to explore how cell types relate to tumor location (enhancing core [ENH] vs. non-enhancing [NE] regions) across sexes. We also investigated sex-specific correlations between MRI radiomics features and cell types in these regions.

RESULTS:
Our findings reveal significant sex-specific cellular heterogeneity in GBM. Male samples showed more pronounced differences in astrocytic and vascular cell types between ENH and NE regions, indicating a tumor environment influenced by vascular remodeling. Female samples exhibited greater involvement of immune cells and certain glioma subtypes associated with proliferation, particularly in the ENH regions. These patterns indicate sex-specific factors may be influencing the tumor microenvironment and behavior. We also observed sex-dependent correlations between imaging (radiomics) features and specific cell subpopulations, suggesting potential for non-invasive yet sex-distinct tumor characterization predictive from imaging. This relationship between cellular composition and imaging characteristics offers opportunities to enhance sex-specific diagnostic and monitoring strategies in GBM that highlight sex as a crucial variable in GBM research and clinical management.

Other Events: Dan She 1 , Lee Curtin 2 , Kyle Singleton 2 , Pamela Jackson 2 Jack Grinband 4,5 , Peter Canoll 3 , Jing Li 1 , Kristin Swanson 2 [Affiliations] 1 H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology, Atlanta, GA, USA 2 Mathematical NeuroOncology Lab, Precision Neurotherapeutics Innovation Program, Mayo Clinic, Phoenix, AZ 3 Department of Pathology and Cell Biology, Columbia University Irving Medical Center, New York, NY, USA 4 Department of Psychiatry, Columbia University Irving Medical Center, New York, NY, USA 5 Department of Radiology, Columbia University Irving Medical Center, New York, NY, USA

16- Microfluidic chips for high-throughput analysis of cellular functions and cell interactions at single cell resolution

Xuewu Liu, Houston Methodist Research Institute

Cellular heterogeneity is a key feature in cancer and immune system, single cell functional analysis is critically important but has always been challenging. High throughput single-cell analysis can be utilized to evaluate tumor cell evolution, treatment effect, heterogeneity of tumor microenvironment, intercellular interaction, immune response, and immunotherapy. Herein, we present two microfluidics-based platforms for high-throughput cell functional analysis at a single cell resolution. First, we develop a hierarchical loading microwell chip (HL-Chip)-based method that efficiently aligns thousands of single cells with multiple antibody-coated microbeads for highly multiplexed detection of secreted proteins from single cells. We demonstrate the applications of this platform in profiling cytokine secretion of single T cells and macrophages after pan- and antigen-specific stimulation. The multiplexed profiling reveals an early but transient cytokine burst and polyfunctional heterogeneity in antigen peptide-stimulated T cell receptor-engineered T (TCR-T) cells. This simple and versatile microwell chip could be used to evaluate the biomarkers from single cell, or two cell pairs with or without treatment or stimulations, find its application in both biological discoveries and clinical diagnosis. Second, we develop a high-throughput single-cell pairing and assaying platform based on microfluidic single-cell trapping and homing. Single cells are efficiently trapped in hydrodynamic traps then transferred to adjacent cell chambers (12,800 in total in a proof-of-concept chip) via centrifugation. Various single-cell, two-cell and three-cell arrays are successfully demonstrated. We envision application of these platform technologies in short-term and long-term single-cell functional assays, drug testing and screening, and cellular interaction studies at a single cell resolution.

Other Authors: Ning Shao, Fransisca Leonard, Milos Kojic

17- Solid tumor cohesion & growth are overcome in phagocytosis by SIRPα-knockout macrophages

Tristan Marchena, University of Pennsylvania

Macrophages are potential effector cells in immunotherapy against solid tumors, but any phagocytosis requires macrophage-cancer cell interactions to out-compete cohesive interactions between tumor cells (e.g. cadherins). It also requires overcoming a ‘don’t eat me’ signal from CD47. We engineered conditionally immortalized macrophages (CIMs) with deletion of the CD47-binding inhibitory receptor SIRPα, and we first show such SIRPαKO-CIMs readily engulf added suspensions of IgG-opsonized melanoma cells in a standard 2D phagocytosis assay. To better model macrophage attack of a solid tumor, we make cohesive melanoma tumoroids on low adhesion plates, and we find exponential growth of the tumoroids over days can be suppressed and even reversed by sufficient SIRPαKO-CIMs. Elimination of tumoroids coincides with formation of macrophage clusters, which others have noted in various tissue contexts but not related to phagocytic activity. SIRPα-KO engineered macrophages thus seem promising for immunotherapies against solid tumors.

Other Authors: Lawrence J. Dooling, Nicholas Ontko, Jason C. Andrechak and Dennis E. Discher

18- Spatially restricted environmental therapy resistance limits competitive release and supports heterogeneity of resistance mechanisms

Andriy Marusyk, H Lee Moffitt Cancer Center

A key open question in understanding the evolutionary dynamics of therapy resistance is whether relapse results from an expansion of pre-existing therapy-resistant subpopulations or from a bona fide gradual evolutionary process. To elucidate this question, we integrated experimental mouse studies with mathematical modeling, interrogating therapeutic responses of therapy-naïve experimental xenograft ALK+ tumors, spiked-in with differentially labeled resistant cells. We found that ALKi treatment induced a rapid expansion of resistant cells, which drastically reduced the magnitude and duration of remissions. Surprisingly, our in silico analyses pointed to the existence of a strong positive ecological interaction between therapy-resistant and therapy-sensitive competitors. Using a combination of spatial analyses, experimental studies, and mathematical modeling, we found that this interaction was mediated by peristromal niches that protected therapy-sensitive tumor cells. Specifically, by limiting tumor regression, the therapy-induced expansion of resistant cells limited the loss of protective peristromal niches, while the subsequent resumption of tumor growth created new peristromal niches capable of supporting the survival and proliferation of therapy-sensitive cells. While this niche-mediated interaction had only a marginal impact on the transition from remission to relapse, enhanced survival of therapy-sensitive cells potentiated their ability to adapt to ALKi, leading to a higher diversity of resistance phenotypes.

In summary, our study has revealed a new type of indirect, niche-mediated ecological interaction between therapy-resistant and therapy-sensitive cells. The rapid expansion of rare pre-existent resistant cells and fast transition to relapse challenge a common assumption of the pre-existence of therapy resistance. Finally, our results highlight the essentiality of TME considerations in understanding the evolutionary dynamic underlying the development of therapy resistance.

Other Authors: Mark Robertson-Tessi; Bina Desai; Tatiana Miti; Pragya Kumar; Sagnik Yarlagadda; Rishi Shah; Robert Vander Velde; Daria Miroshnychenko; Jacob Scott; David Basanta; Alexander Anderson

19- Expression levels of Pip2 binding proteins can predict pan cancer cell response to ECM substrate stiffness and chemistry

Jonathan Nukpezah, University of Pennsylvania

Proteomics analysis has become a pivotal tool for understanding cellular mechanisms by identifying key proteins involved in various biological processes. In this study, we conducted a comprehensive proteomics analysis using intensity-based absolute quantification (iBAQ) mass spectrometry data to identify PIP2-binding proteins and their potential roles in predicting cellular stiffness and motility. We began by quantifying protein expression across multiple experimental conditions using iBAQ mass spectrometry. Differential expression analysis was performed, applying a significance threshold (p < 0.05 and fold change > 1) to identify proteins that exhibit significant changes in expression. From this analysis, a subset of proteins was determined to have a strong binding affinity for phosphatidylinositol 4,5-bisphosphate (PIP2), a lipid known for its involvement in cytoskeletal regulation and cell signaling pathways.

These PIP2-binding proteins were then used as input features in a machine learning model designed to predict key cellular phenotypes, specifically cell stiffness and motility. The model was trained using supervised learning techniques, and its performance was evaluated based on the metric of balanced accuracy. The results demonstrate a strong predictive relationship between the identified PIP2-binding proteins and cellular mechanical properties, with specific proteins emerging as key regulators of these phenotypes.

This approach not only highlights the critical role of PIP2-binding proteins in cellular mechanics but also demonstrates the power of integrating proteomics data with machine learning to predict biophysical properties, offering insights into how molecular components drive cellular behavior.

Other Authors: Jonathan Nukpezah, Kshitz Parihar, Paul Janmey, Ravi Radhakrishnan

20- Theoretical analysis reveals a role for RAF conformational autoinhibition in paradoxical activation

Edward Stites, Yale School of Medicine

RAF kinase inhibitors can, under certain conditions, increase RAF kinase signaling. This process, which is commonly referred to as ‘paradoxical activation’ (PA), is incompletely understood. We use mathematical and computational modeling to investigate PA and derive rigorous analytical expressions that illuminate the underlying mechanism of this complex phenomenon. We find that conformational autoinhibition modulation by a RAF inhibitor could be sufficient to create PA. We find that experimental RAF inhibitor drug dose–response data that characterize PA across different types of RAF inhibitors are best explained by a model that includes RAF inhibitor modulation of three properties: conformational autoinhibition, dimer affinity, and drug binding within the dimer (i.e., negative cooperativity). Overall, this work establishes conformational autoinhibition as a robust mechanism for RAF inhibitor-driven PA based solely on equilibrium dynamics of canonical interactions that comprise RAF signaling and inhibition.

Other Authors: Gaurav Mendiratta

21- Cancer cells co-evolve with retrotransposons to mitigate viral mimicry.

Siyu Sun, Memorial Sloan Kettering Cancer Center

To thrive, cancers must navigate acute immunostimulatory signaling that accompanies oncogenic transformation, such as overexpression of retrotransposable elements. Here we examined the relationship between immunostimulatory repeat expression, tumor evolution and the tumor-immune microenvironment by integrating multimodal data from a cohort of pancreatic ductal adenocarcinoma (PDAC) patients. We find expression of specific Alu repeats predicted to form double-stranded RNA (dsRNAs) and trigger RIG-I-like receptor-associated (RLR) type-I interferon (IFN) signaling. Alu-derived dsRNA formation further anti-correlated with pro-tumorigenic macrophage infiltration. We defined two complementary pathways by which PDAC may adapt to such anti-tumorigenic signaling. In mutant p53 tumors, the autonomous retrotransposon LINE-1, in particular its ORF1 protein, bind Alus and decreases expression of their transcripts, whereas ADAR1 editing reduced dsRNA formation in wild-type p53 tumors. Knock down of both ORF1p and ADAR1 reduced tumor growth in vitro. Finally, we built a mathematical model to predict RLR associated IFN phenotypes as a function of p53 status, which quantifies the relative contributions of L1 and ADAR1 on the suppression of Alu-derived dsRNA. Thus, tumors converge across multiple genetic backgrounds to steer the impact of dark matter repeats away from acute immunogenic signaling towards a more chronic pro-tumorigenic state.

Other Authors: Siyu Sun

22- LINE-1 ORF1p Mimics Viral Innate Immune Evasion Mechanisms in Pancreatic Ductal Adenocarcinoma

Eunae You, Massachusetts General Hospital

Repeat element viral mimicry is a common feature in pancreatic ductal adenocarcinoma (PDAC) that require mechanisms to manage this repeat “viral” load and attenuate innate immune responses. Here, we show that the LINE-1 ORF1 protein (ORF1p) in PDAC cells plays a role in shielding repeat RNAs from activating a pathogen recognition receptor (PRR)-mediated antiviral response that is independent of retrotransposition. Suppression of ORF1p using short hairpin RNA induces innate immune responses through the dsRNA sensors RIG-I and MAVS. Low ORF1p PDAC cell lines have suppressed expression of PRRs demonstrating convergent mechanisms to suppress innate immune signaling. Localization of ORF1p in processing bodies (P-bodies) with the dsRNA helicase MOV10 were found important for these antiviral responses. Loss of ORF1p resulted in significant growth reduction in tumorspheres and mouse xenografts with an enriched epithelial cell state, and high ORF1p expression was associated with worsened survival in a cohort of human PDAC patients.

Other Authors: Bidish K. Patel, Alexandra S. Rojas, Siyu Sun, Natalie I. Ho, Ildiko E. Phillips, Michael J. Raabe, Yuhui Song, Katherine H. Xu, Joshua R. Kocher, Peter M. Richieri, Phoebe Shin, Martin S. Taylor, Linda T. Nieman, Benjamin D. Greenbaum, David T. Ting

23- Synthetic Notch Circuits Enable In Vivo Measurement of Endogenous Forces on B Cell Receptors

Cheng Zhu, Georgia Institute of Technology

Background: Understanding mechanotransduction through immunoreceptors is fundamental to basic and translational research in immunology and immunotherapy. Despite in vitro evidence of endogenous forces immune cells exert on immunoreceptors, in vivo measurement is lacking due to technological limitations.

Methods: Utilizing a synthetic Notch receptor (SynNotch) platform, we re-engineered the wild-type Notch receptor by replacing its ligand-binding domain with an antibody single-chain variable fragment (scFv) targeting CD40 on B cells. This synthetic molecule was designed to induce expression of a reporter (GFP or luciferase) in the sensor (receiver) cell upon binding of CD40 on the target (sender) cell which pulls on the αhCD40-SynNotch molecule, enabling observer to monitor the force signal in vivo. The in vitro characterization and in vivo validation of this SynNotch platform were conducted using Jurkat T cells as sensor cells and B cells as target cells in several models, including co-culture, co-embedded in hydrogel-based organoids, and implanted in animals.

Results: We was successfully demonstrated the ability of αhCD40-SynNotch to measure endogenous forces exerted by B cells on the CD40 in vitro and in vivo. We found that its activation required the application of forces by the target cells as the sensor cells did not exert force on αhCD40-SynNotch. The performance of the Manipulation of the target cells to exert different forces on CD40 resulted in different reporter signals, which correlated with the measurements of these forces by the molecular tension probes (MTP). The αhCD40-SynNotch system worked in similar fashions in simple two-dimensional co-cultures of target cells and sensor cells, when embedded in three-dimensional hydrogel-based organoids, and when implanted in NSG mice.

Conclusion: A technology platform has been designed, fabricated, and tested for in vivo measurement of mechanical forces on specific receptors exerted by cells. Using a specific design of this platform, the αhCD40-SynNotch, we found that in live animals B cells pull on CD40 in similar ways as they did in vitro co-culture with sensor cells. This technology can be applied to various scenarios of mechanobiology research.

Other Authors: Jintian Lyu, Kaitao Li, Ameya Dravid, Deepali Balasubramani, Menglan Li, Amir H.K. Ashkezari, Hyun-Kyu Choi, Ankur Singh, Cheng Zhu

24- Revealing Heterogeneity in Mantle Cell Lymphoma by Integrating Single-cell Biophysical and Transcriptomic Features

Ye Zhang, Koch Institute for Integrative Cancer Research at MIT

Mantle cell lymphoma (MCL) is marked by significant clinical and molecular heterogeneity, complicating the development of standardized treatments. A key challenge is accurately characterizing this heterogeneity for targeted therapies. In this study, we aim to address this question by integrating single-cell biophysical and transcriptomic analyses. Using a novel microfluidic platform, we linked cell buoyant mass and stiffness to gene expression for individual MCL cells from three patient-derived xenograft models. Our findings revealed that genes most positively correlated with buoyant mass were predominantly involved in cellular division, such as CCNB1, CDCA2 and CDK1, while those correlated with stiffness were linked to immune response and B-cell receptor (BCR) signaling, like BLK and CD79A, especially in the highly proliferative models. Our study also showed that cell mass and stiffness were linked to MCL-specific signaling pathways, such as B-cell development and oncogenic signaling. Overexpression of BLK and CD79A in Jeko-1 cells increased both mass and stiffness, corroborating our biophysical-transcriptional findings. Additionally, activation of mature B-cells resulted in distinct changes in buoyant mass and stiffness, underscoring their utility in tracking differentiation states. Notably, our work identified cell buoyant mass as a potential ex vivo biomarker for predicting responses to BCR signaling inhibition in MCL patients, as evidenced by clinical specimens before and after BTK inhibitor therapy. In summary, integrating biophysical profiling with transcriptomic data offers a comprehensive and multi-dimensional readout of tumor cell physiology, underscoring the potential of these properties for understanding MCL cell states and predicting therapeutic outcomes in heterogeneous primary specimens.

Other Authors: Lydie Debaize, Adam Langenbucher, Jenalyn Weekes, Nolawit Mulugeta, Huiyun Liu, Sarah M Duquette, Nezha Senhaji, Liam Hackett, Jeremy Zhang, Paul H Branch, Mingzeng Zhang, Robert A. Redd, Martin Aryee, Matthew Davids, Austin I Kim, Christine Ryan, David M Weinstock, Mark A Murakami, Scott R Manalis