Session R50: Physics of Development and Disease - I

Dynamic force patterns promote coordinated cell movements during embryonic wound repair

Embryos display an outstanding ability to rapidly repair wounds, in a process driven by collective cell movements. Actin and the motor protein non-muscle myosin II become polarized in the cells adjacent to the wound, forming a supracellular cable that contracts to coordinate the movements that drive tissue repair. We showed that, in Drosophila embryos, the cable is heterogeneous, with regions of high and low actin density. Mutants in which actin is uniform around the wound display slower wound closure. However, the mechanisms by which a non-uniform distribution of actin favors rapid repair are unknown. Using laser ablation, we demonstrated that actomyosin-rich segments of the cable sustain higher contractile forces, indicating that cable contraction is non-uniform. We developed a computer model of wound repair, and we found that a heterogeneous actomyosin distribution was favorable for wound closure when myosin assembly at the wound edge was strain-dependent. To test the model prediction, we used a laser-based method to induce ectopic strain on cell boundaries in vivo, and we found that myosin accumulated in response to deformation. Using pharmacological and genetic treatments, we found that stretch-activated ion channels were necessary for rapid embryonic wound repair. Our results suggest that local heterogeneities in supracellular actomyosin networks promote faster wound closure by generating mechanical signals sensed by stretch-activated channels that facilitate myosin assembly and coordinate cell behaviors.   

Probing the role of tissue mechanics in cancer cell dormancy

In the case of breast cancer, patients showing one sub-type may appear disease free for several years before relapsing with brain metastasis. Improved clinical control requires a better understanding of cancer dormancy. Using in vitro models, proteins critical for mechanotransduction regulate emergence from the dormant state. Therefore, we reasoned that the physical microenvironment may regulate reawakening. Here, we mapped the intracellular and extracellular cell and matrix mechanics with near simultaneity as a function of ECM stiffness for PDMS gels and brain mimetic HA gels using optical tweezers to perform active microrheology. The intracellular stiffness of dormant cells embedded in gels was stiffer than the aggressive cells. Comparable measurements to thin sections of excised tumors grown subcutaeneously showed a similar trend. A differential local remodeling of matrix was measured between clones. Taken together, our results suggest that dormant and aggressive cells may respond differently to the brain microenvironment, exhibiting distinct morphology, mechanical properties and ECM remodeling, which bear on their ability to colonize the brain.   

Temporal precision of regulated gene expression

Cellular processes such as cell migration, differentiation, and division often rely on precise timing. Yet, the molecular machinery that regulates timing is inherently noisy. How do cells achieve precise timing with noisy components? We investigate this question using a first-passage-time approach, for an event triggered by a molecule that crosses an abundance threshold and that is regulated by either an accumulating activator or a degrading repressor. We find that the optimal strategy corresponds to a nonlinear increase of the target molecule over time. The optimality arises from a tradeoff between minimizing the extrinsic timing noise of the regulator, and minimizing the intrinsic timing noise of the target molecule itself. Although either activation or repression outperforms an unregulated strategy, we find that repression outperforms activation in the presence of cell division that occurs late in the process. Our results explain the nonlinear increase and low noise of mig-1 gene expression in migrating neuroblast cells during C. elegans development, and suggest that mig-1 regulation is dominated by repression for maximal temporal precision. These findings suggest that dynamic regulation may be a simple and powerful strategy for precise cellular timekeeping.   

Active process driven glass to fluid transition in growing tumor

<!--StartFragment-->We study the dynamics of collection of cancer cells driven by the mechanical interactions as well as division and death of cells using stochastic quantization method. Theory predicts caging effect depending on the strength of adhesion leading to glassy-phase in finite time and birth-death driven fluidization of colony of cells characterized by super-diffusive motion of cells in the long time scale. The theory can be applied to any systems involving division and death characterized by absence of fluctuation-dissipation theorem (FDT). From this unified theory based on stochastic quantization scheme, we understand the similarity between the predictions of super-diffusive motions of soap bubbles, CD$8^+$ T cells and tumor cells. The non-trivial exponent calculated by the theory is in excellent agreement with the exponents obtained by simulations of all three problems indicating all three problems belong to same universality class characterized by the same dynamical exponent in the long time limit.<!--EndFragment-->   

Stochastic and Heterogeneous Cancer Cell Migration: Experiment and Theory

Cell migration is an essential process in the cancer metastasis. A mathematical model for the cell migration would help understand cancer metastasis and might propose a new approach for cancer treatment. Models based on the diffusion equation have been developed because the cell migration looks similar to a random walk. However, recent studies showed that cell migration should be non-Fickian and undergo non-Gaussian diffusion, thus implying that a new model beyond the diffusion equation would be necessary. Here, we propose a mathematical model for anomalous cell migration by introducing two types of heterogeneity in the cell population: cellular heterogeneity and temporal heterogeneity. We investigate 2D migration of A549 cells and find that the migration of A549 cells is non-Gaussian but Fickian diffusion. We employ four different models to elucidate anomalous cell migration: homogeneous (HO), cellular heterogeneity (CH), temporal heterogeneity (TH) and cellular-temporal heterogeneity (CTH) model. We show that only CTH model can describe anomalous cell migration and that considering both cellular and temporal heterogeneity together is crucial to understanding the cell migration.   

Heterogeneity in mechanical phenotypes of cells comprising tissues as an early marker for tumorigenesis

Advances in mechanical characterization of cells have shown specific changes in mechanical properties of cells comprising healthy tissues and malignant tumors– tumor cells are softer, more deformable compared to normal tissue cells. However, this comparison is the comparison between the population means. Data shows that even in healthy tissue populations, there is a wide distribution (heterogeneity) in the mechanical properties of individual cells about the said mean. We use mechanistic computational models to study the influence of this heterogeneity on overall tissue dynamics. Our results show that increasing overall heterogeneity in cellular mechanical properties, complemented by local clustering of cells with similar mechanical phenotypes can drive the population distribution of an initially healthy tissue towards distributions representing malignant tumors. We further investigate the mobility of mechanically different cells within the heterogeneous environment to estimate a timescale over which local clustering of similar cells can occur. These results provide a theoretical estimate for the likelihood and timescale of tumor incidence in healthy tissue environments.
  

Tic-Tac-Toe: Maximum Entropy Analysis of the Spatial T Cell Distribution in the Tumor Microenvironment of Breast Cancer

The effectiveness of cancer immunotherapy is dependent on the infiltration of T cells into the tumor. Using maximum entropy, we analyzed the spatial distribution of T cells in the tumor microenvironment of tissue resected from triple negative breast cancer patients. We find that at longer length scales (> 200 microns), the distribution is self-similar. Furthermore, there is a distinct difference between good and poor clinical outcome. We discuss possible interpretations as well as the prognostic potential of this approach.   

Interstitial Flows Modulate Tumor Cell Invasion using a 3D Microfluidic Model

Malignant tumors are often associated with an elevated fluid pressure due to the abnormal growth of vascular vessels, and thus an increased interstitial flow out of the tumor. However, current in vitro tumor models rarely include fluid flows, and thus, roles of flows in tumor cell invasion and tumor progression is largely unknown. In this talk, I will describe the development of an in vitro tumor model where tumor cells/spheroids are embedded within a 3D biomatrix, and are subjected to well controlled interstitial flows. Our work demonstrated that interstitial flows critically regulate tumor cell/spheroid morphology and migration. This work highlights the importance of biophycal forces in regulating tumor cell invasion in 3D.   

Enhanced blebbing as a marker for metastatic prostate cancer

Experiments with cells flowing through a microfluidic channel show that highly metastatic prostate cancer cells form more plasma membrane blebs than lowly metastatic or normal prostate cells. Increased blebbing of metastatic breast cancer cells flowing through a micropipette was previously attributed to decreased phosphorylated ERM (p-ERM) protein expression – p-ERMs bind the plasma membrane to the actin cortex and increase cell stiffness. Myosin II also influences blebbing as cortical actomyosin contraction, due to myosin II, can increase cellular hydrostatic pressure – leading to cortex rupture and bleb formation. Surprisingly, we find that highly metastatic prostate cells express higher levels of p-ERM and lower levels of myosin II than lowly metastatic or normal prostate cells – suggesting that their level changes do not alter metastatic prostate cell blebbing. Rather, increased blebbing is correlated with reduced cortical actin expression and reduced cell stiffness – already widely accepted as markers of metastatic cancer. These findings indicate that, regardless of p-ERM and myosin II levels, extracted cell blebbing can be used as a simple diagnostic marker for metastatic cancer.   

The Role of Immune-therapy in Delaying the Drug Induced Resistance in Cancer

Drug-induced resistance may occur during a phase of chemotherapy. This type of drug resistance is a major obstacle against cancer treatment. Some current reported results indicate that the strength of the immune system may affect the occurrence of the drug-induced resistance. In this work, the dynamics of tumor cells in the presence of the drug-induced resistance are expressed by a system of coupled differential equations. The model is then modified to include the components of the immune system and immunotherapeutic agents. Under this modification, the evolution of normal and tumor cells is simulated under different experimental conditions. The results indicate that the boosted immune system can create a delay in the occurrence of the drug-resistant tumor cells. Such a delay, which has been clinically observed, may lead to more effective therapeutic strategies and more successful outcomes.

  

Drug resistance mechanisms and combinatorial drug treatments in breast cancer: a network modeling approach

Mechanistic models of within-cell signal transduction can explain how these networks integrate internal and external inputs to give rise to the appropriate cellular response. These models can be fruitfully used in cancer cells, whose aberrant decision-making regarding their survival, death, or proliferation can be connected to errors in the state of nodes or edges of the signal transduction network. Here we present a comprehensive network, and discrete dynamic model, of signal transduction in breast cancer based on the literature of ER+, HER2+, and PIK3CA-mutant breast cancers. The network model recapitulates known resistance mechanisms to PI3K inhibitors, suggests other possibilities for resistance, and reveals known and novel combinatorial interventions that are more effective than PI3K inhibition alone. The use of a discrete dynamics enables the identification of results that are due to the organization of the signaling network, and those that also depend on the kinetics of individual events. Network-based models such as this will play an increasing role in the rational design of high-order therapeutic combinations.   

Analysis of Hierarchical Organization in Gene Expression Networks Reveals Underlying Principles of Collective Tumor Cell Dissemination and Metastatic Aggressiveness of Inflammatory Breast Cancer

Clusters of circulating tumor cells (CTCs), although rare, may account for more than 95% of metastases. Inflammatory breast cancer (IBC) is a highly aggressive subtype that chiefly metastasizes via CTC clusters. Theory suggests that physical systems with hierarchical organization tend to be more adaptable due to their ability to efficiently span the set of available states. We used the cophenetic correlation coefficient (CCC) to quantify the hierarchical organization in the expression of collective dissemination associated and IBC associated genes, and found that the CCC of both gene sets was higher in (a) epithelial cell lines as compared to mesenchymal cell lines and (b) IBC tumor samples as compared to non-IBC breast cancer samples. A higher CCC of both networks was also correlated with a higher rate of metastatic relapse in breast cancer patients. Gene set enrichment analysis could not provide similar insights, indicating that the CCC provides additional information regarding the organizational complexity of gene expression. These results suggest that retention of epithelial traits in disseminating tumor cells as IBC progresses promotes successful metastasis and the CCC may be a prognostic factor for IBC.   

Modularity of the Metabolic Gene Network as a Prognostic Biomarker for Hepatocellular Carcinoma.

Abnormal metabolism is an emerging hallmark of cancer. Cancer cells utilize both aerobic glycolysis and oxidative phosphorylation for energy production and biomass synthesis. Understanding the metabolic reprogramming in cancer can help design therapies to target metabolism. We have previously argued that more malignant tumors are usually characterized by a more modular expression pattern of cancer-related genes. In this work, we analyzed the expression patterns of metabolism genes in terms of modularity for 371 hepatocellular carcinoma (HCC) samples from the Cancer Genome Atlas (TCGA). We found that higher modularity significantly correlated with glycolytic phenotype, later tumor stages, higher metastatic potential, and cancer recurrence, all of which contributed to poorer overall prognosis. Among patients that recurred, we found the correlation of greater modularity with worse prognosis during early to mid-progression. Furthermore, we developed metrics to calculate individual modularity, which was shown to be predictive of cancer recurrence and patients’ survival and therefore may serve as a prognostic biomarker. Our overall conclusion is that more aggressive HCC tumors, as judged by decreased host survival probability, had more modular expression patterns of metabolic genes.