MS10: Liver as a model system for mechanics, flow, and multiscale mathematical biology

Monday, July 17, 3:00-5:30, Alpine Room

Organizers:


Paul Macklin, Department of Intelligent Systems Engineering, Indiana University Bloomington, IN (macklinp@iu.edu)


In this talk, we briefly outline the key features in liver microanatomy, including the microarchitecture and function of liver sinusoids and lobules, interstitial and microvascular flow, and associated clinical problems in toxicology and cancer. This 5-minute primer will provide the necessary biological background for the entire minisymposium.

Acetaminophen (APAP) overdose causes many thousands of deaths in the USA every year due to liver failure. The primary mechanism of liver damage is production of the metabolite NAPQI through Phase I metabolism in hepatocytes (especially through Cytochrome P450 2E1, but also through a number of other cytochromes). NAPQI binds irreversibly to Glutathione, a buffer of reactive oxygen species and depletion of Glutathione in turn leads to rapid cell death via necrosis. While the liver has great powers of regeneration under many circumstances, the particular pattern of cell death due to APAP poisoning often leads to necrosis of the entire liver rather than recovery. Because many factors including within population variability, age, state of health, consumption of alcohol, medications and some foods can significantly alter the pattern of cytochrome expression within individual hepatocytes, doses of APAP tolerated in a typical individual can be toxic in susceptible subpopulations and doses previously tolerated in a given individual may be toxic after exposure to particular modulators of cytochrome production. We have developed a three-scale virtual-tissue simulation in the open-source CompuCell3D simulation environment, which includes a PBPK model of whole body dosimetry and partitioning after ingestion of APAP, a multicell model of a simplified liver sinusoid including transport of Acetaminophen in the blood and its uptake by hepatocytes and a subcellular SBML reaction-kinetic model of APAP metabolism and NAPQI production in each hepatocyte. We validated a baseline set of parameters by tuning to reproduce clinically-measured serum concentration time-series of APAP and its metabolites, then systematically investigated the sensitivity of the maximum degree of Glutathione depletion on model parameters, providing preliminary insights into factors contributing to population variability in APAP toxicity. We also discuss the effects of liver architecture on the microdosimetry of APAP.

Regenerative medicine has the potential to alleviate severe donor organ shortages for patients with end-stage liver failure. Bioengineered liver constructs could also serve as test platforms for pharmaceutical research, liver disease, metastasis, and development. Organ decellularization, a promising bioscaffolding technique, is the removal of all cellular components from an organ leaving behind intact extracellular matrix (ECM). Since the vasculature is retained, decellularized scaffolds have the potential to generate bioengineered liver constructs at the whole-organ scale. Scaffold perfusion through a cannulated portal vein generates a variety of mechanical forces that act across multiple length scales, from pressure and shear stress in vascular channels to interstitial flow and ECM tension in the parenchyma. Since many liver cell types, including hepatic stem/progenitor cells, hepatocytes, endothelial cells, stellate cells, cholangiocytes, and portal fibroblasts, are sensitive to mechanical stimuli, we hypothesize that providing optimal biomechanical conditions inside the scaffold is important for successful generation of viable tissue. Currently little is known about the biomechanical environment in decellularized tissue. The goal of this research is to quantify the mechanical microenvironment in decellularized liver, for varying organ-scale perfusion conditions, using a combined experimental/computational approach. Needleguided ultra-miniature pressure sensors were inserted into liver tissue to measure parenchymal fluid pressure ex-situ in portal vein-perfused native (n=5) and decellularized (n=7) ferret liver, for flow rates from 3-12 mL/min. Pressures were also recorded at the inlet near the portal vein cannula. Experimental results were fit to a multiscale computational model to simulate perfusion conditions inside native versus decellularized livers for all flow rates. The multiscale model consists of two parts: an organ-scale electrical analog model of liver hemodynamics and a tissue-scale model of pore fluid pressure, pore fluid velocity, and solid matrix stress throughout a 3D hepatic lobule. Distinct models were created for native versus decellularized liver. Results show that vascular resistance decreases several fold as a result of decellularization. Similarly, the hydraulic conductivity of decellularized liver, a measure of tissue permeability, was approximately 5 times that of native liver. In future this modeling platform can be used to guide the optimization of biomechanical conditions in decellularized scaffolds for liver bioengineering. In ongoing research, this model is being applied to study flow patterns in liver lobules with tumor obstructions.

The liver is a common metastasis site for many cancers, especially breast, lung, and colorectal cancer (CRC). Beyond the complexities of tumor cell extravasation to seed new metastases, tissue biomechanics, growth substrate biotransport, porous flow, and tumor-host (parenchymal) cell interactions all shape the growth dynamics of newly seeded metastases. In this talk, we build mechanical and transport constitutive relations based upon detailed poroviscoelastic simulations of single liver lobules, allowing us to build agent-based simulations of square centimeter-scale liver tissues with hundreds of liver lobules. Using this simulation framework, we seed CRC metastases in the virtual liver, explore the impact of different hypotheses on tumor-parenchymal interactions, and discuss future model refinements to better match experimental and clinical observations. This work illustrates that the ordinary model of using agent-based simulations to learn constitutive relations for continuum models can be reversed: detailed continuum models can drive simplifications that make large-scale discrete simulations feasible. Our work is built upon the open source packages BioFVM and PhysiCell, and it will be released as open source at http://MathCancer.org.

Cancers are typically complex, multicellular tissues, and interactions between different cell types and their environment may become even more complex upon treatment. Components contributing to the tumor growth and treatment response include cancer and vascular endothelial cells, immune system and stromal cells, extracellular matrix, and the cellular microenvironment. Interactions within the tissue occur across a wide range of physical scales, from the molecular (nanometer) to the tissue (centimeter) scale. These components and their interactions can significantly affect cancer cell survival and eventually lead to the emergence of drug resistance. Conventional chemotherapy targets single cancer cell populations with drug doses and administration frequencies determined by the maximum tolerated dose to avoid lethal patient toxicity. Nanotherapies aim to enhance targeting of tumor tissue while minimizing toxic side effects. However, the large number of combinational protocols specifying the targeting of multiple tumor cell populations and their microenvironment by chemotherapeutic agents in concomitant or sequential administration may preclude determination of potential clinical options solely through experimental effort. This assessment would benefit from methods and principles typically used in systems analysis, such as in engineering and mathematics. We have employed mathematical modeling and computational simulation to simulate tumor response to conventional as well as nanoparticlebased drug regimens. We develop calibration methods to set model parameters based on experimental data in order to project potential response in vivo. This work represents a step towards the development of a comprehensive virtual system to evaluate tumor drug response, with the goal to more efficiently identify optimal course of treatment with patient tumor-specific data. Future model evaluation of chemotherapy possibilities may help to assess their potential value to obtain sustained tumor regression for particular patients, with the ultimate goal of optimizing the cancer drug response.

Understanding metastatic spread to distant sites is one of the most challenging areas in cancer research today; yet it remains a difficult process to study in the laboratory largely due to discrepancies between cell culture models and tumor growth in vivo. Therefore, we are challenged with creating a reproducible and controllable experimental system that approximates in vivo conditions in order to probe the dynamics of cancer progression. Here we discuss the development of high throughput imaging techniques coupled with computational models to allow one to systematically investigate the relationships between tumor growth dynamics and heterogeneous microenvironments within bioengineered living tissue. We implement tools such as image segmentation, automated model-fitting, and machine learning to simultaneously characterize heterocellular phenotypes in response to relevant, co-occurring environmental stimuli. Specifically we highlight the differences in colon tumor cell behavior and response to therapy in standard tissue culture conditions versus a novel liver scaffold model, which better mimics the in vivo milieu. This image-based platform permits direct observation of cell population dynamics within precisely controlled environments that can be faithfully recapitulated in computational models of cancer progression.