archives@tulane.edu / Osteoarthritis (OA) is a common joint disorder with significant economic and healthcare impact. The knee joint is composed of cartilage and the adjoining bone, a synovial capsule, the infrapatellar fat pad (IPFP), and other connective tissues such as tendons and ligaments. Adipose tissue has recently been highlighted as a major contributor to OA through strong inflammation mediating effects. Type II diabetes mellitus (T2D) is a weight independent risk factor for OA, suggesting a link between OA and adipose dysfunction that has yet to be elucidated. There is a critical need for development of new methodologies to investigate the interaction between an osteochondral interface and extra-synovial tissue. There is also a need for investigating adipose derived stem cells (ASCs) isolated from the IPFP of T2D patients as a potential cell source to model diabetic complications. In this work, we develop a novel 3D printed bioreactor model for incorporation into a previously established osteoarthritic knee microphysiological system. Using our established model, we investigated xenoprotein free (XPF) media as a potential commercial product for the MPS industry. Additionally, differences in inflammatory and adipokine related mRNA expression of IPFP-ASCs isolated from non-diabetic (Non-T2D), pre-diabetic (Pre-T2D), and type II diabetes mellitus (T2D) patient samples were analyzed. After 28 Days of differentiation, 3D printer bioreactors using commercially available AdipoQual media exhibited robust increase in adipokine expression and neutral lipid accumulation. Bioreactors cultured with a novel XPF supplemented AdipoQual had similar adipokine expression but no neutral lipid accumulation. Pre-T2D IPFP-ASCs exhibited a robust decrease in CD90 and CD105 levels with no corresponding increase in markers for potentially contaminating cell types. Cox-2 expression and PGE2 secretion were significantly increased in IL-1β stimulated Pre-T2D IPFP-ASCs compared to Non-T2D and T2D IPFP-ASCs. When the Pre-T2D ASCs were co-cultured with M1-induced macrophages, the macrophages significantly reduced expressions of tumor necrosis factor α (TNFα) and IL-6 compared to M1-induced macrophages co-cultured with Non-T2D and T2D IPFP-ASCs. Taken together, this work has taken significant steps towards establishment of a model for the IPFP and establishment of important phenotypic and genotypic changes of IPFP-ASCs isolated from Pre-T2D patients. / 1 / Benjamen O'Donnell

archives@tulane.edu / 1 / Benjamin Vinson

The human placenta facilitates many key functions during pregnancy, including uterine invasion, vascular remodeling, hormone secretion, immune regulation, and maternal-fetal exchange. Placental research, however, has been limited in part by the unrepresentative nature of traditional models. The objective of this doctoral thesis was to build and characterize novel, in vitro models that reintegrated important anatomical and environmental elements of the human placenta, thus enabling more physiologically-accurate assessments of placental function. In our first model, we manipulated the thickness of the extracellular matrix surface to promote the self-assembly of trophoblast cells into three-dimensional (3D) aggregates that exhibited increased genetic and functional markers of syncytial fusion. In our second model, we established a high-throughput platform to generate 3D trophoblast spheroids that underwent dynamic invasion and migration, expressed transcriptomic profiles redolent of the extravillous trophoblast phenotype, and responded to various drugs relevant to pregnancy. In our third model, we developed a trophoblast-endothelial co-culture model of the placental barrier that underwent syncytial fusion, exhibited size-specific barrier permeability, and functioned under physiologically-relevant oxygen tensions. In conclusion, our models may each serve as valuable tools for researchers, contribute to investigations of different aspects of placental biology, and aid in the screening of drugs and toxins for pregnancy. / Thesis / Doctor of Philosophy (PhD) / The human placenta is an important organ that helps regulate the health of both the mother and fetus during pregnancy. Researchers have traditionally studied the placenta through the use of animals or isolated cells, but these have been criticized for not being similar enough to the human placenta. Our objective was to build models that better resembled the structure and environment experienced by the human placenta within the body, such that we could better study its function. During the course of my doctoral work, I built and analyzed three models of the human placenta using human cells that were grown in three dimensions, in multiple layers, and/or in a specific environment. Our first model demonstrated that placental cell behaviour and function can be controlled by altering the thickness of the surface we grew them on. Our second model grew placental cells in three-dimensions and mimicked the invasion process into the mother’s uterus during early pregnancy. Our third model grew placental cells with blood vessel cells to form the barrier that regulates the passage of all substances between the mother and fetus during pregnancy. We also tested the impact of low oxygen on the placental barrier’s formation and function. Overall, we discovered that placental cells could indeed function more similarly to how we expect them to in the body when we design platforms that better resemble their structure and environment. Our model development work provides new information about placental biology and may serve as valuable tools in research and drug development.

The research presented in this dissertation supports the overall goal of producing sensor functionalized neural microphysiological systems to enable deeper fundamental understandings of disease pathology and to provide drug screening and discovery platforms for improved clinical translation. Towards this goal, work addressing three broad objectives has been completed. The first objective was expanding the manufacturing process capabilities for hydrogels and tissues through augmentation of the 3D printing systems and developing novel modeling capabilities. The second objective was to expand the palette of available materials which exhibit the rheological properties required for 3D printing and the mechanical and biological properties required for neural tissue culture. The third objective was to develop sensing capabilities for both monitoring and control of the manufacturing process and to provide non-destructive assessment of microphysiological systems in real-time to quantify the dynamics of disease progression or response to treatment. The first objective of process improvement was addressed both through modification of the 3D printing system itself and through modeling of process physics. A new manifold was implemented which enabled on-the-fly mixing of bioprinting inks (bioinks) to produce smooth concentration gradients or discrete changes in concentration. Modeling capabilities to understand the transport occurring during both the processing and post-processing windows were developed to provide insight into the relationship between the programmed concentration distribution and its temporal evolution and stability. Vacuum-based pick-and-place capabilities for integration of prefabricated components for sensing and stimulation into the printed hydrogel constructs were developed. Models of the stress profiles, which relate to cell viability, within the printing nozzle during extrusion were produced using parameters extracted from rheological characterization of bioinks. The second objective was addressed through the development hydrogel bioinks which exhibited yield stresses without the use of rheological modifiers (fillers) to enable 3D printing of free-standing neural tissue constructs. A hybrid bioink was developed using the combination of a synthetic polaxamer with biomacromolecules present in native neural tissue. Functionalization of the biomacromolecules with catechol or methacrylate groups enabled two crosslinking mechanisms: chelation and UV exposure. Crosslinked gels exhibited moduli in the range of native neural tissue and enabled high viability culture of multiple neural cell types. The third objective was addressed through the characterization and implementation of physical and electronic sensors. The resonance of millimeter-scale dynamic-mode piezoelectric cantilevers submerged in polymer solutions was found to persist into the gel phase enabling viscoelastic sensing in hydrogels and monitoring of sol-gel transitions. Resonant frequency and quality factor of the cantilevers were related with the viscoelastic properties of hydrogels through both a first principles approach and empirical correlation. Electrode functionalized hollow fibers were implemented as impedimetric sensors to monitor bioink quality during 3D printing. Impedance spectra were collected during extrusion of cell-laden bioinks and the magnitude and phased angle of the impedance response correlated with quality measures such as cell viability, cell type, and stemness which were validated with traditional off-line assays. / Doctor of Philosophy / The research presented in this dissertation supports the overall goal of producing sensor functionalized neural microphysiological systems to enable deeper fundamental understandings of disease pathology and to provide drug screening and discovery platforms for improved clinical translation. Microphysiological systems are miniaturized tissue constructs which strive to mimic the complex conditions present in-vivo within an in-vitro platform. By producing these microphysiological systems with sensing functionality, new insight into the mechanistic progression of diseases and the response to new treatment options can be realized. Towards this goal, work addressing three broad objectives has been completed. The first objective was expanding the manufacturing process capabilities for hydrogels and tissues through augmentation of the 3D printing systems and developing novel modeling capabilities. The second objective was to expand the palette of available materials which exhibit both the properties required for 3D printing and the mechanical and biological properties required for neural tissue culture. The third objective was to develop sensing capabilities for both monitoring and control of the manufacturing process and to provide non-destructive assessment of microphysiological systems in real-time to quantify the dynamics of disease progression or response to treatment. Through these efforts higher quality microphysiological systems may be produced benefitting future researchers, medical professionals, and patients.

The placenta is a highly vascularized, temporary organ developed in pregnancy that is composed of both maternal and fetal cells. It plays a pivotal role in gestational health by facilitating embryo implantation and fostering nutrient exchange between mother and fetus. Placental malformation and the diffusion of harmful exogenous substances through the placental barrier can cause pregnancy complications and, in more severe cases, death of the mother or the fetus. Further, the placenta undergoes profound morphological and functional changes throughout pregnancy. Establishing models to mimic these phenomena at different stages of pregnancy informs prescription drug safety and expedites the development of placental disease treatments. Mouse models are often used to simulate human fetal development despite major interspecies differences. These limitations drive researchers to developing in vitro models consisting of human-derived cells. This thesis presents three 3D vascularized placental models utilizing human placental stem cells (PSCs) and human umbilical vein endothelial cells (HUVECs) which can model multiple placental phenomena across early- and late-stage pregnancy. The first model features a 3D fibrin hydrogel network with self-assembled vasculature and a monolayer of syncytialized human trophoblastic stem cells (STs) serving as a platform for barrier studies at the maternal-fetal interface. By tuning trophoblast differentiation and vascularization of this model to mimic the early- and late-stage placenta, it was revealed that placental barrier permeability was dependent on placental maturity and that the vascular barrier is also a critical determinant of what molecules can be passed from the mother to the fetus. The design and manufacturing of this model were then streamlined to meet the demands of large-scale drug studies in the second placental barrier model. Placental invasion into the maternal decidua is carefully orchestrated by multiple cell types to prevent over- and under-invasion, both of which can be dangerous to the mother and fetus. Understanding the biochemical and environmental cues that permit this healthy invasion can allow for improved diagnostics and treatments of placental diseases, such as preeclampsia and placenta accreta. Thus, the third model presented herein is a placental invasion model with chorionic villus-like structures seeded with invasive extravillous cytotrophoblasts (EVTs) and a perfusable vascular channel. Collectively, these models facilitate the exploration of placental morphogenesis and function throughout various stages of pregnancy. They offer a valuable tool for probing placental dysfunctions and assessing drug safety, ultimately contributing to advancements in fetal-maternal health. / Thesis / Doctor of Philosophy (PhD) / The placenta is an essential organ in pregnancy and is responsible for a variety of phenomena that assure the survival of the fetus. However, many women suffer from negative pregnancy outcomes due to placental disorders, such as preeclampsia, or due to the crossing of unsafe compounds through the placenta to the fetus. Trophoblasts are the most notable placental cell type originating from the fetus and they have the capacity to mature into more specialized subtypes that are responsible for placental barrier function and placental development via invasion into the maternal tissue. In this work, we have designed three systems that either model placental barrier function or trophoblast invasion by culturing primary endothelial cells with differentiated trophoblast cells on a gel-based device. Using the barrier models, it is possible to assess the rate of transport of different compounds that may be present in the mother’s blood to the fetus, to assess their safety. Whereas the invasion model has the capacity to model the genesis of the placenta and therefore may be used to shed light on the causes for placental dysfunctions at the early stage of pregnancy.

<p>Endothelial dysfunction is the predominant pathophysiological state prior to the onset of atherosclerosis. Currently, treatments for endothelial dysfunction are evaluated in vitro using two-dimensional (2D) cell culture assays or in vivo animal models. Microphysiological systems are small-scale three-dimensional (3D) tissue models that recapitulate the native tissue structure and function. An ideal microphysiological system is comprised of human cells embedded within a 3D matrix introduced to physiological fluid perfusion. Immune challenge in the form of cytokines or immune cells further recapitulates the native microenvironment.</p><p>A vascular microphysiological system was developed from a small-diameter tissue engineered blood vessel (TEBV) in a perfusion culture circuit. TEBVs were created from collagen gels embedded with human neonatal dermal fibroblasts and plastically compressed to yield collagen constructs with high fiber densities. TEBVs are rapidly producible and can be directly introduced into perfusion culture immediately after fabrication. Endothelium-independent vasoconstriction in response to phenylephrine and endothelium-dependent vasodilation in response to acetylcholine were used to analyze the health and function of the endothelium non-destructively over time.</p><p>Endothelial dysfunction was induced through introduction of the pro-inflammatory cytokine tumor necrosis factor – α (TNF-α). Late-outgrowth endothelial progenitor cells derived from the peripheral blood of coronary artery disease patients (CAD EPCs) were evaluated as a potential endothelial source for autologous implantation in both a two-dimensional (2D) direct co-culture model as well as a 3D model as an endothelial source for a tissue engineered blood vessel. CAD EPCs demonstrated similar adhesive properties to a confluent, quiescent layer of smooth muscle compared to human aortic endothelial cells. Within the TEBV system, CAD EPCs demonstrated the capacity to elicit endothelium-dependent vasodilation. CAD EPCs were compared to adult EPCs from young, healthy volunteers. Both CAD EPCs and healthy volunteer EPCs demonstrated similar endothelium-dependent vasoactivity in response to acetylcholine; however, in response to TNF-α, CAD EPCs demonstrated a reduced response to phenylephrine at high doses.</p><p>The treatment of TEBVs with statins was explored to model the drug response within the system. TEBVs were treated with lovastatin, atorvastatin, and rosuvastatin for three days prior to exposure to TNF-α. In all three cases, statins prevented TNF-α induced vasoconstriction in response to acetylcholine within the TEBVs, compared to TEBVs not treated with statins. Overall, this work characterizes and validates a novel vascular microphysiological system that can be tested in situ in order to determine the effects of various patient populations and drugs on endothelial health and function under healthy and inflammatory conditions.</p> / Dissertation

京都大学 / 新制・課程博士 / 博士(工学) / 甲第24610号 / 工博第5116号 / 新制||工||1978(附属図書館) / 京都大学大学院工学研究科マイクロエンジニアリング専攻 / (主査)教授 土屋 智由, 教授 横川 隆司, 教授 安達 泰治, 教授 田畑 修(京都先端科学大学) / 学位規則第4条第1項該当 / Doctor of Philosophy (Engineering) / Kyoto University / DFAM

This research developed methods for culturing self-assembling capillaries in an injectable gel as a potential method for vascularizing tissue-on-a-chip models to mimic physiological drug delivery. Additionally, a mathematical model was developed as a tool for understanding nutrient delivery and comparison of potential delivery systems. Organs-on-a-chip provide novel platforms for studying biology and physiology in 3D, allow exploration of tissue engineering on a manageable scale, and serve as models for drug screening and drug-delivery testing. Methods were first developed for co-culture of endothelial cells and fibroblasts (3T3s or HDFs) in 2D, evaluating culture time, seeding density and ratio of HUVECs and fibroblasts, and immunostaining with a HUVEC-specific marker. Cells formed large sheets with no signs of vessel formation in 2D; therefore, the setup was translated to 3D culture to further induce stress and release of angiogenetic factors, using fibrin gel to suspend cells in 3D. After 9 days of culture, HUVECs had extensive network formation with a high degree of complexity in the experimental cell ratios (especially with 5:1 HUVECs:HDFs). Therefore, these parameters can be used as a starting point for further development of vascularized tissue constructs. A mathematical model was also successfully developed to assess the impact of cell concentration, consumption, and mode of nutrient delivery on 3D cellular constructs which can be used to predict the spatial distribution of glucose over time. Although the model shows flow introduced through a device is sufficient to maintain nutrient levels for cell growth, developing perfusable capillaries is still a critical part of creating physiologically representative tissues.

<p>  </p> <p>Diabetes, a chronic condition characterized by elevated blood glucose levels, arises when pancreatic β-cells lose capacity to produce a robust, dynamic glucose-stimulated insulin secretion (GSIS) response. Accurate measurement of β-cell health and function <em>ex vivo</em> is thus fundamental to diabetes research, including studies evaluating disease mechanisms, novel drug candidates, and replacement β-cell populations. However, present-day dynamic GSIS assays typically represent end-point measurements, involve expensive commercial perifusion machines, and require time-consuming enzyme-linked immunosorbent assays (ELISA) for insulin detection. Microfluidic devices developed as accessible, low-cost alternatives still rely on secondary ELISAs and suspend islets in liquid medium, limiting their survival <em>in vitro</em>. Here, we present a novel, 3D-printed microphysiological system (MPS) designed to recreate components of <em>in-vivo</em> microenvironments through encapsulation in fibrillar type I collagen and restoration of favorable molecular transport conditions. Following computational-informed design and rapid prototyping, the MPS platform sustained collagen-encapsulated mouse islet viability and cytoarchitecture for 5 days and supported <em>in-situ</em> measurements of dynamic β-cell function. To rapidly detect insulin secretion from β-cells in the MPS, we then developed a highly sensitive electrochemical sensor for zinc (Zn2+), co-released with insulin, based on glassy carbon electrodes modified with bismuth and indium and coated with Nafion. Finally, we validated sensor detection of Zn2+ released from glucose-stimulated INS-1 β-cells and primary mouse islets, finding high correlation with insulin as measured by standard ELISA. Together, the 3D MPS and Zn2+ sensor developed in this dissertation represent novel platforms for evaluating β-cell health and function in a low-cost, user-friendly, and physiologically-relevant manner. </p>