Research
Our long-term goal is to resolve the contributions of cell- and tissue-scale mechanics and protein dysregulation in tumor initiation, growth, and metastasis. Our lab consists of an interdisciplinary team of mechanical, manufacturing, materials, and biomedical engineers as well as biophysicists and biologists. Together, we study the mechanical properties of soft biomaterials and tissues, which informs our engineering of in vitro three-dimensional matrix models. We use the models to investigate cell-matrix interactions and the biomechanics of tumor growth.
Each tissue in the body, from the skeleton to the brain, has unique mechanical properties that are dictated by function and closely linked with health. Disease is sometimes associated with abnormal tissue mechanical properties, like the stiffening of solid tumors. Accurate mechanical characterization of tissues is necessary to decipher the impact of particular mechanical properties on cell behavior and disease progression. Compression and indentation are methods well-suited for mechanical characterization of soft materials and amorphous tissues. We use milli-scale compression and indentation for “whole tissue” testing and, due to mechanical heterogeneity at the much smaller length scales of the tissue’s cell and extracellular matrix (ECM) components, atomic force microscopy for indentation on the micro- and nano-scale.
Collagen bundles (3D projection from Gong et al. 2020) embedded in a nanoporous agarose hydrogel (scale bar 100 μm), provide a model of the two-phase gel + fiber structure of the ECM. Whereas the fiber-embedded gel has a slight, but not significant, increase in indentation modulus compared to a pure agarose gel when measured using a 3 mm-diameter spherical indenter, the stiff collagen fibers can be readily mapped in 15 μm × 15 μm moduli maps with a 10 μm diameter AFM tip. Jamie Gearhart master’s thesis, not yet published.
Three-dimensional in vitro models allow for cell-cell and cell-matrix interplay with an added dimension that better simulates the native environment. However, many 3D in vitro models are simply cells or aggregates embedded in uniform matrices that do not account for the morphological and/or mechanical properties of the tissue or tissue-scale architecture, all of which affect cell organization and function. In order to study the effect of these on cell behavior, we engineer 3D in vitro matrix models that recapitulate features of the ECM structure and mechanical properties.
Often a goal when developing in vitro models is to facilitate investigations of the biomechanical interactions of cells with the ECM. However, few in vitro ECM models have taken into account the composite gel and fiber structure together with the length-scales of the fiber architecture. In collaboration with Prof. Johnson Samuel (MANE, RPI), we have developed a continuous hybrid manufacturing process to make fiber-reinforced composite hydrogels to mimic both the gel-like and fibrous components of the ECM with random 3D deposition of fibers instead of a layer-by-layer mat structure. A far field electrospinning process was combined with a droplet-based system to form these in vitro models. We demonstrated how our matrix model provides specific morphological cues to cancer cells.
Collagen structure, including fiber and pore size as well as alignment, profoundly influences cell behavior and function. In turn, cells remodel the collagen structure, a phenomenon that, in the tumor microenvironment, is associated with increased fiber bundle sizes and increased cross-linking. The majority of current collagen-based in vitro tumor models are gels composed of reticular, isotropic nanofibers that do not fully recapitulate in vivo tumor stromal collagen with fibers that present at several microns in diameter.
Tissue-scale architecture, which defines, for example, tissue interfaces and curvature, can affect cell organization and function. We have developed, and continue to engineer, a method to construct organotypic models of tubular organs, such as lymphatic vessels and the mammary duct. With the ability to use a wide variety of matrix-mimicking hydrogels, we are addressing questions about the influence of the architecture and matrix biomechanical properties on the growth of tumor emboli in the vasculature and ductal carcinoma in situ, early stage breast cancer.
The fibrous protein network of the ECM not only supports cells but influences their behaviors such as migration, matrix remodeling, and proliferation. The fibers span diameters from the nanometer range up to microns, creating a diverse microarchitecture for the cells and providing specific topographical and mechanical cues. The process of identifying (mechanosensing) and responding (mechanotransduction) to the fibrous morphologies is essential for many cells to maintain homeostasis. Abnormal ECM or failure of the cells to appropriately respond to the ECM can lead to negative outcomes, such as the formation of tumors. We investigate cell mechanosensing and mecahanotransduction, particularly tumor-associated fibroblasts and metastatic cancer cells, in three models:
An increase in fiber thickness or pore size could provide a metastatic cell with an easy pathway to migrate out of a primary tumor. Our collagen bundles provide a way to compare cancer cell migration on collagen bundles, similar in dimension to fibers surrounding a tumor, to nanofibrous collagen gels that are typically used as in vitro models. The collagen bundles were aligned using a simple microfluidic device and metastatic breast cancer cells were tracked migrating on and in between the bundles. Their migration speed and directedness were compared to that within isotropic nanofibrous collagen gels, aligned nanofibrous collagen gels, and on collagen-coated glass slides. The directedness of migration on the bundles was greater than on the isotropic fibers and the speed was greater on the bundles than in the aligned fibers. In the 3D environment, the combined increase in directedness and speed may translate to an increased likelihood of tumor cells escaping the tumor if they are adjacent to thick and aligned collagen bundles.
Patients with neurofibromatosis type I (NF1) or neurofibromatosis type II (NF2) may present with one or more of several tumor phenotypes of the peripheral nervous system. Neoplastic Schwann cells or Schwann cell precursors are necessary but not sufficient for the tumor development, which is shown to additionally depend on a haploinsufficient (NF1+/- or NF2+/-) stromal environment. The resulting depletion of neurofibromin or merlin, respectively, which are part of important mechanosensing pathways, has been shown to alter how cells interact with the ECM. This is especially important for tumor-associated fibroblasts that have the capacity to alter the ECM surrounding a tumor.
A practical challenge in studying cell-matrix interactions is longitudinal monitoring of behavioral variations within a population of cells in 3D matrices to make statistically powerful assessments. Population-level measurements mask heterogeneity in cell responses, and large-scale individual cell measurements are often performed in a one-time, snapshot manner after removing cells from their matrix. We developed an easy and low-cost method for large-scale, longitudinal studies of heterogeneous cell behavior in 3D hydrogel matrices (Gong et al. 2018). Using a platform we term the drop-patterning chip, thousands of cells may be simultaneously transferred from microwell arrays and fully embedded, only using the force of gravity, in precise patterns in several hydrogel types. This method allows for throughputs approaching 2D patterning methods that lack phenotypic information on cell-matrix interactions, and does not rely on special equipment and cell treatments. With a large and yet well-organized group of cells captured in 3D matrices, we have the capability of locating selected individual cells and monitoring cell division, migration, and proliferation for multiple days. This platform is vital to ongoing work identifying combinations of gene expression and matrix properties that may combine to promote tumor cell invasion and metastasis.
Tissue growth is a complex process involving protein signaling, cell-cell and cell-ECM interactions, and a myriad of biochemical cues and gradients. During growth cells divide, leading to tissue expansion against the surrounding ECM that constrains it, which induces an internal compression in the tissue. How this compressive stress affects cell division and tissue growth is not understood. But we do know that diseases like cancer, which lead to the seemingly chaotic growth of tumors, often initiate and flourish in environments that have significantly abnormal mechanical properties. Thus, a greater understanding of the effects of compressive stress is critical.
© 2015 Kristen Mills. Statements.