Page 42 - ASME_NEMB_2016_Program
P. 42

Technical Program                                 TRACK 3





        process. The 3-D, layered geometry that it generates is a good representa-  analyze. Collagen-based microtissues could be the ideal 3D culture platform
        tion of the actual scaffold geometry, and the fidelity of this geometric repre-  to study 3D cellular behavior, so we designed and optimized a method to
        sentation represents a significant improvement over current models.  rapidly generate uniform collagen gels using straightforward methods that
                                                                are compatible with high throughput technologies.
        The network fibers are modeled as Cosserat rods, in which director vectors
        are used to keep track of the orientation of the fiber cross section. This   Methods:
        approach allows large, 3-D rotations and axial stretches to be handled ac-  Microtissues were constructed of collagen, with or without cells encapsulat-
        curately. Proper treatment of these deformations is important since the large   ed within the hydrogel and/or adhered to the hydrogel surface. A solution
        macroscopic strains experienced by tissues during normal function result in   of 2-6 mg/mL collagen and 8 x 10^5 - 2 x 10^6 cells/mL was perfused into a
        large rotations and stretches of the network fibers at the microscale. The in-  PDMS microfluidic droplet generator at 150 µl/hr, and fluorocarbon oil with
        dividual fibers were connected using constraints at the points of intersection   2% EA surfactant was perfused at 1000 µl/hr to overcome the jetting to drip-
        and at the periodic boundary. Periodic boundary conditions are applied to   ping transition and produce droplets. Droplets were formed at 4°C, and po-
        complete the specification of the network boundary value problem.   lymerized for 30 minutes in a 37°C incubator.  Encapsulated cell (NIH 3T3 fi-
                                                                broblasts or MDA-MB-231eGFP breast cancer cells) behavior and/or adhered
        We first developed an understanding of basic fibrous network phenomena   cell (HUVECs or SUM 149mCherry cancer cells) behavior was observed in a
        for initially isotropic network geometries. The effects of periodic box size,   droplet capture device for 24-72 hours post-polymerization.
        number of samples, and number of layers on the uncertainty in geometric
        and mechanical QOIs was quantified. Predictions were found to be reason-  Results:
        ably accurate for moderate box and sample sizes. Multilayer scaffolds were   We found that polymerized microtissues had a narrow distribution of size
        found to be accurately modeled by a single network layer, despite the fact   with a coefficient of variance ranging from 8.4% - 13.9%, which is sufficiently
        that the monolayer contains one third the number of fiber-to-fiber contacts   low that the size variance does not interfere with subsequent analyses. We
        as the real scaffold. This finding suggests that the fiber interactions are   also observed that encapsulated cell viability was high (79% - 95%) and re-
        secondary to the effects of fiber volume fraction and ODF. This conclusion   mained consistent for the experimental trials (fluctuations were at most 6%).
        is also supported by the finding that the bending energy contribution to the   In assessing cell-matrix interactions, we observed that cancer cells spread
        total strain energy in the network is negligible.       out and sent projections throughout the collagen matrix, indicating that
                                                                these cells are able to adhere and interact with the 3D environment of the
        The effects of fiber alignment, tortuosity, and material model, both individual-  collagen microtissues. For microtissues with encapsulated fibroblasts, we
        ly and in combination, on the macroscopic stretch vs. stretch behavior were   observed and quantified serum-induced gel compaction (25% in 2 mg/mL
        explored. For moderately tortuous networks of linear fibers, increasing fiber   collagen, 21% in 4 mg/mL collagen, and no significant contraction in 6 mg/
        alignment significantly increases mechanical anisotropy, and significantly in-  mL collagen), which shows a trend already reported in analogous bulk gel
        creases nonlinearity under stress-controlled loading. For isotropic networks   experiments: that less gel compaction is observed at higher collagen con-
        of linear fibers, increasing tortuosity moderately increases nonlinearity at   centrations. Finally, we characterized successful co-culture conditions with
        small strains. A linear fiber was compared with a fiber modeled as an incom-  our microtissues via confocal microscopy.
        pressible Yeoh hyperelastic material. Large differences in the nonlinearity of
        fiber response translate into large differences in the network response. For   Discussion:
        networks of aligned fibers using a Yeoh model, increasing tortuosity signifi-  Our easy-to-use method to miniaturize collagen-based tissue constructs
        cantly increases material nonlinearity under stress controlled loading. These   maintains the 3D in vitro environment, while alleviating several obstacles
        simulations allow the effects of geometry and fiber material properties on   associated with larger avascular tissue constructs. The microtissues formed
        macroscale behavior to be studied systematically in a way that is virtually   with our platform are uniform in size and are generated rapidly (>>4500/
        impossible using current experimental techniques.       hour), which facilitates large-scale experiments with large numbers of treat-
                                                                ment groups and hundreds of replicates. Our method uses only common
        Overall, the simulations presented represent an important advancement   laboratory temperatures (4°C and 37°C) and requires no on-chip temperature
        in the field of tissue engineering. The accurate representation of detailed   control or temperature gradients, which means this technology can be easily
        fibrous network geometry permits a greater understanding of the complex   disseminated and can be integrated within a high throughput pipeline. Out-
        mechanisms underlying the macroscopic behavior unique to these bioma-  side of the method development, we have also demonstrated the biological
        terials. Furthermore, simulations of scaffold fiber networks form the basis for   potential for this platform via novel patterning of cell co-cultures within and
        understanding how interactions with cellular and ECM phases contribute to   around the microtissues. Finally, we have shown, for the first time to the au-
        the growth, remodeling, and ultimate mechanical and biological behavior   thor’s knowledge, quantification of gel compaction for tissues on this small
        of the entire engineered tissue. Insights gained from such simulations can   size scale (~200 µm). Overall, this is a highly versatile platform for 3D cell
        significantly aid the process of designing scaffold network geometries that   culture that is engineered to facilitate quantitative studies of drugs, cellular
        result in engineered tissues that function as well as or better than the native   interactions, and biological mechanisms in high throughput applications.
        tissues they are intended to replace.
                                                                10:22am 3D cell culture and osteogenic and chondrogenic dif-
        10:06am Rapid Generation of Collagen-based Microtissues to   ferentiation of human mesenchymal stem cells plated onto an
        Study Cell-matrix Interactions                          electrospun scaffold

        Technical Presentation. NEMB2016-5972                   Technical Presentation. NEMB2016-6069

        Alexandra L Crampton, Marie-Elena Brett, David K Wood, Univer-  Laura Pandolfi, Naama Toledano, Houston Methodist Research
        sity of Minnesota, Minneapolis, MN, United States       Institute, Houston, TX, United States, Francesca Taraballi, Houston
                                                                Methodist Research Inst., Houston, TX, United States, Ennio Tasci-
        Introduction:                                           otti, The Methodist Hospital Research Institute, Houston, TX, United
        2D in vitro models have long been the workhorses of drug discovery; how-  States
        ever, 3D culture models are better able to recapitulate the complex 3D envi-
        ronment of tissues in vitro and have thus come to the forefront of tissue en-  Regenerative processes in living tissues draw from reservoirs of stem cells.
        gineering. One of the most common material choices for 3D tissue scaffolds   Among the plethora of stem cells, mesenchymal stem cells (MSC) have been
        is collagen, which is typically cast as relatively large (~mm) bulk gels to study
   42   embedded or adhered cellular behavior. But fabrication of bulk gels is labor   shown to be multipotent progenitors, able to provide an optimal microenvi-
                                                                ronment at the site of injury. Due to their high plasticity, immunosuppressive
        and time intensive, and the large gels have significant diffusion limitations for   potential and immunomodulatory properties, MSC hold promises in a diver-
        nutrients and signaling molecules, and they are difficult and cumbersome to
   37   38   39   40   41   42   43   44   45   46   47