Microfluidic platform for biomimetic tissue design and multiscale rheological characterization

TL;DR

This work presents a microfluidic platform to assemble GUV-based prototissues with tunable inter-vesicle adhesion and to characterize their rheology across multiple length scales. By implementing two adhesion chemistries—streptavidin–biotin and DNA linkers—the authors create prototissues with a broad range of cohesion and study their deformation under controlled aspiration, analyzing both local GUV responses and global tissue-like behavior. The results show that weak DNA-mediated adhesion yields predominantly viscous, diffusion-driven rearrangements with plasticity limited to local reorganization, while strong SA–biotin adhesion drives viscoelastic responses with substantial GUV deformation and partial irreversibility under stress. The work demonstrates that microfluidic confinement enables reproducible prototissue fabrication and a quantitative, multiscale description of tissue mechanics, providing a robust framework to study how cell–cell adhesion influences flow dynamics in soft tissues.

Abstract

The way living tissues respond to external mechanical forces is crucial in physiological processes like embryogenesis, homeostasis or tumor growth. Providing a complete description across length scales which relates the properties of individual cells to the rheological behavior of complex 3D-tissues remains an open challenge. The development of simplified biomimetic tissues capable of reproducing essential mechanical features of living tissues can help achieving this major goal. We report in this work the development of a microfluidic device that enables to achieve the sequential assembly of biomimetic prototissues and their rheological characterization. We synthesize prototissues by the controlled assembly of Giant Unilamellar Vesicles (GUVs) for which we can tailor their sizes and shapes as well as their level of GUV-GUV adhesion. We address a rheological description at multiple scales which comprises an analysis at the local scale of individual GUVs and at the global scale of the prototissue. The flow behavior of prototissues ranges from purely viscous to viscoelastic for increasing levels of adhesion. At low adhesion the flow response is dominated by viscous dissipation, which is mediated by GUV spatial reorganizations at the local scale, whereas at high adhesion the flow is viscoelastic, which results from a combination of internal reorganizations and deformation of individual GUVs. Such multiscale characterization of model biomimetic tissues provides a robust framework to rationalize the role of cell adhesion in the flow dynamics of living tissues.

Figures (7)

• Figure 1: (a) GUV-GUV bonding strategies: an example of a GUV-doublet is shown. Index 1 and 2 refer to the GUV membrane and the GUV-GUV patch respectively. GUV-GUV adhesion can be mediated by the streptavidin-biotin binding (top arrow) or the hybridization of complementary DNA strands (bottom arrow). The DNA anchor and linker sequences are displayed in red and green, respectively. (b) Scheme (not to scale) of the microfluidic device including the microfluidic chip, a device for the control of the applied pressure, a microscope and an acquisition system. (c) Detail of the microfluidic chip used (not to scale) for the prototissue assembly and aspiration tests. Different colors correspond to different parts of the chip used to achieve different tasks: GUV trapping (red), adhesion of GUV-GUV adhesion by the flow of a solution containing fluorescent streptavidin molecules (yellow), release of the prototissue from the trap by reversing the direction of the flow (green) and aspiration test in the microfluidic constriction (purple). The scale bars corresponds to 100 $\mu$m.
• Figure 2: (a) Example of two GUV-prototissues obtained in the microfluidic trap using SA-biotin binding (red) or DNA complementary strands (green). (b) Angularity ($\Phi$) values obtained for an ensemble of GUVs. Different colors correspond to different adhesion molecules (SA-biotin in red, DNA in green and black in the absence of linker molecules). The values corresponding to GUVs contained within prototissues located inside the trap or after being released are compared. The displayed values correspond to the mean obtained over 10 GUVs (located away from the trap borders) and the error bars to the standard deviation. (c). Example of different prototissues obtained using microfluidic traps of different sizes (left 180 $\mu$m $\times$ 220 $\mu$m and right 290 $\mu$m $\times$ 330 $\mu$m). (d) Size distribution of the GUVs obtained after electroformation (red) and of the GUVs contained in the microfluidic trap (yellow). The scale bars correspond to 100 $\mu$m.
• Figure 3: (a). Snapshot of the microfluidic channel. The identification of individual vesicles contained within the prototissue was achieved based on image segmentation methods. Vesicles belonging to different regions of the channel are labeled with different colors: (1) constriction (black); (2) narrower part of the nozzle, next to the constriction (red); (3) wider part of the nozzle, next do the main channel (green); (4) main channel (blue). (b-c). Mean eccentricity values obtained for each channel region after averaging over all the GUVs located in a given region and over time. The error bars corresepond to the standard deviation. GUV-GUV assembly was mediated by DNA (panel (b)) or SA-biotin binding (panel (c)). (d). Evolution of the mean eccentricity of GUVs over time, obtained during an aspiration sequence for a SA-biotin-prototissue. Each color codes for a different region. The applied pressure signal is also displayed with a purple dashed line (right axis).
• Figure 4: (a) Representative snapshots of a microfluidic aspiration experiment of a vesicle prototissue obtained with SA-biotin binding, including the aspiration and relaxation phases. (b) Identification and tracking of individual vesicles contained within the prototissue obtained using the FastTrack software and a home-made image segmentation routine. Each GUV is represented with a different color. (c) Geometrical mesh used to determine the inter-GUV displacement. Vertexs of the mesh correspond to the center of each GUV contained in the prototissue, and dots highlight the mid-position of all GUV-GUV distances (represented in blue). (d) Spatiotemporal map displaying the eccentricity values obtained for GUVs constituting the prototissue over an aspiration phase. The scale bars in all panels correspond to 100 $\mu$m.
• Figure 5: (a-b) Pressure (top panels), inter-GUV displacements (medium panels) and intra-GUV deformations (bottom panels) as a function of time. Different curves (shown in different colors) correspond to five different pairs of GUVs. Mean values are shown with a dashed line. Panel (a) corresponds to a prototissue assembled with DNA strands and (b) with SA-biotin binding. (c) Intra-GUV deformation represented as a function of the inter-GUV displacement for DNA and SA-biotin prototissues. The area explored by five different GUV-pairs is shown with shaded regions and the mean is represented with a solid line.