We present a method for microfluidic deposition of patterned genipin and fibronectin on PDMS substrates, allowing extended viability of vascular smooth muscle cell-dense tissues. This tissue fabrication method is combined with previous vascular muscular thin film technology to measure vascular contractility over disease-relevant time courses.
The chronic nature of vascular disease progression requires the development of experimental techniques that simulate physiologic and pathologic vascular behaviors on disease-relevant time scales. Previously, microcontact printing has been used to fabricate two-dimensional functional arterial mimics through patterning of extracellular matrix protein as guidance cues for tissue organization. Vascular muscular thin films utilized these mimics to assess functional contractility. However, the microcontact printing fabrication technique used typically incorporates hydrophobic PDMS substrates. As the tissue turns over the underlying extracellular matrix, new proteins must undergo a conformational change or denaturing in order to expose hydrophobic amino acid residues to the hydrophobic PDMS surfaces for attachment, resulting in altered matrix protein bioactivity, delamination, and death of the tissues.
Here, we present a microfluidic deposition technique for patterning of the crosslinker compound genipin. Genipin serves as an intermediary between patterned tissues and PDMS substrates, allowing cells to deposit newly-synthesized extracellular matrix protein onto a more hydrophilic surface and remain attached to the PDMS substrates. We also show that extracellular matrix proteins can be patterned directly onto deposited genipin, allowing dictation of engineered tissue structure. Tissues fabricated with this technique show high fidelity in both structural alignment and contractile function of vascular smooth muscle tissue in a vascular muscular thin film model. This technique can be extended using other cell types and provides the framework for future study of chronic tissue- and organ-level functionality.
Vascular diseases, such as cerebral vasospasm1,2, hypertension3, and atherosclerosis4, develop slowly, are typically chronic in nature, and involve dysfunctional force-generation by vascular smooth muscle cells (VSMCs). We aim to study these slow-progressing vascular dysfunctions using in vitro methods with finer control of experimental conditions than in in vivo models. We have previously developed vascular muscular thin films (vMTFs) for measuring functional contractility of in vitro engineered cardiovascular tissues5, but this method has been limited to relatively short-term studies. Here, we present a substrate modification technique that expands our previous vMTF technique for long-term measurements.
While the endothelium is also critical in overall vascular function, engineered arterial lamellae provide a useful model system for assessing changes in vascular contractility during disease progression. To engineer a functional vascular disease tissue model, both the structure and function of the arterial lamella, the basic contractile unit of the vessel, must be recapitulated with high fidelity. Arterial lamellae are concentric, circumferentially-aligned sheets of contractile VSMCs separated by sheets of elastin6. Microcontact printing of extracellular matrix (ECM) proteins onto polydimethylsiloxane (PDMS) substrates has been previously used to provide guidance cues for tissue organization to mimic aligned cardiovascular tissue5,7-10. However, tissues patterned using microcontact printing can lose integrity after 3-4 days in culture, limiting their applicability in chronic studies. This protocol provides a solution to this issue by replacing previous microcontact printing techniques with a new microfluidic deposition technique.
Genchi et al. modified PDMS substrates with genipin and found prolonged viability of myocytes up to one month in culture11. Here, we use a similar approach to extend culture of patterned vascular smooth muscle cells on PDMS. Genipin, a natural hydrolytic derivative of the gardenia fruit, is a desirable candidate for substrate modification due to its relatively low toxicity compared to similar crosslinking agents and its increasing use as a biomaterial in the fields of tissue repair12,13 and ECM modification14,15. In this protocol, fibronectin is utilized as a cell guidance cue, as in previous microcontact printing methods; however, genipin is deposited onto PDMS substrates prior to fibronectin patterning. Thus, as cells degrade the patterned matrix, newly synthesized ECM from attached VSMCs can bind to the genipin-coated PDMS substrate.
This protocol utilizes a microfluidic delivery device for two-step genipin and ECM deposition. The design of the microfluidic device mimics microcontact printing patterns used for engineered arterial lamellae in previous studies16. Thus, we expect this protocol to yield arterial lamellae mimics that successfully recapitulate the highly-aligned in vivo structure and contractile function of arterial lamellae. We also evaluate tissue contractility to confirm that genipin is a suitable substrate modification compound for long-term in vitro vascular disease models.
Note: The goal of this protocol is to construct and utilize a vascular muscular thin film (vMTF) with the structure shown in Figure 1 to assess contractility during extended culture of vascular smooth muscle cells (VSMCs) on PDMS substrates. To prolong VSMC viability, we utilize the crosslinker compound genipin. The substrates for these vMTFs are designed to analyze tissue contractility as developed by Grosberg et al.8 Other vMTF methods5 may also be used, with subtle changes to the presented substrate fabrication protocol.
1. Substrate Fabrication
2. Microfluidic Patterning for Engineering Tissues
Figure 1. Microfluidic Protein Delivery Device. (A) Taped off coverslip for PIPAAm coating. Red dotted circle: cutting path to release coverslip. (B) Representative AutoCAD drawing of tissue microfluidic mask pattern. Inset: Detail of binary branching to alternating 10 µm x 10 µm tissue pattern. (C) Placement of microfluidic device on a coverslip substrate with inlet and outlet indicated. (D) Schematic of microfluidic protein patterning and delivery. Left-to-right: scanning electron microscope image of microfluidic channels (scale bar: 50 µm); Detailed schematic of method for protein deposition; Immunohistochemistry stained fibronectin (scale bar: 50 µm); Cell seeding with vascular smooth muscle cells. (E) Schematic of fabricated tissue. 1st inset: Detail of layered construct. 2nd inset: Detail of genipin modification of PDMS substrate after microfluidic deposition. © IOP Publishing. Reproduced and/or modified with permission. All rights reserved.19 Please click here to view a larger version of this figure.
3. Tissue Function Analysis with vMTF Contractility Assay
Note: The MTF contractility assay presented here is modeled after the technique developed in Grosberg et al.8
The primary goal of this work was to extend the viability of micropatterned VSMCs on hydrophobic PDMS substrates. This was accomplished by incorporating a microfluidic delivery system to deposit patterned genipin and fibronectin on PDMS (Figure 1). Deposition of ECM proteins using microfluidic delivery yielded high fidelity transfer of the channel pattern with bare PDMS between lines of genipin and fibronectin (Figure 1D). The attached cells (Figure 1E) form confluent monolayers mimicking the in vivo structure of arterial lamellae (Figure 2), similar to previous microcontact printing methods5,10,16. These tissues yielded responsive, contractile constructs, whose stress was measured using vMTF technology (Figure 3).
Qualitative assessment of tissue viability over the course of two weeks showed minimal deterioration on genipin-modified substrates (Figure 2A). Tissue confluence and alignment were maintained over two weeks (Figure 2B, 2C and 2D). Two key stress values were calculated for every vMTF: 1) basal tone and 2) induced contractility (Figure 3C and 3D). Basal tone is the stress maintained by unstimulated VSMCs at equilibrium. Induced contractility is the additional stress induced by stimulation with endothelin-1. Both basal tone and induced contractility showed consistent behavior over the two-week time course, demonstrating vasoactive tissues throughout (Figure 3E and 3F). The slight drop in tissue contractility at the end of the assay is the direct result of the reduced number of cells composing the tissue, since serum-starved VSMCs do not proliferate19. The addition of a minimal basal level of serum to culture medium may alleviate this result in future work.
Figure 2. Tissues Remain Viable and Successfully Mimic in vivo Arterial Lamellar Structure for Two Weeks on Genipin-modified Substrates. (A) Representative phase contrast images of tissues at sacrifice time points throughout the course of two weeks (scale bar: 200 µm). (B) Representative immunohistochemistry images of tissues fabricated on genipin-modified substrates fixed at Day 1, Day 4, and Day 10 after serum starvation (green: f-actin filaments, blue: nuclei (shown to establish presence of cells), scale bar: 100 µm). (C) Percent confluence measured by f-actin coverage (error bars: standard error, n = 3 – 7). (D) Tissue alignment measured by f-actin orientation order parameter (OOP)22 (error bars: standard error, n = 3 – 7). © IOP Publishing. Reproduced and/or modified with permission. All rights reserved. All rights reserved.19 Please click here to view a larger version of this figure.
Figure 3. Tissue Contractility is Maintained Over the Course of Two Weeks. (A) Representative vMTF cutting scheme. (B) Schematic of cut vMTF. Upon cooling below 32 °C, PIPAAm dissolves, releasing vMTF. Stress in the active cell layer causes the passive PDMS layer to bend. Measurement of projection length can be converted to a radius of curvature according to methods in Grosberg et al.8 Radius of curvature is used to calculate the average cross-sectional stress in the tissue. (C) Sequential transmitted light images of representative contractility assay (scale bar: 1 mm). Tissues reach equilibrium, are stimulated with endothelin-1, then treated with HA-1077 to allow complete relaxation. Bottom: Schematic of side view of idealized tissue during the course of the assay. (D) Representative stress curve for the contractility assay. Two key stress values are calculated. Basal tone is the difference between the equilibrium stress state and the relaxed stress state. Induced contractility is the change in stress from the equilibrium state to the endothelin-1 stimulated state. (E) Basal tone (error bars: standard error, n = 5 – 12). (F) Induced contractility (error bars: standard error, n = 5 – 12). © IOP Publishing. Reproduced and/or modified with permission. All rights reserved.19 Please click here to view a larger version of this figure.
Here, we present a protocol that builds upon previously developed vMTF technology, allowing extended experiment times more typical of chronic vascular disease pathways1,23,24. To accomplish this, we micropattern genipin, which has previously been shown to provide long-term functionalization of PDMS substrates11, using a microfluidic deposition technique to yield engineered arterial lamellae with improved vascular tissue viability for use in MTF contractility experiments. McCain et al. developed an alternative micromolded gelatin hydrogel substrate for extended culture of engineered cardiac tissues for several weeks in a related MTF model25.
This protocol yielded vMTFs that successfully mimicked arterial lamella structure (Figure 2) and function (Figure 3) over the course of two weeks. While successful completion of the presented protocol yields the desired result of extended culture times for disease-relevant time course (e.g., the pathologic effects of cerebral vasospasm persisting for up to 14 days26), a few common pitfalls arise. Repeated use of PDMS microfluidic devices results in deleterious damage that can result in partial or complete blockage of branches in the device. Thus, new devices must be used for each experiment. Another issue, while not as common, is unexpected and inconsistent delamination of tissues. These occurrences seemed to be random in nature and rare in occurrence. By observation, tissues form an arch- or lumen-like structure before delaminating. We also observed this behavior in tissues fabricated using microcontact printing techniques. Thus, we believe this issue to be the result of either over-seeding or a phenotypic switch in the cultured VSMCs that results in abnormal behavior and not a direct result of the fabrication methods presented here.
The current microfluidic design requires linear flow patterns and as such, is limited to tissues with alignment in a single principal direction. Application to tissues requiring more complex patterning, such as the “brick wall” pattern of ventricular myocardial structure27, will require more elaborate microfluidic design. We have not seeded genipin-modified substrates with other cell types. However, Genchi et al. showed extended viability of isotropic skeletal muscle on genipin-modified PDMS11. Thus, we feel confident that a minimally-modified version of this protocol has widespread applicability to other organ systems in the future development of organ-on-a-chip technologies.
The stress calculation method for vMTFs based on previous methods8 is limited by the relative contractility in the tissues and corresponding cut length of the thin films. If cut too long, films will curl on themselves. If cut too short, films will not bend. Either case prevents proper analysis of contractile properties due to model limitations. Proper cut length should be refined through repeated experiments. Implementation of a laser engraving system, as in Agarwal et al., can improve upon repeatability and quality of MTF cutting28.
The ability to better recapitulate native tissue structure and function in a tightly-controlled, self-contained experimental system is a key pursuit in the field of bioengineering. This pursuit has led to the development of several in vitro tissue mimics and multi-functional, yet simplified model organs-on-a-chip, advancing the understanding of fundamental physiologic and pathologic behaviors of organ systems29,30. Using deposition of genipin onto PDMS substrates, we have demonstrated extended cell viability and maintenance of vascular smooth muscle function over two weeks. This is a significant improvement over previous fabrication techniques, which can lead to delamination of tissues and cell death after 4 – 7 days in culture19, and may aid development of more robust artery-on-a-chip methods. Due to the chronic nature of most vascular diseases, this advance provides the framework for a variety of future investigations into the contractile mechanisms involved in specific vascular pathologies.
The authors have nothing to disclose.
We acknowledge financial support from the American Heart Association Scientist Development Grant, 13SDG14670062 (PWA) and the University of Minnesota Doctoral Dissertation Fellowship (ESH). We also acknowledge the microfabrication resources of the Minnesota Nano Center (MNC) and the image processing resources of the University Imaging Centers (UIC), both at the University of Minnesota. Parts of this work were carried out in the Characterization Facility, University of Minnesota, which receives partial support from NSF through the MRS program.
Coverslip staining rack | Electron Microscopy Sciences | www.emsdiasum.com/ | 72239-04 | Alternative coverslip rack may be used |
Microscope cover glass – 25 mm | Fisher Scientific, Inc. | www.fishersci.com | 12-545-102 | Alternative brand and size may be used; Microscope slides may also be substituted as substrate base |
Poly(N-iso-propylacrylamide) (PIPAAm) | Polysciences, Inc. | www.polysciences.com/ | #21458 | Sigma-Aldrich makes an alternate compound, but we have not tested it for use with this protocol; Compound gives strong odor, use proper ventilation |
1-butanol | Sigma-Aldrich | www.sigmaaldrich.com | 360465 | Hazard: flammable (store stock solution in flammable cabinet); flash point is 37 °C, avoid heating; alternative product may be used |
Spincoater | Specialty Coating Systems, Inc. | www.scscoatings.com | SCS G3P8 Model; Alternative brand and/or model may be used | |
Polydimethylsiloxane (PDMS) | Ellsworth Adhesives (Dow Corning) | www.ellsworth.com | 184 SIL ELAST KIT 0.5KG | Alternative distributor may be used |
Fluorescent microbeads | Polysciences, Inc. | www.polysciences.com/ | 17151 | Alternative brand and/or larger size may be used |
Silicon wafers | Wafer World, Inc. | www.waferworld.com | 2398 | Alternative brand and/or size may be used |
Photoresist | MicroChem Corp. | www.microchem.com | SU-8 3025 allows 20-25-µm feature height | |
Contact mask aligner | Suss MicroTec | www.suss.com | MA6 contact mask aligner; alternative brand and/or model may be used for wafer exposure | |
Developer | MicroChem Corp. | www.microchem.com | SU-8 Developer; Hazard: flammable | |
Tridecafluro-trichlorosilane | UCT Specialties, Inc. | www.unitedchem.com | T2492 | Silane for non-stick coating of patterned silicon wafers (CAUTION: Tridecafluro-trichlorosilane is a flammable and corrosive liquid. Proper personal protective equipment and local exhaust is necessary for use. ) |
Surgical biopsy punch | Integra LifeSciences Corp. | www.miltex.com | 33-31AA-P/25 | Alternative brand and/or size may be used |
Genipin | Cayman Chemical | www.caymanchem.com | 10010622 | Sigma-Aldrich (G4796-25MG) makes an alternate compound, but we have not tested it for use with this protocol |
1X phosphate buffered saline | Mediatech, Inc. | www.cellgro.com | 21-031-CV | Alternative brand may be used |
Fibronectin | Corning, Inc. | www.corning.com | 356008 | Sigma-Aldrich (F1056) makes an alternate compound, but we have not tested it for use with this protocol |
Penicillin/streptomycin | Life Technologies, Inc. | www.lifetechnologies.com | 15140-122 | Alternative brand and/or size may be used, as long as concentration is the same |
Umbillical artery smooth muscle cells | Lonza | www.lonza.com | CC-2579 | Alternative cell types may be used for alternative applications. Media should be modified accordingly |
Tyrode's solution components | Sigma-Aldrich | www.sigmaaldrich.com | various | Alternative brand may be used for mixing solution |
Stereomicroscope | Zeiss | www.zeiss.com | 4350020000000000 | SteREOLumar V12; Alternative brand/type of stereomicroscope may be used |
Temperature-controlled platform | Warner Instruments | www.warneronline.com | 641659; 640352; 641922 | |
Endothelin-1 | Sigma-Aldrich | www.sigmaaldrich.com | E7764-50UG | Alternative amount may be purchased, as long as treatment concentration is maintained |
HA-1077 | Sigma-Aldrich | www.sigmaaldrich.com | H139-10MG | Alternative amount may be purchased, as long as treatment concentration is maintained |