Slowly degradable porous silk microfabricated scaffolds for vascularized tissue formation

Abstract

There is critical clinical demand for tissue-engineered (TE), three-dimensional (3D) constructs for tissue repair and organ replacements. Current efforts toward this goal are prone to necrosis at the core of larger constructs because of limited oxygen and nutrient diffusion. Therefore, critically sized 3D TE constructs demand an immediate vascular system for sustained tissue function upon implantation. To address this challenge the goal of this project was to develop a strategy to incorporate microchannels into a porous silk TE scaffold that could be fabricated reproducibly using microfabrication and soft lithography. Silk is a suitable biopolymer material for this application because it is mechanically robust, biocompatible, slowly degrades in vivo, and has been used in a variety of TE constructs. We report the fabrication of a silk-based TE scaffold that contains an embedded network of porous microchannels. Enclosed porous microchannels support endothelial lumen formation, a critical step toward development of the vascular niche, while the porous scaffold surrounding the microchannels supports tissue formation, demonstrated using human mesenchymal stem cells. This approach for fabricating vascularized TE constructs is advantageous compared to previous systems, which lack porosity and biodegradability or degrade too rapidly to sustain tissue structure and function. The broader impact of this research will enable the systemic study and development of complex, critically-sized engineered tissues, from regenerative medicine to in vitro tissue models of disease states.

Keywords: biodegradable; microfabrication; silk; tissue engineering; vascularization.

Figure 1

Schematic of micropatterned porous silk scaffold fabrication. (A) Microfluidic channel patterns are transferred from a silicon wafer to polydimethyl siloxane (PDMS) using standard photolithography and soft lithography techniques. The PDMS mold is trimmed to the desired shape and size. (B) A porous silk film is cast over the PDMS mold and water annealed to induce β-sheet crystallinity. An aqueous-derived salt leached silk scaffold is assembled over the silk film. The scaffold is cured and removed from the PDMS mold. SEM images show patent microchannels on the surface of the scaffold and 100–350 um diameter interconnected pores in the bulk of the scaffold. (C) The scaffold thickness is trimmed by swiping a blade across the top of a positioning mold, which removes the bit of exposed scaffold. This mold allows for tunable control over the resulting thickness of the scaffold by changing the height of the positioning block.

Figure 2

Morphology of porous silk microchannels. (A) Channel profiles can be half-moon or rectangular (scale bars = 50 μm). (B) The optimal microchannel height is 25 μm. Microchannels with heights greater than 25 μm tear during removal from the PDMS mold as demonstrated by 125 μm high channels (scale bars = 50 um). (C) After the microchannels are removed from the PDMS mold there is a 1–10% increase in channel width and 20–34% decrease in channel depth. (D) Channel widths range from 25 μm to 300 μm and versatile branching patterns can be achieved (scale bar = 100 μm). (E) The microchannels contain pores that are open to the bulk space and are 2.2 um (1.3 +/− standard deviation) in diameter (scale bar = 20 μm).

Figure 3

Enclosed microchannels within a porous silk platform. (A) Schematic of the bonding procedure in which an adhesive solution is pipetted into the flat surface, platforms are stacked, and the adhesive is cured. (B–C) H&E histological cross sections of bonded microfluidic channels, showing the channel profile surrounded by a porous scaffold. (B) Microchannels bonded with a silk-based adhesive. The silk-based adhesive is prepared by mixing 6% wt v⁻¹ silk solution with 0.3 M HCl in a 9:1 volume ratio. (C) Microchannels bonded with a fibrin adhesive. Fibrin adhesive is prepared by mixing 1–20 mg ml⁻¹ fibrinogen solution with a 5 U ml⁻¹ in a 4:1 volume ratio. B & C scale bars = 50 μm.

Figure 4

Representative CSLM fluorescence and transmission images of human microvascular endothelial cells (hMVECs) cultured in open microchannel platforms for seven days. hMVECs were stained with calcein AM to determine cell localization and viability within the michrochannels. (A) hMVECs proliferated to confluence in the microchannels but were sparsely observed in the scaffold bulk. Scale bars = 300 μm. (B) CSLM images of top-down and cross-sectional views showed that the hMVECs proliferated on the bottom and side walls of the channels. Scale bars = 300 μm.

Figure 5

Representative histological cross-sections of hMVECs cultured in enclosed microchannel platforms. (A) Schematic of seeding and culturing procedure. hMVECS were cultured in an open channel platform for two days before enclosing the channels with fibrin adhesive. (B) H&E stained sections of enclosed channel platforms. Platforms were cultured either statically or dynamically in a spinner flask. Cells were sparsely observed in the microchannels of statically cultured platforms. In contrast, lumen-like structures were found in the microchannels of dynamically cultured platforms, indicated by black arrows. Scale bars = 100 μm. (C) Histological sections of platforms cultured in spinner flasks and probed for VE-cadherin. Intact lumen structures were observed at day one and four time points. At day seven, lumen-like structures appeared less intact. Scale bars = 100 um. (D) At day four, lumen-like structures were observed in 50 μm, 100 μm, and 200 μm wide channels. Lumen structures were not found in 25 μm wide channels. Scale bars = 50 um.

Figure 6

hMVECs co-cultured with human mesenchymal stem cells (hMSCs) in the microchannel platform. (A) Schematic of the seeding protocol. hMVECS were seeded first on the patterned scaffold surface. After two days of static culture, the microchannels were enclosed and hMSCs were seeded in the platform bulk space. (B) After the bonding step and one day of dynamic culture, the hMVECs (red) were observed in the microchannel space and the hMSCs (blue) were observed in the bulk space. After seven days of dynamic culture the hMVECs proliferated in the microchannel space, while hMSCs proliferated in the platform bulk (scale bars = 300 μm). (C) Higher magnification of a microchannel after seven days of dynamic culture shows hMVECs lining the microchannel while hMSCs populate the space around the microchannel (scale bar = 50 μm).