A glycosaminoglycan based, modular tissue scaffold system for rapid assembly of perfusable, high cell density, engineered tissues

Abstract

The limited ability to vascularize and perfuse thick, cell-laden tissue constructs has hindered efforts to engineer complex tissues and organs, including liver, heart and kidney. The emerging field of modular tissue engineering aims to address this limitation by fabricating constructs from the bottom up, with the objective of recreating native tissue architecture and promoting extensive vascularization. In this paper, we report the elements of a simple yet efficient method for fabricating vascularized tissue constructs by fusing biodegradable microcapsules with tunable interior environments. Parenchymal cells of various types, (i.e. trophoblasts, vascular smooth muscle cells, hepatocytes) were suspended in glycosaminoglycan (GAG) solutions (4%/1.5% chondroitin sulfate/carboxymethyl cellulose, or 1.5 wt% hyaluronan) and encapsulated by forming chitosan-GAG polyelectrolyte complex membranes around droplets of the cell suspension. The interior capsule environment could be further tuned by blending collagen with or suspending microcarriers in the GAG solution These capsule modules were seeded externally with vascular endothelial cells (VEC), and subsequently fused into tissue constructs possessing VEC-lined, inter-capsule channels. The microcapsules supported high density growth achieving clinically significant cell densities. Fusion of the endothelialized, capsules generated three dimensional constructs with an embedded network of interconnected channels that enabled long-term perfusion culture of the construct. A prototype, engineered liver tissue, formed by fusion of hepatocyte-containing capsules exhibited urea synthesis rates and albumin synthesis rates comparable to standard collagen sandwich hepatocyte cultures. The capsule based, modular approach described here has the potential to allow rapid assembly of tissue constructs with clinically significant cell densities, uniform cell distribution, and endothelialized, perfusable channels.

Figure 1. Bottom-up vs. top-down approaches in tissue engineering.

The traditional, top-down approach (right) involves seeding cells into full sized porous scaffolds to form tissue constructs. This approach poses many limitations such as slow vascularization, diffusion limitations, low cell density and non-uniform cell distribution. In contrast, the modular or bottom-up approach (left) involves assembling small, non-diffusion limited, cell-laden modules to form larger structures and has the potential to eliminate the shortcomings of the traditional approach.

Figure 2. Microencapsulation through complex coacervation and modular assembly.

(A) Droplets of cells suspended in a polyanionic solution were dispensed into a stirred chitosan solution. Ionic interactions between the oppositely charged polymers formed an insoluble ionic complex membrane at the droplet-solution interface, thus encapsulating the suspended cells. Capsule were washed surface-stabilized with a suitable anionic polymer solution, and transferred to culture. (B) Cell laden capsules can be assembled in a packed bed fashion with interconnected endothelialized channels that may enable perfusion of fluids such as blood with limited adverse reactions.

Figure 3. Histology of microencapsulated cultures of human trophoblasts (HTBs) in various GAG-chitosan capsule formulations.

HTBs in CSA/CMC capsules on days (A) 5, (B) 10, (C) 15, (D) 20, (E) 25, (F) 30. (G) Hyaluronan/CMC capsules. (H) HA capsules. (I) Dextran sulfate/CMC capsules quickly ruptured due to osmotic swelling.

Figure 4. HUVECs seeded on CSA/CMC capsules after surface coating with collagen.

(A) Phase contrast image of a capsule coated with a monolayer of HUVECs, 24 hours after cell seeding. (B,C) CellTracker Green fluorescence images of HUVECs seeded on the outer surface shown in A. (D,E) SEM images of HUVEC seeded capsule surfaces after 1 hour (E) and 24 hours (D). (F) Non-seeded capsule surface.

Figure 5. Specific growth rates of aortic smooth muscle cells in HA and CSA/CMC capsules.

Specific growth rates were calculated using DNA measurements. Error bars represent standard deviations of 3–5 independent measurements. Significant differences are denoted by single or double asterix (*  =  p<0.05; **  =  p<0.10).

Figure 6. Albumin permeability measurements and mass transfer characteristics of HA and CSA/CMC capsules.

(A) Representative plots of the concentration factor ln(Q) vs. time for three replicate runs with the CSA/CMC capsule formulation. (B) Plots of permeability coefficient (P), overall mass transfer coefficient (K) and wall thickness (δ) for the HA and CSA/CMC capsule formulations. Error bars represent the standard deviation of three replicate measurements.

Figure 7. Tuning the inner capsule microenvironment with a collagen gel matrix and microcarriers.

(A–B) HTBs in CSA/CMC capsules with a collagen type-I gel after one week of static culture. (A) H&E histology. (B) Phase contrast image. (C–E) SMCs co-encapsulated with gelatin coated dextran (Cytodex-3) microcarriers in HA capsules. (C) 60 min after encapsulation. (D) Day 14 of culture. (E) Calcein-AM stained fluorescence images on day 14 (green  =  live cells, red  =  microcarriers).

Figure 8. Vascular smooth muscle cells in collagen-containing capsules.

(A–C) SMCs in CSA/CMC capsules with a 1 mg/ml collagen gel. (A) 60 min after encapsulation, SMCs are well dispersed in the internal collagen matrix. (B) After 24 hours of culture, the cells had contracted the internal collagen gel and formed a dense cell-matrix mass. (C) Calcein-AM fluorescence of contracted cell mass. Inset shows phase contrast image. (D–F) SMC encapsulated in HA capsules with 1 mg/ml collagen-I gel. (D) 60 min after encapsulation, cells are well dispersed in the internal collagen matrix. (E) After 24 hours of culture, the cells contracted the internal collagen gel, simultaneously collapsing the entire capsule structure to form a denser module with a convoluted surface membrane. (F) Calcein-AM fluorescence of contracted cell mass. Inset shows phase contrast image. (G–I) Histology of contracted capsules. H&E (G,H) staining showing compacted capsule structure with minimal void volume. (I) Masson's Trichrome staining of contracted capsule, showing the distribution of collagen (blue) within the structure.

Figure 9. Effects of HA and collagen concentrations on cell-mediated capsule contraction.

Maximal cell-mediated contraction was seen in the formulation with a final concentration of 0.33 wt% HA/1.33 mg/ml collagen-I. Capsules without cells exhibited an insignificant reduction in capsule diameter. Error bars represent standard deviation from at least 10 capsule measurements. Asterix denote statistically significant differences (p<0.05).

Figure 10. Cell growth in encapsulated cocultures of SMC and AEC.

Cocultures of encapsulated SMCs with AECs on the external surfaces of HA/Collagen capsules exhibited increased SMC proliferation compared to encapsulated SMC monocultures. (A) Encapsulated SMCs only, day 7. (B) Encapsulated SMC with AECs, day 7.

Figure 11. Modular assembly of GAG based microcapsules by fusion.

Modular constructs were fabricated by perfusing packed capsules in a chamber of desired dimensions with diluted polymer solutions. This method yielded self-supporting constructs with uniform porosity. (A) Individual capsules in buffer solution before fusion. (B) Capsules being perfused with polymer solution in a perfusion chamber. Arrow indicates direction of flow. (C) Fused construct after removal from perfusion chamber.

Figure 12. Endothelialized, interconnected channels in a fused modular construct.

(A) Phase contrast image of CSA/CMC capsules, seeded externally with HUVECs and fused 48 hours after seeding. (B) Combined confocal image stack of the modular construct shown in A with HUVECS visualized via CellTracker Green staining. (C) SEM image of an axially sectioned, modular construct assembled from fused empty capsules showing interconnected channels.

Figure 13. H&E staining of modular constructs based on hepatocytes in HA/collagen capsules.

(A, B). Fused construct with reduced fluid volume and porosity due to centrifugation of capsules during the fusion process. (C, D) Construct formed by fusion of capsules settled under unit gravity, resulting in significantly greater fuid volume inside capsules and larger intercapsular spaces suitable for perfusion culture.

Figure 14. Albumin and urea synthesis rates of hepatocytes in encapsulated perfusion cultures.

Primary rat hepatocytes were encapsulated in CSA/CMC capsules with a with 1 mg/ml collagen gel, at a density of 2×10⁷ cells/ml of CSA/CMC/collagen solution. (A) Control collagen sandwich dish culture. (B) Encapsulated hepatocytes aggregated into spheroids during culture as either (C) individual capsules in a fluidized bed bioreactor, or as a (D) fused modular construct in a packed bed bioreactor. (E–H) Albumin and urea synthesis rates by the hepatocytes in the three culture conditions. (E,F) Control collagen sandwich cultures. (G, H) Perfusion cultures. Error bars denote standard deviations from 3 replicate measurements.