Multi-layered silk film coculture system for human corneal epithelial and stromal stem cells

Abstract

With insufficient options to meet the clinical demand for cornea transplants, one emerging area of emphasis is on cornea tissue engineering. In the present study, the goal was to combine the corneal stroma and epithelium into one coculture system, to monitor both human corneal stromal stem cell (hCSSC) and human corneal epithelial cell (hCE) growth and differentiation into keratocytes and differentiated epithelium in these three-dimensional tissue systems in vitro. Coculture conditions were first optimized, including the medium, air-liquid interface culture, and surface topography and chemistry of biomaterial scaffold films based on silk protein. The silk was used as scaffolding for both stromal and epithelial tissue layers because it is cell compatible, can be surface patterned, and is optically clear. Next, the effects of proliferating and differentiating hCEs and hCSSCs were studied in this in vitro system, including the effects on cell proliferation, matrix formation by immunochemistry, and gene expression by quantitative reverse transcription-polymerase chain reaction. The incorporation of both cell types into the coculture system demonstrated more complete differentiation and growth for both cell types compared to the corneal stromal cells and corneal epithelial cells alone. Silk films for corneal epithelial culture were optimized to combine a 4.0-μm-scale surface pattern with bulk-loaded collagen type IV. Differentiation of each cell type was in evidence based on increased expression of corneal stroma and epithelial proteins and transcript levels after 6 weeks in coculture on the optimized silk scaffolds.

Keywords: coculture; cornea; silk; tissue engineering.

Figure 1. Schematic of the tissue engineered cornea experimental design

(A) Medium was optimized for growth of both hCEs and hCSSCs by testing combinations of hCSSC medium and hCE medium, as well as hCSSC medium with hCE growth factors (GF) added. (B) hCEs and hCSSCs were cultured in Transwell® systems to study the effects of the co-culture on the proliferation and differentiation of each cell type. (C) Silk film topography and chemistry were optimized for hCE growth. Different micron-scale groove sizes and extracellular matrix proteins were tested to enhance cell attachment and growth. (D) Tissue engineered corneas containing hCE films and hCSSC films were cultured long term in optimized medium.

Figure 2. AlamarBlue® results for co-culture medium optimization

Proliferation activity was measured using fluorescence intensity and normalized to day 1 and compared using two-way ANOVA. (A) hCEs showed significantly higher proliferation activity in all media compared to human keratinocyte serum free medium (Ep). Growth factor added media (40, 70, 100%) were compared to stem cell proliferation medium (Stem). hCE proliferation activity in 100% supplemented medium showed no significant differences to stem cell proliferation medium, but lower proliferation activity was seen at Day 9 for the 70% and 40% supplemented media (p<0.05*, p<0.0001****, respectively). When comparing to the 100% supplemented medium, hCE proliferation activity was lower at Day 9 for the 40% supplemented medium (p<0.01++). (B) hCSSCs showed higher proliferation activity at Day 9 with the hCSSCs cultured in the 100% supplemented medium when compared to stem cell proliferation medium (p<0.0001****). When compared to the cells cultured in 100% supplemented medium, hCSSCs cultured in 70% and 40% supplemented media showed lower proliferation activity (p<0.001+++ and p<0.0001, respectively). Normal stem cell proliferation medium and human keratinocyte serum free medium were used as control media. Error bars are shown as ± standard deviation.

Figure 3. alamarBlue® results for the Transwell® co-culture system

Proliferation activity was measured using fluorescence intensity and normalized to day 1 and compared using two-way ANOVA. (A) hCSSCs in co-culture showed decreased proliferation compared to hCSSCs cultured independently (p<0.01**, p<0.0001****), without the switch to differentiation medium plus growth factors. (B) After the medium was switched to differentiation medium plus growth factors (day 5), there was a decrease in proliferative activity of both hCEs and hCSSCs in co-culture.

Figure 4

(A) Immunohistochemical staining of hCEs in culture using connexin 43 (red), involucrin (green), and DAPI (blue). At day 14, hCEs in co-culture expressed more of the marker present in differentiated epithelium than hCEs cultured independently. (B) Immunochemical staining of hCSSCs in culture using BCRP (green), and DAPI (blue). At day 14, hCSSCs in culture expressed less of the stem cell marker.

Figure 5. Cell Adhesion assay results for surface topography and ECM addition, showing statistical differences between time points and samples

Cell adhesion results were analyzed using two-way ANOVA. (A) 3 hours after seeding with hCEs, the films with 4.0 µm grooves retained the most cells compared to the other groove types (p<0.05+). (B) 3 hours after seeding with hCEs, the films prepared with collagen type IV retained the most cells compared to the other topography chemistries (p<0.05*+).

Figure 6. hCE differentiation in 3D culture

(A) hCEs cultured in 3D with hCSSCs for 6 weeks showed high protein expression of involucrin and connexin 43 through immunostaining. (B) RT-qPCR results were analyzed relative to a housekeeping gene and normalized to day 1. One-way ANOVA was used to compare transcript expression to day 1 values. hCEs cultured in 3D with hCSSCs for 6 weeks were evaluated for keratin 3 (KRT3), n=4, involucrin (IVL), n=3, and gap junction protein alpha 4 (GJA4), n=4, and showed a significant increase in mRNA expression of KRT3 over the course of 6 weeks (p<0.0001****).

Figure 7. hCSSC differentiation in 3D culture

(A) hCSSCs cultured in 3D with hCEs for 6 weeks showed high protein expression of keratin sulfate, keratocan, decorin, and lumican. (B) RT-qPCR results were analyzed relative to a housekeeping gene and normalized to day 1. One-way ANOVA was used to compare transcript expression to day 1 values. hCSSCs cultured in 3D with hCEs for 6 weeks were evaluated for stroma markers lumican (Lum), pyruvate dehydrogenase lipoamide kinase isozyme 4 (PDK4), lipocalin prostaglandin D synthase (PTDGS), aldehyde dehydrogenase 3 family member A1 (ALDH3A1), and alpha smooth muscle actin (aSMA), and showed a significant increase in mRNA expression of ALDH3A1 over the course of 6 weeks (p<0.001***).