Development of stromal differentiation patterns in heterotypical models of artificial corneas generated by tissue engineering

Abstract

Purpose: We carried out a histological characterization analysis of the stromal layer of human heterotypic cornea substitutes generated with extra-corneal cells to determine their putative usefulness in tissue engineering. Methods: Human bioartificial corneas were generated using nanostructured fibrin-agarose biomaterials with corneal stromal cells immersed within. To generate heterotypical corneas, umbilical cord Wharton's jelly stem cells (HWJSC) were cultured on the surface of the stromal substitutes to obtain an epithelial-like layer. These bioartificial corneas were compared with control native human corneas and with orthotypical corneas generated with human corneal epithelial cells on top of the stromal substitute. Both the corneal stroma and the basement membrane were analyzed using histological, histochemical and immunohistochemical methods in samples kept in culture and grafted in vivo for 12 months in the rabbit cornea. Results: Our results showed that the stroma of the bioartificial corneas kept ex vivo showed very low levels of fibrillar and non-fibrillar components of the tissue extracellular matrix. However, in vivo implantation resulted in a significant increase of the contents of collagen, proteoglycans, decorin, keratocan and lumican in the corneal stroma, showing higher levels of maturation and spatial organization of these components. Heterotypical corneas grafted in vivo for 12 months showed significantly higher contents of collagen fibers, proteoglycans and keratocan. When the basement membrane was analyzed, we found that all corneas grafted in vivo showed intense PAS signal and higher contents of nidogen-1, although the levels found in human native corneas was not reached, and a rudimentary basement membrane was observed using transmission electron microscopy. At the epithelial level, HWJSC used to generate an epithelial-like layer in ex vivo corneas were mostly negative for p63, whereas orthotypical corneas and heterotypical corneas grafted in vivo were positive. Conclusion: These results support the possibility of generating bioengineered artificial corneas using non-corneal HWJSC. Although heterotypical corneas were not completely biomimetic to the native human corneas, especially ex vivo, in vivo grafted corneas demonstrated to be highly biocompatible, and the animal cornea became properly differentiated at the stroma and basement membrane compartments. These findings open the door to the future clinical use of these bioartificial corneas.

Keywords: Wharton’s jelly mesenchymal stem cells; basement membrane; cornea; stroma; tissue engineering.

FIGURE 1

Ophthalmologic analysis of rabbit eyes grafted with the orthotypical (OAC) and heterotypical (HAC) artificial corneas and control eyes (CTR). Macroscopic photographs, slit lamp observations and optical coherence tomography images (OCT) are shown for each type of sample after 3 and 12 months of in vivo follow-up.

FIGURE 2

Histological evaluation of orthotypical (OAC) and heterotypical (HAC) artificial corneas and controls. The top panel corresponds to the analysis using Masson’s trichrome staining (MAS), whereas the lower panel shows the immunohistochemical analysis of p63. In both cases, the area in which the in vivo histological images were obtained is labeled with a blue square in the illustration representing the rabbit cornea shown at the upper right corner. Bioartificial corneas kept ex vivo for 7, 14 and 21 days, and corneas grafted in vivo for 3 and 12 months are shown. Controls (CTR) correspond to native, normal human corneas. Scale bars: 50 µm (applicable to all images). The histogram corresponding to the p63 panel shows the results of the quantitative analysis of the percentage of positive cells in each group. Asterisks represent statistically significant differences as compared to CTR corneas.

FIGURE 3

Identification of fibrillar components of the corneal stroma of orthotypical (OAC) and heterotypical (HAC) artificial corneas and controls. PSR: picrosirius red histochemistry for collagen fibers detection; VER: Verhoeff histochemistry for elastic fibers; RET: Gomori’s reticulin histochemistry for reticular fibers. The area in which the in vivo histological images were obtained is labeled with a blue square in the illustration representing the rabbit cornea shown at the upper right corner. Bioartificial corneas kept ex vivo for 7, 14 and 21 days, and corneas grafted in vivo for 3 and 12 months are shown. Controls (CTR) correspond to native, normal human corneas. Scale bars: 50 µm (applicable to all images). Histograms to the right correspond to the quantitative analysis of signal intensity quantified in intensity units (i.u.) of the ImageJ software. Asterisks represent statistically significant differences as compared to CTR corneas.

FIGURE 4

Identification of non-fibrillar components of the corneal stroma of orthotypical (OAC) and heterotypical (HAC) artificial corneas and controls. AB: alcian histochemistry for global proteoglycan detection; DCN, KER and LUM: immunohistochemical detection of the proteoglycans decorin, keratocan and lumican, respectively. The area in which the in vivo histological images were obtained is labeled with a blue square in the illustration representing the rabbit cornea shown at the upper right corner. Bioartificial corneas kept ex vivo for 7, 14 and 21 days, and corneas grafted in vivo for 3 and 12 months are shown. Controls (CTR) correspond to native, normal human corneas. Scale bars: 50 µm (applicable to all images). Histograms to the right correspond to the quantitative analysis of signal intensity quantified in intensity units (i.u.) of the ImageJ software. Asterisks represent statistically significant differences as compared to CTR corneas.

FIGURE 5

Analysis of the basement membrane in orthotypical (OAC) and heterotypical (HAC) artificial corneas and controls. PAS: periodic acid of Schiff histochemistry for glycosaminoglycans of the basement membrane; NID: nidogen-1 immunohistochemistry. In both cases, the area in which the in vivo histological images were obtained is labeled with a blue square in the illustration representing the rabbit cornea shown at the upper right corner. TEM: transmission electron microscopy with white arrowheads pointing to the basement membrane and yellow arrows highlighting the presence of the lamina lucida (LL), lamina densa (LD) and hemidesmosomes (HD). Bioartificial corneas kept ex vivo for 7, 14 and 21 days, and corneas grafted in vivo for 3 and 12 months are shown for the histochemistry and immunohistochemistry analyses. Controls (CTR) correspond to native, normal human corneas. Scale bars: 50 µm for the histochemistry and immunohistochemistry and 1 µm for the TEM analysis (applicable to all images). Histograms to the right correspond to the quantitative analysis of PAS and NID signal intensity quantified in intensity units (i.u.) of the ImageJ software.