Application of retinoic acid improves form and function of tissue engineered corneal construct

Abstract

Retinoic acid has recently been shown to control the phenotype and extracellular matrix composition of corneal stromal cells cultured in vitro as monolayers. This study set out to investigate the effects of retinoic acid on human corneal keratocytes within a 3D environment. Human corneal keratocytes were encapsulated in collagen gels, which were subsequently compressed under load, and cultured in serum-free media supplemented with 10 µM retinoic acid or DMSO vehicle for 30 days. Cell proliferation was quantified on selected days, while the expression of several important keratocytes markers was evaluated at day 30 using RT-PCR and immunoblotting. The weight and size of the collagen constructs were measured before and after hydration and contraction analyses. Retinoic acid enhanced keratocyte proliferation until day 30, whereas cells in control culture conditions showed reduced numbers after day 21. Both gene and protein expressions of keratocyte-characteristic proteoglycans (keratocan, lumican and decorin), corneal crystallins and collagen type I and V were significantly increased following retinoic acid supplementation. Retinoic acid also significantly reduced the expression of matrix metalloproteases 1, 3 and 9 while not increasing α-smooth muscle actin and fibronectin expression. Furthermore, these effects were also correlated with the ability of retinoic acid to significantly inhibit the contractility of keratocytes while allowing the build-up of corneal stromal extracellular matrix within the 3D constructs. Thus, retinoic acid supplementation represents a promising strategy to improve the phenotype of 3D-cultured keratocytes, and their usefulness as a model of corneal stroma for corneal biology and regenerative medicine applications.

Keywords: 3D model; Cornea; collagen gel; in vitro; keratocytes; retinoic acid; tissue engineering.

FIGURE 1.

Effects of RA supplementation on the proliferation of keratocytes embedded in compressed collagen gels. Quantification was performed using AlamarBlue® assay, and cell numbers were normalized as a percentage of cells initially seeded. Data (mean ± SD) were obtained from 3 independent experiments (n = 3) and compared using t-test (*corresponds to P < 0.05).

FIGURE 2.

Contraction assay. (A) Contraction assay of 3D collagen constructs embedded with corneal keratocytes and cultured in serum-free medium (with RA supplementation or DMSO control) or standard medium (with serum; FBS). The outer dotted line represents the size of the gel at the beginning of the experiment. Scale bar = 10 mm. (B) The gel size (in percentage) was determined by calculating the difference in constructs' surface area between the initial day of experiment (white bar) and day 30 (colored bar). Data (mean ± SD) were obtained from 3 independent experiments (n = 3) and compared using t-test (** and *** correspond to P < 0.01 and <0.001, respectively).

FIGURE 3.

Effects of RA supplementation on the hydration of cell-encapsulating compressed collagen gels. The weight of all gels was measured before and after the 30 d culture period to find the differences in the wet weight. The dry weight was quantified following freeze-drying of the gels. The percentage of hydration denotes the loss of bound water from the collagen gels with or without supplementation of RA. Data (mean ± SD) were obtained from 3 independent experiments (n = 3) and compared using t-test (*** correspond to P <0.001).

FIGURE 4.

Expression of keratocyte markers at the transcriptional level. Total mRNA from cells embedded in compressed collagen gels cultured for 30 d in control (gray) and +RA (orange) was extracted and analyzed by RT-PCR. Gene expression was normalized relative to that of control group. Data (mean ± SD) were obtained from 3 independent experiments (n = 3) and compared using t-test (*, ** and *** correspond to P < 0.05, <0.01 and <0.001, respectively).

FIGURE 5.

Expression of keratocyte markers at the protein level. (A) Lysates from corneal stroma (CS), corneal fibroblasts cultured in FBS (FBS), or 3D stromal constructs supplemented with RA (RA) or DMSO (Ctl) were extracted and analyzed by reducing SDS-PAGE followed by immunoblotting. (B-E) Quantification of protein expression from control (gray) and RA (orange) was performed by immunoblot densitometry for all markers. Protein expression was normalized relatively to that of control group. Data (mean ± SD) were obtained from 3 independent experiments (n = 3) and compared using t-test (* correspond to P < 0.05 respectively).

FIGURE 6.

Immunofluorescence staining of ECM proteins characteristic of the corneal stroma, expressed by encapsulated keratocytes within compressed collagen gels with or without RA supplementation. All photographs were taken at x200 magnification. Scale bar = 50 µm.