Limbal stem cells as a potential source of pancreatic beta cells

BACKGROUND/RATIONAL Beta cell replacement therapy represents an alternative approach to conventional insulin regimen for the management of type 1 diabetes. Stem cells theoretically represent a suitable and renewable source of surrogate beta cells. Recently, it has been demonstrated that corneal epithelium contains stem cells in the basal layer of the limbus. Our aim was to isolate and characterize limbal stem cells (LSCs) from adult corneal-limbal biopsies, and to assess their capability to differentiate into insulin-producing cells. METHODS 2-3 mm2 biopsies were obtained from limbus and central cornea of four patients undergoing surgery for ocular diseases not involving conjuctive or corneal surface. Tissues were incubated with DMEM containing 250 U/ml dispase II at 37°C for 1hr and then treated with 0.25% trypsin/1 mM EDTA at 37°C for 10-20 min to isolate single cells. Cells were cultured in F12/DMEM medium, 10% stem cell qualified FBS, 4 ng/ml bFGF, ITS 1X, glutamine and antibiotics. Analysis of stem cell markers SSEA4 and Oct-4 was assessed by flow cytometry. Cells were sorted for SSEA4 by magnetic separation and efficiency of enrichment was established. The stem cell markers Sox2, KLF4, c-Kit, Nanog, THY-1 and the putative LSC marker ABCG2 were analysed in the SSEA4+ sorted population by qRT-PCR. Sorted LSCs were cultured in beta cell differentiation medium (DMEM/F12 supplemented with 10% stem cell qualified FBS, 1% N2, 2% B27, 1 mM nicotinamide, 10 ng/ml bFGF, glutamine and antibiotics) for 21 days. In the last week 50 ng/ml exendin-4 were added. Cells were then analysed for stem and beta cell markers (Pdx-1, Foxa2, Beta2NeuroD, Ngn3, Isl-1, Glut2,insulin) at different time points by qRT-PCR. RESULTS After digestion, the expression of stem cell markers SSEA4 and Oct-4 was 52.1 ± 10.7 and 47.9 ± 8.3 %, respectively. SSEA4 enrichment with magnetic separation showed a purity of 91.5 ± 1.9 %. Stem cell marker analysis by qRT-PCR in SSEA4+ cells confirmed the expression of Oct-4 and showed positivity for Sox2, KLF4, c-Kit, THY-1 and ABCG2. On the contrary Nanog was always negative. During the differentiation protocol, expression of stem cell markers decreased progressively at each time points (day 0, 7, 14, 21), in particular in one sample. Interestingly, in the same sample the appearance of Foxa2 mRNA was observed at day 14, and its expression increased at day 21, when also Pdx-1 became detectable. The other beta cell markers analyzed, including insulin, were negative. CONCLUSION The existence of LSCs has been proposed. Our study confirms that stem cells can be easily isolated from limbal biopsies and that can be driven into Pdx-1+/Foxa2+ beta cell precursors. The absence of other mature beta cell markers, including insulin mRNA, suggest that producing functional beta cells from LSCs is not straightforward and improvement in the differentiation protocols is necessary. However, the accessibility of the tissue and the presence of a relatively large proportion of uncommitted cells compared to other adult tissues, make the limbus a competitive and renewable source of autologous stem cells that can be used in diabetic patients.