RSPO3 Promotes Proliferation and Self-Renewal of Limbal Epithelial Stem Cells Through a WNT/β-Catenin-Independent Signaling Pathway

Abstract

Purpose: R-spondin3 (RSPO3), a mammalian-specific amplifier of WNT signaling pathway, maintains the homeostasis of various adult stem cells. However, its expression at the limbus and the effect on limbal epithelial stem cells (LESCs) remains unclear. We investigated the impact of RSPO3 on the proliferation and self-renewal of LESCs and explored its molecular mechanisms.

Methods: The expression of four RSPO subtypes at the limbus were detected. Co-cultured with RSPO3 in vitro, the cell outgrowth area and cell density of human LESCs (hLESCs) were measured, along with EdU assay and evaluation of biomarkers of cell proliferation (Ki67) and stemness (△Np63 and ABCG2). The expression of key molecules in WNT/β-catenin signaling pathway were investigated in RSPO3-co-incubated hLESCs and controls. The effect of RSPO3 on corneal epithelium wound recovery in vivo was investigated in a mouse model of corneal epithelium injury.

Results: Among four subtypes of RSPO protein, only the RSPO3 isoform was stably expressed at the human limbus. RSPO3 promoted the proliferation and stemness maintenance of hLESCs in vitro in a dose-dependent manner when its concentration ≤ 100 ng/mL, and this effect was not impaired when the activation of β-catenin was inhibited by XAV939, indicating that the effect of RSPO3 on hLESCs was not dependent on canonical WNT/β-catenin signaling pathway. Exogenous RSPO3 accelerated epithelial wound healing by enhancing the proliferation and self-renewal of residual LESCs.

Conclusions: RSPO3 promotes the proliferation and self-renewal of LESCs through a WNT/β-catenin-independent signaling pathway which might have translational significance in the treatment of corneal epithelium injury and limbal stem cell deficiency.

Figure 1.

Endogenous expression of RSPO subtypes at the human limbus. (A–C) Although the expression of RSPO3 and RSPO4 were found by qRT-PCR (A), only RSPO3 was stably expressed at protein level, as confirmed by WB (B) and IF staining (C). Scale bar = 100 µm (C).

Figure 2.

The effect of RSPO3 on the proliferation and stemness maintenance of hLESCs in vitro. Images and quantification of cell outgrowth area (A, B) on days 6, 8, and 14 and cell density (C) on days 6, 8, 10, 12, and 14 showed that compared to the control group, RSPO3 promoted cell outgrowth and increased cell density in a dose-dependent manner, which reached the peak at 100 ng/mL. The effect of RSPO3 on hLESCs in vitro was confirmed by the quantification of EdU⁺, Ki67⁺, △Np63⁺, and ABCG2⁺ cells, all of which reached the highest level at 100 ng/mL, and the expression of K14 was not affected by RSPO3 (D, E). Data were collected from three independent experiments. In each independent experiment, the hLESCs were obtained from the same donor and were used to minimize donor variation. The boundaries of cell outgrowth were outlined with red or blue color. Scale bars = 200 µm (A) and 50 µm (D). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 3.

The outcome of β-catenin inhibition on the effect of RSPO3. Images and quantification of the cell outgrowth area (A, B) on days 6, 8, and 14 and cell density (C) on days 6, 8, 10, 12, and 14 showed that RSPO3 promoted the proliferation of hLESCs no matter if XAV939 was present or not. IF staining (D) and quantification (E) of EdU⁺, Ki67⁺, △Np63⁺, and ABCG2⁺ cells confirmed that the effect of RSPO3 was not impaired even when the activation of β-catenin was inhibited by XAV939. Data were collected from three independent experiments. In each independent experiment, the hLESCs were obtained from the same donor and were used to minimize donor variation. The boundaries of hLESCs outgrowth were indicated with red or blue lines. Scale bars = 200 µm (A) and 50 µm (D). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 4.

The expression of key molecules in WNT/β-catenin signaling pathway with the treatment of RSPO3 and/or XAV939. The treatment of XAV939 caused reduced expression of total β-catenin (A, B), cytoplasmic β-catenin (C, D), nuclear β-catenin (E, F), and p-GSK3β (G, L), as well as an elevated level of Axin1 (G, J) and GSK3β (G, K), all of which could not be compensated with RSPO3. Meanwhile, RSPO3 promoted the expression of LGR4 (G, H) and FZD7 (G, I) no matter if XAV939 was present or not. Data were from three independent experiments. In each independent experiment, the hLESCs were obtained from the same donor and were used to minimize donor variation. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001.

Figure 5.

RSPO3 promotes mice corneal epithelium wound healing. Only a trace amount of RSPO3 was detected at the limbus of WT C57BL/6 mice by qRT-PCR (A) and WB (B, C). Among 3 siRNA to knockdown the expression of endogenous RSPO3, siRSPO3-3 was the most effective at both mRNA (D) and protein (E) levels. After the establishment of a corneal epithelial injury model, RSPO3 significantly promoted epithelium recovery, whereas siRSPO3 caused delayed epithelial healing (F–G). The percentages of cells expressing the biomarker of cell proliferation (P63, n = 3) and stemness (ABCG2, n = 3) were significantly higher in the RSPO3 group compared with the control and the siRSPO3 groups. Moreover, they were much lower in the siRSPO3 group than in the siControl group (H–J). H&E staining (K) showed that the corneal epithelium in RSPO3 group was consisted of three to four layers of regularly arranged epithelial cells (n = 3). However, that in the siRSPO3 group was characterized with loosely arranged and disordered cells as well as vacuole-like structures. Scale bars = 50 µm (H, K). *, control versus RSPO3; #, RSPO3 versus siRSPO3; $, siControl versus siRSPO3. *, #, $P < 0.05; **, ##, $$P < 0.01; ***, ###, $$$P < 0.001; ****, ####, $$$$P < 0.0001.