TGF-β induces corneal endothelial senescence via increase of mitochondrial reactive oxygen species in chronic corneal allograft failure

Abstract

The corneal endothelium (CE) dysfunction impairs optical transparency and leads to corneal allograft failure. Morphologically, CE cells are characterized by premature senescence at the late stage of corneal graft. However, the detailed molecular mechanisms are largely unknown. Here we found that transforming growth factor-β (TGF-β) is elevated in the CE of late graft failure. In addition, senescence-associated gene p21 and p16 are increased as well, which is consistent with their elevation upon TGF-β treatment in human corneal endothelial cell B4G12. Furthermore, TGF-β treatment leads to high positive ratio of SA-β-gal, indicating B4G12 cells undergo cellular senescence. Mechanistically, we demonstrated that TGF-β could induce mitochondrial ROS (mtROS) production and mtROS scavenger could rescue CE cell senescence upon TGF-β treatment. Our study provides new evidence that elevated TGF-β plays a crucial role in the CE cell senescence and loss in chronic corneal graft failure, which could be potential targets for clinical treatment.

Keywords: cellular senescence; chronic corneal graft failure; corneal endothelium; mitochondrial reactive oxygen species; transforming growth factor-β.

Figure 1

Increased senescence associated markers and TGF-β1 in the human corneal buttons CE with chronic graft rejection. (A) Clinical evaluation of corneal endothelial cells with chronic failure using alizarin red staining. (B-C) The mRNA expression of p16, p21 (B) and TGF-β1 (C) in the CE with or without chronic graft failure (**P<0.01,*P<0.05). (D-E) Representative photographs (D) and histopathology scores (E) for the IHC staining of p16 and TGF-β1 in the CE with chronic failure.

Figure 2

TGF-β1 induces cellular senescence in B4G12 cells. (A-B) The G1 phase arrest was induced by TGF-β1 treatment. Control and TGF-β1–treated B4G12 cells were subjected to cell cycle analysis after 48h-treatment. A representative flow cytometric analysis of the DNA content was shown in (A) and the values are mean±SD in (B). (C-D) SA-β-Gal activity was measured in B4G12 cells treated with 10 ng/ml TGF-β1 alone, or in combination with NAC (10mM) for 72h. NAC treatment alone (10mM) did not affect the cell cycle status and SA-β-Gal activity. Bar graphs represent mean±SD. *P<0.05, **P<0.01. All the experiments were independently repeated at least three times.

Figure 3

TGF-β1 treatment of B4G12 cells causes the induction of p16 and p21. B4G12 cells were treated with 10 ng/ml TGF-β1 alone, or in combination with NAC (10mM) or NAA (10mM) for 72h. The mRNA (A) and protein (B-C) expression of p16 and p21 were induced by TGF-β1. NAC or NAA treatment alone (10mM) did not change the levels of p16 and p21. Bar graphs represent mean±SD. *P<0.05, **P<0.01.

Figure 4

mtROS production in B4G12 cells after exposure to 10ng/ml TGF- β 1 for 48 hours. (A-B) Control and TGF-β1–treated B4G12 cells were subjected to MitoSOX Red Indicator staining after 48h of culture. A representative flow cytometric analysis of mtROS was shown in (A) and the values are mean±SD (B). (C) The MitoSOX Red and peroxy orange 1 fluorescence was imaged with a fluorescent microscope. (D) SOD2 protein expression was tested in CE cells upon 10 ng/ml TGF-β1 treatment. *P<0.05, **P<0.01. All the experiments were independently repeated at least three times.

Figure 5

Premature senescence and elevated TGF-β1 in the murine CE with chronic failure. (A-C) Corneal graft phenotype in syngenic and chronic failure mouse model (D-E) p21, p16 (D) and TGF-β1 (E) expression was measured at transcription level in the murine CE from syngenic and chronic failure group. (F) Representative photographs for the IF staining of TGF-β1 in the CE with chronic graft failure. **P<0.01,*P<0.05.

Figure 6

Leukocytes in proximity to murine CE of chronic graft failure. (A) H&E staining showing immune cell infiltration in the murine CE. (B)The expression of senescence-associated secretory phenotypefactor in RNA pooled from murine CE from chronic failure group compared with syngenic group. *P<0.05, **P<0.01. All the experiments were independently repeated at least three times.