Senescence Mediated by p16^INK4a Impedes Reprogramming of Human Corneal Endothelial Cells into Neural Crest Progenitors

Abstract

Purpose : To further explore the mechanism of reprogramming human corneal endothelial (HCEC) monolayers into neural crest (NC)-like progenitors by knockdown with p120-catenin (p120)-Kaiso siRNAs.

Methods : HCEC monolayers derived from collagenase digestion of stripped Descemet membrane were cultured to 7 days and treated with 100 nM of each p120 and Kaiso siRNAs with or without various inhibitors or siRNAs in MESCM for up to 5 weeks. Before termination, cells were further treated with 10 μM BrdU for 4 h. Transcript expression of NC, cell cycle and senescence markers, JAKs and STAT3 was determined by RT-qPCR. Immunostaining was performed to monitor cytolocalization of BrdU, NC, cell cycle and signaling markers. Western blotting was used to measure relative protein level.

Results : We have reported contact-inhibition of HCECs can be delayed from Day 21 to Day 42 by switching the culture medium from SHEM to MESCM, a serum-free and LIF/bFGF-containing medium, by transient activation of LIF-JAK1-STAT3 signaling with concomitant inhibition of eventual nuclear translocation of p16^INK4a. Using the latter system, we have created a novel tissue engineering technique by implementing 5 weekly knockdowns with p120-Kaiso siRNAs to achieve effective expansion of HCEC monolayers to a transplantable size with a normal density by reprogramming HCECs into neural crest progenitors through activating p120-Kaiso-RhoA-ROCK-canonical BMP signaling. Interestingly, a single knockdown with p120-Kaiso siRNAs at Day 42 failed to achieve such reprogramming when contact inhibition transitioned to senescence with nuclear translocation of p16^INK4a. In contrast, 5 weekly knockdowns with p120 and Kaiso siRNAs precluded senescence mediated by p16^INK4a by induction of nuclear translocation of Bmi1 that downregulated p16^INK4a and prohibited nuclear translocation of p16^INK4a. Such nuclear translocation of Bmi1 was resulted from sustained activation of JAK2-STAT3 signaling downstream of p120-Kaiso-RhoA-ROCK signaling. In addition, STAT3 or Bmi1 siRNA impeded nuclear exclusion of p16^INK4a and suppressed BrdU labeling and nuclear translocation of Oct4, Sox2 and Nanog induced by p120-Kaiso siRNAs.

Conclusions : We have discovered another important mechanism of the reported engineering strategy of HCECs that lies in prevention of senescence mediated by nuclear translocation of p16^INK4a.

This is an abstract that was submitted for the 2016 ARVO Annual Meeting, held in Seattle, Wash., May 1-5, 2016.