Several clinical researches believed that vitreomacular interface abnormality (VMIA) might be related to AMD’s occurrence and progress by mechanical stretch from vitreous to RPE layers through the whole retina4–7. Our laboratory found mechanic stretch induced by Flexcell system could induce oxidative stress by mitochondrial and NADPH oxidase, induce changes in morphological, actin-cytoskeleton and raise the expression level of cytokines in RPE cells which refers to apoptosis8–11. At the same time, Autophagy in RPE cells had proven as a vital mechanism for age-related macular degeneration for degrading reactive oxygen species by oxidative stress 18and photoreceptor outer segment phagocytosis19,20. Its abnormality might induce redundant inpaired cargo and lipofuscin accumulation, and reduced disturbed waste clearance21. And it could be induced by mechanical stretch in recent researches. Xu et al. found autophagy could keep bone homeostasis by resistance to cyclic stretch22. Besides, Lin et al. believed autophagy involved in cardiac hypertrophy through AT1 receptor-mediated activation of p38MAPK by cyclic stretch models23. In a word, autophagy possesses the peculiarity of mechanical sensitivity, and cyclic stretch might alter autophagic capacity in RPE cells.
Indeed, in our study, autophagic flex, which is represented by the expression level of LC3II, was raised at 6h and keep rising until to 48h in a high level. Meanwhile, Beclin-1, which was participated in autophagic induction and phagophore, and LAMP-1, which was represented for process of autophagy fusion, were both up-regulated and MTOR, which was considered as a main regulatory target for autophagy, declined, while SQSTM1/p62 showed no change during the cyclic stretch process. We further showed that the induction of autophagy in RPE cells by cyclic stretch could be inhibited by 3-Methyladenine (3-MA), and a rising protein expression level of VEGF was appeared with the inhibitor after stretch. Furthermore, in condition of autophagic stimulator (Rapamycin)cultivation combined with cyclic stretch, VEGF and NOX4 were down-regulated opposed to control group. To our knowledge this is the first study on analysis for induction of autophagy in RPE cells by cyclic stretch. Flexcell Tension System was applied for mimic mechanical stretch to RPE layers, 20% elongation and 1HZ were applied broadly in recent studies. Qi et al. found cyclic stretch could induce autophagy elevation in vascular smooth muscle cells by mTOR (mammalian target of Rapamycin) pathway24, King et al reported breast cancer cells suffered from mechanical compression by mTOR pathway25,26 also, while Liton et al. found no change in trabecular meshwork cells27. Based on previous reports, autophagic induction is not only due to mechanical signal stimulation direct to cytoskeleton, but also due to damaged components by excessive mechanical stress indirectly. In our previous studies, actin cytoskeleton, IL-8, IL-6, VEGF, mitochodrial and NADPH oxidase were all induced by cyclic stretch8–11. Therefore, we should have ample reason to assumed that cyclic stretch could induce autophagy in RPE cells.
In our study, indeed, mTOR was suppressed by cyclic stretch at 12h until 48h. mTOR, which is a vital regulator for autophagy, could phosphorylate autophagic proteins for autophagy imitation and autophagosome nucleation and nuclear translocation prevention28. Meanwhile, as proven by Vion et al29, autophagy was induced by mechanical stretch in mTOR and AMPK pathway, which identically elevated in RPE cells by mechanical stretch in our study. It’s worth noting that SQSTM1/p62 didn’t decrease during the period, which be similar to Liton’s15. They also found no change for SQSTM1/p62 after mechanical stretch, for the reason that SQSTM1/p62 decrease seemed to less relevant to LC3-II accumulation, though SQSTM1/p62 was also a critical indicator for autophagy machinery. Additionally, LC3-associated phagocytosis (LAP) might be also considered due to: 1. The autophagic proteins have a conspicuous non-canonical role for endocytic/phagocytic pathway which called LC3- associated phagocytosis (LAP), LAP play a vital role in the homeostasis and phagocytosis for RPE cells30; 2. Evidence in recent studies showed no relationship between SQSTM1 and LC3 expression when LAP happens31. However, cyclic stretch induces whether macroautophagy or LAP or other forms of autophagy such as chaperone-mediated autophagy (CMA) needs further study. Therefore, mechanical stretch could induce autophagy in RPE cells in mTOR and AMPK pathway. The high expression of autophagic flux represented stress protective function of RPE cells from traction.
Autophagy is considered to be an important mechanism for maintaining cellular homeostasis, and it is now generally believed that under conditions of exposure to adverse external cellular stimuli, cellular autophagy levels are first upregulated in response to external stress conditions, enhancing the enzymatic resolution of misfolded proteins within the cell21. When external stress is extended, the cellular autophagy level decreases instead due to the excessive accumulation of autophagosome, thus leaving the cell lacking the ability to remove cellular wastes32. Especially in the case of retinal pigment epithelial cells, for example, its autophagy level also assumes the function of phagocytosis of photoreceptor outer segments33, and when a decrease in autophagy level occurs, this decrease in phagocytic digestion is manifested by lipid deposits and thus derived to the early stages of age-related macular degeneration20,30. However, the impact of altered autophagy in cells under stress interventions is still controversial. Some studies have suggested that autophagic upregulation of cells under stress conditions are necessary for disease development34,35, while others have suggested that the upregulation is a self-protective mechanism of cells15,36. To investigate the effects of increased autophagy in RPE cells by mechanical stretch, on the one hand, the autophagy inhibitor 3-MA was used to inhibit autophagy, and a significant increase in VEGF expression was found in RPE cells; on the other hand, the autophagy activator Rapamycin was used, and the cells were given 24 hours of traction after pre-raising the autophagy level of RPE cells, and it was found that the NOX4 and VEGF expression was significantly decreased in RPE cells compared to the control group In our previous study, NOX411 and VEGF8, which were two vital cytokines for AMD progression, were both up-regulated by cyclic stretch. Although RPE cells which pre-treated by Rapamycin still showed less upgraded expression of NOX4 and VEGF after 24h stretch, lower expression of them compared to control ones was also demonstrated by Western Blot, which means moderately elevated autophagy for RPE cells might reduce the negative influence by cyclic stretch. Kauppinen et al. found Resvega, which was seen as an activator for autophagy, could reduce VEGF release by IL-1β diminish level37. Mitter et al. believed that autophagy for RPE cells was crucial for resistance to oxidative stress, autophagic imperfection might aggravate damage in AMD38; Chen et al. found enhancing autophagic activity can alleviate oxidative stress in AMD and protect RPE and photoreceptor cells from progressive degenerations39. It is believed that NFE2L2, PGC-1, AMPK and mTOR pathways is vital to improve our comprehending of the regulatory mechanisms in autophagy that alleviate oxidative stress and assuage the progress of AMD18. Therefore, Rapamycin might act on mTOR or AMPK pathways for improving resistance to cyclic stretch, so as to VEGF and NOX4 lower expression compared with control group. However, the molecular mechanism for how autophagy affects VEGF secretion and NOX4 expression needs additional study.
This study is still having several limitations. First, ARPE-19 cells are only a cellular model for research, it has less representation for RPE cells in healthy people. Second, due to the limitation of light transmission of Flexcell plates, transmission electron microscope for autophagosome was lack, and gene upregulation/down regulation by qPCR for VEGF and ROS was lack either. Third, the regulatory mechanism between mechanical stretch, autophagy and AMD-related cytokines required further study.