Advanced glycation end products (AGEs) are post-translational modifications formed from the reaction of reactive carbonyl compounds with amino groups in proteins. Our laboratory has previously shown that AGEs in extracellular matrix (ECM) proteins promote TGFβ2 (transforming growth factor-beta 2)-mediated epithelial-to-mesenchymal transition (EMT) of lens epithelial cells (LECs), which could play a role in fibrosis associated with posterior capsule opacification. We have also shown that αB-crystallin plays an important role in TGFβ2-mediated EMT of LECs. Here, we investigated the signaling mechanisms by which ECM-AGEs enhance TGFβ2-mediated EMT in LECs. We found that in LECs cultured on AGE-modified basement protein extract (AGE-BME), TGFβ2 treatment up-regulated the mesenchymal markers α-SMA (α-smooth muscle actin) and αB-crystallin and down-regulated the epithelial marker E-cadherin more than LECs cultured on unmodified BME and treated with TGFβ2. Using a Multiplex Assay, we found that AGE-BME significantly up-regulated the noncanonical pathway by promoting phosphorylation of ERK (extracellular signal-regulated kinases), AKT, and p38 MAPK (mitogen-activated protein kinases) during TGFβ2-mediated EMT. This EMT response was strongly suppressed by inhibition of AKT and p38 MAPK phosphorylation. The AKT inhibitor LY294002 also suppressed TGFβ2-induced up-regulation of nuclear Snail and reduced phosphorylation of GSK3β. Inhibition of Snail expression suppressed TGFβ2-mediated α-SMA expression. αB-Crystallin was up-regulated in an AKT-dependent manner during AGE-BME/TGFβ2-mediated EMT in LECs. The absence of αB-crystallin in LECs suppressed TGFβ2-induced GSK3β phosphorylation, resulting in lower Snail levels. Taken together, these results show that ECM-AGEs enhance the TGFβ2-mediated EMT response through activation of the AKT/Snail pathway, in which αB-crystallin plays an important role as a linker between the TGFβ2 and AGE-mediated signaling pathways.
According to the National Eye Institute, the prevalence of cataracts is ∼20 million (17.2%) among adults 40 years and older in the U.S.A., and this number is expected to increase to nearly 40 million by 2030 . By the age of 80 years, ∼70% of Americans have cataracts . Many epidemiological studies have shown that diabetes is a defined risk factor for developing cataracts [2–4]. The increasing number of diabetic patients worldwide has coincided with the increasing incidence of cataracts, and diabetics develop cataracts at an earlier age than nondiabetics . The lens is enclosed by a basement membrane known as a capsule, and a single layer of epithelial cells [lens epithelial cells (LECs)] is attached to the capsule on the anterior side. The remaining lens is composed of differentiated fiber cells, which are organelle-free at the center (nucleus) of the lens.
During cataract surgery, a portion of the anterior capsule is excised (capsulorhexis) to remove the cloudy lens (via a process known as phacoemulsification), and the remaining capsule is left intact to allow implantation of an intraocular lens (IOL). During this procedure, most capsule-adhering anterior epithelial cells peripheral to capsulorhexis are removed, but some cells remain attached despite the best efforts of surgeons to remove them. These residual LECs initiate a wound-healing response, which is characterized by cell proliferation and migration, with an attempt to differentiate into fiber cells . Such differentiation is usually incomplete, resulting Elschnig's pearl formation at the lens equator . After cells reach the posterior capsule, they undergo epithelial–mesenchymal transition (EMT) and synthesize α-smooth muscle actin (α-SMA), resulting in contraction and wrinkle formation in the capsule. This EMT response eventually results in fibrosis, blocking the visual axis in a process known as posterior capsule opacification (PCO) or secondary cataract formation. PCO is a common post-cataract surgery complication and occurs 2–5 years after cataract surgery in ∼20–40% of patients . To remedy this problem, a Neodymium : yttriumaluminum-garnet (Nd : YAG) laser is routinely used, which obliterates fibrotic tissue and clears the visual axis . Although this procedure is safe in the majority of cases, in some cases, it leads to complications, such as retinal tears, hemorrhages, and endophthalmitis [10,11]. Thus, efforts are underway to prevent PCO by improving the design of IOLs to prevent LEC migration to the posterior capsule and by developing inhibitors of EMT.
Several recent studies have investigated the mechanisms of PCO and have implicated cytokines and growth factors, such as transforming growth factor (TGF), fibroblast growth factor, hepatocyte growth factor, interleukins, and epithelial growth factor [12–15]. However, the major driver appears to be TGFβ . Three TGFβ isoforms have been identified in mammals: TGFβ1, TGFβ2, and TGFβ3. In the eye, TGFβ2 is the dominant form and is mainly produced in ciliary and lens epithelia . Many studies have shown that TGFβ2-mediated EMT involves a morphological change of epithelial cells from a cuboidal to an elongated spindle shape, with a decrease in expression of epithelial markers, such as E-cadherin and zonula occludin-1, and enhanced expression of mesenchymal markers, such as α-SMA and fibronectin .
There are two main modes of TGFβ-signaling pathways. Canonical TGFβ signaling is mediated by Smads . In this pathway, TGFβ-induced activation of the receptor leads to activation of Smad2 and Smad3 through direct C-terminal phosphorylation by Type I serine/threonine kinase receptors (TβRI) . Phosphorylated Smad2 and Smad3 then form a heterotrimer with Smad4 and translocate into the nucleus, where they associate and co-operate with other DNA-binding transcription factors to activate or repress target gene transcription . Noncanonical signaling is directly activated by TβRI either through phosphorylation or direct interaction with various MAP kinase pathways, such as ERK (extracellular signal-regulated kinases), JNK, and p38 MAPK (mitogen-activated protein kinases), Rho-like GTPase, and phosphatidylinositol-3-kinase/AKT [21–24]. The contribution of two pathways to EMT in other tissues also has been observed [25,26], but it is unclear for the lens.
αB-Crystallin is a major protein in the lens; it is also present in significant amounts in the heart, kidneys, and skeletal muscles [27–29]. Several previous studies have shown that αB-crystallin plays a promotional role in EMT of epithelial cells in other tissues [30–32]. We recently showed that in LECs, αB-crystallin is obligatory for TGFβ2-mediated EMT . In that study, we found that αB-crystallin was essential for phosphorylation of AKT, Smad2/3, and their translocation to the nucleus. Similarly, translocation of Snail to the nucleus was also dependent on αB-crystallin. In addition, αB-crystallin bound to both p-Smad2/3 and Snail during their translocation to the nucleus. Furthermore, the chaperone activity of αB-crystallin was found to be required for EMT promotional activity. However, whether extracellular matrix-bound advanced glycation end products (AGEs) have any effect on the αB-crystallin levels was not investigated in that study.
AGEs are post-translational modifications in proteins formed from the reaction of reactive carbonyl compounds with protein amino groups, mostly on lysine and arginine residues. AGEs have been implicated in both age- and diabetes-related eye diseases . Our laboratory has previously shown that AGEs generated in basement membrane extract (BME) or lens capsules further promote TGFβ2-mediated EMT of LECs . Our studies also showed that AGEs in the capsule engage RAGE, a receptor for AGEs, in LECs during the promotion of EMT . However, in the latter study, after a 24 h treatment with TGFβ2, there was no change in either Smad or ERK phosphorylation during AGE-promoted EMT. This raised the possibility that AGEs may up-regulate TGFβ2 signaling pathways at an earlier time period (earlier than 24 h). The aim of this study was to investigate the mechanisms by which extracellular AGEs alter TGFβ2 signaling during EMT and to determine whether αB-crystallin has any role in these mechanisms.
Human LEC-line FHL124 cells were kindly provided by Dr Michael Wormstone (School of Biological Sciences, University of East Anglia, U.K., originally from Dr John Reddan, Oakland University, MI). Cell authentication was performed using an Affymetrix gene microarray, which found 99.5% homology between FHL124 cells and native human lens epithelium in the expression pattern of 22 270 genes , and by analysis of short tandem repeats . FHL124 cells up to passage 20 were used in this study. FHL124 cells were cultured in MEM (Life Technologies, Carlsbad, CA) containing 5% (v/v) heat-inactivated fetal bovine serum (FBS, Gibco, Grand Island, NY) and gentamycin/l-glutamate (1 : 100, Sigma–Aldrich, St. Louis, MO) and maintained in a humidified incubator at 37°C containing 5% CO2.
Mouse LECs from wild-type (WT, 129Sv strain from Jackson Laboratories) or αB-crystallin knockout (KO, 129Sv strain, originally from Dr Eric Wawrousek, NEI, re-derived) mice were isolated and cultured in MEM containing 20% (v/v) heat-inactivated FBS as recently described . FHL124 cells and mouse LECs were both serum-starved overnight and stimulated with TGFβ2 (Peprotech, Inc., Rocky Hill, NJ, 10 ng/ml) for the indicated times. Specific inhibitors were treated for 1 h prior to being co-treated with 10 ng/ml TGFβ2 for stimulation.
AGE modification of BME
AGE modification of BME (Cultrex® BME, Trevigen, Gaithersburg, MD) was performed as recently described . Briefly, after coating in 60-mm culture dishes (Falcon,Cat#353002) with BME (50 μg/ml), the plate was incubated at 37°C with a sterile-filtered (0.2 μm filter) glycating mixture that consisted of 2 mM ascorbate, 25 mM d-glucose, and 250 μM methylglyoxal (all from Sigma) in 50 mM sodium phosphate buffer for 1 week at pH 7.4. After rinsing the plate with PBS, FHL124 cells were seeded on AGE-modified (AGE-BME) or unmodified BME (BME) with MEM containing 5% FBS for 24 h.
Quantitative real-time PCR
RNA was extracted from cells using the RNeasy Plus Mini Kit (Qiagen, Valencia, CA). The quantities of isolated RNA were determined using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Rockford, IL). RNA was reverse-transcribed to synthesize cDNA using the QuantiTect Reverse Transcription Kit (Qiagen). Quantitative PCR for analyzing mRNA gene expression was performed with iTaq™ Universal SYBR® Green Supermix (Bio-Rad, Richmond, CA, U.S.A.) using an iCycler iQ5 Real-Time PCR Detection System (Bio-Rad). The primers used are listed in Table 1.
|Gene .||Primer sequence .|
|Gene .||Primer sequence .|
Total cell lysates were prepared using the Mammalian Protein Extraction Reagent (Thermo scientific) containing a protease and phosphatase inhibitor cocktail (Sigma–Aldrich, St. Louis, MO, 1 : 100 dilution). Cells were separated into cytosolic and nuclear fractions using a NucBuster™ protein extraction kit (Novagen, Madison, WI) according to the manufacturer's protocol. Proteins were separated on SDS–PAGE, transferred to a nitrocellulose membrane, and blocked using 5% nonfat dry milk in TBST buffer (10 mM Tris, 150 mM NaCl, and 0.1% Tween 20, pH 7.5) for 1 h. The membranes were incubated at 4°C overnight with a primary antibody against the following proteins: α-SMA (Sigma–Aldrich, Cat#A5228), αB-crystallin (Developmental Studies Hybridoma Bank, University of Iowa, IA), E-cadherin (Cat#3195), β-actin (Cat#4970), Snail (Cat#3879), Slug (Cat#9585), zinc-finger E-box-binding (ZEB) (Cat#3396), histone H3 (Cat#9715), GSK3β (Cat#12456), p-GSK3β (Cell Signaling Technology, Inc., Beverly, MA) (Cat#5558), and Twist (Santa Cruz Biotechnology, Santa Cruz, CA) (Cat#sc-81471). HRP-conjugated secondary antibodies (Cell Signaling Technology) were incubated at RT for 1 h and detected with the SuperSignal West Pico or Femto Kit (Pierce Chemicals, Rockford, IL). The band intensities were measured using Image J software (National Institutes of Health), and the results are expressed as fold change over controls.
Bead-based multiplex assays
The TGFβ signaling pathway was assessed using intracellular bead-based multiplex assay kits from the EMD Millipore Corporation (Billerica, MA) according to the manufacturer's instructions. After the treatment of TGFβ2 for 1 h, cells were lysed in the provided lysis buffer and particulate matter was removed by filtration. The protein concentration in the filtered samples was measured using the BCA assay. Cell lysates (equal to 10 μg protein) were added to the wells of a 96-well plate containing conjugated beads and incubated on a plate shaker (800 rpm) in the dark at 4°C overnight. After washing, the respective detection antibody was added to each well and the plate was incubated for 1 h at RT. The detection antibody was removed by decantation, streptavidin-R-phycoerythrin was incubated for 15 min and amplification buffer was added to each well. The beads were analyzed using the Luminex Magpix system (Luminex Corp., Austin, TX). The beads detected changes in phosphorylated Smad2 (Ser465/Ser467), Smad3 (Ser423/Ser425), ERK (Thr185/Tyr187), and AKT (Ser473), as well as the total protein levels of TGFβRII and Smad4 in cell lysates.
Transfection of siRNA for Snail
FHL124 cells were transfected with nonsilencing or Snail-specific siRNA (Qiagen) using Lipofectamine 2000 (Invitrogen, Carlsbad, CA) in accordance with the manufacturer's protocol. On the day before transfection, cells were seeded in a six-well plate at a density of 5 × 104 cells/well. The next day, cells were transfected for 4 h with 20 nM siRNA and Opti-MEM including 7.5 μl of Lipofectamine 2000, and the media were replaced with fresh serum-free media. After 24 h, the cells were used in experiments. The target sequences of the siRNAs were siSnail, GAGGTGTGACTAACTATGCAA, and nonsilencing siRNA, AATTCTCCGAACGTGTCACGT, as a negative control.
GraphPad Prism software version 7 (GraphPad Prism Software, Inc., San Diego, CA) was used for all statistical analyzes. Data are expressed as the means ± standard deviation (SD) of three independent experiments. We used Tukey's multiple comparison test for significant differences among treatment groups. A P-value of <0.05 was considered statistically significant.
AGE-BME enhances the TGFβ2-induced EMT
To determine whether AGE-BME enhances TGFβ2-mediated EMT of FHL124 cells, cells were treated with 10 ng/ml TGFβ2. The α-SMA, αB-crystallin, and E-cadherin levels were compared at 24 and 48 h using qRT-PCR and Western blotting. We found that AGE-BME promoted TGFβ2-mediated up-regulation of α-SMA and αB-crystallin (Figure 1A,B). TGFβ2 decreased the E-cadherin mRNA and protein levels by 47 and 60% compared with untreated BME control cells, respectively. The combination of AGE-BME with TGFβ2 decreased the E-cadherin mRNA and protein levels by 63 and 78% compared with the AGE-BME control, respectively. The levels of α-SMA and αB-crystallin were higher in AGE-BME than in BME, but they were not statistically significant.
AGE-BME enhances TGFβ2-mediated EMT in FHL124 cells.
AGE-BME promotes the noncanonical signaling pathway in FHL124 cells
Stimulation of Smad signaling by TGFβ is well known . In response to TGFβ, Smads translocate from the cytoplasm to the nucleus and regulate target gene transcription . Our data show that TGFβ2 treatment rapidly elevated the levels of the canonical pathway molecules p-Smad2 and p-Smad3, but the levels of TGFβRII and Smad4 were unchanged after 1 h (Figure 2A). However, there was no significant difference in Smad signaling between BME and AGE-BME in response to TGFβ2. In the noncanonical signaling pathway of TGFβ, the role of ERK, p38 MAPK, and PI3K (phosphoinositide 3-kinase)/AKT has been well established . The addition of TGFβ2 to cells cultured on AGE-BME for 1 h increased the noncanonical signaling pathway, as indicated by an increase in the levels of p-ERK and p-AKT (Figure 2B). In the case of AKT, the increase was significantly greater than for cells cultured on BME alone. Since we used a commercial Multiplex Assay Kit that did not include p-p38 MAPK, we measured p-p38 MAPK by Western blotting in the same cell lysate used for the Multiplex Assay. The addition of TGFβ2 to cells cultured on AGE-BME for 1 h enhanced the p-p38 MAPK levels (Figure 2C).
AGE-BME promotes MAPK phosphorylation in FHL124 cells.
Inhibitors of the TGFβ2-mediated noncanonical pathway suppress the EMT response
We evaluated the contribution of AGE-BME to the noncanonical TGFβ signaling pathway using specific inhibitors. FHL124 cells were treated with inhibitors of ERK1/2 (U0126, 10 μM), p38 MAPK (SB202190, 10 μM), or AKT (LY294002, 10 μM) for 1 h followed by treatment with TGFβ2 for 48 h (in the presence of inhibitor). Expression of α-SMA, αB-crystallin, and E-cadherin was determined by Western blot analysis. The inhibitors for AKT and p38 MAPK phosphorylation suppressed the expression of α-SMA and αB-crystallin in cells on AGE-BME treated with TGFβ2 (Figure 3A,B), but not from TGFβ2 alone. These results suggest that TGFβ2 induces the EMT mainly through the canonical signaling pathway, whereas AGE-BME through the noncanonical pathway. In addition, the inhibitors of ERK and AKT phosphorylation blocked the TGFβ2-mediated decrease in E-cadherin levels in cells on AGE-BME (Figure 3C).
AGE-BME enhances TGFβ2-induced EMT of FHL124 cells through activation of MAPK pathway.
AGE-BME promotes TGFβ2-mediated up-regulation of nuclear Snail and Slug
The levels of the EMT-inducing transcription factors Snail, Slug, Twist, and ZEB were measured. Snail and Slug were both significantly higher in the nuclear proteins of TGFβ2-treated cells on AGE-BME compared with TGFβ2-treated cells on BME (Figure 4A). However, the Twist and ZEB levels were unchanged. These results suggest that AGE-BME promotes TGFβ2-mediated EMT through the activation of Snail and Slug. To define which noncanonical TGFβ signaling pathway is associated with these transcription factors, specific inhibitors were pretreated for 1 h and cells were induced with TGFβ2 for 6 h. Blocking of the AKT pathway with LY294002 strongly suppressed the nuclear levels of Snail and Slug (Figure 4B).
AGE-BME activates transcription factors in TGFβ2-induced FHL124 cells.
Glycogen synthase kinase 3β (GSK3β) has been reported to enhance degradation of Snail via ubiquitination and is inactivated by phosphorylation . As shown in Figure 4C, TGFβ2 increased GSK3β phosphorylation in FHL124 cells after 1 h. The addition of TGFβ2 to cells cultured on AGE-BME for 1 h further enhanced GSK3β phosphorylation. The addition of a specific inhibitor of AKT decreased this GSK3β phosphorylation.
Snail is upstream of α-SMA but not αB-crystallin
To determine whether Snail is upstream of either α-SMA or αB-crystallin, cells were transfected using Snail-specific siRNA. We determined the efficiency of Snail siRNA using qRT-PCR. Snail mRNA expression was decreased by 60% (Figure 5), and both the mRNA and protein levels of α-SMA also strongly decreased. However, the level of αB-crystallin did not change. These results indicate that Snail is upstream of α-SMA, but not αB-crystallin. Interestingly, when the Snail levels were reduced, the mRNA levels, but not the protein levels, of αB-crystallin increased relative to scrambled siRNA treated LECs. The reason for this is not known.
Snail siRNA suppresses α-SMA levels in FHL124 cells.
αB-crystallin promotes GSK3β phosphorylation and up-regulates Snail
We determined the role of αB-crystallin in Snail nuclear translocation using WT and αB-crystallin KO mouse LECs. Phosphorylation of GSK3β was higher after treatment with TGFβ2 for 2 h in WT, resulting in higher levels of Snail. However, a similar TGFβ2 treatment in αB-crystallin KO mouse LECs did not change the p-GSK3β levels. The Snail levels in KO mouse LECs were significantly lower than WT LECs after TGFβ2 treatment (Figure 6B,C). Together, these data indicated that αB-crystallin regulates Snail translocation via GSK3β phosphorylation.
The absence of αB-crystallin reduces phosphorylation of GSK3β and decreases Snail expression.
The capsule secreted by the lens epithelium is the thickest basement membrane (BM) in the human body . The proteins in the BM are generally long-lived and are targets for post-translational modification during aging. Our recent study showed that with aging, AGEs accumulate in the human lens capsule . That study also showed that TGFβ2-mediated EMT signaling positively correlated with the AGE levels in the human lens capsule and suggested that AGEs promote TGFβ2-mediated EMT in LECs . The AGE levels in AGE-BME are comparable to those in aged human lens capsules, as shown in our recent study . In a subsequent study, we showed that AGEs in the capsule engage a receptor for AGEs, known as RAGE, in LECs during EMT . Even though FHL124 cells are derived from a fetal lens, our recent study showed that those cells are similar to primary human LECs in their response to AGE-BME and TGFβ2 . We have also recently shown that αB-crystallin is essential for TGFβ2-mediated EMT in LECs . However, it is unknown whether AGE-BME promotes TGFβ2-mediated EMT in LECs through up-regulation of αB-crystallin. The purpose of this study was to investigate the mechanisms by which AGE-BME promotes TGFβ2 signaling during EMT of LECs and to determine whether αB-crystallin has a role in these mechanisms. The major findings are: (1) in early TGFβ2-mediated signaling, AGE-BME activates the noncanonical pathway; (2) AGE-BME up-regulates αB-crystallin through activation of the AKT pathway; and (3) AGE-BME promotes TGFβ2-mediated accumulation of Snail in the nucleus through inactivation of GSK3β.
Previous studies have shown that phosphorylation of Smad2 occurs to a maximum extent after ∼2 h of TGFβ2 treatment in cells [39–41]. However, in our recent study, we found that the p-Smad2 levels remained at significantly higher levels at 24 h in cells grown on AGE-BME compared with cells grown on unmodified BME . In the present study, after 1 h of TGFβ2 treatment, there was no difference in the p-Smad2/3 levels between the two groups. This indicated that the difference in the p-Smad2 levels between the two groups occurred between 1 and 24 h of TGFβ2 treatment. A similar difference was also seen for ERK phosphorylation. While a recent study showed no difference between the two groups in the p-ERK1/2 levels after 24 h of TGFβ2 treatment , the present study showed a significant difference after 1 h of treatment. These observations led us to investigate the mechanisms for an early increase in p-ERK1/2 and sustained up-regulation of p-Smad2 by AGEs in conjunction with TGFβ2.
We found that in the presence of AGEs, TGFβ2 activated noncanonical signaling after 1 h, while the p-Smad2 levels remained indistinguishable between TGFβ2-treated cells on BME or AGE-BME. This suggested that noncanonical signaling by TGFβ2 is triggered early by AGEs. Thus, AGE-BME exacerbated TGFβ2-mediated EMT of LECs by enhancing signaling via the noncanonical pathway.
In addition, cross-talk between the canonical and the noncanonical TGFβ2 signaling pathways can occur during TGFβ2-mediated EMT of LECs in the presence of AGEs. In fact, several studies have shown that TGFβ-induced Smad signaling and ERK pathway cross-talk during EMT in other cells [26,42,43]. For example, Bakin et al.  showed that PI3K/AKT activation is directly involved in TGFβ1-mediated Smad2 phosphorylation during EMT of tumor cells. However, the mechanism by which TGFβ activates the PI3/AKT signaling pathway is unclear. We believe, based on our results, that αB-crystallin has a role in TGFβ-mediated PI3/AKT activation.
The role of αB-crystallin in EMT is becoming clearer with recent studies. Ishikawa et al.  reported that αB-crystallin is essential to the formation of subretinal fibrosis after EMT in retinal pigment epithelium cells. Other studies have established that αB-crystallin favors TGFβ1 signaling via Smad4 activation during fibrosis [44,45]. In our recent study, we showed that the absence of αB-crystallin results in inhibition of EMT of LECs by TGFβ2 and in lensectomized mouse lens capsule-adherent LECs . In that study, we demonstrated that the chaperone activity of αB-crystallin is required and that it physically interacts with Smads and Snail during TGFβ2-mediated EMT of LECs. The lack of αB-crystallin reduced activation of ERK, p38 MAPK, and AKT signaling. In this study, we demonstrated that AGE-BME promotes TGFβ2-mediated up-regulation of αB-crystallin and that inhibition of AKT phosphorylation reduces this up-regulation.
Several transcription factors regulate EMT, including zinc-finger proteins of the Snail/Slug family, Twist, and ZEB . It is established that TGFβ2-induced Smad2/3 transcription factors complex with the promoter of transcriptional regulators, such as Snail and Slug [46,47]. We have recently shown that αB-crystallin up-regulates p-Smad2/3, Smad4, and Snail expression after TGFβ2 treatment both in the cytoplasm and in the nucleus and that αB-crystallin physically interacts with Smads and Snail . The present data demonstrate that AGE-BME further up-regulates TGFβ2-induced αB-crystallin, resulting in the promotion of Snail accumulation in the nucleus. As Snail expression was stronger than Slug expression, we suspect that Snail is the primary driver of EMT by AGE-BME/TGFβ2. Inactivation of GSK3β through phosphorylation is a well-established phenomenon during TGFβ-mediated EMT , and inactivation of GSK3β is directly linked to increased Snail expression and translocation to the nucleus [48,49]. Our results show that AGE-BME phosphorylates Ser9 of GSK3β via AKT activation and results in Snail accumulation in the nucleus, supporting the idea that AKT-mediated GSK3β phosphorylation may trigger Snail accumulation in the nucleus. Phosphorylation of GSK3β in response to TGFβ2 was inhibited in αB-crystallin KO mouse LECs, which suggests its role in GSK3β phosphorylation. Analogous to our results, Zhou et al.  found that αB-crystallin protects against retinal degeneration, in part through AKT phosphorylation and GSK3β phosphorylation in RPE cells. Taken together, both canonical and noncanonical pathways up-regulate αB-crystallin, which is essential for the enhancement of EMT of LECs by AGE-BME.
In conclusion, our study showed that AGE-BME enhances TGFβ2-induced EMT of LECs by enhancing activation of the AKT/Snail pathway, in which αB-crystallin plays an important role as a linker between the TGFβ2- and AGE-mediated signaling pathways (Figure 7). As αB-crystallin is an important driver of EMT in LECs, we propose that the AKT/GSK3β pathway and αB-crystallin are additional targets for the prevention of PCO.
Possible signal transduction pathways by which AGE-BME enhances TGFβ2-induced EMT of LECs.
advanced glycation end products
basement membrane extract
extracellular signal-regulated kinases
fetal bovine serum
glycogen synthase kinase 3β
lens epithelial cells
mitogen-activated protein kinases
posterior capsule opacification
a receptor for AGEs
transforming growth factor-beta
Type I serine/threonine kinase receptors
α-smooth muscle actin
R.H.N. and M.-H.N. conceived, designed, and analyzed all experiments. M.-H.N. performed all experiments. The authors reviewed the results and approved the final version of the manuscript.
This work was supported by the National Institutes of Health Grants [EY022061 and EY023286 (to R.H.N.)] and Research to Prevent Blindness, NY.
We thank Dr Michael Wormstone for providing FHL124 cells. We thank Drs J. Rankenberg, R.B. Nahomi and S.K. Nandi for critical reading of the manuscript.
The Authors declare that there are no competing interests associated with the manuscript.