Effects of rapamycin on body weight and random blood glucose levels in diabetic rats

The body weight and random blood glucose levels of the rats were examined after treatment with rapamycin for 0 w (before treatment), 4 w and 8 w. The mean body weights of the three diabetic groups (DC, DMSO, and RAPA group) were obviously lower than that of the control group, and the body weight in the RAPA group was lower than that in the DC and DMSO groups (Fig. 1A). Before treatment with rapamycin (0 w), at 4 w and at 8 w, the random blood glucose levels measured in the three diabetic groups were higher than those in the normal group, while no significant differences were observed among the three diabetic groups (Fig. 1B).

Fig. 1

Rapamycin prevented the formation of cataracts in diabetic rats. Diabetic rats (DC) were induced by STZ. Rats in the normal control group (NC) were injected with the corresponding volume of citric acid-sodium citrate buffer. Diabetic rats were treated with DMSO solvent (DMSO) and rapamycin (RAPA) respectively. Body weight (A, n = 10) and random blood glucose levels (B, n = 10) of rats were examined after treatment with rapamycin for 0 (before treatment), 4 and 8 w. (C, D) Lenses of rats in each group (n = 20) were observed and photographed under a slit lamp microscope at 0 w, 4 w and 8 w after treatment with rapamycin. E, F H&E staining showed histological changes in the lens (n = 4). ns indicates P > 0.05; **P < 0.01 and ****P < 0.0001 vs. NC group; #P < 0.05 and ###P < 0.001 vs. DC group; ★P < 0.05 and ★★P < 0.01 vs. DMSO group

Rapamycin prevented cataract formation in diabetic rats

The lenses of rats in each group were observed and photographed under a slit lamp microscope at 0 w, 4 w, and 8 w (Fig. 1C, D). The lenses in the NC group remained transparent for 8 w. At 4 w, the lenses in the three diabetic groups showed various degrees of opacification. The degree of opacity of most lenses in the DC group (11/20) and the DMSO group (12/20) was Grade 2, the vacuoles at the equator of the lenses gradually expanded toward the center, and light cloud-like opacity appeared in the nuclear area of the lenses. In the RAPA group, the degree of opacity of most lenses was Grade 1, and vacuoles only appeared around the equator of the lenses. No significant difference in the degree of opacity was observed between the DC group and the DMSO group (adjusted P > 0.05), while the degree of lens opacity in the RAPA group was significantly lower than that in the DC group and the DMSO group (all adjusted P < 0.05). At 8 w, lens opacity in the three diabetic groups continued to increase. The degree of opacity of most lenses in the DC group (14/20) and the DMSO group (13/20) was Grade 3, and the vacuoles became denser and extended to the nuclear area. In the RAPA group, the degree of opacity of most lenses (13/20) was Grade 2. The statistical analysis revealed that the results were consistent with those obtained at 4 w. Based on these results, rapamycin delayed and prevented the development of cataracts in diabetic rats.

Histological changes in the lenses are showed in Fig. 1E, F. In the NC group, the LECs underneath the center anterior lens capsule were arranged in a single layer, and they showed a physiological bow arrangement at the equatorial region. The lens fibers in the superficial cortex displayed an orderly and tight arrangement. In the DC and DMSO groups, the lens epithelium was arranged in multiple layers, and some LECs migrated into the superficial cortex. Lens fibers in the cortex showed edema and vacuole formation. After treatment with rapamycin (RAPA group), the lens epithelium was basically maintained in a single flat structure, and the edema of lens fibers in the cortex was reduced.

Rapamycin activated autophagy and inhibited EMT in the lens of diabetic rats

At 8 w, the proteins in the LECs from the rats in each group were extracted, and autophagy marker proteins were detected using Western blot assays (Fig. 2A-C). Compared with the NC group, the expression of LC3 II/I in the LECs of the DC group was significantly reduced (P < 0.0001), while the expression of SQSTM1/p62 was increased (P < 0.0001). The expression levels of LC3 II/I and SQSTM1/p62 were not significantly different between the DC and DMSO groups (P > 0.05), indicating that autophagic activity in the lens epithelium of diabetic rats was inhibited. After treatment with rapamycin, the expression of LC3 II/I increased in the RAPA group (P < 0.0001) and the expression of SQSTM1/p62 decreased compared with that in both the DC group and the DMSO group (all P < 0.001), but the RAPA and NC groups showed significant differences in the expression of LC3II/I and SQSTM1/p62 (all P < 0.01).

Fig. 2

Rapamycin activated autophagy and inhibited the EMT in LECs from diabetic rats. Diabetic rats (DC) were induced by STZ. Rats in the normal control group (NC) were injected with the corresponding volume of citric acid-sodium citrate buffer. Diabetic rats were treated with DMSO solvent (DMSO) and rapamycin (RAPA) for 8 weeks respectively. Levels of the LC3 and SQSTM1/p62 proteins in the LECs of rats in each group were detected using Western blotting at 8 w (A). Statistical analysis of the relative expression of LC3II/I B and SQSTM1/p62 (C), n = 3. D Representative TEM images of autophagosomes (arrows) in different groups, bar = 0.5 µm. E Quantification and statistical analysis of the number of autophagosomes, n = 5. F E-cadherin and α-SMA protein levels in LECs from rats in each group were detected using Western blotting at 8 w. Statistical analysis of the relative expression of E-cadherin G and α-SMA (H), n = 3. I Immunofluorescence staining for E-cadherin (red) and α-SMA (green) in LECs from the different groups, n = 3. Bar = 10 µm. ns indicates P > 0.05; *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. NC group; #P < 0.05, ##P < 0.01, ###P < 0.001 vs. DC group; ★P < 0.05, ★★P < 0.01 and ★★★P < 0.001 vs. DMSO group

Autophagosomes were observed under a TEM to further detect the changes in autophagic activity in the LECs of diabetic rats (Fig. 2D, E). The number of autophagosomes in the DC group was significantly reduced compared with that in the NC group (P < 0.0001). However, the number of autophagosomes increased in the RAPA group. Collectively, this evidence suggested that rapamycin partially rescued the reduced autophagic activity in the LECs of diabetic rats.

EMT marker proteins were detected using Western blot assays (Fig. 2F–H). The expression of E-cadherin was significantly downregulated (P < 0.001) and the expression of α-SMA was significantly upregulated (P < 0.001) in the DC group compared with the NC group. In addition, the expression of E-cadherin and α-SMA showed no differences between the DC and DMSO groups (P < 0.05), indicating that the EMT occurred in the lens epithelium of diabetic rats. Compared with the DC and DMSO groups, the expression of E-cadherin increased (all P < 0.05) and the expression of α-SMA decreased (all P < 0.05) in the RAPA group, suggesting that rapamycin not only increased autophagic activity but also partially prevented the EMT in the lens epithelium of diabetic rats. Alterations in both E-cadherin and α-SMA expression were also detected in whole-mount lens epithelium using immunofluorescence staining, and the results were consistent with those from the Western blot assays (Fig. 2I).

Therefore, autophagy was suppressed but the EMT was activated in the lens epithelium of diabetic rats. The EMT was effectively prevented by enhancing autophagic activity with rapamycin, resulting in delayed formation of diabetic cataracts.

Rapamycin activated autophagy by inhibiting the mTOR/ULK1 signaling pathway in high-glucose-stimulated HLE-B3 Cells

We conducted additional in vitro experiments to explore whether activation of autophagy regulated the EMT of LECs under high-glucose conditions. HLE-B3 cells were stimulated with high glucose concentrations and treated with rapamycin. The expression of LC3 II/I was decreased (P < 0.001) and SQSTM1/p62 expression was increased (P < 0.0001) after stimulation with high glucose concentrations compared with those in the NC cells. After treatment with rapamycin, the expression of LC3 II/I was increased (P < 0.001) and SQSTM1/p62 expression was decreased (P < 0.0001) compared with the levels detected in cells cultured under high-glucose conditions (Fig. 3A–C), but these parameters still differed from those in the NC group. Under TEM, the number of autophagosomes in HLE-B3 cells in the HG group was significantly reduced compared with the number in the NC group (P < 0.0001). However, the number of autophagosomes increased in the RAPA group (Fig. 3G, H). Altered protein expression in the signaling pathway regulating autophagic activity in HLE-B3 cells was also observed using Western blot assays. The levels of p-AKT, p-mTOR, and p-ULK1 were all increased (all P < 0.0001) in HLE-B3 cells cultured under high-glucose conditions (in both the HG and DMSO groups) compared with those in NC cells, but the expression of AKT, mTOR, and ULK1 was not significantly altered (P > 0.05). After treatment with rapamycin, the levels of both p-AKT and AKT (P > 0.05), mTOR, and ULK1 were not significantly changed (P > 0.05), whereas the levels of p-mTOR and p-ULK1 were significantly decreased (P < 0.0001) compared with those in cells cultured under high-glucose conditions, and they were not significantly different from those in the NC group (Fig. 3A, D–F). Therefore, rapamycin activated autophagy by inhibiting the mTOR/ULK1 signaling pathway.

Fig. 3

Rapamycin activated autophagy by inhibiting the mTOR/ULK1 signaling pathway in HLE-B3 cells cultured under high-glucose conditions. Cells were treated with high glucose (HG, 30 mM glucose), high glucose + DMSO (equal volume solvent of rapamycin) and high glucose + 200 nM rapamycin (RAPA) for 24 h, respectively. A The expression of molecules in the autophagy signaling pathway and autophagy marker proteins. Statistical analysis of the relative levels of LC3II/I B and SQSTM1/p62 (C), p-AKT/AKT (D), p-mTOR/mTOR E and p-ULK1/ULK1 (F), n = 3. G Representative TEM images of autophagosomes (arrows) in different groups. H Quantification and statistical analysis of the number of autophagosomes, n = 5. ns indicates P > 0.05; *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. NC group; #P < 0.05, ##P < 0.01, ###P < 0.001 and ####P < 0.0001 vs. HG group; ★P < 0.05, ★★P < 0.01, ★★★P < 0.001 and ★★★★P < 0.0001vs. DMSO group

In summary, autophagy was inhibited in HLE-B3 cells cultured under high-glucose conditions, and rapamycin enhanced autophagic activity through the mTOR/ULK1 signaling pathway.

Rapamycin inhibited the EMT of HLE-B3 cells treated with high glucose concentrations through the notch signaling pathway

HLE-B3 cells were stimulated with high glucose concentrations and treated with rapamycin. The expression levels of E-cadherin and ZO-1 were downregulated (P < 0.001) and α-SMA and Fibronectin were significantly upregulated (P < 0.0001) in the HG group compared with those in the NC group. After treatment with rapamycin (RAPA group), the expression levels of E-cadherin and ZO-1 were increased (P < 0.05), and α-SMA and Fibronectin expression levels were significantly decreased (P < 0.05) compared with those in the high glucose groups (HG and DMSO groups), yet the significant differences were still observed between the RAPA group and NC group (all P < 0.05) (Fig. 4A–E).

Fig. 4

Rapamycin inhibited the EMT of HLE-B3 cells cultured under high-glucose conditions by modulating the Notch signaling pathway. HLE-B3 cells were stimulated with 30 mM glucose (HG group), and then treated with DMSO (DMSO group, equal volume solvent of rapamycin), 200 nM rapamycin (RAPA group) for 24 h, respectively. A The expression of E-cadherin, ZO-1, α-SMA and Fibronectin in HLE-B3 cells from each group was detected using Western blotting at 24 h. Statistical analysis of the relative expression of E-cadherin (B), ZO-1 (C), α-SMA D and Fibronectin (E), n = 3. Cell migration after exposure to different treatments for 24 h was detected by performing a transwell assay and scratch wound assay. Representative images and the quantification of the transwell assay (F, G, n = 3, bar = 100 μm) and scratch wound assay (H, I, n = 3, bar = 200 μm). J Expression of proteins in the Notch signaling pathway in HLE-B3 cells at 24 h. Quantification of the relative levels of Jagged1 (K), Notch1 (L), NICD (M) and Snail (N), n = 3. ns indicates P > 0.05; *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. NC group; ##P < 0.01 and ###P < 0.001 vs. HG group; ★P < 0.05, ★★★P < 0.001 and ★★★★P < 0.0001 vs. DMSO group

Cell migration was determined by performing transwell and scratch wound assays. Under high-glucose conditions, cell migration was significantly increased compared with that of normal cultured cells (all P < 0.0001). After treatment with rapamycin, migration was inhibited compared with that in the high-glucose and DMSO-treated cells, suggesting that rapamycin partially restrained the EMT of HLE-B3 cells exposed to high glucose concentrations (Fig. 4F–I).

The Notch signaling pathway consisting of Jagged1/Notch1/NICD/Snail is a classic signaling pathway regulating the EMT and is activated by TGFβ2 during the EMT of LECs (Han et al. 2019; Chen et al. 2017). Western blot assays (Fig. 4J–N) showed higher levels of Jagged1, Notch1, NICD and Snail in the HG and DMSO groups than those in the NC group (all P < 0.001). Compared with the HG and DMSO groups, Jagged1 expression was not significantly altered after treatment with rapamycin (P > 0.05), but the expression levels of Notch1, NICD and Snail were significantly decreased (all P < 0.001), although they were still higher than those in the NC group (all P < 0.05). Thus, the activation of autophagy partially blocked the activation of Notch signaling pathway stimulated by high glucose concentrations.

Immunofluorescence staining showed the colocalization of Notch1 and LC3 (Fig. 5A) and of Notch1 and SQSTM1/p62 (Fig. 5B) in the cytoplasm of HLE-B3 cells. Co-IP showed that Notch1 interacted with LC3 and SQSTM1/p62 (Fig. 5C), and LC3 interacted with SQSTM1/p62 (Fig. 5D). Based on these results, autophagy might mediate the selective degradation of Notch1 through interactions with LC3 and SQSTM1/p62, thus negatively regulating the Notch signaling pathway.

Fig. 5

Autophagy promoted Notch1 degradation through the interactions of Notch1, LC3 and SQSTM1/p62. HLE-B3 cells were cultured in DMEM (5.5 mM glucose) for 24 h. A Immunofluorescence staining for LC3 (red) and Notch1 (green) in HLE-B3 cells, bar = 40 μm. B Immunofluorescence staining for SQSTM1/p62 (red) and Notch1 (green) in HLE-B3 cells, bar = 40 μm. C, D Co-IP of Notch1, LC3 and SQSTM1/p62. IP: immunoprecipitation, WB: Western blot, n = 3

Crosstalk between the EMT and autophagy signaling pathways

We employed not only rapamycin but also DAPT, a γ-secretase inhibitor that blocks the final step of Notch cleavage and activation and then inhibits the Notch signaling pathway in vivo and in vitro (Park et al. 2015; Jiao et al. 2014), to evaluate the crosstalk between autophagy signaling pathways mediated by the mTOR/ULK1 and the EMT signaling pathways mediated by Jagged1/Notch1/NICD/Snail in HLE-B3 cells.

Western blot assays showed no significant differences in the expression of E-cadherin, ZO-1, α-SMA and Fibronectin between the DAPT and RAPA groups, although the levels were significantly different from those in the high-glucose groups (HG and DMSO groups). However, after combined treatment with rapamycin and DAPT (R + D group), the alterations in the levels of EMT marker proteins were more significant than those in the groups treated with RAPA and DAPT alone (Fig. 6A–E). In addition, transwell and scratch wound assays revealed that either rapamycin or DAPT effectively inhibited cell migration induced by high glucose concentrations, and the combination of rapamycin and DAPT prevented cell migration much more significantly than either of treatment alone (Fig. 6F–I).

Fig. 6

Effects of rapamycin combined with DAPT on autophagy and the EMT in HLE-B3 cells cultured under high-glucose conditions.. HLE-B3 cells were stimulated with 30 mM glucose (HG), then treated with DMSO solvent (DMSO), 200 nM rapamycin (RAPA), 5 μM DAPT (DAPT), and combination of 200 nM rapamycin with 5 μM DAPT (R + D) for 24 h, respectively. A The expression of E-cadherin, ZO-1, α-SMA and Fibronectin in HLE-B3 cells from each group was detected using Western blotting at 24 h. Statistical analysis of the relative expression of E-cadherin (B), ZO-1 (C), α-SMA D and Fibronectin (E), n = 3. Cell migration after exposure to different treatments for 24 h was detected by performing a transwell assay and scratch wound assay. Representative images and results of the transwell assay (F, G, n = 3, bar = 100 μm) and scratch wound assay (H, I, n = 3, bar = 200 μm). J The expression of molecules in the autophagy signaling pathway and autophagy marker proteins. Statistical analysis of the relative levels of LC3II/I (K), SQSTM1/p62 (L), p-mTOR/mTOR (M) and p-ULK1/ULK1 (N), n = 3. O The expression of proteins in the Notch signaling pathway. Quantification of the relative levels of Jagged1 (P), Notch1 (Q), NICD (R) and Snail (S), n = 3. ns indicates P > 0.05; *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. NC group; #P < 0.05, ##P < 0.01, ###P < 0.001 and ####P < 0.0001 vs. DMSO group; ★P < 0.05, ★★P < 0.01, ★★★P < 0.001 and ★★★★P < 0.0001 vs. RAPA group; ◆P < 0.05, ◆◆P < 0.01, ◆◆◆P < 0.001 and ◆◆◆◆P < 0.0001 vs. DAPT group

The detection of the expression of LC3 and SQSTM1/p62 proteins showed that either DAPT or rapamycin significantly increased autophagic activity in HLE-B3 cells stimulated with high glucose concentrations, and treatment with the combination of rapamycin and DAPT exerted a stronger effect than either compound alone (Fig. 6J–L). Among the proteins that regulate autophagy, p-mTOR, mTOR, p-ULK1 and ULK1 were assessed. Rapamycin effectively decreased the levels of p-mTOR and p-ULK1, while DAPT had no obvious effects on the levels of p-mTOR but significantly reduced the levels of p-ULK1 compared with those in the HLE-B3 cells cultured under high-glucose conditions (HG and DMSO groups). After treatment with the combination of rapamycin and DAPT, p-mTOR levels were not decreased further than those observed after rapamycin treatment, but p-ULK1 levels showed a greater reduction (Fig. 6J, M, N). These results suggested that DAPT regulated the autophagic activity of HLE-B3 cells by modulating the phosphorylation of ULK1 at Ser757. As a γ-secretase inhibitor, the NICD level was reduced after DAPT treatment in HLE-B3 cells. We speculated that NICD regulated the phosphorylation of ULK1 at Ser757.

Among the key proteins in the Notch signaling pathway, Jagged1, Notch1, NICD and Snail were all upregulated in HLE-B3 cells cultured under high-glucose conditions. As expected, DAPT reduced the levels of NICD and Snail but had no significant effects on the expression of Notch1, while rapamycin significantly reduced the level of Notch1, NICD and Snail. After treatment with the combination of rapamycin with DAPT, the levels of NICD and Snail were reduced to a greater extent than those detected in the groups treated with each compound alone, but Notch1 expression was not reduced to a greater extent than that in the RAPA group (Fig. 6O–S).

Thus, the activation of autophagy by rapamycin inhibited the EMT through the Notch1/NICD/Snail signaling pathway, while the inhibition of the EMT with DAPT inhibited the NICD/Snail signaling pathway and simultaneously activated autophagy by decreasing p-ULK1 levels. Therefore, the crosstalk between the EMT and the autophagy signaling pathways was mediated by NICD signaling to p-ULK1.

Enhanced interaction of NICD and ULK1 in HLE-B3 cells stimulated with high glucose concentrations

We found that DAPT inhibited the release of NICD and then prevented the phosphorylation of ULK1 stimulated by high glucose, thereby activating autophagy. Immunofluorescence staining showed that NICD and ULK1 colocalized in the cytoplasm of HLE-B3 cells, and this colocalization increased significantly after stimulation with high glucose concentrations. Moreover, NICD levels were increased and entered the nucleus (Fig. 7A). Co-IP assays proved an interaction between NICD and ULK1 in HLE-B3 cells stimulated with high glucose concentrations (Fig. 7B, C).

Fig. 7

The Notch signaling pathway regulated autophagic activity through NICD-induced phosphorylation of ULK1. A Immunofluorescence staining for NICD (red) and ULK1 (green) in HLE-B3 cells cultured under normal-glucose (NC) and high-glucose conditions (HG), bar = 40 μm. B, C Co-IP of NICD and ULK1 in normal cultured HLE-B3 cells. IP immunoprecipitation, WB Western blot. D Verification of NICD overexpression plasmids, and detection of the levels of autophagy marker proteins and p-ULK1 after transfected with pEGFP-C3 and pEGFP-C3-NICD plasmids for 36 h in normal cultured HLE-B3 cells (NC) by Western blotting. Quantification of the relative levels of NICD (E), p-ULK1/ULK1 (F), LC3II/I G and SQSTM1/p62 (H), n = 3. ns indicates P > 0.05; *P < 0.05, **P < 0.01 and ****P < 0.0001 vs. NC group; #P < 0.05, ##P < 0.01 and ####P < 0.0001 vs. HG + pEGFP-C3 group. I Verification of ULK1 and ULK1 (Ser 757) overexpression plasmids, and detection of the levels of proteins in the Notch signaling pathway and p-ULK1 after transfected with pcDNA3.1, pcDNA3.1-ULK1, and pcDNA3.1-ULK1 (S758A) plasmids for 36 h, respectively, in HLE-B3 cells cultured in high-glucose conditions (HG) by Western blot assays. Quantification of the relative levels of ULK1 (J), p-ULK1/ULK1 (K), Notch1 (L), NICD M and Snail (N), n = 3. ns indicates P > 0.05; *P < 0.05, **P < 0.01, ***P < 0.001 and ****P < 0.0001 vs. NC group; #P < 0.05, ##P < 0.01 vs. HG + pcDNA group; ★★P < 0.01 vs. HG + pcDNA-ULK1 group

We overexpressed NICD by transfecting pEGFP-C3-NICD into HLE-B3 cells to further confirm that ULK1 phosphorylation is regulated by NICD, which resulted in increased levels of p-ULK1 and SQSTM1/p62 and the downregulation of LC3II/I expression (Fig. 7D–H), indicating the inhibition of autophagy.

Then, we overexpressed ULK1 and the ULK1 mutant at S758A using the pcDNA3.1 vector to verify the role of ULK1 phosphorylation in the Notch signaling pathway. The transfection was confirmed by performing Western blot assays. A significant difference in ULK1 expression was not observed between these two ULK1 overexpression groups; however, ULK1 phosphorylation at Ser 757 was significantly higher in the HG + pcDNA-ULK1 group than in the HG + pcDNA-ULK1 (S758A) group (Fig. 7I–K). Overexpression of ULK1, but not ULK (S758A), significantly increased the expression of proteins in the Notch pathway under high-glucose conditions (Fig. 7I, L–N), indicating that the phosphorylation of ULK1 at Ser 757 activated the Notch signaling pathway in HLE-B3 cells under high-glucose conditions.

Thus, the Notch signaling pathway was not activated in normally cultured HLE-B3 cells, as the amount of activated NICD in the cytoplasm was very low, its effects on ULK1 were weak, and autophagy proceeded normally. Under high-glucose conditions, the Notch signaling pathway was activated, and the NICD level in the cytoplasm increased substantially. The interaction of NICD and ULK1 increased the phosphorylation of ULK1 and inhibited the activity of ULK1, thereby blocking the initiation of autophagy and inhibiting the degradation of Notch1 by autophagy.