3.1 High glucose induced excessive proliferation of RECs
To evaluate the effect of high glucose on the proliferation of RECs, the RECs were cultured with different concentrations of high glucose, and the viability of RECs were measured at varied incubation times, which were compared with normal-glucose group. As shown in Fig. 1A, the RECs viability in HG-1 was significantly elevated compared with NG at the 24-hour period (1.12 ± 0.04 vs. 1.00 ± 0.00, P < 0.01). Meanwhile, the RECs viability in HG-2 was also greatly elevated compared with NG at the 24-hour, 48-hour and 72-hour period, respectively (1.21 ± 0.04 vs. 1.00 ± 0.00, P < 0.01; 1.11 ± 0.02 vs. 1.00 ± 0.00, P < 0.01; 1.08 ± 0.03 vs. 1.00 ± 0.00, P < 0.01). For other comparisons, such as HG-3 vs. NG and HG-4 vs. NG at the 24-hour, 48-hour, 72-hour period, respectively, all the findings suggested that the viability of RECs had a more significant relation with the increase of high glucose concentration. These results indicated that high glucose could induce excessive proliferation of RECs.
3.2 Hyperoside inhibited excessive proliferation of RECs in high glucose
We evaluated the effects of hyperoside against high-glucose-induced excessive viability of RECs. Excessive proliferation of RECs is an early process of BRB destruction and DR occurrence. Inhibition of hyperoside on over-proliferation of RECs under high glucose condition would significantly improve DR. As presented in Fig. 1B, all RECs viability levels in HG at different incubation periods (24, 48, and 72 h) were higher than for NG groups (1.11 ± 0.04 vs. 1.00 ± 0.00, P < 0.05; 1.26 ± 0.07 vs. 1.00 ± 0.00, P < 0.01; 1.10 ± 0.03 vs. 1.00 ± 0.00, P < 0.05). Meanwhile, all RECs viability levels in HG + H100 at different treatment periods were lower than HG groups (1.03 ± 0.02 vs. 1.11 ± 0.04, P < 0.05; 1.05 ± 0.11 vs. 1.26 ± 0.07, P < 0.01; 0.94 ± 0.04 vs. 1.10 ± 0.03, P < 0.01). The similar difference was also seen in comparisons of HG + H400 vs. HG and HG + H400 vs. HG + H100. The results showed that both low-concentration (100 ug/mL) and high-concentration (400 ug/mL) hyperoside could significantly inhibit RECs viability under high glucose condition, and the stronger inhibition of RECs viability was along with the increase of hyperoside concentration. These data suggested that hyperoside could dose-dependently inhibit excessive proliferation of RECs in high glucose.
3.3 Hyperoside inhibited excessive migration and tube formation of RECs in high glucose
It is well known that DR may include not only RECs proliferation, but also significantly increased migration and tube formation of RECs. Therefore, we measured the role of hyperoside for the migration and tube formation of RECs in high glucose. As shown in Fig. 2A,C, the number of migrated RECs in HG was significantly higher than that in NG (375.7 ± 10.4 vs. 133.3 ± 9.6, P < 0.01), and also higher than HG + H100 and HG + H400 (375.7 ± 10.4 vs. 234.7 ± 6.5, P < 0.01; 375.7 ± 10.4 vs. 137.0 ± 9.0, P < 0.01). Similar results were also seen in the tube formation assay as presented in Fig. 2B,D, the number of branch points of RECs in HG was significantly higher than that in NG (38.0 ± 4.0 vs. 9.3 ± 1.5, P < 0.01), meanwhile, the numbers in HG + H100 and HG + H400 were significantly decreased compared with HG (21.0 ± 3.6 vs. 38.0 ± 4.0, P < 0.01; 10.3 ± 1.5 vs. 38.0 ± 4.0, P < 0.01). These data suggested that hyperoside could inhibit excessive migration and tube formation of RECs induced by high glucose.
3.4 TGF-β1/miR-200b/VEGF pathway contributed to over-proliferation of RECs in high glucose
To estimate the role of TGF-β1/miR-200b/VEGF pathway in the over-proliferation of RECs induced by high glucose, we, respectively, transfected miR-200b mimic and miR-200b inhibitor into RECs in high-glucose and normal-glucose condition. As shown in Fig. 3, miR-200b inhibitor in NG + MI could significantly elevate RECs viability (1.20 ± 0.15 vs. 1.00 ± 0.00, P < 0.05, Fig. 3A) and expressions of VEGF mRNA and protein (1.76 ± 0.27 vs. 1.00 ± 0.00, P < 0.05; 2.96 ± 0.39 vs. 1.00 ± 0.00, P < 0.01, Figs. 3B,C), but reduce VEGF miR-200b expression (0.57 ± 0.12 vs. 1.00 ± 0.00, P < 0.05, Fig. 3B) compared with NG. However, miR-200b mimic in HG + MM could significantly down-regulate RECs viability (0.95 ± 0.15 vs. 1.22 ± 0.10, P < 0.01, Fig. 3A) and VEGF mRNA and protein levels (0.94 ± 0.16 vs. 2.19 ± 0.58, P < 0.01; 1.59 ± 0.13 vs. 3.70 ± 0.35, P < 0.01, Figs. 3B,C), but up-regulate VEGF miR-200b expression (4.91 ± 1.00 vs. 0.45 ± 0.18, P < 0.01, Fig. 3B) compared with HG. There were no significant differences of TGF-β1 mRNA and protein expressions between NG and NG + MI. Also, there were no significant differences of TGF-β1 mRNA and protein expressions between HG and HG + MM. The expressions of TGF-β1 mRNA and protein were notably enhanced by high glucose but not regulated by miR-200b mimic or inhibitor (Figs. 3B,C). These results indicated that the high-glucose-induced activation of TGF-β1/miR-200b/VEGF pathway played a positive role in excessive proliferation of RECs.
3.5 Hyperoside regulated TGF-β1/miR-200b/VEGF pathway in high-glucose-cultured RECs
To elucidate the protective mechanism of hyperoside against RECs over-proliferation in HG, we analyzed the variation of TGF-β1, miR-200b, and VEGF levels in different groups. Compared with NG, HG could remarkably induce high mRNAs and proteins expressions of TGF-β1 and VEGF (3.08 ± 0.35 vs. 1.00 ± 0.00, P < 0.01; 1.80 ± 0.09 vs. 1.00 ± 0.00, P < 0.01 & 4.82 ± 1.08 vs. 1.00 ± 0.00, P < 0.01; 2.25 ± 0.16 vs. 1.00 ± 0.00, P < 0.01, Figs. 4A,B), and obviously inhibit miR-200b expressions (0.39 ± 0.13 vs. 1.00 ± 0.00, P < 0.05, Fig. 4A). However, hyperoside reversed the expressions of TGF-β1/miR-200b/VEGF under high glucose condition. As presented in Fig. 4, the low concentration of hyperoside in HG + H100 could significantly inhibit mRNA and protein levels of TGF-β1 and VEGF (1.54 ± 0.14 vs. 3.08 ± 0.35, P < 0.01; 1.39 ± 0.08 vs. 1.80 ± 0.09, P < 0.01 & 2.82 ± 0.43 vs. 4.82 ± 1.08, P < 0.01; 1.92 ± 0.09 vs. 2.25 ± 0.16, P < 0.05, Figs. 4A,B), and up-regulate miR-200b levels (1.06 ± 0.10 vs. 0.39 ± 0.13, P < 0.05, Fig. 4A) compared with HG. The similar difference was also seen in comparisons of HG + H400 vs. HG and HG + H400 vs. HG + H100. There was a stronger regulation of TGF-β1/miR-200b/VEGF pathway when hyperoside concentration increased. Therefore, these data indicated that hyperoside could decrease TGF-β1 and VEGF, but increase miR-200b of RECs under high glucose condition in a dose-dependent manner.
3.6 Body mass and fasting blood glucose of rats in different groups
We observed BM and fasting blood glucose (FBG) of rats at different periods including at 4 weeks before STZ injection (-4 w), 0 week before STZ injection (0 w), 8 weeks after STZ injection (8 w), and 16 weeks after STZ injection (16 w). As shown in Fig. 5, there were no significant differences of BM and FBG at -4 w in all groups. After 4 weeks of high-fat diet, BM and FBG results in all DR groups (DR, DR + L-HY, DR + H-HY & DR + CD) were significantly higher than NC. At 8 w and 16 w, BM results in all DR groups were significantly lower than NC, while FBG results were significantly higher than NC. However, there were no significant difference of BM and FBG in DR, DR + L-HY, DR + H-HY or DR + CD groups at any time. It could be speculated that the primary effect of hyperoside and Calcium Dobesilate was not aimed at hyperglycemia.
3.7 Hyperoside improved pathology of retinal tissue
To further analyze the possible role of hyperoside on retinal injury in diabetic rats, we observed the pathological changes of retinal tissues in NC, DR, DR + L-HY, DR + H-HY and DR + CD groups. As shown in Fig. 6A, in NC group, the ganglion cell layer (GCL) in retinal tissue was orderly, and the inner nuclear layer (INL) and outer nuclear layer (ONL) were closely arranged. In DR group, the arrangement of GCL was disordered, and INL and ONL appeared sparser and less compact than NC. However, these pathological changes were alleviated in DR + L-HY group, and improved in DR + H-HY and DR + CD groups as marked by black arrows in Fig. 6A. These results indicated that hyperoside could alleviate retinal tissue damage in DR rats.
3.8 Hyperoside improved retinal vasculopathy
To confirm the improvement of hyperoside on diabetic retinal injury, we also evaluated the effect of hyperoside on retinal vessels in DR rats. Retinal trypsin digest and retinal vascular staining were performed in NC, DR, DR + L-HY, DR + H-HY and DR + CD groups. In NC group, the retinal capillaries distributed regularly with smooth vascular branches, uniform diameter and very few acellular capillaries. In DR group, the retinal capillary network remained dense, disorderly and twisted, with uneven diameter, more acellular capillaries (AC), more ghost-pericytes (GP), and proliferative RECs. However, these retinal vasculopathy were alleviated in DR + L-HY group, and improved in DR + H-HY and DR + CD groups as shown in Fig. 6B. These preliminary results showed the improvement of hyperoside on diabetic retinal vasculopathy.
3.9 Comparisons of RVQ and E/P to evaluate the degrees of retinal vasculopathy
As mentioned above, RVQ and E/P were further calculated to quantitatively assess the alleviation of hyperoside in diabetic retinal vascular lesions. As shown in Table 1, RVQ and E/P in DR groups were significantly higher than NC (30.33 ± 3.83 vs. 14.83 ± 2.04, P < 0.01; 2.43 ± 0.22 vs. 1.03 ± 0.12, P < 0.01), but all RVQ and E/P in DR + L-HY, DR + H-HY and DR + CD groups were significantly lower than DR. In addition, RVQ and E/P in DR + H-HY were significantly lower than DR + L-HY (19.83 ± 2.32 vs. 24.50 ± 2.88, P < 0.05; 1.52 ± 0.18 vs. 1.83 ± 0.13, P < 0.05). These results indicated that hyperoside could alleviate retinal vasculopathy severity of DR rats, and the alleviation effect was more strengthened with the increase of hyperoside dose.
Table 1
Comparisons of RVQ and E/P of retinal capillaries in different groups
| NC (n = 6) | DR (n = 6) | DR + L-HY (n = 6) | DR + H-HY (n = 6) | DR + CD (n = 7) |
---|
RVQ | 14.83 ± 2.04 | 30.33 ± 3.83** | 24.50 ± 2.88## | 19.83 ± 2.32## Δ | 19.86 ± 1.57## |
E/P | 1.03 ± 0.12 | 2.43 ± 0.22** | 1.83 ± 0.13## | 1.52 ± 0.18## Δ | 1.27 ± 0.07## |
RVQ: retinal vascular quantity; E/P: endothelial cells to pericytes ratio. Data are presented as mean ± standard deviation. P < 0.05 is statistically significant. ** vs. NC, P < 0.01; ## vs. DR, P < 0.01; Δ vs. DR + L-HY, P < 0.05.
3.10 Hyperoside regulated TGF-β1/miR-200b/VEGF pathway in retinal tissues of DR rats
Finally, we compared the expressions of TGF-β1, miR-200b and VEGF in retinal tissues of all groups by qRT-PCR, WB and IF. The mRNAs and proteins expressions of TGF-β1 and VEGF in DR were significantly higher than NC (4.25 ± 0.72 vs. 1.00 ± 0.00, P < 0.01; 3.41 ± 0.39 vs. 1.00 ± 0.00, P < 0.01 & 3.97 ± 0.51 vs. 1.00 ± 0.00, P < 0.01; 4.93 ± 0.53 vs. 1.00 ± 0.00, P < 0.01, Figs. 7A,B), while the miR-200b in DR was significantly lower than NC (0.19 ± 0.07 vs. 1.00 ± 0.00, P < 0.01, Fig. 7A). However, hyperoside reversed the expressions of TGF-β1/miR-200b/VEGF in DR group. As shown in Fig. 7A,B, the low-dose hyperoside (in DR + L-HY), high-dose hyperoside (in DR + H-HY), and Calcium Dobesilate (in DR + CD) all significantly inhibited mRNAs and protein levels of TGF-β1 and VEGF, and up-regulated miR-200b levels compared with DR groups (0.48 ± 0.10 vs. 0.19 ± 0.07, P < 0.05; 0.77 ± 0.14 vs. 0.19 ± 0.07, P < 0.01; 1.09 ± 0.08 vs. 0.19 ± 0.07, P < 0.01, Fig. 7A). The comparison between DR + L-HY and DR + H-HY further showed that the stronger regulations of TGF-β1, VEGF, and miR-200b were along with the increase of hyperoside dose. Meanwhile, the similar differences of TGF-β1 and VEGF in all groups were also shown in Fig. 7C. In DR group, the green fluorescence intensities representing TGF-β1 and VEGF expression levels were significantly stronger than for NC group. However, the green fluorescence intensity of DR + L-HY group was significantly weaker than for DR group. In DR + H-HYand DR + CD group, the green fluorescence intensities were further weakened. Therefore, these data suggested that hyperoside could down-regulate TGF-β1 and VEGF, but up-regulate miR-200b in retinal tissues of DR rats.