Table 1.
Dietary dosage of DIAVIT.
Fig 1.
DIAVIT prevents albuminuria and increased glomerular water permeability in diabetic db/db mice, whilst having no effect on blood glucose.
A) db/db mice develop increased blood glucose at 6–7 weeks of age (week 1 of treatment), compared to lean controls (*p<0.05), which is not affected by DIAVIT consumption (ns; p>0.05; n = 6–9 mice; Two-way ANOVA based on the average values over 14 weeks). B) When assessing the urinary albumin creatinine ratio (uACR), db/db mice develop progressive albuminuria at 10 weeks of age (week 4 of treatment), compared to lean controls (*p<0.05), which is significantly rescued by DIAVIT (†p<0.05; n = 6–9 mice; Two-way ANOVA based on the average values over 14 weeks). C) Glomeruli from db/db mice have an increased glomerular water permeability (LpA/Vi) when compared to lean control glomeruli (*p<0.05), which is significantly rescued in glomeruli from DIAVIT-treated db/db mice (†p<0.05; n = 6 mice, 15–25 glomeruli; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). D) Plasma creatinine levels remained unchanged between groups (ns; p>0.05; n = 6–9 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs).
Fig 2.
Diabetic db/db mice develop glomerular fibrosis, which is prevented in DIAVIT treated db/db mice.
A) Periodic Acid Schiff (PAS) staining indicated structural abnormalities in db/db glomeruli, including mesangial matrix expansion and the presence of vacuoles (scale bar 40 μm). These appeared to be less frequent in DIAVIT-treated db/db glomeruli. Trichrome blue staining showed an increase in fibrosis (blue collagen staining) in the db/db kidney cortex, which was lower in the DIAVIT-treated db/db kidneys (scale bar 40 μm). More specifically, immunofluorescence for collagen IV and fibronectin showed an increase in glomerular fibrosis in db/db mice, compared to lean controls, which was prevented by DIAVIT treatment of the diabetic mice (quantified in B; *p<0.05 lean vs db/db; †p<0.05 db/db vs db/db + DIAVIT; n = 4 mice; 15–20 glomeruli per mouse; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs; scale bar is 40 μm). C) Western blotting of protein from the kidney cortex confirms an increase in the protein expression of collagen IV and fibronectin in db/db mice, which is prevented with DIAVIT treatment. Analysis of the Western blots is summarised in D and E (*p<0.05 lean vs db/db; †p<0.05 db/db vs db/db + DIAVIT; n = 3–4 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs).
Fig 3.
Diabetic db/db mice develop an endothelial insult, which is prevented by DIAVIT treatment of diabetic db/db mice.
A) Immunofluorescence for nephrin and podocin showed no change in the expression of these podocyte markers between all groups (n = 4 mice; scale bar 40 μm). This was confirmed by Western blotting of protein extracted from the kidney cortex (B); analysis is summarised in (C) (ns; p>0.05; n = 3–4 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). D) Immunofluorescence for the endothelial marker PECAM-1 showed a reduction in PECAM-1 expression in db/db glomeruli, which was prevented when db/db mice were treated with DIAVIT; analysis shown in (E) (*p<0.05 lean vs db/db; †p<0.05 db/db vs db/db + DIAVIT; n = 3–4 mice; 14–20 glomeruli per mouse; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs; scale bar 40 μm). F) Western blotting for the endothelial marker VEGF receptor 2 (VEGFR2) showed a decrease in the db/db kidney cortex, further indicating endothelial loss, which was prevented, and even increased relative to controls, in DIAVIT-treated db/db mice. Analysis is shown in (G) (*p<0.05 lean vs db/db; †p<0.05 lean and db/db vs db/db + DIAVIT; n = 3–4 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs).
Fig 4.
DIAVIT protects against diabetes-induced glomerular ultra-structural changes.
A) Representative glomerular electron micrographs from lean, db/db, and db/db + DIAVIT mice. Diabetic db/db glomeruli shows evidence of MME, which is not apparent in lean or db/db + DIAVIT glomeruli. Diabetic db/db glomeruli developed an increased GBM width (B), decreased number of endothelial fenestrations (C) and podocyte foot processes (D) per μm length, and an increased average podocyte foot process width (E) compared to lean controls (*p<0.05 vs lean). DIAVIT prevented the changes to the GBM width (B), number of endothelial fenestrations per μm length (c), and average podocyte foot process width (E) in db/db + DIAVIT glomeruli (†p<0.05 vs db/db). DIAVIT had no effect on the podocyte foot processes per μm length (D) (*p<0.05 vs lean) (n = 3 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs).
Fig 5.
DIAVIT alters VEGF-A splicing to increase VEGF-A165b.
A) DIAVIT treatment (1 mg/ml) of podocytes for 48 hrs resulted in an increased protein expression of VEGF-A165b relative to total VEGF-A165 (quantified in B; *p<0.05; n = 3 biological repeats; T-test; A—the same blot was first probed with VEGF-A165b before stripping and reprobing with panVEGF-A). C) This switch in splicing to increase the VEGF-Axxxb/VEGF-Axxx ratio was also observed at the mRNA level (*p<0.05; n = 4 biological repeats; T-test). D) DIAVIT inhibited angiogenesis in HUVECs when plated on to Matrigel (*p<0.05 vs control), which was partially prevented when an antibody specific for VEGF-A165b was added to the treatment (†p<0.05 vs DIAVIT). Measurments were taken in the form of the relative number of branch points (Ei) and the relative tubule length (Eii) (n = 4; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). F) Diabetic db/db mice showed switches in the spicing of VEGF-A in the kidney cortex; VEGF-A120 and VEGF-A164 were increased relative to lean controls (Gi; *p<0.05 and *p<0.01, VEGF-A120 and VEGF-A164, respectively; all VEGF-A isoforms were detected on the same blot), whereas VEGF-A165b was down-regulated (Gii; *p<0.05). Treatment of the diabetic mice with DIAVIT resulted in no significant increases in VEGF-A120 (Gi); although no effect was observed on total VEGF-A164 expression compared to un-treated db/db mice, DIAVIT did cause a shift in splicing to up-regulate VEGF-A165b relative to VEGF-A164 (Gii; †p<0.05 vs diabetic; n = 3–6 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). H) Diabetes did not alter the expression of the podocyte marker and transcription factor WT1. However, DIAVIT did result in an increase in WT1 expression in the kidney cortex of db/db mice (I) (*p<0.05; n = 4 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs).
Fig 6.
Diabetic mice develop an increase in the activation of pro-angiogenic and pro-fibrotic factors in the kidney, which is prevented with DIAVIT treatment.
A) Western blotting of protein extracted from the kidney cortex shows an increased phosphorylation and expression of AktSer473 (A, B), an increase in the phosphorylation and expression of ERK1/2 (a, c), and an increase in the expression of COX-2 in diabetic db/db mice (A, D; phosphorylation *p<0.01; expression *p<0.05 vs lean). DIAVIT prevented the increased activation and expression of these factors in the diabetic mice (A-D; †p<0.05 vs diabetic; n = 3–8 mice; all proteins were detected on the same blot). In addition, diabetic db/db mice had an increased expression of p65-NFĸB in the kidney cortex, which was singifcantly rescued by DIAVIT treatment (E, F; p<0.05; n = 3–4 mice; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs).
Fig 7.
Delphinidin alters VEGF-A splicing to increase VEGF-A165b and decrease total VEGF-A expression.
A) Treatment of podocytes with delphinidin chloride (10 μg/ml) under normal glucose (NG; 5 mM glucose + 25 mM mannitol) and high glucose (HG; 30 mM glucose, 1 ng/ml TNFα, 1 ng/ml IL-6, and 100 nM insulin) for 48 hrs increased the protein expression of VEGF-A165b relative to total VEGF-A165 (quantified in B; *p<0.05 vs NG, †p<0.05 vs HG; n = 3 biological repeats; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs; A—the same blot was first probed with VEGF-A165b before stripping and reprobing with panVEGF-A). C) Under both NG and HG condition, delphinidin significantly decreased the protein expression of total VEGF-A165 (*p<0.05; n = 3 biological repeats; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). D) Analysis at the mRNA level shows an increase in the VEGF-Axxxb/VEGF-Axxx ratio after treatement with delphinidn (*p<0.05; n = 4 biological repeats; T-test). E) Delphinidin inhibited angiogenesis in HUVECs when plated on to Matrigel (*p<0.05 vs control). Addition of an antibody specific for VEGF-A165b did not alter the effect of delphinidin. Measurments were taken in the form of the relative number of branch points (Fi) and the relative tubule length (Fii) (n = 4; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). G) Cells pre-treated with DMSO/delphinidin were stimulated with NG/HG for 30 or 60 mins. After 30 min, there was no significant effect of delphinidin on the phosphorylation of SRSF6 or SRSF1 under either condition (H). After 60 min, delphinidin significantly increased the phosphorylation of SRSF6 after stimulation with HG (*p<0.05 vs NG and HG controls; n = 3 biological repeats; One-way ANOVA with Bonferroni post-hoc test for comparison between pairs). There was no significant effect on phospho-SRSF1 at 60 min.
Fig 8.
Flow diagram describing the potential mechanism of DIAVIT and Delphinidin in type II DN.
High glucose and an increase in oxidative stress (ROS; reactive oxygen species) in the kidney results in increased phosphorylation of Akt. In turn this leads to activation of p65-NFĸB, resulting in the increased expression of pro-angiogenic and pro-fibrotic factors, which feed back to further increase the activation of Akt. Akt activation also results in the phosphorylation of serine protein kinase 1 (SRPK1), leading to proximal splice site (PSS) selection in exon 8 of the VEGF-A gene (pro-angiogenic VEGF-A165). DIAVIT acts to inhibit the phosphorylation of Akt and increase the expression of the SRPK1 inhibitor WT1. Furthermore, delphinidin acts to switch VEGF-A splicing, promoting distal splice site selection through the activation of SRSF6. This results in a reduction in fibrosis and angiogenesis in the diabetic kidney through p65-NFĸB and VEGF-A splicing.