GSK3β Inhibition Reduced Vascular Calcification in Ins2Akita/+ Mice

Endothelial–mesenchymal transition (EndMT) drives the endothelium to contribute to vascular calcification in diabetes mellitus. In our previous study, we showed that glycogen synthase kinase-3β (GSK3β) inhibition induces β-catenin and reduces mothers against DPP homolog 1 (SMAD1) to direct osteoblast-like cells toward endothelial lineage, thereby reducing vascular calcification in Matrix Gla Protein (Mgp) deficiency. Here, we report that GSK3β inhibition reduces vascular calcification in diabetic Ins2Akita/wt mice. Cell lineage tracing reveals that GSK3β inhibition redirects endothelial cell (EC)-derived osteoblast-like cells back to endothelial lineage in the diabetic endothelium of Ins2Akita/wt mice. We also find that the alterations in β-catenin and SMAD1 by GSK3β inhibition in the aortic endothelium of diabetic Ins2Akita/wt mice are similar to Mgp−/− mice. Together, our results suggest that GSK3β inhibition reduces vascular calcification in diabetic arteries through a similar mechanism to that in Mgp−/− mice.

The Ins2 Akita/wt mouse results from the Akita mutation, which largely reduces mature insulin by disruption of the two disulfide bonds of A and B chains [16]. Ins2 Akita/wt mice become spontaneously diabetic at 3-4 weeks of age and are recognized as a model for type I diabetes mellitus (DM1) [17]. Previous studies have shown that Ins2 Akita/wt mice develop vascular calcification and provide not only a monogenic diabetic model but also a model of diabetic calcific vasculopathy [5].
Glycogen synthase kinase 3 (GSK3) is a serine/threonine kinase that is constitutively activated in unstimulated cells [18]. The activity of GSK3 is regulated by serine phosphorylation in response to extracellular signals [19]. GSK3 plays different roles in osteogenic and endothelial differentiation. GSK3 promotes osteogenic differentiation [20], and GSK3 deficiency disrupts the maturation of osteoblasts, resulting in the reduction of bone formation [21]. In contrast, GSK3 prevents endothelial differentiation, and inhibition of GSK3 promotes the differentiation, proliferation, and migration of ECs [22,23]. GSK3 has two isoforms, GSK3a and GSK3β. SB216763 is a small molecule compound that specifically inhibits the activity of the GSK3 isoforms in an ATP-competitive manner [24]. SB216763 has been commonly used to probe the function of GSK3 inhibition [18].
In a previous study, our results suggested that the aortic osteoblast-like cells were redirected back to endothelial differentiation by the SB216763 treatment in Mgp −/− mice [25]. We also showed that osteoblast-like cells with EC origin contributed to aortic calcification in Ins2 Akita/+ mice [7]. Here, we hypothesize that GSK3β inhibition ameliorates vascular calcification in diabetic Ins2 Akita/+ mice.

GSK3β Inhibition Reduced Aortic Calcification in Ins2 Akita/+ Mice
To determine whether GSK3β inhibition reduces aortic calcification in Ins2 Akita/+ mice, we treated Ins2 Akita/+ mice with SB216763 (5 µg/g daily) or saline control at 36 weeks of age for 4 weeks. Alizarin red staining showed a robust decrease of aortic calcification in the SB216763-treated Ins2 Akita/+ mice (Figure 1a,b). The quantification of total aortic calcium confirmed the reduction of calcium in the mice (Figure 1c). Immunoblotting of whole aortic tissues showed that SB216763 reduced the expression of osteogenic markers, osterix and osteocalcin (Figure 2a,b). Together, the results suggested that SB216763 reduced the calcification in diabetic Ins2 Akita/+ mice.   GSK3 inhibition reduces aortic calcification of Ins2 Akita/+ mice. (a) Alizarin red staining of aortas of Ins2 Akita/+ mice and wild-type mice treated with or without SB216763 (n = 8). (b,c) Calcified aortic area and total aortic calcium of Ins2 Akita/+ mice treated with or without SB216763 (n = 10). Scale bar, 50 µm. b was analyzed for statistical significance by unpaired 2-tailed Student's t-test. c was analyzed for statistical significance by ANOVA with post hoc Tukey's analysis. *** p < 0.001.

GSK3β Deletion Limited Aortic Calcification in Ins2 Akita/+ Mice
Previously, we performed lineage-tracing using Col1α1 CreERT2 mice and identified osteoblast-like cells in Mgp −/− aortic tissues [25]. We showed that osteoblast-specific deletion of GSK3β reduced aortic calcification in Mgp −/− mice [25]. To determine if osteoblast-specific deletion of GSK3β ameliorates vascular calcification in diabetes mellitus, we treated mice at 34 weeks of age with tamoxifen (75 mg/kg, daily) for 5 days to delete GSK3β in Figure 2. Immunoblotting (a) with densitometry analysis, and (b) of aortic tissues of Ins2 Akita/+ mice and wild-type mice treated with or without SB216763 (n = 3). The numbers represent each sample. OSX, osterix; OC, osteocalcin. Densitometry was analyzed for statistical significance by ANOVA with post hoc Tukey's analysis. *** p < 0.001.

GSK3β Deletion Limited Aortic Calcification in Ins2 Akita/+ Mice
Previously, we performed lineage-tracing using Col1α1 CreERT2 mice and identified osteoblast-like cells in Mgp −/− aortic tissues [25]. We showed that osteoblast-specific deletion of GSK3β reduced aortic calcification in Mgp −/− mice [25]. To determine if osteoblastspecific deletion of GSK3β ameliorates vascular calcification in diabetes mellitus, we treated mice at 34 weeks of age with tamoxifen (75 mg/kg, daily) for 5 days to delete GSK3β in osteoblast-like cells as previously described [25]. At 40 weeks of age, we examined the aortic tissues. Alizarin red staining showed reduced calcification in the Col1a1 CreERT2 GSK3β flox/flox Ins2 Akita/+ mice (Figure 3a,b). Total aortic calcium was also significantly decreased in the mice with GSK3β deletion (Figure 3c). Immunoblotting of whole aortic tissues revealed the reduction of osteogenic markers in Col1a1 CreERT2 GSK3β flox/flox Ins2 Akita/+ mice after GSK3β deletion (Figure 4a,b). The results suggested that osteoblast-specific deletion of GSK3β reduced the calcification in diabetic Ins2 Akita/+ mice.
osteoblast-like cells as previously described [25]. At 40 weeks of age, we examined the aortic tissues. Alizarin red staining showed reduced calcification in the Col1a1 CreERT2 GSK3β flox/flox Ins2 Akita/+ mice (Figure 3a,b). Total aortic calcium was also significantly decreased in the mice with GSK3β deletion (Figure 3c). Immunoblotting of whole aortic tissues revealed the reduction of osteogenic markers in Col1a1 CreERT2 GSK3β flox/flox Ins2 Akita/+ mice after GSK3β deletion (Figure 4a,b). The results suggested that osteoblast-specific deletion of GSK3β reduced the calcification in diabetic Ins2 Akita/+ mice.   The numbers represent each sample. Densitometry was analyzed for statistical significance by ANOVA with post hoc Tukey's analysis. *** p < 0.001.

GSK3β Inhibition Redirected Osteoblast-like Cells toward Endothelial Differentiation in Ins2 Akita/+ Mice
To determine if GSK3β inhibition directed EC-derived osteoblast-like cells in Ins2 Akita/+ mice to revert endothelial differentiation, we generated VE-cadherin creERT2 Rosa tdTomato Ins2 Akita/+ mice. At 18 weeks of age, we treated the mice with tamoxifen (75 mg/kg, daily) for 5 days to label the aortic ECs as previously described [25]. At 20, 30, and 40 weeks of age, we isolated tdTomato-positive aortic cells and examined the endothelial and osteogenic markers (Figure 5a). Real-time PCR showed a decrease of endothelial markers with an increase of osteogenic markers in the tdTomato-positive cells of Ins2 Akita/+ mice (Figure 5b). We treated the mice with SB216763 (5 µg/g daily) at 36 weeks of age for 4 weeks and isolated tdTomato-positive cells (Figure 5c). The results showed that SB216763 prevented the decrease of endothelial markers and inhibited the increase of osteogenic markers ( Figure  5d), suggesting that GSK3β inhibition directed EC-derived osteoblast-like cells back to endothelial differentiation in Ins2 Akita/+ mice.

GSK3β Inhibition Redirected Osteoblast-like Cells toward Endothelial Differentiation in Ins2 Akita/+ Mice
To determine if GSK3β inhibition directed EC-derived osteoblast-like cells in Ins2 Akita/+ mice to revert endothelial differentiation, we generated VE-cadherin creERT2 Rosa tdTomato Ins2 Akita/+ mice. At 18 weeks of age, we treated the mice with tamoxifen (75 mg/kg, daily) for 5 days to label the aortic ECs as previously described [25]. At 20, 30, and 40 weeks of age, we isolated tdTomato-positive aortic cells and examined the endothelial and osteogenic markers (Figure 5a). Real-time PCR showed a decrease of endothelial markers with an increase of osteogenic markers in the tdTomato-positive cells of Ins2 Akita/+ mice (Figure 5b). We treated the mice with SB216763 (5 µg/g daily) at 36 weeks of age for 4 weeks and isolated tdTomato-positive cells (Figure 5c). The results showed that SB216763 prevented the decrease of endothelial markers and inhibited the increase of osteogenic markers (Figure 5d), suggesting that GSK3β inhibition directed EC-derived osteoblast-like cells back to endothelial differentiation in Ins2 Akita/+ mice.

Discussion
This study provides evidence that GSK3β inhibition reduces vascular calcification in diabetes mellitus. The role of GSK3 in diabetes mellitus has been well investigated in recent studies. GSK3 activity was found to regulate insulin sensitivity, which directly affects glycogen synthesis and glucose metabolism [26]. Several signaling pathways are involved in these processes, such as the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) signaling pathway [26]. Interestingly, inhibition of GSK3 improves the activity of glycogen synthase and glucose uptake, pointing to GSK3 inhibition as a potential antidiabetic strategy [27]. However, the activation of GSK3 in diabetic vascular calcification has never been reported, although GSK3 has been shown to promote osteogenic differentiation and bone formation [20,21]. GSK3β is one of the GSK3 isoforms. Here, we find that limiting GSK3β reduces the calcification in diabetic mice and reveals that the GSK3β inhibitor SB216763 redirects osteoblast-like cells to endothelial differentiation, similar to our previous observations in Mgp −/− mice [25]. These findings might bring more attention to GSK3 inhibition as a strategy to limit diabetes and calcification.
EndMTs have been observed to contribute to vascular calcification in diabetes mellitus. Previous studies demonstrated that ECs lose their normal morphology but express mesenchymal stem cell markers to migrate through a degraded internal elastic lamina into the arterial media and contribute to calcification [6,7,10]. The studies showed that excess BMP activity induces a number of serine proteases, such as elastases and kallikreins, to activate Sry-box 2(Sox2) expression in ECs and trigger EndMTs toward osteogenic differentiation [28]. A recent study constructed a systematic screen to explore the possibility of re-directing osteoblast-like cells in vascular calcification back to endothelial differentiation. The GSK3β inhibitor SB216763 was identified to have this capacity and decreased vascular calcification in Mgp −/− mice [25]. In this study, the GSK3β inhibitor SB216763 also reduced vascular calcification in diabetic Ins2 Akita/+ mice. Our results suggest that GSK3β inhibition prevents EndMTs and reduces calcification in diabetes.

Discussion
This study provides evidence that GSK3β inhibition reduces vascular calcification in diabetes mellitus. The role of GSK3 in diabetes mellitus has been well investigated in recent studies. GSK3 activity was found to regulate insulin sensitivity, which directly affects glycogen synthesis and glucose metabolism [26]. Several signaling pathways are involved in these processes, such as the phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) signaling pathway [26]. Interestingly, inhibition of GSK3 improves the activity of glycogen synthase and glucose uptake, pointing to GSK3 inhibition as a potential antidiabetic strategy [27]. However, the activation of GSK3 in diabetic vascular calcification has never been reported, although GSK3 has been shown to promote osteogenic differentiation and bone formation [20,21]. GSK3β is one of the GSK3 isoforms. Here, we find that limiting GSK3β reduces the calcification in diabetic mice and reveals that the GSK3β inhibitor SB216763 redirects osteoblast-like cells to endothelial differentiation, similar to our previous observations in Mgp −/− mice [25]. These findings might bring more attention to GSK3 inhibition as a strategy to limit diabetes and calcification.
EndMTs have been observed to contribute to vascular calcification in diabetes mellitus. Previous studies demonstrated that ECs lose their normal morphology but express mesenchymal stem cell markers to migrate through a degraded internal elastic lamina into the arterial media and contribute to calcification [6,7,10]. The studies showed that excess BMP activity induces a number of serine proteases, such as elastases and kallikreins, to activate Sry-box 2(Sox2) expression in ECs and trigger EndMTs toward osteogenic differentiation [28]. A recent study constructed a systematic screen to explore the possibility of re-directing osteoblast-like cells in vascular calcification back to endothelial differentiation. The GSK3β inhibitor SB216763 was identified to have this capacity and decreased vascular calcification in Mgp −/− mice [25]. In this study, the GSK3β inhibitor SB216763 also reduced vascular calcification in diabetic Ins2 Akita/+ mice. Our results suggest that GSK3β inhibition prevents EndMTs and reduces calcification in diabetes.

Pre-Sorting of tdTomato Positive Cells
The pre-sorting of aortic tdTomato positive cells was performed as previously described [25]. The aortas were perfused briefly with dispase and enzymatically dispersed. Then, the aortas were dissected into small pieces and incubated for 45 min prior to cell isolation, fixation, staining, and FACS analysis.

Quantification of Aortic Calcium
Total aortic calcium was measured using a calcium assay kit (Bioassay) as previously described [25].

Alizarin Red Staining
Slides were stained with Alizarin red solution (2% Alizarin red in distilled water) for 2 min. Then, excess solution was removed. The sections were dehydrated in acetone, followed by acetone-xylene (1:1) solution. After that, the sections were cleared by xylene and mounted with mounting medium.

Lesion Quantification
The mice were euthanized, and then perfusion fixed with 10% buffered formalin via the left ventricle for 4 min. The proximal aorta was excised. The specimen was embedded in OCT (Tissue-Tek, Fisher Scientific, Waltham, MA, USA), frozen on dry ice, and stored at −80 • C until sectioning. Serial cryosections were prepared. From then on, every fifth 10-µm section was collected on poly-D-lysine-coated slides. Sections were stained with hematoxylin and Alizarin red. Slides were examined by light microscopy, and the lesion area was quantified with computer-assisted image analysis (Image-Pro Plus, Media Cybernetics, Bethesda, MD, USA).

Statistical Analysis
Data were analyzed for statistical significance by ANOVA with post hoc Tukey's analysis. The analyses were performed using GraphPad Instat ® , version 3.0 (GraphPad Software). Data represent mean ± SD. p < 0.05 was considered significant, and experiments were performed a minimum of three times.

Conclusions
Together, our results show the importance of GSK3 inhibition in diabetic vascular calcification and provide new information for developing new treatment strategies.