Vasohibin-1 rescues erectile function through up-regulation of angiogenic factors in the diabetic mice

Neovascularization of the erectile tissue emerges as a beneficial curative approach to treat erectile dysfunction (ED). Here we for the first time report the unexpected role of vasohibin-1 (VASH1), mainly known as an anti-angiogenic factor, in restoring erectile function in diabetic mice. A diabetic patient has lower cavernous VASH1 expression than in the potent man. VASH1 was mainly expressed in endothelial cells. There were significant decreases in cavernous endothelial cell and pericyte contents in VASH1 knockout mice compared with those in wild-type mice, which resulted in impairments in erectile function. Intracavernous injection of VASH1 protein successfully restored erectile function in the diabetic mice (~ 90% of control values). VASH1 protein reinstated endothelial cells, pericytes, and endothelial cell–cell junction proteins and induced phosphorylation of eNOS (Ser1177) in the diabetic mice. The induction of angiogenic factors, such as angiopoietin-1 and vascular endothelial growth factor, is responsible for cavernous angiogenesis and the restoration of erectile function mediated by VASH1. Altogether, these findings suggest that VASH1 is proangiogenic in diabetic penis and is a new potential target for diabetic ED.

Neovascularization of the erectile tissue emerges as a beneficial curative approach to treat erectile dysfunction (ED). Here we for the first time report the unexpected role of vasohibin-1 (VASH1), mainly known as an anti-angiogenic factor, in restoring erectile function in diabetic mice. A diabetic patient has lower cavernous VASH1 expression than in the potent man. VASH1 was mainly expressed in endothelial cells. There were significant decreases in cavernous endothelial cell and pericyte contents in VASH1 knockout mice compared with those in wild-type mice, which resulted in impairments in erectile function. Intracavernous injection of VASH1 protein successfully restored erectile function in the diabetic mice (~ 90% of control values). VASH1 protein reinstated endothelial cells, pericytes, and endothelial cell-cell junction proteins and induced phosphorylation of eNOS (Ser1177) in the diabetic mice. The induction of angiogenic factors, such as angiopoietin-1 and vascular endothelial growth factor, is responsible for cavernous angiogenesis and the restoration of erectile function mediated by VASH1. Altogether, these findings suggest that VASH1 is proangiogenic in diabetic penis and is a new potential target for diabetic ED.

Ang1
Angiopoietin-1 Ang4 Angiopoietin www.nature.com/scientificreports/ some population of the diabetic patients 3 . The main pathophysiologic causes of diabetic ED are structural and functional derangements of cellular components of erectile tissue, such as pericytes, smooth muscle cells, and endothelial cells 4,5 . The degree of cavernous microangiopathy and a decrease in bioavailable endogenous nitric oxide (NO) from diabetes is related to low responsiveness to oral phosphodiesterase-5 inhibitors 6,7 . Vasculogenic ED demands an effective therapeutic approach to rebuild the structural and functional integrity of cavernous microvasculature. Previous animal studies revealed that local delivery of angiogenic factors, such as basic fibroblast growth factor (bFGF), angiopoietin-1 (Ang1), vascular endothelial growth factor (VEGF), and angiopoietin-4 (Ang4), rescued the erectile function in diabetic condition [8][9][10][11][12] . These findings provide us a proof-of-concept of therapeutic angiogenesis as a potential therapeutic strategy for diabetic ED.
Vasohibin-1 (VASH1) is originally discovered as an endothelium-derived inhibitor of angiogenesis. In the animal model of diabetic nephropathy and ischemic retinopathy, the upregulation of VASH1 blocked aberrant angiogenic activity induced by VEGF 13,14 . Meanwhile, VASH1 expression in tumor endothelial cells is positively correlated with microvascular density in a variety of cancers, such as cervical cancer 15 , colorectal cancer 16 , renal cell carcinoma 17 , and prostate cancer 18 , which suggests that VASH1 enhances angiogenesis in the tumor environment.
Vascular complications in the diabetic nephropathy model were more severe in VASH1 heterozygous knockout (VASH1 +/− ) mice than in wild-type mice 19 . Moreover, it was also reported in the cisplatin-induced acute kidney injury model that the peri-tubular capillary loss and the increase in serum creatinine concentrations were more prominent in VASH1 +/− mice than in wild-type mice 20 . These findings suggest that endogenous VASH1 is required for the protection of renal blood vessels in pathologic conditions.
In this study, we studied function of VASH1 on cavernous vascular integrity and erectile function by using VASH1 homozygous-knockout (VASH1 −/− ) mice or the streptozotocin (STZ)-induced diabetic mice.

Results
Physiologic and metabolic parameters. Bodyweight, mean systolic blood pressure (MSBP), and blood glucose level of each experimental group were summarized in Supplementary Table 1-3. VASH1 is mainly expressed in cavernous endothelial cells. The cDNA microarray of mouse corpus cavernosum tissue was performed as a screening approach for the prediction of new therapeutic target(s) or candidate(s) to treat diabetic ED. We found that several growth factors, such as VASH1, angiopoietin-like protein 4, dickkopf3, and angiopoietin-like protein 7, are relatively abundantly expressed in the erectile tissue. Of those growth factors, we focused on VASH1 that shows exceedingly high expression in the corpus cavernosum (Fig. 1a).
Next, the differential expression of VASH1 in human erectile tissue was determined. Immunohistochemical staining revealed a reduction of VASH1 expression in the corpus cavernosum of the diabetic patient compared to the control group. Co-labeling of the cavernous tissue with antibodies to IB4 (an endothelial cell marker) and VASH1 demonstrated that VASH1 was co-localized mostly with cavernous endothelial cells (Fig. 1b).
Downregulation of VASH1 decreases cavernous endothelial cell and pericyte contents, and deteriorates penile erection. Immunohistochemical staining of cavernous tissue with antibody to an endothelial cell maker PECAM-1 or a pericyte marker neuron-glial antigen 2 (NG2) revealed that the cavernous endothelial cell and pericyte content was depressed significantly in VASH1-knockout mice compared with wildtype (WT) mice ( Fig. 2a-c). These results imply that VASH1 has a potential protective role in penile vasculature. Therefore, we analyzed the role of target deletion of VASH1 gene on penile erection. In order to analyze the physiological significance of the downregulation of VASH1, erectile function assessed by in vivo cavernous nerve stimulation. In line with the findings of immunohistochemistry, demonstrating a decline in cavernous endothelial cell and pericyte content in the VASH1-knockout mice, the proportion of maximal intracavernous pressure (ICP) to MSBP, or total ICP to MSBP was significantly lower in VASH1-knockout mice than in WT group ( Fig. 2d-f). VASH1 protein transfer restores erectile function in the diabetic mice. We further defined the impact of VASH1 protein treatment in diabetic mice. The ratios of maximal and total ICP to MSBP in PBStreated diabetic mice were profoundly decreased compared with that in the control group. The repeated intracavernous injections of VASH1 (4 μg/20 μL) protein (days − 3 and 0) significantly restored erectile function in the diabetic mice. However, VASH1 at a dosage of 1 μg/20 μL didn't show significant improvements in erectile function (Fig. 3). VASH1 protein transfer restores cavernous endothelial cell and pericyte contents in the diabetic mice. Immunofluorescent staining of cavernous tissue with antibodies to PECAM-1 and NG2 was performed in control or diabetic mice 2 weeks after treatment. We observed lower endothelial cell and pericyte contents in the PBS-treated diabetic group than in the control group. Intracavernous administration of VASH1 protein (days − 3 and 0; 4 μg/20 μL) completely restored cavernous endothelial cell and pericyte contents in the diabetic mice (Fig. 4).

VASH1 protein transfer restores cavernous endothelial cell to cell junction proteins in the diabetic mice.
Endothelial cells adhere to the adjacent ones by endothelial cell to cell junctional proteins, which is important to regulate vascular permeability 21 . Derangements in endothelial cell junctions are known to be a distinctive pathophysiologic mechanism involved in diabetic ED 22  www.nature.com/scientificreports/ junction proteins (occludin and claudin-5) in the cavernous tissue was decreased in the diabetic group treated with PBS compared with that in the control mice. VASH1 protein recovered cavernous endothelial cell to cell junction proteins in diabetic mice, which are equivalent to the levels of the age-matched controls ( Fig. 5a-c).

VASH1 protein transfer induces phosphorylation of endothelial nitric oxide synthase (eNOS)
in the diabetic mice. The phosphorylation of eNOS (Ser1177) is a crucial step to produce NO from endothelial cells 23 . eNOS phosphorylation was significantly impaired in the diabetic mice and VASH1 protein successfully induced phosphorylation of eNOS in the corpus cavernosum of diabetic mice ( Fig. 5d-g).

VASH1 protein transfer increases the expression of angiogenic factors in the diabetic mice.
We performed immunohistochemical staining and Western blot to examine whether the production of angiogenic factors was enhanced by VASH1 protein. The expression of Ang1 and VEGF was significantly lower in the corpus cavernosum of diabetic mice treated with PBS than in the control group. VASH1 protein increased the expression of Ang1 and VEGF in the diabetic mice (Fig. 6).

Inhibition of Ang1 and VEGF abolishes VASH1 protein-mediated angiogenesis and restoration of erectile function in the diabetic mice.
We further determined whether the inhibition of Ang1 or VEGF affects the beneficial effects of VASH1. Nerve-induced erectile function studies revealed that inhibition of Ang1 or VEGF with soluble Tie2-Fc or VEGF-trap diminished the improvement in the erectile function by VASH1 in the diabetic mice ( Fig. 7a-c). Furthermore, immunohistochemical staining of cavernous tissues www.nature.com/scientificreports/ demonstrated that the restoration of endothelial cell and pericyte content induced by VASH1 protein also significantly declined after treatment with soluble Tie2-Fc or VEGF-trap ( Fig. 7d-f).

Discussion
Here, we studied the role of VASH1 in a mouse model of diabetic ED. Local delivery of VASH1 protein into the corpus cavernosum of diabetic mice rescued erectile function by enhancing eNOS phosphorylation and by restoring endothelial cells, pericytes, and endothelial cell to cell junction proteins. The major mechanism accountable www.nature.com/scientificreports/ for VASH1-mediated promotion of cavernous angiogenesis and restoration of erectile function in diabetic mice is the induction of angiogenic factors, such as Ang1 and VEGF. It was reported that the expression of VASH1 is induced in endothelial cells by angiogenic factors at the site of angiogenesis 24 . However, VASH1 is also known to be detectable in endothelial cells in quiescent state 25 . Similar to this finding, we demonstrated the expression of VASH1 in cavernous endothelial cells. Moreover, the cavernous expression of VASH1 was profoundly downregulated in diabetic conditions.
In the present study, diminished cavernous endothelial area was found in the diabetic group than in the control group. Local administration of VASH1 protein into the corpus cavernosum completely reinstated cavernous endothelial cell content in the diabetic group. The previous research reported that knockdown of VASH1 in endothelial cells induces premature senescence through an increase in the production of reactive oxygen species and a decrease in antioxidant enzyme 26 . The senescent endothelial cells are more prone to cell death in pathological conditions in vitro, such as exposure to H 2 O 2 and serum deprivation 27 . On the contrary, overexpression of VASH1 in endothelial cells showed resistance to premature cellular senescence and cell death 26 , supporting the role of VASH1 in stress tolerance and vascular protection. www.nature.com/scientificreports/ The lack of pericyte coverage in microvessels is one of the major mechanisms responsible for diabetic angiopathies, such as diabetic retinopathy 28 and diabetic ED 5 . Although the effects of VASH1 on mural cells were described in tumor vessels 29 , i.e., the normalization of tumor vasculature by intensifying pericyte coverage, its role in the non-tumor blood vessel is still unraveled. In the present study, VASH1 protein also significantly restored pericyte content in the diabetic mice.
Endothelial cell-cell barrier-genesis is critical for regulating vascular permeability and plays a crucial function during blood vessel formation and maturation 21 . Premature endothelial cell senescence induced by knockdown of VAHS1 shows impairments in the integrity of cell-cell junctions 30 . Knockdown of VASH1 decreased the expression of zonula occludens-1, a tight junction protein, in endothelial cells and increased transmigration of Lewis lung carcinoma cells across the endothelial cell monolayer. Moreover, the overexpression of VASH1 rescued the expression of zonula occludens-1 30 . Comparable to these results, we demonstrated that intracavernous treatment with VASH1 protein significantly reestablished cavernous endothelial cell-cell junction proteins in the diabetic mice. However, it remains to be further elucidated how VASH1 regulates endothelial cell-cell junction integrity.
In this study, VASH1 restored the expression of Ang-1 and VEGF, which were downregulated in the diabetic penis. Similar to our finding, Ang1 expression was significantly decreased in the kidney of diabetic mice compared with the control mice and was further decreased in the diabetic VASH1 heterozygous knockout (VASH1 +/− ) mice 14 .
Although VEGF is a potent angiogenic factor, but by itself initiates leaky and unstable blood vessel formation 31 . In comparison, Ang1, the ligand for tyrosine kinase with immunoglobulin and epidermal growth factor homology domain-2 (Tie2), has a major role in vascular maturation and stabilization 32 . Ang1 also ameliorates VEGF induced inflammation while having an additive influence on angiogenesis 33 . We previously reported in an animal model of hypercholesterolemic ED that adenoviral-mediated combined Ang1 and VEGF gene therapy exerted a synergism on angiogenesis and on recovery of erectile function compared with that of either therapy alone 34 . Moreover, our recent study demonstrated in diabetic mice that synthetic Ang1 protein also restored the integrity of endothelial cell-cell junction 22 . These findings suggest that induction of angiogenic factors is one of the major mechanisms by which VASH1 induces cavernous angiogenesis and restores erectile function in the diabetic mice. Inhibition studies with soluble Tie2-Fc or VEGF-trap further supported the role of VASH1-mediated upregulation of angiogenic factors on diabetic ED. www.nature.com/scientificreports/

Conclusions
Our findings demonstrate an unanticipated and unique function of VASH1 in the diabetic ED. The intracavernous injection of VASH1 protein enhanced cavernous angiogenesis and restored erectile function by boosting angiogenic factors expression in diabetic mice. As far as we know, this is the first study demonstrating that VASH1 is proangiogenic in the diabetic penis. Further investigations are necessary to discover the efficacy of VASH1 in different disease models for ED.

Materials and methods
Study design. The primary aim of this study is to explore the role of VASH1 on cavernous vascular integrity and erectile function. For this purpose, we used VASH1 homozygous-knockout (VASH1 −/− ) mice or VASH1 protein. 1 M citrate buffer, pH 4.5) for 5 sequential days as described previously 35 . Animals were considered as diabetic if the nonfasting glucose levels are greater than 300 mg/dL. Eight weeks after diabetes was induced, the mice were anesthetized with intramuscular injections of ketamine (100 mg/kg) and xylazine (5 mg/kg) and placed supine on a thermoregulated surgical table. The penis was www.nature.com/scientificreports/ exposed by use of a sterile technique 34,35 . To examine the effectiveness of VASH1 protein, the mice were then distributed into four groups (N = 6 per group): age-matched controls and STZ-induced diabetic mice receiving sequential intracavernous injections of PBS (days − 3 and 0; 20 μL) or VASH1 protein (days -3 and 0; 1 µg or 4 µg in 20 μL of PBS, respectively). A 30-gauge insulin syringe was used to deliver PBS or VASH1 protein into the midportion of the corpus cavernosum. The incision was closed with 6-O Vicryl (polyglactin 910) sutures. Erectile function was evaluated by electrical cavernous nerve stimulation at 2 weeks after treatment. A separate group of animals was used for histologic examination and biochemical study.
Human corpus cavernosum tissue. Human corpus cavernosum tissue samples were obtained from a 21-year-old patient with congenital penile curvature who had normal erectile function during reconstructive penile surgery and a 56-year-old patient with diabetic ED during penile prosthesis implantation. All tissue donors provided informed consent, and the experiments were approved by the internal review board of Inha University.
Measurement of erectile function. The mice from each group were anesthetized with ketamine (100 mg/ kg) and xylazine (5 mg/kg) intramuscularly. Erectile function was measured as described previously 35 . Briefly, the bladder and prostate exposed through a midline abdominal incision. The major pelvic ganglion and cavernous nerve were identified posterolateral to the prostate on one side, and bipolar platinum wire electrodes were placed around the cavernous nerve for electrical stimulation. The penis skin was denuded and a 26-gauge needle filled with 250 U/mL of heparin inserted into one side of the corpus cavernosum for monitoring ICP with a Statham P23 pressure transducer connected to a computerized system for data acquisition (Biopac Systems, Goleta, CA, USA), which was interfaced to a personal computer for recording and data analysis 35 . Stimulation parameters were 5 V at a frequency of 12 Hz, a pulse width of 1 ms, and a duration of 1 min. During tumescence, the maximal ICP was recorded. The total ICP was determined by the area under the curve from the beginning of cavernous nerve stimulation to a point 20 s after stimulus termination. Systemic blood pressure was measured by using a noninvasive tail-cuff system (Visitech systems, Apex, NC, USA). The ratios of maximal ICP and total ICP (area under the curve) to MSBP were calculated to adjust for variations in systemic blood pressure 35 . Histological examinations. The penis tissue (N = 6 per group) was fixed in 4% paraformaldehyde for 24 h at 4 °C as described previously 10  Western blot analysis. Equal amounts of protein (40 µg per lane) were electrophoresed on sodium dodecylsulfate-polyacrylamide gels (8-15%), transferred to polyvinylidene difluoride membrane. After the gel transferred to a blot, razor blade and ruler were used to cut the blot into strips corresponding to the size of the protein of interest then probed with antibodies to phospho-eNOS (Ser1177, Cell Signaling; 1:1000), eNOS (BD biosciences, California, USA; 1:1000), Ang1 (Abcam; 1:1000), VEGF (Santa Cruz Biotechnology; 1:1000), or β-actin (Santa Cruz Biotechnology; 1:6000). The results were quantified by densitometry (N = 4 per group).