Angiotensin II Induces Aortic Rupture and Dissection in Osteoprotegerin-Deficient Mice.

Background The biological mechanism of action for osteoprotegerin, a soluble decoy receptor for the receptor activator of nuclear factor-kappa B ligand in the vascular structure, has not been elucidated. The study aim was to determine if osteoprotegerin affects aortic structural integrity in angiotensin II (Ang II)-induced hypertension. Methods and Results Mortality was higher (P<0.0001 by log-rank test) in 8-week-old male homozygotes of osteoprotegerin gene-knockout mice given subcutaneous administration of Ang II for 28 days, with an incidence of 21% fatal aortic rupture and 23% aortic dissection, than in age-matched wild-type mice. Ang II-infused aorta of wild-type mice showed that osteoprotegerin immunoreactivity was present with proteoglycan. The absence of osteoprotegerin was associated with decreased medial and adventitial thickness and increased numbers of elastin breaks as well as with increased periostin expression and soluble receptor activator of nuclear factor-kappa B ligand concentrations. PEGylated human recombinant osteoprotegerin administration decreased all-cause mortality (P<0.001 by log-rank test), the incidence of fatal aortic rupture (P=0.08), and aortic dissection (P<0.001) with decreasing numbers of elastin breaks, periostin expressions, and soluble receptor activator of nuclear factor-kappa B ligand concentrations in Ang II-infused osteoprotegerin gene-knockout mice. Conclusions These data suggest that osteoprotegerin protects against aortic rupture and dissection in Ang II-induced hypertension by inhibiting receptor activator of nuclear factor-kappa B ligand activity and periostin expression.

angiotensin II

receptor activator of nuclear factor‐kappa B ligand

Clinical Perspective

What Is New?

This study demonstrates that the osteoprotegerin−proteoglycan complex might contribute to the pathogenesis of aortic rupture and dissection and that the complex may be associated with the inhibition of receptor activator of nuclear factor‐kappa B ligand activity and periostin expression.

What Are the Clinical Implications?

The supplementation of osteoprotegerin could be useful for decreasing the risk of aortic rupture and dissection.

Aortic dissection occurs at a rate of 3.5 to 5 per 100 000 annually , and has a high in‐hospital overall mortality rate (22% for type A versus 14% for type B). Hypertension is the most important risk factor for increasing the incidence of aortic dissection after adjusting for age and sex. Fragmentation of elastic fibers and reduction in numbers of smooth muscle cells accompanied by an accumulation of proteoglycans are observed in dissected human aorta. Matrix metalloproteinases degrade elastic fibers that endow the aortic wall with resilience and collagen fibers that endow the wall strength. Proteoglycans with negatively charged glycosaminoglycan generate an interstitial swelling pressure that may be mechanically disruptive to fibrillar proteins. In addition, the numbers of heparin‐binding proteins binding to glycosaminoglycan might be sufficient to affect disease progression. , , However, it remains unknown which molecules binding to them might help initiate aortic rupture and dissection.

Osteoprotegerin is a soluble glycoprotein belonging to the tumor necrosis factor receptor superfamily and consists of 380 amino acids containing 7 domains: 4 cysteine‐rich N‐terminal domains (domains 1–4), 2 death‐domain homologous regions (domains 5 and 6), and a C‐terminal heparin‐binding domain (domain 7). Osteoprotegerin works as a decoy receptor for the receptor activator of nuclear factor‐kappa B ligand (RANKL) through domains 1 to 4 to inhibit binding to its receptor RANK, leading to regulation of osteoclasts maturation. , RANKL is mainly released from bone and primary/secondary lymphoid organs. Osteoprotegerin and RANK are widely distributed in vascular walls. , , , Deletion or loss of function of the osteoprotegerin gene leads to marked activation of RANKL signaling, which results in a deformity of bone condition from childhood named juvenile Paget disease. Some patients with this disorder show giant bilateral cavernous carotid artery aneurysms and angioid streaks by breaking the elastic fibers. , We hypothesized that osteoprotegerin not only regulates bone metabolism but also has a role in maintaining the structural integrity of vascular walls. To test our hypothesis, we administered a vasoconstrictive peptide, angiotensin II (Ang II), to the wild‐type (WT) and homozygotes of osteoprotegerin gene‐knockout (OPG−/−) mice to explore the pathological roles of osteoprotegerin in the vasculature.

METHODS

Ethical Considerations

All protocols for the animal experiments were reviewed and approved by the University of the Miyazaki Institutional Animal Care and Use Committee (approval number 2018‐516) and Genetic Modification Safety Committee of Miyazaki University (approval number 600). All studies were conducted in accordance with the National Institute of Health Guide for the Care and Use of Laboratory Animals (Revised 2015). The data that support the findings of this study are available from the corresponding author upon reasonable request.

Animal Experiments

Mice were housed in a temperature‐ and light‐controlled room (25°C±1°C; 12/12‐hour light/dark cycle) with free access to normal chow and water. OPG−/− mice (genetic background of C57BL/6) were generated by targeted disruption of the gene, and OPG−/− mice were bred with OPG−/− mice. WT mice (C57BL/6J) were purchased from Japan Clea Co. (Tokyo, Japan). The phenotype of each mouse was verified at every mating by polymerase chain reaction analysis of the tail DNA. Eight‐week‐old male OPG−/− mice and age‐matched WT mice (Japan Clea Co.) were used in this study. OPG−/− and age‐matched WT mice were anesthetized by injecting 0.75 mg/kg medetomidine, 4 mg/kg midazolam, and 5 mg/kg butorphanol intraperitoneally. Either Ang II (1000 ng/kg per minute dissolved in 0.9% saline) or vehicle (0.9% saline) was infused subcutaneously for 28 days by using the implanted mini‐osmotic pump (Alzet, Model 1004; DURECT Co.). In another setting of experiments, 10 mg/kg of human recombinant osteoprotegerin (hrOPG)/osteoclastogenesis inhibitory factor, which was (poly)PEGylated at the C‐terminus region to interact with heparin sulfate proteoglycan, dissolved in 10 mmol/L Na‐Pi, 0.15 mol/L NaCl, and 0.01% polysorbate 80 (pH 7.4) (provided by Daiichi‐Sankyo Co., Ltd, Tokyo, Japan) was administered intraperitoneally to the Ang II‐infused OPG−/− mice every 2 days from day 1 to 28. At day 28, we anesthetized the mice by injecting 0.75 mg/kg medetomidine, 4 mg/kg midazolam, and 5 mg/kg butorphanol intraperitoneally and removed the aorta. If the mice died during the experimental periods, we performed a necropsy to determine a possible cause of death. A finding of blood clots in the chest or abdominal cavity was judged to be aortic rupture, and transparent effusion in the chest cavity was judged to be heart failure. We defined aortic dissection in survived mice as the presence of blood clots microscopically in the false lumen.

Sample Collection

Blood

At day 28, blood samples were collected from the left ventricle of mice under anesthesia and mixed with 10 μL of 10 mg/mL EDTA‐2Na and 0.7 mg/mL aprotinin. The mixture was centrifuged at 830g at 4°C for 10 minutes. The plasma was then stored at −80°C until use.

Aorta

The mice aortae were harvested from the aortic arch to the common iliac artery bifurcation by perfusion fixation with 4% paraformaldehyde, followed by immersion in 4% paraformaldehyde overnight for the histological analysis, or perfused with phosphate‐buffered saline and immersed in RNAlater (Ambion) at 4°C overnight for polymerase chain reaction (PCR) array analysis.

Blood Pressure

Systolic blood pressure was measured 3 times by tail‐cuff plethysmography (BP‐98A; Softron, Tokyo, Japan) in a conscious situation at day 28. The results were averaged.

Morphological Analysis

The excised aortic tree was photographed, after which the external aortic diameters were measured at the aortic arch, thoracic aorta, suprarenal aorta, and abdominal aorta using ImageJ (National Institutes of Health, Bethesda, MD). An aneurysm was defined based on a ≥50% increase in the external aorta width compared with controls (WT mice, 0.76 mm; OPG−/− mice, 0.73 mm in diameter, mean value obtained from 11 and 6 sham mice, respectively).

Histology and Immunohistochemistry

The cross‐sections (4 µm) of the aortic arch, descending thoracic aorta, suprarenal aorta, and abdominal aorta were stained with hematoxylin–eosin, Sirius red (for collagen), Victoria blue (for elastin), and alcian blue (for proteoglycan). The semi‐quantitative deposition of collagen, elastin, and proteoglycan was evaluated by using WinROOF 2018 (Mitani Co., Tokyo, Japan). Images covering the entire field of the aortic sections were captured under the same lighting conditions at ×100 magnification (one image) for medial and adventitial thickness and ×200 magnification (4–6 images) for elastin, collagen, and alcian blue. All images were analyzed under the same threshold and averaged. The medial and adventitial thicknesses were calculated as mean values for the 4 measurements taken orthogonal to each other in the cross‐sections. The numbers of elastic lamina breaks (focal dissection) in the elastic layers were counted from the luminal to abluminal side , , and then summed. Histological evaluation was performed by 3 observers in a blind manner (T. T., Aya Kawano, Yukiko Kawagoe). To detect the immunoreactivity for osteoprotegerin, deparaffinized tissue sections were autoclaved at 121°C for 20 minutes in 10 mmol/L citrate buffer (pH 6.0). The staining protocol followed the manufacturer’s instructions (Anti‐Goat HRP‐DAB Cell & Tissue Staining Kit, catalog number CTS008, R&D Systems). In brief, tissue sections were immersed in 3% hydrogen peroxide for 5 minutes and incubated with serum blocking reagent for 15 minutes. Then, the sections were incubated with avidin for 15 minutes and then with biotin blocking reagents for 15 minutes. Tissue sections were incubated with the polyclonal primary antibody against mouse osteoprotegerin (5 µg/mL, catalog number AF459, R&D Systems) in Can Get Signal immunostaining solution A (TOYOBO, Osaka, Japan) at 4°C overnight. The slides were incubated with biotinylated secondary antibody for 30 minutes, followed by incubation with high sensitivity streptoavidin conjugated to horseradish peroxidase for 30 minutes. The immunoreactivity was visualized with 0.05% 3, 3′‐diaminobenzidine containing hydrogen peroxide and counterstained with hematoxylin. The slide sections were dehydrated in xylene and coverslipped. We scanned the slides at ×40 magnification using an Olympus BX53F microscope (Olympus, Tokyo, Japan). Negative control staining was performed by omitting the first antibody (data not shown).

Micro‐Computed Tomography

After sacrificing the mouse, the right‐lower extremity was taken and examined by micro‐computed tomography scan (ScanXmate‐L090H; Comscantecno, Kanagawa, Japan), as previously described. , The trabecular bone microstructure of the tibiae was analyzed by using a 3‐dimensional image analysis system (TRI/3D‐BON; Ratoc System Engineering Co. Ltd., Tokyo, Japan). We examined the secondary trabecular area at the proximal tibia metaphysis (2.0‐mm trimming) and determined the bone volume/tissue volume (%).

RANKL Concentration

The RANKL concentration was measured in mouse plasma using Quantikine ELISA kit (R & D Systems, Minneapolis, MN, USA).

RT2 Profiler PCR Array Assay and Volcano Plot Analysis

The whole aortic tree was pulverized in Isogen (Nippon Gen), and total RNA was extracted using RNeasy Mini Kit (QIAGEN, Hilden, Germany). Thereafter, 1 µg of total RNA was used to synthesize complementary DNA using RT2 First Strand Kit (QIAGEN), and Extracellular Matrix and Adhesion Molecules RT2 profiler PCR array (QIAGEN) was performed in a 96‐well plate using a Real‐Time PCR System (ABI Prism 7300 Sequence Detector, Applied Biosystems) according to the manufacturer’s protocol. Data were presented as (∆CT=CT in each gene of interest − average CT in reference gene) in an Excel‐based PCR Array data template provided by QIAGEN and CT in the reference gene was determined by averaging β‐actin, β2‐microglobin, GAPDH, and glucuronidase β. The distribution of gene expression fold changes by plotting the logarithm and P values by plotting the negative logarithm were analyzed by Free web‐based RT2 Profiler PCR Array Data Analysis software and presented as volcano plots.

Real‐Time Quantitative PCR

A total of 3 µg of total RNA was used as a template for synthesizing complementary DNA with Invitrogen SuperScript II reverse transcriptase. Thereafter, complementary DNA was amplified using oligonucleotide primers and probes labeled with 6‐carboxy‐fluorescein as reporter fluorescence and with 6‐carboxy‐tetramethyl‐rhodamine as quencher fluorescence via real‐time quantitative PCR. The oligonucleotide sequences of probes and primers for mouse periostin were purchased from Thermo Fisher Scientific (TaqMan Gene Expression Assays; Mm01284919_m1). The gene expression levels were normalized relative to the level of 18S ribosomal RNA.

Statistical Analysis

Data were analyzed in GraphPad Prism 8 (La Jolla, CA, USA). Two variables were compared by performing Student t‐test (effects of PEGylated hrOPG treatment), and multiple variables were compared by performing 2‐way analysis of variance (genotype×Ang II), followed by the Bonferroni post‐hoc test. Survival rate was analyzed by log−rank test. The incidence of aortic rupture and dissection were analyzed using Chi‐square test in Excel 2010 software. Size of samples and detailed statistical tests were described in figure legends. We presented the data as means±SEM, and P<0.05 was considered significant.

RESULTS

OPG−/− Mice Administered Ang II Displays High Mortality Attributable to Aortic Rupture

OPG−/− mice showed severe osteoporosis up to adolescence. , Subcutaneous Ang II administration to these mice fed with normal chow promoted higher mortality relative to that of other groups over the 28‐day experimental period (P<0.0001 by log‐rank test, Figure 1A). The causes of death were pleural hemorrhage, peritoneal hemorrhage, heart failure, and undetermined in 4, 6, 1, and 4 mice, respectively. Representative images of fatal ruptured aorta with hemorrhage outside the vessel wall are shown in Figure 1B. The incidences of fatal aortic rupture (Chi‐square, 8.024, P=0.005, Figure 1C) and aortic dissection (Chi‐square, 4.285, P=0.038, Figure 1D) were greater in OPG−/− mice than in WT mice administered with Ang II. Ang II infusion promoted significantly greater aortic diameter at the suprarenal aorta in WT and OPG−/− mice than that in sham mice (Figure 1E), with no difference in the incidence of aneurysm formation (WT mice, 5/15; OPG−/− mice, 7/11, Chi‐square, 2.345, P=0.126). Among the surviving mice, Ang II raised systolic blood pressure equivalently at day 28 in WT and OPG−/− mice (P=0.512 by unpaired t‐test) (Figure 1F).

Figure 1

Survival, fatal aortic rupture, aortic dissection, aortic diameter, and systolic blood pressure in wild‐type (WT) and OPG−/− mice with and without angiotensin II (Ang II) administration for 28 days.

A, Survival rate. WT mice (Sham), n=42; WT mice (Ang II), n=33; OPG−/− mice (Sham), n=23; OPG−/− mice (Ang II), n=47. B, Example of an OPG−/− mouse administered Ang II that died because of a fatal aortic rupture (B). The aorta shown in the square area was cut vertically and magnified under a microscope. Bar, 200 µm. H indicates hemorrhage; and L, lumen. C and D, Incidence of fatal aortic rupture (C), and aortic dissection (D) in the surviving mice at day 28. Numbers in the panel indicate the animal numbers used (denominator) and those with aortic rupture or aortic dissection (numerator). The Chi‐squared test gave a value of 8.024, P=0.005 for aortic rupture, and 4.285, P=0.038 for aneurysm formation in the WT and OPG−/− mice stimulated by Ang II. E, Aortic diameter at the aortic arch, (descending) thoracic aorta, suprarenal aorta, and abdominal aorta at 28 days with and without Ang II administration in WT and OPG−/− mice. F, Systolic blood pressure at 28 days with and without Ang II administration in WT and OPG−/− mice (Sham: WT, n=28, OPG−/−, n=16; Ang II: WT, n=21, OPG−/−, n=30). Data are presented as a dot plot with means±SEM and analyzed by 2‐way analysis of variance, followed by the Bonferroni post‐hoc test. Ang II indicates angiotensin II; OPG−/−, osteoprotegerin gene‐knockout; and WT, wild‐type. ***P<0.001.

Histological Alterations in OPG−/− and WT Mice Administered Ang II

Ang II infusion showed increased media thicknesses at the descending thoracic aorta and suprarenal aorta of WT and OPG−/− mice, which was greater in former than in the latter OPG−/− mice (Figure 2A through 2D). In addition, Ang II infusion promoted increased adventitia thicknesses at the suprarenal aorta in WT and OPG−/− mice, which was greater in the former than in the latter (Figure 2E through 2H). Ang II infusion decreased the percentage of collagen (Figure 2I through 2L) occupying the media to a similar extent between the WT and OPG−/− mice. Elastin occupying the media decreased in OPG−/− regardless of Ang II infusion (Figure 3A through 3D). However, Ang II infusion promoted a greater number of elastin breaks at the suprarenal aorta in OPG−/− mice than in WT mice (Figure 3E through 3H). Ang II infusion tended to increase the area for proteoglycan at the medial layer of aortic arch, thoracic aorta, and suprarenal aorta in WT and OPG−/− mice (Figure 3I through 3L). Figure 4A through 4H shows a representative collagen (Figure 4A and 4E), elastin (Figure 4B and 4F), proteoglycan (Figure 4C and 4G), and osteoprotegerin (Figure 4D and 4H) in WT and OPG−/− mice administered Ang II. The intense staining of osteoprotegerin was not limited to the inner side of aortic wall (Figure 4D) but throughout the medial layer (Figure S1A through S1H), and the distribution was consistent with that of proteoglycan (Figure 4C) in WT mice. Osteoprotegerin immunoreactivity shown by proteoglycan was absent in OPG−/− mice (Figure 4G and 4H).

Figure 2

Histological characteristics of wild‐type (WT) and osteoprotegerin gene‐knockout (OPG−/−) mice with and without angiotensin II (Ang II) infusion at 28 days.

A through D, Media thickness (aortic arch, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=5, OPG−/−, n=8; thoracic aorta, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=8; suprarenal aorta, Sham: WT, n=11, OPG−/−, n=11; Ang II: WT, n=13, OPG−/−, n=16; abdominal aorta, Sham: WT, n=4, OPG−/−, n=3; Ang II: WT, n=9, OPG−/−, n=8); (E through H) adventitia thickness (aortic arch, Sham: WT, n=5, OPG−/−, n=3; Ang II: WT, n=8, OPG−/−, n=3; thoracic aorta, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=7, OPG−/−, n=8; suprarenal aorta, Sham: WT, n=11, OPG−/−, n=11; Ang II: WT, n=14, OPG−/−, n=17; abdominal aorta, Sham: WT, n=4, OPG−/−, n=3; Ang II: WT, n=8, OPG−/−, n=8); (I through L) percent collagen (aortic arch, Sham: WT, n=5, OPG−/−, n=5; Ang II: WT, n=9, OPG−/−, n=9; thoracic aorta, Sham: WT, n=5, OPG−/−, n=5; Ang II: WT, n=9, OPG−/−, n=9; suprarenal aorta, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=9; abdominal aorta, Sham: WT, n=5, OPG−/−, n=2; Ang II: WT, n=9, OPG−/−, n=8). Data are presented as a dot plot with means±SEM and analyzed via 2‐way analysis of variance, followed by the Bonferroni post‐hoc test. Ang II indicates angiotensin II; OPG−/−, osteoprotegerin gene‐knockout; and WT, wild‐type. *P<0.05, **P<0.01, ***P<0.001, ****P<0.001.

Figure 3

Histological characteristics of wild‐type (WT) and osteoprotegerin gene‐knockout (OPG−/−) mice with and without angiotensin II (Ang II) infusion at 28 days.

A through D, Percent elastin (aortic arch, Sham: WT, n=4, OPG−/−, n=5; Ang II: WT, n=4, OPG−/−, n=10; thoracic aorta, Sham: WT, n=4, OPG−/−, n=4; Ang II: WT, n=4, OPG−/−, n=10; suprarenal aorta, Sham: WT, n=4, OPG−/−, n=4; Ang II: WT, n=4, OPG−/−, n=10; abdominal aorta, Sham: WT, n=4, OPG−/−, n=4; Ang II: WT, n=4, OPG−/−, n=10); (E through H) elastin breaks (aortic arch, Sham: WT, n=5, OPG−/−, n=2; Ang II: WT, n=9, OPG−/−, n=3; thoracic aorta, Sham: WT, n=6, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=9; suprarenal aorta, Sham: WT, n=4, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=7; abdominal aorta, Sham: WT, n=5, OPG−/−, n=3; Ang II: WT, n=10, OPG−/−, n=8); (I through L) alcian blue (aortic arch, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=9; thoracic aorta, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=9; suprarenal aorta, Sham: WT, n=5, OPG−/−, n=4; Ang II: WT, n=9, OPG−/−, n=9; abdominal aorta, Sham: WT, n=4, OPG−/−, n=2; Ang II: WT, n=6, OPG−/−, n=6). Data are presented as a dot plot with means±SEM and analyzed via 2‐way analysis of variance, followed by the Bonferroni post hoc test. *P<0.05, **P<0.01, ***P<0.001. Ang II indicates angiotensin II; OPG−/−, osteoprotegerin gene‐knockout; and WT, wild‐type.

Figure 4

Representative images of collagen (A and E), elastin (B and F), proteoglycan (C and G), and osteoprotegerin (D and H) are shown.

The asterisks in (F) indicate the part of elastin break (focal dissection). Triangles in (C and G) indicate the distribution of proteoglycan assessed by alcian blue; bar 50 µm. OPG−/− indicates osteoprotegerin gene‐knockout; and WT, wild‐type.

Increased Periostin Expression in Ang II‐Infused OPG−/− Mice

To elicit the candidate molecule to link the RANKL and osteoprotegerin, we used PCR array analysis, focusing on the extracellular matrix and adhesion molecule in the aortae of WT and OPG−/− mice with and without Ang II administration. In the sham‐treatment groups, the aortae of OPG−/− mice were characterized by increasing thrombospondin 1 (2.59‐folds, P=0.0002), matrix metalloproteinase 8 (2.63‐folds, P=0.045), and osteopontin (2.9‐folds, P=0.039) levels and by decreasing collagen type III, α1 (−2.94‐folds, P=0.026) levels compared with WT mice (Figure 5A). In the 28‐day‐Ang II treatment group, periostin expression was 2.5‐fold higher (P=0.007) in OPG−/− mice than in WT mice (Figure 5B). Real‐time quantitative PCR confirmed that the magnitude of increase in periostin expression was higher in OPG−/− mice than in WT mice (Figure 5C). All the genes identified by this analysis are listed in Table S1.

Figure 5

Volcano plots graphs of mouse aorta assesed by polymerase chain reaction array.

Gene expressions between osteoprotegerin gene‐knockout (OPG−/−) and wild‐type (WT) mice groups without (A) and with (B) angiotensin II infusion (Sham: WT, n=4, OPG−/−, n=4; Ang II: WT, n=4, OPG−/−, n=6). Fold‐regulation threshold >2, and P<0.05 from the Student t‐test. Each gene >2 was highlighted in red, whereas <2 was highlighted in blue. Thbs1, thrombospondin 1; Mmp8, matrix metalloproteinase 8; Spp1, osteopontin; Col3a1, collagen type III, α1; Postn, periostin expression. The expression was normalized by averaging the Ct in the reference gene (β‐actin, β2‐microglobulin, and GAPDH). C, Real‐time quantitative PCR analysis for periostin in WT mice and OPG−/− mice without and with Ang II infusion for 28 days (Sham: WT, n=4, OPG−/−, n=7; Ang II: WT, n=6, OPG−/−, n=6). Data are presented as dot plot with means±SEM and analyzed via 2‐way analysis of variance, followed by the Bonferroni post‐hoc test. Ang II indicates angiotensin II; OPG−/−, osteoprotegerin gene‐knockout; and WT, wild‐type. *P<0.05.

Recombinant Osteoprotegerin Administration Reversed All‐Cause Mortality, Aortic Rupture, and Aortic Dissection in OPG−/− Mice

PEGylated hrOPG administered to Ang II‐stimulated OPG−/− mice (n=39) promoted a decrease in all‐cause mortality (−22%), aortic rupture (−13%), and aortic dissection (−23%) relative to those in mice not receiving the same treatment (Figure 6A). The administration did not affect systolic blood pressure (Figure 6B, P=0.0931). The treatment increased bone volume/tissue volume at the tibia metaphysis (Figure 6C, P<0.0001) and decreased the concentration of soluble RANKL (Figure 6D, P=0.0012) and the periostin expression (Figure 6E, P=0.0011). During morphological analysis, the administration decreased the aortic diameter at the suprarenal aorta (Figure 6F, P<0.0001), the thickness of the adventitia at the aortic arch (Figure 6H, P=0.0296) and suprarenal aorta (Figure 6H, P=0.0274), and the numbers of elastin breaks (Figure 6J, P=0.0539) at the suprarenal aorta. However, the administration did not affect the media thickness (Figure 6G) or elastin occupying the media (Figure 6I).

Figure 6

(A) Effects of poly(PEG)‐human recombinant osteoprotegerin (hrOPG) treatment on the incidence of all‐cause mortality, aortic rupture, and aortic dissection; (B) systolic blood pressure; (C) bone volume/tissue volume at the tibial metaphysis; (D) soluble receptor activator of nuclear factor‐kappa B ligand concentration; and (E) periostin expression; (F) aortic diameter; (G through J) histological analyses (G, media thickness; H, adventitia thickness; I, percent elastin; J, numbers of elastin breaks) in angiotensin II‐infused osteoprotegerin gene‐knockout mice.

Ten mg/kg of PEGylated human recombinant osteoprotegrin (hrOPG) was administered intraperitoneally to the Ang II‐infused osteoprotegerin gene‐knockout mice every 2 days from day 1 to 28. A, Numbers in the panel indicate the animal numbers used (denominator) and the incidences of all‐cause mortality, aortic rupture, or aortic dissection (numerator). The Chi‐square test was used for statistical analysis. B through J, Data are presented as a dot plot with means±SEM without and with PEGylated hrOPG treatment. Data are analyzed by Student t‐test. BV/TV indicates bone volume/tissue volume; SBP, systolic blood pressure; Postn, periostin expression; and RANKL, receptor activator of nuclear factor‐kappa B ligand.

DISCUSSION

Our data suggested that osteoprotegerin plays an important role in maintaining the structural integrity of the aorta in mice with Ang II‐induced hypertension. The incidences of aortic rupture and dissection were greater in OPG−/− mice than in WT mice. However, recombinant osteoprotegerin supplementation in OPG−/− mice reversed the increase in morbidity, supporting the evidence that osteoprotegerin functions to protect the aortic structure. Ex‐vivo analysis demonstrated that the suprarenal aorta selectively increased the diameter mostly because of the dissection of the aneurysm in WT and OPG−/− mice. However, a hematoma was found in the chest cavity at equivalent prevalence to peritoneal hemorrhage in OPG−/− mice. The initiation of the medial tear and the subsequent adventitial dissection occurred near the intercostal orifice of the thoracic aorta. Micro‐computed tomography scan used for bone morphometrical analysis in this study did not reach the visualizing fine vascular structure. Moreover, we could not determine the exact dissecting and rupture sites, and our ex‐vivo histological feature was not validated via micro‐computed tomography imaging.

Degradation of elastin in the aorta is associated with increase in the wall thickness and stiffness with aging. The passive mechanical force and other chemical alterations enhanced the fragility of elastin, and alteration in the smooth muscle cell phenotype contributed to the pathogenesis of aortic rupture and dissection. In this experiment, Ang II‐infused OPG−/−mice showed exaggerated elastin fragmentation at the suprarenal aorta regardless of using thinner media when compared with WT mice, suggesting that inappropriate pulsatile wall strain may cause elastin fragmentation mechanically. Moreover, the effects of osteoprotegerin deficiency on factors weakening the aortic wall may involve the following points. We previously reported that the soluble RANKL concentration was elevated in OPG−/−mice. The present study showed that recombinant osteoprotegerin administration to these mice decreased the RANKL concentration. RANKL stimulates the activity of elastin degrading enzyme, matrix metalloproteinase‐9, , which might explain the increase in elastin breaks in OPG−/− mice. In support of our findings, Tanaka et al reported that neutralization with the RANKL antibody inhibited the dissecting aneurysm in Ang II‐ infused apolipoprotein E‐deficient mice. In this study, Ang II‐infused OPG−/− mice significantly increased periostin mRNA expression, with values greater than that observed in the other groups. By contrast, hrOPG supplementation decreased periostin mRNA expression. Bonet et al reported a novel mechanism of RANKL to increase the bone fragility in a periostin‐dependent fashion. Periostin is a matricellular protein that regulates extracellular matrix formation and is upregulated in human abdominal aortic aneurysm. Our data suggested that RANKL links to periostin alters the vascular structure when osteoprotegerin is deficient. Adventitial layer consists predominantly of type I and III collagens and contributes in the prevention of aortic rupture at high blood pressures, which is accompanied by changing the type I and III ratio. , , The newly synthesized type III collagen, which is susceptible to proteolysis, is increased in human abdominal aortic aneurysm. OPG−/− mice shows the discordant expression of type III collagen in homogenates of whole aorta with its histological distribution at the adventitia, suggesting the distinct synthesis and metabolism in 3 layers of the aortic wall. The Ang II‐infused OPG−/− mice demonstrated a trend of higher collagen type III than WT mice, however, it did not reach at significant difference of collagen type I and III ratio (Figure S2). Meanwhile, this study shows that the adventitial layer was thinner at the suprarenal aorta in the Ang II‐infused OPG−/− mice than in the WT mice. Thus, structural alternation at the adventitial layer may be another cause of aortic rupture induction in Ang II‐infused OPG−/− mice.

Osteoprotegerin was positively stained in areas where proteoglycan is distributed in the aortic wall of Ang II‐infused WT mice. Osteoprotegerin binds to proteoglycan via its heparin‐binding domain (domain 7) , , , , and activates focal adhesion kinase and extracellular signal‐regulated kinase, , , , which are crucial for vascular smooth muscle cell survival. , Moreover, recombinant osteoprotegerin stimulates the proliferation and extracellular matrix synthesis independent of RANKL activity in vascular smooth muscle cells. , , , It remains inconclusive whether osteoprotegerin plays a protective role in rodent models of abdominal aortic aneurysm. , , , Vorkapic et al reported that human recombinant full‐length osteoprotegerin (0.24–0.36 mg/kg, subcutaneous administration with implanted mini‐pump for 28 days) did not affect the development of abdominal aortic aneurysm in Ang II‐infused apolipoprotein E‐deficient mice. To address whether this was a matter of the concentration of recombinant osteoprotegerin interacting with the heparin‐binding domain and/or it depended on rodent models to develop atherosclerosis, we used the modification of hrOPG with PEGylation at the C‐terminus region (heparin‐binding domain), which failed to interact with heparin sulfate proteoglycan. Thus, our results suggested that hrOPG prevented the rupture of mice aorta by blocking the bioactivity of RANKL in Ang II‐infused non‐atherosclerosis mice. We used the experimental OPG−/− data from the first half of the study as reference for hrOPG administration. Accordingly, we confirmed the efficacy of hrOPG by decreasing the circulating RANKL concentration and increasing the bone density; however, we acknowledged that the inadequate vehicle control greatly reduced the robustness and reproducibility of our evidence. Only a few studies have investigated the association between osteoprotegerin and aortic dissection in humans; therefore, further studies are needed to clarify the role of osteoprotegerin as a biomarker and therapeutic target for aortic rupture and dissection.

In summary, the study findings indicated that osteoprotegerin might protect against aortic rupture and dissection in Ang II‐induced hypertension by inhibiting RANKL activity and periostin expression.

Sources of Funding

This study was supported by grants‐in‐aid for Scientific Research (26461076, 19K08521 to Tsuruda) from the Japan Society for the Promotion of Science, and Clinical Research from Miyazaki University Hospital.

Disclosures

None.

Table S1

Click here for additional data file.

Figures S1–S2

Click here for additional data file.