Runx2 (Runt-Related Transcription Factor 2)-Mediated Microcalcification Is a Novel Pathological Characteristic and Potential Mediator of Abdominal Aortic Aneurysm
Abdominal aortic aneurysms (AAAs) are highly lethal diseases without effective clinical predictors and therapeutic targets. Vascular microcalcification, as detected by fluorine-18-sodium fluoride, has recently been recognized as a valuable indicator in predicting atherosclerotic plaque rupture and AAA expansion. However, whether vascular microcalcification involved in the pathogenesis of AAA remains elusive.
Approach and Results:
Microcalcification was analyzed in human aneurysmal aortas histologically and in AngII (angiotensin II)-infused ApoE−/− mouse aortas by fluorine-18-sodium fluoride positron emission tomography and X-ray computed tomography scanning in chronological order in live animals. AAA patients’ aortic tissue showed markedly enhanced microcalcification in the aortic media within the area proximal to elastic fiber degradation, compared with non-AAA patients. Enhanced fluorine-18-sodium fluoride uptake preceded significant aortic expansion in mice. Microcalcification-positive mice on day 7 of AngII infusion showed dramatic aortic expansion on subsequent days 14 to 28, whereas microcalcification-negative AngII-infused mice and saline-induced mice did not develop AAA. The application of hydroxyapatite, the main component of microcalcification, aggravated AngII-induced AAA formation in vivo. RNA-sequencing analysis of the suprarenal aortas of 4-day-AngII–infused ApoE−/− mice and bioinformatics analysis with ChIP-Atlas database identified the potential involvement of the osteogenic transcriptional factor Runx2 (runt-related transcription factor 2) in AAA. Consistently, vascular smooth muscle cell–specific Runx2 deficiency markedly repressed AngII-induced AAA formation in the ApoE−/− mice compared with the control littermates.
Our studies have revealed microcalcification as a novel pathological characteristic and potential mediator of AAA, and targeting microcalcification may represent a promising strategy for AAA prevention and treatment.
Using real-time fluorine-18-sodium fluoride positron emission tomography and X-ray computed tomography scanning in chronological order in live aneurysmal mice, we observed microcalcification preceded significant aortic expansion in ApoE−/− mice.
Hydroxyapatite augments AngII (angiotensin II)-induced abdominal aortic aneurysms, accompanied by profound macrophage infiltration.
Vascular smooth muscle cell–specific Runx2 (runt-related transcription factor 2) deletion inhibited AngII-induced microcalcification and abdominal aortic aneurysm formation in the ApoE−/− mice.
Abdominal aortic aneurysm (AAA) is defined as the irreversible expansion of the abdominal aorta to over 1.5× the adjacent vessel diameter or ≥30 mm. AAA is usually asymptomatic unless ruptured. Once ruptured, the mortality rate is as high as 85% to 90%.1 Although a growing range of surgical options for AAA has emerged, thus far, there is no effective drug therapy available due to limited understanding of the pathogenesis of AAA. In parallel, the diagnosis of AAA and the indications for surgical intervention are exclusively dependent on the maximum aortic diameters monitored by ultrasonography or computed tomography (CT) angiography, although we now know that the pathogenesis and progression of AAA are nonlinear and that AAA rupture does not infrequently occur in small aneurysms below 55 mm in diameter.2 Thus, we need to thoroughly understand the pathological characteristics and underlying molecular mechanisms of AAA formation, particularly during the early disease stage.
Vascular calcification, that is, the aberrant deposition of calcium-phosphate crystals in the vessel wall as an actively regulated process, contributes directly to cardiovascular morbidity and mortality among aging individuals and patients with chronic renal failure, atherosclerosis or diabetes mellitus.3–5 Nevertheless, the relationship between vascular calcification and AAA remains elusive. To date, a limited number of retrospective studies indicate that the calcium score determined by CT scans is positively correlated with AAA expansion and rupture risk,6,7 whereas controversy also exists.8–10 Recently, a prospective study using fluorine-18-sodium fluoride (18F-NaF) positron emission tomography and X-ray CT (PET-CT) imaging of atherosclerosis plaques identified microcalcification (<50 μm) as being associated with subsequent high-risk plaque rupture in patients.11,12 Compared with macrocalcification (≥50 μm), microcalcification generally reflects the earliest and most active nanocrystalline stages of mineralization associated with inflammation and necrosis, which cannot be detected by conventional CT scans.13–16 Interestingly, the latest SoFIA3 study (Sodium Fluoride Imaging of Abdominal Aortic Aneurysms) also identified 18F-NaF uptake localized to areas of active microcalcification in AAA patients, and the degree of 18F-NaF uptake was positively associated with aneurysm expansion.17 Therefore, the present studies sought to determine whether microcalcification promotes the AAA formation or simply coexists in the AAA lesions. By monitoring the progression of AngII (angiotensin II)-induced AAA in mice with 18F-NaF PET-CT, we uncovered that microcalcification predicts the aneurysmal expansion. Furthermore, we demonstrated that smooth muscle cell (SMC)–specific ablation of Runx2 (runt-related transcription factor 2) abrogated microcalcification and inhibited AngII-induced AAA formation in the ApoE−/− mice. These results suggest that microcalcification is a novel pathological characteristic and possibly a potential mediator of AAA, and thus, targeting microcalcification may represent a promising strategy for AAA prevention and treatment.
Materials and Methods
The data that support the findings of this study are available from the corresponding author on reasonable request.
AngII (A9525) and Pluronic F-127 gel (P2443) were purchased from Sigma Aldrich (MO). [18F]Fluoride was obtained from HTA, Co, Ltd (Beijing, China). Hydroxyapatite (MB3039) was purchased from Dalian Meilun Biotechnology, Co, Ltd. Antibodies against β-actin (4967), Runx2 (8486), NLRP3 (NLR family pyrin domain-containing 3; 15101), and Caspase1 (2225) were from Cell Signaling Technology (Boston, MA). The antibody against CD45 (550539) was purchased from BD Biosciences (San Diego, CA). Antibodies against CD68 (ab125212) and NLRP3 (ab214185) for immunohistochemical staining were from Abcam (Cambridge, United Kingdom).
Human AAA specimens from the anterior region of the aortic wall were obtained from 20 patients undergoing open aortic repair at Fuwai Hospital, National Center for Cardiovascular Diseases, Beijing, China. The patients were diagnosed with AAA by repeated ultrasonography or CT angiography. Non-AAA aortic tissues were resected from aortic roots of 19 patients who underwent heart transplantation for dilated cardiomyopathy, hypertrophic cardiomyopathy, or restrictive cardiomyopathy and did not have an AAA. Specimens from patients with collagen vascular disease or severe chronic kidney disease (estimated glomerular filtration rate<30 mL/[min·1.73 m2]) were excluded. After being harvested, the aortic tissues were immediately placed in formalin, embedded in paraffin, and made into serial sections (7 μm thick) for further histological analysis. The study was conducted in accordance with the ethical guidelines of the 1975 Declaration of Helsinki, and the study protocol was approved by the Scientific and Ethics Committee of the Fuwai Hospital. Informed consent was obtained from all study participants. All experiments were performed in accordance with relevant guidelines and regulations.
All animal studies followed the guidelines of the Animal Care and Use Committee of Peking University. As male mice are more susceptible than female mice to AAA, male mice were used for AAA induction.18 C57BL/6J mice (male, 8–10 weeks) and ApoE−/− mice (male, 4 months) were obtained from the Department of Laboratory Animal Science, Peking University Health Science Center.
The SMC-specific Runx2-deficient ApoE−/− mice were generated as previously described.19 In short, Runx2 exon 8 floxed mice with a C57BL/6J background were bred with SM22α (Tagln [transgelin])-Cre transgenic mice in a C57BL/6J background to generate smooth muscle-specific Runx2-deficient mice (Runx2Δ/Δ SMC). Runx2Δ/Δ SMC mice were further crossed with ApoE−/− mice to generate Runx2Δ/Δ SMC:ApoE−/− mice and the Runx2f/f:ApoE−/− littermates. Primers for genotyping were as follows: Runx2: F-5′-GCAAGATCATGACTAGGGATTG-3′ and R-5′-ATCAGTTCCCAATGGTACCCG-3′; SM22α-Cre: F-5′-GCGGTCTGGCAGTAAAAACTATC-3′ and R-5′-GTGAAACAGCATTGCTGTCACTT-3′; ApoE: F-5′-GCCTAGCCGAGGGAGAGCCG-3′, R-5′-TGTGACTTGGGAGCTCTGCAGC-3′ and R-5′-GCCGCCCCGACTGCATCT-3′.
AngII-Induced AAA in Mice
Mouse with AAA was induced by AngII, as previously described.20 After the mice were anesthetized, ALZET osmotic minipumps (Model 2004) were filled with saline or AngII solution and subcutaneously implanted into 4-month-old ApoE−/− male mice. AngII was delivered at a rate of 1000 ng/(kg·min) for 28 days. Systolic blood pressure was obtained using a CODA noninvasive blood pressure system (tail-cuff method, Kent Scientific Corporation) before and after pump implantation. The maximal suprarenal aortic diameter was measured by the Vevo 770 ultrasound system (VisualSonics, Toronto, Canada) as well as ex vivo measurement of dissected aortas. AAA was defined as an increase of 50% or more in the maximal suprarenal aorta width compared with that of its adjacent intact portion of aorta.21
Micro–PET-CT Imaging in Mice
PET-CT images were obtained using a micro–PET-CT scanner (Inveon PET-CT, Siemens, Germany). Approximately 10 MBq of 18F-NaF in 0.1 mL saline via the tail vein was injected for micro–PET-CT imaging.22 At 10 minutes postinjection, the animal was placed into a chamber connected to an isoflurane anesthesia unit. Anesthesia was induced using an airflow rate of 1.0 L/min and ≈3.0% isoflurane. After induction of anesthesia, the animal was immediately placed supine on the scanning bed. The airflow rate was then reduced to ≈0.8 L/min with ≈1.5% isoflurane. The mice were transferred into the attached CT scanner and imaged using the magnification low acquisition settings (acquisition: Step-and-shoot; projection: 180; estimated scan time: 504 s; binning: 4×4; effective pixel size: 105.3 μm; transaxial field of view: 53.9 mm; axial scanning length: 134 mm; voltage: 80 kV; current: 500 μA; filter: 0.5 mm; exposure: 145 ms). After CT acquisition, the mice were placed near the center of the field of view of the PET scanner (Axial scan length: 127 mm) and then allowed to recover in a lead-shielded cage. The regions of interest (ROIs) from PET and CT were selected based on orthogonal optical images provided by the integrated webcams.
Image Reconstruction and Data Processing
Reconstruction of PET data was performed using an OSEM3D/SP-MAP (shifted poisson-maximum a posteriori) algorithm (OESM3D: 2 iterations; SP-MAP: 18 iterations; matrix size: 128×128; image zoom: 1). CT data were reconstructed using a Feldkamp algorithm (Inveon Research Workplace, Siemens). After reconstruction, the PET and CT data were coregistered according to the movement of the robotic stage and resampled to equivalent voxel sizes. A 3-dimensional-Gaussian filter (1.0 mm FWHM [the full width at half maximum]) was applied to smooth noise, and the look-up tables were adjusted for good contrast. The reconstructed images were visualized as sagittal and transversal slices. ROIs were carefully drawn according to the CT images. Anatomic ROIs were drawn around the area of suprarenal aorta (the area between 2 kidneys and <2 mm upon the spine). The mean standardized uptake value (SUVmean) of ROIs was calculated by 2 independent investigators in a blinded manner. For better visualization of regions of microcalcification, PET images were further processed by AMIDE software with using hot metal contour in the colormap tool and thresholding SUV levels (SUV, 0.5–1.3).23
Analysis of Plasma Lipids
Mouse plasma samples were collected by centrifugation from whole blood containing EDTA. Total plasma cholesterol and triglyceride levels were measured with the kits from Zhong Sheng Biotechnology (Beijing, China).
After the mice were euthanized and perfused with 4% paraformaldehyde, the suprarenal abdominal aortas and descending aortas were dissected and prepared as serial cryosections (7 μm thick, 300 μm apart).
An elastic Van Gieson staining kit (BASO Precision Optics, Ltd, Taiwan, China) was used for elastin assessment. Elastin degradation was described as 4 grades: 1, <25% degradation; 2, 25% to 50% degradation; 3, 50% to 75% degradation; or 4, over 75% degradation. The data are presented as the mean of 16 serial sections from each mouse.
Von Kossa staining was performed to analyze the microcalcification in aortic tissue. Briefly, the aortic sections were incubated with 5% silver nitrate for 30 minutes and then exposed to UV light for 20 minutes. To remove unreacted silver, the sections were incubated with 5% sodium thiosulfate for 5 minutes. Nuclei were counterstained by hematoxylin. Stained sections were imaged using an Olympus microscopy system. For calcification/microcalcification quantification, the area of von Kossa staining for each section was analyzed by ImageJ with IHC-Toolbox. The content was presented as the percentage of the area of each section.
To visualize chondrocyte-like cells in abdominal aortic tissue, Movat pentachrome staining kit (Leagene Biotechnology, Beijing, China) was used on the frozen sections of abdominal aortas. Stained sections were imaged using an Olympus microscopy system. Chondrocyte-like cells were defined as large amount of clear cytoplasm surrounded by a lacuna.24,25
Oil Red O staining was applied to assess atherosclerotic plaques. In short, the cryosections of abdominal aorta were incubated with Oil Red O solution for 30 minutes. Nuclei were counterstained by hematoxylin. Images were captured by the Olympus microscopy system.
Abdominal Ultrasonography in Mice
The in vivo ultrasonography of abdominal aortas was performed by the Vevo 770 ultrasound system with a 30 MHz linear transducer (VisualSonics, Toronto, Canada). Mice were anesthetized by 2% isoflurane and placed on a heating pad at 37°C. After removing the hair from abdomen by depilatory cream, mice were transferred and fixed to the heated stage in the supine position. Then warmed ultrasound transmission gel was applied to the abdominal skin. Long-axis ultrasound scans of suprarenal aortas were performed from the aortic hiatus to the left renal artery. Images of abdominal aorta were captured at the midsystole, and >10 dynamic sagittal images were acquired. The maximal aortic internal diameters were measured by 2 independent investigators in a blinded manner.
Hydroxyapatite Application in Mice
Needle-shaped hydroxyapatite (from Dalian Meilun Biotechnology, Co, Ltd, Dalian, China), with a mean size of 0.1 μm, was suspended in 30% Pluronic F-127 gel at a concentration of 0.25 M. After AAA surgery, the abdominal aorta was isolated. Thirty microliters of gel solution were applied periadventitially on the suprarenal aorta and soon became gelatinous. Then, the mice were sutured and recovered for further application.
Microarray Analysis of Mouse Aortas
The suprarenal aortas of 4 groups of mice (saline, saline+hydroxyapatite, AngII, and AngII+hydroxyapatite) were harvested and placed in liquid nitrogen immediately. After cryopulverization by a Biopulverizer (Biospec) and homogenization by Mini-Bead-Beater-16 (Biospec), total RNA was extracted using TRIzol (Invitrogen, Life Technologies), labeled by a Quick Amp Labeling Kit, One-Color (Agilent) and purified by an RNasey Mini Kit (Qiagen). An Agilent Gene Expression Hybridization Kit (Agilent) was used for the hybridization of purified RNA. After microarray washing by Gene Expression Wash Buffer 1 (Agilent) and Gene Expression Wash Buffer 2 (Agilent), the microarrays underwent scanning with an Agilent Microarray Scanner (Agilent). The data were extracted using Agilent Feature Extraction Software.
Quantitative Real-Time Polymerase Chain Reaction Analysis
Total RNA was extracted, and equal amounts (2 μg) were reverse transcribed into cDNA. The SYBR Green 2× polymerase chain reaction (PCR) mix (TransGen Biotech, Beijing, China) was used for quantitative real-time PCR amplification. All samples were normalized to β-actin. The primer sequences for quantitative real-time PCR are shown in Table IX in the Data Supplement.
Western Blot Analysis
Cell or tissue lysates were resolved by 10% SDS-PAGE for Western blot analysis and then transferred onto nitrocellulose membranes. After being blocked with 5% milk in TBST (tris-buffered saline Tween-20), the blots were incubated with primary antibodies at 4°C overnight, followed by secondary antibody incubation for 1 hour. The immunofluorescence images were obtained by Odyssey infrared imaging (LI-COR Biosciences, Lincoln, NE).
The ELISA kits for IL (interleukin)-1β and IL-6 were purchased from R&D (MN). The suprarenal aortic rings of 4 groups of mice (saline, saline+hydroxyapatite, AngII, and AngII+hydroxyapatite) were cultured in DMEM at 37°C in 5% CO2 for 48 hours. The culture media were collected to measure the level of IL-1β by using an ELISA kit. The levels of IL-6 and IL-1β in mouse plasma samples were also measured by ELISA kits.
RNA Sequencing of AngII-Treated Aortas of ApoE−/− Mice
ALZET osmotic minipumps (Model 2001) were filled with saline (n=6) or AngII solution (n=5) and implanted into ApoE−/− male mice. After 4 days of infusion, the aortas were dissected. Using the GeneJet RNA Purification Kit (K0732, Thermo Scientific), total RNA of the aortas was extracted. Using the TruSeq RNA Library Prep Kit V2 (RS-122-2002, Illumina), 1 μg of RNA was used to generate the sequencing library. The libraries were sequenced on the NextSeq500 sequencer (FC-404-2005, Illumina) according to the 35-nucleotide paired-end sequencing protocol.
Bioinformatics Analysis of RNA Sequencing
For RNA-sequencing analysis of AngII-treated aortas of ApoE−/− mice, the clean reads were mapped to mouse reference genome mm9 by Subread aligner tool (R-based, version 1.24.2),26 and the normalized expression values were quantified by limma.27 The differentially expressed genes (DEGs) were identified by requiring log2FC (the base-2 logarithmic expression value of AngII-treated aorta minus that of saline-treated control aorta) value <−0.75 or >0.75. Next, the enriched functions and transcription factor regulations among the DEGs were analyzed by comparison with the function and disease association annotations from gProfiler (https://biit.cs.ut.ee/gprofiler)28 and transcription factor binding site annotations from ChIP-Atlas (http://chip-atlas.org/), respectively. For transcription factor regulation identification, both human and mouse databases in ChIP-Atlas were used. The target gene of a given transcription factor was identified if one of the transcription factor’s ChIP-seq peaks settled within the gene’s promoter region (5 kb around the transcription start site). For enrichment analysis, the significant overlap of each transcription factor’s target genes with the DEGs was further assessed by Fisher exact test followed by Benjamini-Hochberg correction. The significantly overlapping transcription factor was selected if its corrected P value was <0.05 in both human and mouse databases.
Adeno-Associated Virus-Mediated Runx2 Interference In Vivo
Small interfering RNA (siRNA) targeting the mouse Runx2 mRNA was designed with the sequence of sense (5′-3′), 5′-GGCAAGAGUUUCACCUUGATT-3′, and antisense, 5′-UCAAGGUGAAACUCUUGCCTT-3′. Serotype 1 adeno-associated virus (AAV1) carrying the sequence (AAV1-shRunx2 [short hairpin RNA against Runx2]) or the scramble control (AAV1-shNC [short hairpin RNA against negative control]) was generated by Hanbio Biotechnology (China). The abdominal aorta of 4-month-old male ApoE−/− mice was isolated, and 30 μL of F-127 gel (30%) solution containing viruses at 5×1011 vector genomes were applied periadventitially on the suprarenal aorta. Seven days later, mice underwent saline or AngII infusion. The knockdown efficiency of AAV1-shRunx2 was determined by the decreased protein levels of Runx2 in the suprarenal aortas (Figure VB in the Data Supplement).
Culture of Rat Vascular SMCs
Primary vascular SMCs (VSMCs) were isolated from thoracic aortas of 150 to 180 g male Sprague-Dawley rats by collagenase digestion as previously described.29 VSMCs were cultured in DMEM supplemented with 10% FBS at 37°C in 5% CO2. Primary VSMCs at passages 3 to 6 were applied for all experiments.
siRNA against Runx2 was designed by GenePharma (Shanghai, China). The scrambled siRNA served as a negative control. The sequences of siRNA for Runx2 rat genes were sense, 5′-AAATGCCTCTGCTGTTATGAACCTGTCTC-3′, and antisense, 5′-AATACGGAGACGACAATACTTCCTGTCTC-3′. The scramble siRNA sequences were sense, 5′-UUCUCCGAACGUGUCACGUTT-3′, and antisense, 5′-ACGUGACACGUUCGGAGAATT-3′. VSMCs cultured in 5well plates were transfected with 20 nmol/L siRNA per well using Lipofectamine RNAiMAX Reagent (Invitrogen, CA).
For immunohistochemical staining, cryosections of mouse suprarenal aortas were incubated with 3% hydrogen peroxide. After blocking with 5% normal blocking serum, frozen sections were incubated with primary antibodies against CD68 (1:200; Abcam), NLRP3 (1:200; Abcam), or CD45 (1:50; BD Biosciences) overnight at 4°C. Sections incubated with rabbit IgG (for CD68 and NLRP3) or rat IgG (for CD45) alone were used as negative controls. After incubation with a horseradish peroxidase–labeled anti-rabbit IgG secondary antibody (for CD68 and NLRP3) or anti-rat IgG secondary antibody (for CD45), sections were stained with a DAB (3,3′-diaminobenzidine) Kit (ZSGB-BIO, Beijing, China). Nuclei were counterstained with hematoxylin.
All results are presented as mean±SD. Statistical analyses were performed using GraphPad Prism 8.0 software (GraphPad Software, San Diego, CA). For statistical comparisons, we first evaluated whether the data were normally distributed using Shapiro-Wilk normality test. Then, we applied Student t test, with Welch correction if equal SDs are not assumed through an F test, for 2-group comparisons of normally distributed data. In addition, the Brown-Forsythe test was used to assess equal variances among data from >2 groups; we applied ordinary ANOVA or Welch ANOVA for equal variances assumed or not, respectively. Nonparametric tests were used when the data were not normally distributed. In all cases, a statistically significant difference was present when the 2-tailed probability was <0.05. The details of the statistical analysis applied to each experiment are presented in the corresponding figure legends. The Fisher exact test was performed to compare the incidence of microcalcification between AAA-positive, AAA-negative, and saline-induced mice; the incidence of AAA between microcalcification-positive, microcalcification-negative, and saline-induced mice; and the incidence of AAA Runx2f/f:ApoE−/− and Runx2Δ/Δ SMC:ApoE−/− mice with saline or AngII induction. The Mann-Whitney test was applied to analyze the difference in calcified area between AAA tissues and non-AAA tissues in patients and the difference of CD68 positive area in adventitia of the suprarenal abdominal aortas from AngII-induced AAAs with or without hydroxyapatite application. The paired Student t test (2-tailed) was performed to identify the gene-silencing efficiency of Runx2 siRNA compared with scramble siRNA, as well as AAV1-shRunx2 compared with AAV1-shNC. The nonparametric Kruskal-Wallis test with Dunn post hoc test was used to compare the elastin degradation grade. One-way ANOVA followed by the Student-Newman-Keuls test for post hoc comparison was applied to analyze the effect of AngII or hydroxyapatite on AAA-related genes’ expression and the effect of AngII on Runx2 and its downstream genes in VSMCs. Two-way ANOVA followed by the Bonferroni test was used to compare the maximum suprarenal aortic diameter between ApoE−/− mice or Runx2f/f:ApoE−/− and Runx2Δ/Δ SMC:ApoE−/− mice with saline or AngII induction, as well as the SUVmean of ROIs between microcalcification-positive, microcalcification-negative and saline-induced mice.
Microcalcification Is a Morphological Characteristic of Human and Mouse AAAs
To study aortic calcification in the human AAA aorta, we collected abdominal aortic tissue from 20 patients with AAA (abdominal aortas >40 mm diameter) undergoing open aortic repair and aortas from 19 control subjects who underwent heart transplantation but without AAA (abdominal aortas <30 mm diameter). We performed von Kossa staining for histological analysis of calcification since it has been repeatedly applied for detecting microcalcification in breast cancer, atherosclerosis lesion, and tuberculosis.22,30,31 The 2 groups of patients were matched with regard to gender frequency (percentage of males: 73.7% versus 85.0%, P=0.4506), age distribution (45.6±10.99 versus 47.5±11.16, P=0.6613), and the presence of other clinical characteristics including hypertension, diabetes mellitus, coronary heart disease, stroke, dyslipidemia, and respiratory disease (Table I in the Data Supplement). We observed 2 types of aortic calcification in AAA aortic sections: macrocalcification (sheet-like or segmented, ≥50 μm) and microcalcification (granular or spotty, <50 μm; Figure 1A, Figure I in the Data Supplement). Macrocalcification was only detected in one AAA patient (Figure I in the Data Supplement, AAA No. 6), whereas microcalcification was much more prevalent in aortic tissues from AAA patients than non-AAA patients (the incidence of microcalcification determined by von Kossa staining: 100% versus 21.1%, P<0.0001; Figure 1A and 1B, Figure I in the Data Supplement). The AAA tissues also showed a significant increase in calcified areas compared with those in the control subjects (Figure 1C), suggesting a positive correlation of microcalcification with AAA.
We further determined microcalcification in the mouse model of AAA in the ApoE−/− mice infused with AngII (1000 ng/[kg·min]) for 28 days. Similar to the observation in the human tissues, microcalcification deposits, as determined by von Kossa staining, were specifically present in the media and adventitia of the aneurysmal aorta, but not in the aortas of saline-infused ApoE−/− mice or the nondilated segment of AngII-infused mice (aneurysmal aorta versus saline-infused aorta: 100% versus 0%; aneurysmal aorta versus nondilated segment of AngII-infused aorta: 100% versus 0%; n=6 per group, P=0.0022; Figure 1D through 1E, Table II in the Data Supplement). Further Movat staining showed that cartilaginous matrices consisting of collagen- (yellow) and proteoglycan- (blue) rich extracellular matrix embedded with chondrocyte-like cells (green arrows) characterized by the relatively large amount of clear cytoplasm surrounded by a lacunar rim were found in the aneurysmal aortas but barely in the nondilated aortas (Figure 1D). Meanwhile, microcalcification in aneurysmal tissue didn’t colocalize with the area of intima plaque (Figure II in the Data Supplement).
Microcalcification was assessed in vivo by PET-CT using 18F-NaF, a specific radioactive tracer that has been shown to detect microcalcification in humans.1218F-NaF (10 MBq) was injected intravenously into ApoE−/− mice infused with saline (n=4) or AngII (n=9) for 28 days, respectively. After 10 minutes of injection, 18F-NaF uptake was mainly observed in the skeleton, kidneys, bladder, and regions of microcalcification (mainly located near the spine) by PET-CT in mice (Figure 2A). For a clear demonstration of microcalcification signals to avoid interference by physiological signals, PET images were further processed by AMIDE software with thresholding SUV levels (SUV, 0.5–1.3).23 In the processed PET images, the strong uptake of 18F-NaF in skeletal bone was subtracted while the signal of the edges of bones and microcalcification was highlighted (Figure 2A). We selected the regions between 2 kidneys and <2 mm upon the spine as ROIs since the AngII-induced AAA in mice exclusively occurs in the suprarenal aorta.32 Notably, the signal of CT-detected macrocalcification was absent in areas of 18F-NaF uptake on PET (Figure 2A). The aortic diameters were monitored by ultrasound analysis and ex vivo measurements after the sacrifice of the mice. In accordance with previous reports,33 AngII infusion caused a 44% (4/9) incidence of AAA in the ApoE−/− mice, whereas saline infusion did not show any effect (Figure IIIA and Table III in the Data Supplement). Of interest, upon 28-day AngII infusion, the 18F-NaF signal in ROIs of PET-CT images was only present in AAA-positive mice but not in AAA-negative mice (Figure 2A and 2B) or saline-infused mice (Figure IIIB in the Data Supplement). In line with the PET-CT data, von Kossa staining of microcalcification only existed in the area proximal to elastin fragmentation of AAA-positive mice but not in negative mice (Figure 2C). Taken together, these results suggest that microcalcification is a morphological characteristic of AAA and positively associates with AAA formation.
Microcalcification Precedes Marked Aortic Aneurysmal Expansion in Mice
The spatiotemporal correlation of microcalcification and AAA was further determined in the AngII-infused ApoE−/− mice (Figure 3A). Of note, 10.5% of mice with AngII infusion (2/19) died of dissection within the first week, and only the survivals were included in the following study. During the period of 28-day infusion, 18F-NaF PET-CT scanning was performed on days 7, 14, and 28 after AngII (n=17) or saline (n=6) infusion in mice. In parallel, the suprarenal aortic diameters were monitored by ultrasonography at the indicated time points. As early as day 7, 59% of mice (10/17) treated with AngII infusion were positive for microcalcification (microcalcification positive [MicroC+]), as evidenced by the increased uptake of 18F-NaF, whereas 41% of mice (7/17) with AngII infusion and 100% of saline-infused mice (6/6) showed no microcalcification (microcalcification negative [MicroC−]; Figure 3B, Table IV in the Data Supplement). The MicroC+ mice maintained a markedly greater SUV of 18F-NaF than that of MicroC− AngII-infused ApoE−/− mice (0.802±0.1193 versus 0.458±0.0826, P<0.0001) or saline-infused mice (0.802±0.1193 versus 0.375±0.0699, P<0.0001) from day 7 to day 28, whereas the latter 2 groups of mice did not differ from each other (Figure 3B and 3C). Intriguingly, the AngII-induced MicroC+ aortas did not exhibit significant dilation under ultrasound monitoring at day 7 of AngII infusion (Figure 3B and 3D). Nevertheless, the MicroC+ mice showed dramatic aortic expansion at subsequent day 14 to day 28, whereas the MicroC− and saline groups did not. More importantly, all MicroC+ mice developed AAA by the time point of sacrifice as revealed aortic expansion by ex vivo measurement and severe degradation of elastin fibers by elastic Van Gieson staining. In contrast, none of the MicroC− mice and saline-infused mice manifested AAAs (Figure 3D through 3G). In accordance, MicroC+ aortas were positive for von Kossa staining in the area adjacent to elastin degradation, whereas MicroC− aortas and saline-treated aortas were negatively stained (Figure 3H). Together, these data suggest that microcalcification may precede aortic dilation during AngII-induced AAA formation.
Hydroxyapatite Aggravates AAA Formation In Vivo
The predominant crystal found in vascular microcalcification is hydroxyapatite, a type of basic calcium phosphate crystal and the target of the 18F-NaF tracer, which represents the initial and active stage of calcification formation.13 To investigate the role of microcalcification in AAA formation, hydroxyapatite nanoparticles (0.25 M) were suspended in 30% Pluronic F-127 gel34 and applied periadventitially on the suprarenal aortas of 4-month-old ApoE−/− male mice infused with AngII for 28 days (Table V in the Data Supplement). The animals were divided into 4 groups: saline infusion with F-127 gel application (Saline), saline infusion with hydroxyapatite application (Saline+hydroxyapatite), AngII infusion with F-127 gel application (AngII), and AngII infusion with hydroxyapatite application (AngII+hydroxyapatite). Hydroxyapatite alone did not cause dramatic aortic expansion at day 28 but enhanced the incidence of AngII-induced AAA formation (AngII+hydroxyapatite versus AngII: 100% versus 60%, P=0.1818; Figure 4A and 4B). AngII+hydroxyapatite-treated ApoE−/− mice exhibited a greater maximal aortic diameter compared with AngII-infused ApoE−/− mice (AngII+hydroxyapatite versus AngII: 2.18±0.285 mm versus 1.57±0.326 mm, P=0.0018; Figure 4C). Similarly, elastin Van Gieson staining showed dramatically increased elastin fragmentation in the abdominal aortas of AngII+hydroxyapatite mice compared with those of AngII mice (Figure 4D). Further histological analysis revealed markedly increased CD68 positive macrophage infiltration in hydroxyapatite+AngII ApoE−/− mice compared with AngII ApoE−/− mice (Figure IV in the Data Supplement). Thus, hydroxyapatite, the structural component of microcalcification, enhances AngII-induced development of AAA in mice.
Runx2, the Master Transcriptional Regulator of Microcalcification, Mediates AAA Formation
To explore the key mediator of microcalcification and subsequent AAA pathogenesis, we performed RNA-sequencing in abdominal aortic tissue from 4-month-old ApoE−/− male mice upon 4-day AngII or saline treatment (n=5–6 for each group, Table VI in the Data Supplement). By using the gProfiler software,28 the DEGs of RNA sequencing were further analyzed based on the Human Phenotype Ontology35 and Online Mendelian Inheritance in Man.36 The enrichment analysis showed that the DEGs were correlated with the pathogenesis of aortic dissection and Ehlers-Danlos syndrome (Figure 5A). Interestingly, these upregulated genes were particularly associated with bone or joint diseases, such as osteogenesis imperfecta, joint laxity, and osteoporosis (Figure 5A), suggesting the upregulation of osteogenesis-related genes were involved during the early stage of AAA, which reinforced our early observation that microcalcification is involved in AAA formation.
We further compared the DEGs with known transcription factor target genes in humans and mice from the public ChIP-seq database ChIP-Atlas (http://chip-atlas.org/). More specifically, for each transcription factor, if one of its ChIP-seq peaks settled within the promoter region (5 kb around transcription start site) of a gene, this gene was considered as its target gene. The significant overlap of one transcription factor’s target genes with the DEGs was further assessed by Fisher’s exact test followed by Benjamini-Hochberg correction. Five transcription factors showing a corrected P value <0.05 in both humans and mice were finally selected as the significantly overlapping transcription factors, including Runx2, Esrra (estrogen-related receptor alpha), Jarid2 (jumonji AT-rich interactive domain 2), Tcf4 (transcription factor 4), and Pparg (peroxisome proliferator–activated receptor gamma; Figure 5B). Among them, Runx2 has been consistently identified as an essential transcription factor for both skeletal development37 and vascular calcification.19,38 Moreover, VSMCs stimulated by AngII (1 μM) for 24 hours showed markedly upregulated mRNA expression of Runx2 and its downstream target genes, including bone gamma-carboxyglutamate protein 2 (BGLAP2), secreted phosphoprotein 1 (SPP1), Sp7 transcription factor 7 (SP7) and receptor activating NF-κB ligand-L (RANKL; Figure VA in the Data Supplement). Therefore, we generated an AAV1-expressing short hairpin RNA against Runx2 (Figure VB in the Data Supplement). In the AngII-induced ApoE−/− mice model, AAV1-shRunx2 or control AAV (AAV1-shNC) was periadventitially applied to the suprarenal aortas (Table VII in the Data Supplement). After 28 days of AngII infusion in ApoE−/− mice, AAV-mediated Runx2 knockdown significantly suppressed aortic expansion (1.85±0.621 mm, n=9 versus 1.03±0.721 mm, n=7; P=0.0077) as well as the incidence of AAA (66.7% versus 14.3%, P=0.0361; Figure VC through VE).
As microcalcification mainly involves vascular SMCs (VSMCs), we further determined the effects of SMC-specific Runx2 deficiency on AngII-induced AAA formation in the ApoE−/− mice (Runx2Δ/Δ SMC:ApoE−/−, Table VIII in the Data Supplement).19 After 28 days of AngII induction, 55.6% (10/18) of the control littermates (Runx2f/f:ApoE−/−) developed AAA, whereas the incidence of AAA was reduced to 20% (2/10) in Runx2-deficient mice (Runx2Δ/Δ SMC:ApoE−/−; Figure 5C and 5D). The aortic rupture rate showed no difference between 2 groups of mice with AngII infusion (Runx2f/f:ApoE−/− versus Runx2Δ/Δ SMC:ApoE−/−: 11.1% versus 10.0%, P=0.9274). Compared with AngII-induced control littermates, Runx2Δ/Δ SMC:ApoE−/− mice with AngII induction also showed decreased AngII-induced aortic expansion (1.20±0.334 mm versus 1.77±0.69 mm, P=0.0322; Figure 5E) and alleviated degradation of elastin fibers (Figure 5F). The aortas of Runx2f/f:ApoE−/− mice exhibited a 55.6% incidence of microcalcification; in contrast, no microcalcification was detected in the Runx2Δ/Δ SMC:ApoE−/− mice (P=0.0033; Figure 5G). These results support a possible causative effect of Runx2-dependent microcalcification on AngII-induced AAA formation.
Microcalcification Promotes Vascular Inflammation
We then investigated the underlying mechanism of microcalcification-aggravated AAA by performing an mRNA microarray analysis of aortas from 7-day saline, saline+hydroxyapatite, AngII, and AngII+hydroxyapatite-treated ApoE−/− mice (Table V in the Data Supplement). The genes representing metalloproteases (matrix metalloproteinase [MMP]-9, MMP-12, MMP-19 and ADAM metallopeptidase with thrombospondin type 1 motif 4 [ADAMTS-4]), extracellular matrix remodeling (thrombospondin-1 [TSP-1] and lysyl oxidase [LOX]), inflammatory cytokines (MCP-1, CXCR4, TNFα, and interleukin 6 receptor alpha [IL6RA]), and macrophage activation (CD68, acid phosphatase 5 [ACP-5] and triggering receptor expressed on myeloid cells 2 [TREM2]) were significantly upregulated and were all indicated as contributing to AAA formation in vivo39–43 (Figure 6A). The upregulation of gene expression by hydroxyapatite was confirmed by real-time PCR (Figure 6B). We further verified the observation by using ex vivo aortic rings treated with saline, saline+hydroxyapatite, AngII, and AngII+hydroxyapatite. Real-time PCR revealed that AngII plus hydroxyapatite synergistically enhanced aortic inflammatory cytokines IL-6, TNF (tumor necrosis factor)-α, and MMP-9 expression (Figure 6C).
Previous studies suggested that hydroxyapatite was capable of activating the NLRP3 inflammasome in arthropathy, leading to the activation of IL-1β and IL-18 and subsequent inflammation.44 NLRP3 inflammasome pathway, including NLRP3, Caspase1, PYCARD (PYD and CARD domain containing), and IL-18, was also upregulated in hydroxyapatite-induced mice and further amplified in AngII+hydroxyapatite treated mice or aortic rings (Figure 6A, 6D through 6F). Accordingly, AngII+hydroxyapatite caused greater IL-1β secretion in aortic supernatants compared with AngII or hydroxyapatite alone (Figure 6G). These results suggested that hydroxyapatite aggravates AngII-induced AAA formation, possibly involving increased inflammatory cytokine and MMPs expression as well as NLRP3 inflammasome activation.
Furthermore, the role of Runx2 in mediating AngII-induced inflammation was determined in aorta from the Runx2-deficient mice and cultured VSMCs with Runx2 knockdown. Runx2Δ/ΔSMC:ApoE−/− mice exhibited downregulated IL-1β and IL-6 in plasma during AngII infusion, less CD45 positive leukocyte infiltration, and lower expression level of NLRP3 in the aortas than AngII-induced control littermates (Figure 6H through 6J). In accordance, AngII-upregulated VSMC expression of inflammatory cytokines (MCP-1 and IL-6), MMP-9, and inflammasome-related genes (NLRP3, Caspase1, and IL-1β) were significantly inhibited by Runx2 silencing compared with scramble siRNA interference (Figure VIA through VIC in the Data Supplement). Together, our data suggest that AngII induces Runx2 that leads to microcalcification, which may further enhance inflammatory cytokines/MMPs/inflammasome activation and thus ultimately leading to AAA formation (Figure 7).
To date, we know little about the pathogenesis of AAA and have few clinical and pharmacological choices for the early diagnosis and prevention of AAA. The present study provides the first evidence that vascular microcalcification aggravates aortic expansion and AAA formation by inducing aortic inflammation. Moreover, we identified Runx2, the master regulator of bone and ectopic calcification, as the key transcriptional regulator during AAA formation and progression.
Currently, the diagnosis of AAA depends solely on the morphological measurement of the abdominal aortic diameter, which is supposed to correlate with the rate of aortic expansion and rupture. However, AAA involves complex pathological alterations, including matrix degradation, vascular inflammation, and VSMC degeneration, and ≈20% of ruptures occur in small AAAs with diameters <5.5 cm.2,45 Therefore, we need to explore more pathological characteristics of AAA for the early detection and prevention of AAA. Vascular calcification has been identified as an independent risk factor and preclinical indicator for atherosclerosis,46–48 whereas its role in AAA remains an enigma. Although in vitro mechanical monitoring of AAA tissue has indicated that calcification decreases the wall strength in the aneurysmal area and may lead to aortic rupture,7,49 several retrospective studies have shown controversial observations regarding the correlation between calcification and AAA existence and expansion.8–10
In the latest SoFIA3 study,17 PET-CT scanning of AAA patients revealed that 18F-NaF, an indicator of active microcalcification, localized to the area of aortic aneurysm. More intriguingly, the level of 18F-NaF uptake correlated with AAA expansion, as well as clinical events (AAA repairs and ruptures), indicating the strong association of microcalcification with AAA disease. However, it remains unclear whether a cause-effect relationship exists between microcalcification and AAA or whether microcalcification is just a bystander during AAA formation. By monitoring real-time 18F-NaF uptake and aortic dilation in the aortas of an AngII-infused ApoE−/− mouse AAA model, we demonstrated that microcalcification did correlate with AAA, which is in accord with SoFIA3 study. Moreover, our data revealed that microcalcification occurred before significant aortic dilation. More importantly, only abdominal aortas with early microcalcification signals developed subsequently into aortic aneurysm and vice versa. Together with the complementary data showing that hydroxyapatite aggravates AngII-induced AAA formation and that VSMC-specific deletion of the master osteogenic transcriptional factor Runx2 prevents AAA formation in mice, the study pointed out a potential causative effect of microcalcification on AAA pathogenesis. Our data have shed light on the promising prognostic value of microcalcification for the early detection of AAA in addition to the canonical aortic diameter measurement.
Of note, microcalcification and macrocalcification were distinctly based on the size of the calcified micronodules, <50 and ≥50 μm in diameter. Whereas macrocalcification normally confers a noninflamed and healing response, microcalcification is usually associated with early-stage active calcification, macrophage infiltration, and inflammation.13,50 Although the regular methods, including CT, can only detect macrocalcification (≥200 μm), increasing evidence supports that 18F-NaF as a marker of microcalcification.11,12,51 Findings on 4 levels, including autoradiography, light microscopy, in vivo clinical PET-CT and ex vivo micro–PET-CT, show that 18F-NaF is a highly specific ligand with high affinity for the detection of pathologically high-risk microcalcification.1218F-NaF preferentially binds to the surface of nanocrystalline hydroxyapatite and was associated with histological markers of osteogenic activity.51 Notably, according to the MESA (Multi-Ethnic Study of Atherosclerosis), vascular calcification is present and increases with age in a healthy population that is free of clinical cardiovascular disease and treated diabetes mellitus.52 The lack of histological characterization of calcification in human aortas with or without AAA prompted us to compare the calcification morphology in AAA. We observed an increased prevalence of microcalcification in aneurysmal aortas than in non-AAA aortas, whereas macrocalcification was not obvious in neither the AAA nor the non-AAA human samples we analyzed. The latter finding may be due to the middle-age (40–49 years) status of patients with an abdominal aortic calcification rate <19%.53,54 In addition, both human and mouse aneurysmal aortas exhibited multiple dot-like rather than sheet-like calcium crystals, which is in line with our current 18F-NaF PET-CT data and the recent SoFIA3 findings. Moreover, although AAA is epidemiologically associated with atherosclerosis and 18F-NaF PET-CT specifically recognized microcalcification in vulnerable plaques,11,12,51 our histological analysis of the abdominal aortas of AngII-infused 4-month-old ApoE−/− mice fed with standard laboratory diet (1032; Huafukang Bioscience, Beijing, China) revealed that microcalcification was predominantly located in the media area adjacent to expanded and destroyed vessels but not within the intima plaque. Thus, our results showing early detection of microcalcification by 18F-NaF PET-CT that preceded aortic dilation and AAA formation suggest that microcalcification is a novel pathological characteristic of AAA.
How microcalcification promotes AAA pathogenesis remains elusive. We found that the application of hydroxyapatite, the major component of calcification, significantly aggravated aortic expansion in mouse AAA formation in vivo. In our preexperiment of hydroxyapatite application in mice without removing the F-127 gel and adventitial tissue, we observed hydroxyapatite surrounded by numerous cells as well as the phenomenon of internalization, indicating hydroxyapatite could be absorbed into cells and causes inflammation.44,55 The microarray and ex vivo experiments also indicated that hydroxyapatite treatment induced vascular inflammation (IL-6, MCP-1, TNFα, NLRP3 inflammasome, MMP-9), which all have been demonstrated to contribute to AAA development.56,57 Our study is in line with the previous observation that ectopic deposition of hydroxyapatite crystals in joints activates macrophage NLRP3 inflammasome and renders arthropathy in mice.44 In addition to inflammation, microcalcification may also decrease vessel wall stability,58,59 which is capable of inducing the differentiation of mononuclear cells into osteoclast-like cells and promoting the secretion of MMPs,60 and may ultimately lead to aneurysm formation.61 Our observation is also in agreement with a previous report that bisphosphonate, an inhibitor of hydroxyapatite formation, inhibits AAA formation in mice.61,62
Consistently with these pathophysiological changes during the early stage of AAA formation, the nonbiased RNA sequencing and bioinformatics analysis revealed upregulation of the osteogenic factors and identified the master osteogenic transcription factor, Runx2, as the key mediator orchestrating microcalcification with AAA. Previous studies have determined that Runx2 promotes osteogenic differentiation of VSMCs that leads to vascular calcification.19,38 The present studies identified that SMC-specific Runx2 deletion inhibited AngII-induced microcalcification and AAA formation in the ApoE−/− mice. Although there are no SMCs in the adventitia, the microcalcification in adventitia is also possibly influenced by SMC-Runx2 deficiency via cell-cell communication. For instance, microcalcification-aggravated inflammation can reciprocally promote ectopic calcification beyond media and create a vicious circle, which can be possibly blocked by the absence of Runx2. Increased Runx2 has been reported in human AAA tissues,63 and a gene network analysis also indicated Runx2 as a highly interconnected gene in AngII-infused mice at day 7.64 Our studies provide the genetic proof of a causative effect of Runx2-dependent microcalcification on AngII-induced AAA formation. Furthermore, we identified that Runx2 mediates AngII-induced elevation of inflammatory factors and MMPs expression in AAA, supporting an integrated role of Runx2 in regulating AngII-induced AAA formation.
We acknowledge that there are limitations to our study. Thoracic aortas that we used may not be appropriate control samples for the AAA study because thoracic aortas and abdominal aortas show differences in structure, pathological genetics, and origin of SMC.65,66 Although 18F-NaF PET-CT is well recognized as the noninvasive measurement for the quantification of microcalcification and von Kossa staining is commonly applied for in vitro histological staining of microcalcification/calcification, both of them alone might not guarantee the accuracy. Of note, the 18F-NaF uptake might result from spill-over effect of PET signal from vascular macrocalcification in adjacent slices, even in the absence of CT uptake.67 Other alternative methods, such as Alizarin red and tissue calcium concentration, also have limitations in specificity and detection range, respectively. In the current study, we combined 18F-NaF PET-CT, von Kossa, and Movat staining to monitor microcalcification in aorta. Besides, we applied ultrasonography to real-time monitoring aortic expansion in live mice, which is commonly used clinically in AAA diagnosis but may not be accurate enough for mouse aortic measurements. CT angiography would be an alternative instrument to monitor the aortic diameter and has the strength to visualize the colocalization of microcalcification with aortic expansion in vivo. Nevertheless, the combination of CT angiography and PET in live small animals is technically difficult. Meanwhile, although the application of the major component of calcification hydroxyapatite mimics the microcalcification to some extent, it may cause nonspecific effect. Additionally, although the AngII infusion aneurysmal mice model is well established, it does not entirely mimic the pathology of AAA in humans.68,69 Of note, the SoFIA3 study identified 18F-NaF uptake associated with aneurysm rupture. In the AngII-infused ApoE−/− mice model, the dissection or aortic rupture usually occurred at the early stage upon AngII stimulation with low incidence (around 10%).70 All the survival mice we monitored by micro–PET-CT and ultrasound between day 7 to day 28 were excluded from dissection or aortic rupture. Thus, our data only reveals an association between microcalcification and subsequent AAA formation but not the relationship between microcalcification and dissection or AAA rupture. In spite of this, we still could not draw a definitive conclusion of their cause-effect relationship. Specifically, it is challenging to use a single molecule or probe to fully describe the process of microcalcification during AAA pathogenesis. More studies are warranted to explore the authentic relation and molecular mechanism underlying the microcalcification process during AAA formation.
Taken together, results from the present study suggest microcalcification is not only a pathological characteristic of AAA but also a potential predictor and mediator of AAA formation. Besides, we identified an important role of the osteogenic factor Runx2 in mediating AngII-induced microcalcification and aortic inflammation, which provides novel insights into the mechanism underlying the development of AAA and sheds light on early diagnosis, prevention and drug development for AAA.
abdominal aortic aneurysm
estrogen-related receptor alpha
jumonji AT-rich interactive domain 2
Multi-Ethnic Study of Atherosclerosis
NLR family pyrin domain-containing 3
positron emission tomography and X-ray computed tomography
peroxisome proliferator–activated receptor gamma
regions of interest
runt-related transcription factor 2
standardized uptake value
transcription factor 4
vascular smooth muscle cell
Sources of Funding
This research was supported by funding from the National Natural Science Foundation of P. R. China (NSFC, 81730010, 91839302, 81921001, and 31930056).
Kent KC. Clinical practice. Abdominal aortic aneurysms.N Engl J Med. 2014; 371:2101–2108. doi: 10.1056/NEJMcp1401430CrossrefMedlineGoogle Scholar
Nicholls SC, Gardner JB, Meissner MH, Johansen HK. Rupture in small abdominal aortic aneurysms.J Vasc Surg. 1998; 28:884–888. doi: 10.1016/s0741-5214(98)70065-5CrossrefMedlineGoogle Scholar
Schiffrin EL, Lipman ML, Mann JF. Chronic kidney disease: effects on the cardiovascular system.Circulation. 2007; 116:85–97. doi: 10.1161/CIRCULATIONAHA.106.678342LinkGoogle Scholar
Bastos Gonçalves F, Voûte MT, Hoeks SE, Chonchol MB, Boersma EE, Stolker RJ, Verhagen HJ. Calcification of the abdominal aorta as an independent predictor of cardiovascular events: a meta-analysis.Heart. 2012; 98:988–994. doi: 10.1136/heartjnl-2011-301464CrossrefMedlineGoogle Scholar
Juutilainen A, Lehto S, Suhonen M, Rönnemaa T, Laakso M. Thoracoabdominal calcifications predict cardiovascular disease mortality in type 2 diabetic and nondiabetic subjects: 18-year follow-up study.Diabetes Care. 2010; 33:583–585. doi: 10.2337/dc09-1813CrossrefMedlineGoogle Scholar
Buijs RV, Willems TP, Tio RA, Boersma HH, Tielliu IF, Slart RH, Zeebregts CJ. Calcification as a risk factor for rupture of abdominal aortic aneurysm.Eur J Vasc Endovasc Surg. 2013; 46:542–548. doi: 10.1016/j.ejvs.2013.09.006CrossrefMedlineGoogle Scholar
O’Leary SA, Mulvihill JJ, Barrett HE, Kavanagh EG, Walsh MT, McGloughlin TM, Doyle BJ. Determining the influence of calcification on the failure properties of Abdominal Aortic Aneurysm (AAA) tissue.J Mech Behav Biomed Mater. 2015; 42:154–167. doi: 10.1016/j.jmbbm.2014.11.005CrossrefMedlineGoogle Scholar
Lindholt JS. Aneurysmal wall calcification predicts natural history of small abdominal aortic aneurysms.Atherosclerosis. 2008; 197:673–678. doi: 10.1016/j.atherosclerosis.2007.03.012CrossrefMedlineGoogle Scholar
Hendy K, Gunnarsson R, Cronin O, Golledge J. Infra-renal abdominal aortic calcification volume does not predict small abdominal aortic aneurysm growth.Atherosclerosis. 2015; 243:334–338. doi: 10.1016/j.atherosclerosis.2015.07.027CrossrefMedlineGoogle Scholar
Rai D, Wisniowski B, Bradshaw B, Velu R, Tosenovsky P, Quigley F, Walker PJ, Golledge J. Abdominal aortic aneurysm calcification and thrombus volume are not associated with outcome following endovascular abdominal aortic aneurysm repair.Eur Radiol. 2014; 24:1768–1776. doi: 10.1007/s00330-014-3185-yCrossrefMedlineGoogle Scholar
Joshi NV, Vesey AT, Williams MC, Shah AS, Calvert PA, Craighead FH, Yeoh SE, Wallace W, Salter D, Fletcher AM,. 18F-fluoride positron emission tomography for identification of ruptured and high-risk coronary atherosclerotic plaques: a prospective clinical trial.Lancet. 2014; 383:705–713. doi: 10.1016/S0140-6736(13)61754-7CrossrefMedlineGoogle Scholar
Irkle A, Vesey AT, Lewis DY, Skepper JN, Bird JL, Dweck MR, Joshi FR, Gallagher FA, Warburton EA, Bennett MR,. Identifying active vascular microcalcification by (18)F-sodium fluoride positron emission tomography.Nat Commun. 2015; 6:7495. doi: 10.1038/ncomms8495CrossrefMedlineGoogle Scholar
Aikawa E, Nahrendorf M, Figueiredo JL, Swirski FK, Shtatland T, Kohler RH, Jaffer FA, Aikawa M, Weissleder R. Osteogenesis associates with inflammation in early-stage atherosclerosis evaluated by molecular imaging in vivo.Circulation. 2007; 116:2841–2850. doi: 10.1161/CIRCULATIONAHA.107.732867LinkGoogle Scholar
Pugliese G, Iacobini C, Blasetti Fantauzzi C, Menini S. The dark and bright side of atherosclerotic calcification.Atherosclerosis. 2015; 238:220–230. doi: 10.1016/j.atherosclerosis.2014.12.011CrossrefMedlineGoogle Scholar
Chen O, Dontineni N, Nahlawi G, Bhumireddy GP, Han SY, Katri Y, Gulkarov IM, Ciaburri DG, Tortolani AJ, Lazzaro RS,. Serial cardiac magnetic resonance imaging of a rapidly progressing liquefaction necrosis of mitral annulus calcification associated with embolic stroke.Circulation. 2012; 125:2792–2795. doi: 10.1161/CIRCULATIONAHA.111.043182LinkGoogle Scholar
Nadra I, Mason JC, Philippidis P, Florey O, Smythe CD, McCarthy GM, Landis RC, Haskard DO. Proinflammatory activation of macrophages by basic calcium phosphate crystals via protein kinase C and MAP kinase pathways: a vicious cycle of inflammation and arterial calcification?Circ Res. 2005; 96:1248–1256. doi: 10.1161/01.RES.0000171451.88616.c2LinkGoogle Scholar
Forsythe RO, Dweck MR, McBride OMB, Vesey AT, Semple SI, Shah ASV, Adamson PD, Wallace WA, Kaczynski J, Ho W,. 18F-sodium fluoride uptake in abdominal aortic aneurysms: the SoFIA3 study.J Am Coll Cardiol. 2018; 71:513–523. doi: 10.1016/j.jacc.2017.11.053CrossrefMedlineGoogle Scholar
Robinet P, Milewicz DM, Cassis LA, Leeper NJ, Lu HS, Smith JD. Consideration of sex differences in design and reporting of experimental arterial pathology studies-statement from ATVB council.Arterioscler Thromb Vasc Biol. 2018; 38:292–303. doi: 10.1161/ATVBAHA.117.309524LinkGoogle Scholar
Sun Y, Byon CH, Yuan K, Chen J, Mao X, Heath JM, Javed A, Zhang K, Anderson PG, Chen Y. Smooth muscle cell-specific runx2 deficiency inhibits vascular calcification.Circ Res. 2012; 111:543–552. doi: 10.1161/CIRCRESAHA.112.267237LinkGoogle Scholar
Daugherty A, Manning MW, Cassis LA. Angiotensin II promotes atherosclerotic lesions and aneurysms in apolipoprotein E-deficient mice.J Clin Invest. 2000; 105:1605–1612. doi: 10.1172/JCI7818CrossrefMedlineGoogle Scholar
Kanematsu Y, Kanematsu M, Kurihara C, Tsou TL, Nuki Y, Liang EI, Makino H, Hashimoto T. Pharmacologically induced thoracic and abdominal aortic aneurysms in mice.Hypertension. 2010; 55:1267–1274. doi: 10.1161/HYPERTENSIONAHA.109.140558LinkGoogle Scholar
Wilson GH, Gore JC, Yankeelov TE, Barnes S, Peterson TE, True JM, Shokouhi S, McIntyre JO, Sanders M, Abramson V,. An approach to breast cancer diagnosis via PET imaging of microcalcifications using (18)F-NaF.J Nucl Med. 2014; 55:1138–1143. doi: 10.2967/jnumed.114.139170CrossrefMedlineGoogle Scholar
Loening AM, Gambhir SS. AMIDE: a free software tool for multimodality medical image analysis.Mol Imaging. 2003; 2:131–137. doi: 10.1162/153535003322556877CrossrefMedlineGoogle Scholar
MOVAT HZ. Demonstration of all connective tissue elements in a single section; pentachrome stains.AMA Arch Pathol. 1955; 60:289–295.MedlineGoogle Scholar
Fu Y, Gao C, Liang Y, Wang M, Huang Y, Ma W, Li T, Jia Y, Yu F, Zhu W,. Shift of macrophage phenotype due to cartilage oligomeric matrix protein deficiency drives atherosclerotic calcification.Circ Res. 2016; 119:261–276. doi: 10.1161/CIRCRESAHA.115.308021LinkGoogle Scholar
Liao Y, Smyth GK, Shi W. The Subread aligner: fast, accurate and scalable read mapping by seed-and-vote.Nucleic Acids Res. 2013; 41:e108. doi: 10.1093/nar/gkt214CrossrefMedlineGoogle Scholar
Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, Smyth GK. limma powers differential expression analyses for RNA-sequencing and microarray studies.Nucleic Acids Res. 2015; 43:e47. doi: 10.1093/nar/gkv007CrossrefMedlineGoogle Scholar
Reimand J, Kull M, Peterson H, Hansen J, Vilo J. g:Profiler–a web-based toolset for functional profiling of gene lists from large-scale experiments.Nucleic Acids Res. 2007; 35(web server issue):W193–W200. doi: 10.1093/nar/gkm226CrossrefMedlineGoogle Scholar
Chamley-Campbell J, Campbell GR, Ross R. The smooth muscle cell in culture.Physiol Rev. 1979; 59:1–61. doi: 10.1152/physrev.19188.8.131.52CrossrefMedlineGoogle Scholar
Nakahara T, Dweck MR, Narula N, Pisapia D, Narula J, Strauss HW. Coronary artery calcification: from mechanism to molecular imaging.JACC Cardiovasc Imaging. 2017; 10:582–593. doi: 10.1016/j.jcmg.2017.03.005CrossrefMedlineGoogle Scholar
Ordonez AA, DeMarco VP, Klunk MH, Pokkali S, Jain SK. Imaging chronic tuberculous lesions using sodium [(18)F]fluoride positron emission tomography in mice.Mol Imaging Biol. 2015; 17:609–614. doi: 10.1007/s11307-015-0836-6CrossrefMedlineGoogle Scholar
Lu H, Howatt DA, Balakrishnan A, Moorleghen JJ, Rateri DL, Cassis LA, Daugherty A. Subcutaneous angiotensin II infusion using osmotic pumps induces aortic aneurysms in mice.J Vis Exp. 2015; 103:53191. doi: 10.3791/53191Google Scholar
Trachet B, Fraga-Silva RA, Jacquet PA, Stergiopulos N, Segers P. Incidence, severity, mortality, and confounding factors for dissecting AAA detection in angiotensin II-infused mice: a meta-analysis.Cardiovasc Res. 2015; 108:159–170. doi: 10.1093/cvr/cvv215CrossrefMedlineGoogle Scholar
Zhou AJ, Peel SA, Clokie CM. An evaluation of hydroxyapatite and biphasic calcium phosphate in combination with Pluronic F127 and BMP on bone repair.J Craniofac Surg. 2007; 18:1264–1275. doi: 10.1097/scs.0b013e318158cb1aCrossrefMedlineGoogle Scholar
Robinson PN, Köhler S, Bauer S, Seelow D, Horn D, Mundlos S. The Human Phenotype Ontology: a tool for annotating and analyzing human hereditary disease.Am J Hum Genet. 2008; 83:610–615. doi: 10.1016/j.ajhg.2008.09.017CrossrefMedlineGoogle Scholar
Amberger JS, Bocchini CA, Schiettecatte F, Scott AF, Hamosh A. OMIM.org: Online Mendelian Inheritance in Man (OMIM®), an online catalog of human genes and genetic disorders.Nucleic Acids Res. 2015; 43(database issue):D789–D798. doi: 10.1093/nar/gku1205CrossrefMedlineGoogle Scholar
Komori T. Roles of Runx2 in skeletal development.Adv Exp Med Biol. 2017; 962:83–93. doi: 10.1007/978-981-10-3233-2_6CrossrefMedlineGoogle Scholar
Lin ME, Chen T, Leaf EM, Speer MY, Giachelli CM. Runx2 expression in smooth muscle cells is required for arterial medial calcification in mice.Am J Pathol. 2015; 185:1958–1969. doi: 10.1016/j.ajpath.2015.03.020CrossrefMedlineGoogle Scholar
Nosoudi N, Nahar-Gohad P, Sinha A, Chowdhury A, Gerard P, Carsten CG, Gray BH, Vyavahare NR. Prevention of abdominal aortic aneurysm progression by targeted inhibition of matrix metalloproteinase activity with batimastat-loaded nanoparticles.Circ Res. 2015; 117:e80–e89. doi: 10.1161/CIRCRESAHA.115.307207LinkGoogle Scholar
Liu Z, Morgan S, Ren J, Wang Q, Annis DS, Mosher DF, Zhang J, Sorenson CM, Sheibani N, Liu B. Thrombospondin-1 (TSP1) contributes to the development of vascular inflammation by regulating monocytic cell motility in mouse models of abdominal aortic aneurysm.Circ Res. 2015; 117:129–141. doi: 10.1161/CIRCRESAHA.117.305262LinkGoogle Scholar
Onoda M, Yoshimura K, Aoki H, Ikeda Y, Morikage N, Furutani A, Matsuzaki M, Hamano K. Lysyl oxidase resolves inflammation by reducing monocyte chemoattractant protein-1 in abdominal aortic aneurysm.Atherosclerosis. 2010; 208:366–369. doi: 10.1016/j.atherosclerosis.2009.07.036CrossrefMedlineGoogle Scholar
Eagleton MJ. Inflammation in abdominal aortic aneurysms: cellular infiltrate and cytokine profiles.Vascular. 2012; 20:278–283. doi: 10.1258/vasc.2011.201207CrossrefMedlineGoogle Scholar
Raffort J, Lareyre F, Clément M, Hassen-Khodja R, Chinetti G, Mallat Z. Monocytes and macrophages in abdominal aortic aneurysm.Nat Rev Cardiol. 2017; 14:457–471. doi: 10.1038/nrcardio.2017.52CrossrefMedlineGoogle Scholar
Jin C, Frayssinet P, Pelker R, Cwirka D, Hu B, Vignery A, Eisenbarth SC, Flavell RA. NLRP3 inflammasome plays a critical role in the pathogenesis of hydroxyapatite-associated arthropathy.Proc Natl Acad Sci U S A. 2011; 108:14867–14872. doi: 10.1073/pnas.1111101108CrossrefMedlineGoogle Scholar
Powell JT, Gotensparre SM, Sweeting MJ, Brown LC, Fowkes FG, Thompson SG. Rupture rates of small abdominal aortic aneurysms: a systematic review of the literature.Eur J Vasc Endovasc Surg. 2011; 41:2–10. doi: 10.1016/j.ejvs.2010.09.005CrossrefMedlineGoogle Scholar
Rumberger JA, Simons DB, Fitzpatrick LA, Sheedy PF, Schwartz RS. Coronary artery calcium area by electron-beam computed tomography and coronary atherosclerotic plaque area. A histopathologic correlative study.Circulation. 1995; 92:2157–2162. doi: 10.1161/01.cir.92.8.2157LinkGoogle Scholar
Sangiorgi G, Rumberger JA, Severson A, Edwards WD, Gregoire J, Fitzpatrick LA, Schwartz RS. Arterial calcification and not lumen stenosis is highly correlated with atherosclerotic plaque burden in humans: a histologic study of 723 coronary artery segments using nondecalcifying methodology.J Am Coll Cardiol. 1998; 31:126–133. doi: 10.1016/s0735-1097(97)00443-9CrossrefMedlineGoogle Scholar
O’Rourke RA, Brundage BH, Froelicher VF, Greenland P, Grundy SM, Hachamovitch R, Pohost GM, Shaw LJ, Weintraub WS, Winters WLAmerican College of Cardiology/American Heart Association Expert Consensus Document on electron-beam computed tomography for the diagnosis and prognosis of coronary artery disease.J Am Coll Cardiol. 2000; 36:326–340. doi: 10.1016/s0735-1097(00)00831-7CrossrefMedlineGoogle Scholar
Barrett HE, Cunnane EM, Hidayat H, O’Brien JM, Moloney MA, Kavanagh EG, Walsh MT. On the influence of wall calcification and intraluminal thrombus on prediction of abdominal aortic aneurysm rupture.J Vasc Surg. 2018; 67:1234.e2–1246.e2. doi: 10.1016/j.jvs.2017.05.086CrossrefGoogle Scholar
New SE, Goettsch C, Aikawa M, Marchini JF, Shibasaki M, Yabusaki K, Libby P, Shanahan CM, Croce K, Aikawa E. Macrophage-derived matrix vesicles: an alternative novel mechanism for microcalcification in atherosclerotic plaques.Circ Res. 2013; 113:72–77. doi: 10.1161/CIRCRESAHA.113.301036LinkGoogle Scholar
Creager MD, Hohl T, Hutcheson JD, Moss AJ, Schlotter F, Blaser MC, Park MA, Lee LH, Singh SA, Alcaide-Corral CJ,. 18F-fluoride signal amplification identifies microcalcifications associated with atherosclerotic plaque instability in positron emission tomography/computed tomography images.Circ Cardiovasc Imaging. 2019; 12:e007835. doi: 10.1161/CIRCIMAGING.118.007835LinkGoogle Scholar
Bild DE, Detrano R, Peterson D, Guerci A, Liu K, Shahar E, Ouyang P, Jackson S, Saad MF. Ethnic differences in coronary calcification: the Multi-Ethnic Study of Atherosclerosis (MESA).Circulation. 2005; 111:1313–1320. doi: 10.1161/01.CIR.0000157730.94423.4BLinkGoogle Scholar
Pepe J, Diacinti D, Fratini E, Nofroni I, D’Angelo A, Pilotto R, Savoriti C, Colangelo L, Raimo O, Cilli M,. High prevalence of abdominal aortic calcification in patients with primary hyperparathyroidism as evaluated by Kauppila score.Eur J Endocrinol. 2016; 175:95–100. doi: 10.1530/EJE-15-1152CrossrefMedlineGoogle Scholar
Liu P, Zhang S, Gao J, Lin Y, Shi G, He W, Touyz RM, Yan L, Huang H. Downregulated serum 14, 15-epoxyeicosatrienoic acid is associated with abdominal aortic calcification in patients with primary aldosteronism.Hypertension. 2018; 71:592–598. doi: 10.1161/HYPERTENSIONAHA.117.10644LinkGoogle Scholar
Lebre F, Sridharan R, Sawkins MJ, Kelly DJ, O’Brien FJ, Lavelle EC. The shape and size of hydroxyapatite particles dictate inflammatory responses following implantation.Sci Rep. 2017; 7:2922. doi: 10.1038/s41598-017-03086-0CrossrefMedlineGoogle Scholar
Golledge J. Abdominal aortic aneurysm: update on pathogenesis and medical treatments.Nat Rev Cardiol. 2019; 16:225–242. doi: 10.1038/s41569-018-0114-9CrossrefMedlineGoogle Scholar
Usui F, Shirasuna K, Kimura H, Tatsumi K, Kawashima A, Karasawa T, Yoshimura K, Aoki H, Tsutsui H, Noda T,. Inflammasome activation by mitochondrial oxidative stress in macrophages leads to the development of angiotensin II-induced aortic aneurysm.Arterioscler Thromb Vasc Biol. 2015; 35:127–136. doi: 10.1161/ATVBAHA.114.303763LinkGoogle Scholar
Kelly-Arnold A, Maldonado N, Laudier D, Aikawa E, Cardoso L, Weinbaum S. Revised microcalcification hypothesis for fibrous cap rupture in human coronary arteries.Proc Natl Acad Sci U S A. 2013; 110:10741–10746. doi: 10.1073/pnas.1308814110CrossrefMedlineGoogle Scholar
Maldonado N, Kelly-Arnold A, Vengrenyuk Y, Laudier D, Fallon JT, Virmani R, Cardoso L, Weinbaum S. A mechanistic analysis of the role of microcalcifications in atherosclerotic plaque stability: potential implications for plaque rupture.Am J Physiol Heart Circ Physiol. 2012; 303:H619–H628. doi: 10.1152/ajpheart.00036.2012CrossrefMedlineGoogle Scholar
Lange T, Schilling AF, Peters F, Haag F, Morlock MM, Rueger JM, Amling M. Proinflammatory and osteoclastogenic effects of beta-tricalciumphosphate and hydroxyapatite particles on human mononuclear cells in vitro.Biomaterials. 2009; 30:5312–5318. doi: 10.1016/j.biomaterials.2009.06.023CrossrefMedlineGoogle Scholar
Takei Y, Tanaka T, Kent KC, Yamanouchi D. Osteoclastogenic differentiation of macrophages in the development of abdominal aortic aneurysms.Arterioscler Thromb Vasc Biol. 2016; 36:1962–1971. doi: 10.1161/ATVBAHA.116.307715LinkGoogle Scholar
Tsai SH, Huang PH, Peng YJ, Chang WC, Tsai HY, Leu HB, Chen JW, Lin SJ. Zoledronate attenuates angiotensin II-induced abdominal aortic aneurysm through inactivation of Rho/ROCK-dependent JNK and NF-κB pathway.Cardiovasc Res. 2013; 100:501–510. doi: 10.1093/cvr/cvt230CrossrefMedlineGoogle Scholar
Dubis J, Litwin M, Michalowska D, Zuk N, Szczepanska-Buda A, Grendziak R, Baczynska D, Barc P, Witkiewicz W. Elevated expression of runt-related transcription factors in human abdominal aortic aneurysm.J Biol Regul Homeost Agents. 2016; 30:497–504.MedlineGoogle Scholar
Spin JM, Hsu M, Azuma J, Tedesco MM, Deng A, Dyer JS, Maegdefessel L, Dalman RL, Tsao PS. Transcriptional profiling and network analysis of the murine angiotensin II-induced abdominal aortic aneurysm.Physiol Genomics. 2011; 43:993–1003. doi: 10.1152/physiolgenomics.00044.2011CrossrefMedlineGoogle Scholar
Pinard A, Jones GT, Milewicz DM. Genetics of thoracic and abdominal aortic diseases.Circ Res. 2019; 124:588–606. doi: 10.1161/CIRCRESAHA.118.312436LinkGoogle Scholar
Wang G, Jacquet L, Karamariti E, Xu Q. Origin and differentiation of vascular smooth muscle cells.J Physiol. 2015; 593:3013–3030. doi: 10.1113/JP270033CrossrefMedlineGoogle Scholar
Soret M, Bacharach SL, Buvat I. Partial-volume effect in PET tumor imaging.J Nucl Med. 2007; 48:932–945. doi: 10.2967/jnumed.106.035774CrossrefMedlineGoogle Scholar
Trachet B, Piersigilli A, Fraga-Silva RA, Aslanidou L, Sordet-Dessimoz J, Astolfo A, Stampanoni MF, Segers P, Stergiopulos N. Ascending aortic aneurysm in angiotensin II-infused mice: formation, progression, and the role of focal dissections.Arterioscler Thromb Vasc Biol. 2016; 36:673–681. doi: 10.1161/ATVBAHA.116.307211LinkGoogle Scholar
Trachet B, Aslanidou L, Piersigilli A, Fraga-Silva RA, Sordet-Dessimoz J, Villanueva-Perez P, Stampanoni MFM, Stergiopulos N, Segers P. Angiotensin II infusion into ApoE −/− mice: a model for aortic dissection rather than abdominal aortic aneurysm?Cardiovasc Res. 2017; 113:1230–1242. doi: 10.1093/cvr/cvx128CrossrefMedlineGoogle Scholar
Saraff K, Babamusta F, Cassis LA, Daugherty A. Aortic dissection precedes formation of aneurysms and atherosclerosis in angiotensin II-infused, apolipoprotein E-deficient mice.Arterioscler Thromb Vasc Biol. 2003; 23:1621–1626. doi: 10.1161/01.ATV.0000085631.76095.64LinkGoogle Scholar