Perivascular Adipose Tissue–Derived PDGF-D Contributes to Aortic Aneurysm Formation During Obesity

A close relationship has been well established between obesity and vascular diseases, including hypertension, atherosclerosis, and aortic aneurysm (AA) (1–3). In addition to metabolic effects, adipose tissue–derived adipokines perform various important roles in the regulation of vascular disorders such as neointimal formation, angiogenesis, and vascular remodeling (4–6). Virtually all arteries are surrounded by amounts of perivascular adipose tissue (PVAT), which is juxtaposed to the vascular adventitia. Recently, it has become recognized that PVAT is abundantly expanded with an altered production of adipokines and chronic inflammatory response during obesity, which have important effects on vascular disease (7–9).

AA, characterized by chronic vascular inflammation and destructive connective tissue remodeling, tends to expand asymptomatically until a catastrophic event, such as aortic rupture or dissection, occurs (10,11). Multiple risk factors are associated with the incidence of AA, including advanced age, smoking, hyperlipidemia, and hypertension (12,13). Recently, obesity is reported to increase the risk of abdominal AA. A population study of >12,000 men confirmed that an index of obesity (waist circumference and waist-to-hip ratio) independently associates with abdominal AA formation (14). Another cohort study (15) suggested that the risk of abdominal AA increased by 15% per 5-cm increment of waist circumference up to 100 cm for men and 88 cm for women. A previous study (16) has pointed out increased AA formation in leptin-deficient obese mice (ob/ob) after angiotensin II (AngII) infusion. However, little is known about the detailed mechanism of AA formation during obesity.

Vascular inflammation is a crucial cause of AA, which is histologically characterized by medial degeneration and various degrees of adventitial immune cell recruitment (17). The infiltration of inflammatory cells, including macrophages, lymphocytes, and mast cells, is mainly observed in the tunica adventitia of human aneurismal tissues (18). Recent studies showed that adventitial fibroblast activation contributes to AA formation via accelerating immune cell–mediated vascular inflammation in mice (19,20). It is also well known that chronic low-grade inflammation within the adipose tissue is an important causal factor for obesity-related vascular disorders (21–23). Therefore, we hypothesized that PVAT-derived factor mediates adventitial inflammation and promotes AA formation in AngII-infused obese mice.

In the current study, by performing cDNA microarray analysis, we detected highly expressed platelet-derived growth factor-D (PDGF-D) in the PVAT of AngII-induced ob/ob mice. PDGF-D stimulated adventitial fibroblast migration and proliferation, as well as inflammatory factor expression. Blockade of PDGF-D function prevented AA formation in obese mice with AngII infusion. More importantly, we further demonstrated the pivotal role of PDGF-D in AA formation by using an adipocyte-specific PDGF-D transgenic mouse model.

Male 8-week-old ob/ob mice and wild-type (WT) mice were purchased from Slac Laboratory Animal Co., LTD. High-fat diet (HFD)–induced obese mice were fed a diet containing 60% kcal from fat for 4 months before and during the infusion of AngII. Mice in the normal diet group (low-fat diet) were age matched to the HFD mice to control for effects of aging. Mice were infused with PBS or 1,000 ng/kg/min AngII using ALZET Mini-Pumps for 14 days. For in vivo PDGFR-D inhibition experiment, mice were treated via oral gavages (CP673451; 40 ng/kg/day) for 14 days during AngII infusion. Systolic blood pressure was measured by the tail-cuff method, and the average of three pressure readings was obtained. All animals had free access to water and a standard laboratory diet. All animal procedures were approved in accordance with the institutional guidelines established by the Committee of Ethics on Animal Experiments at the Chinese Academy of Sciences.

Whole-mouse genome microarray 4 × 44 K was purchased from Agilent. Data were scanned by an Agilent Microarray Scanner using the default settings and were analyzed by the Shanghai Biotechnology Corporation.

The aortic internal diameter of mice was assessed by a VisualSonics Vevo770 Ultrasound biomicroscope (VisualSonics Inc., Toronto, ON, Canada) with a 30-MHz linear array ultrasound transducer, and the inner lumen diameter at the maximal expanded portion of the aorta was quantified as the internal maximal diameter.

A mouse pdgfd open reading frame was inserted into the PiggyBac transposon gene expression vector after the adipocyteprotein-2 promoter. Then, transgenic mice were generated by vector microinjection to the zygote. The pups were screened by PCR assay. Of 52 pups screened, 13 were identified positive (F0). F1 was generated by each F0 mating with WT mice with the same C57BL/6 background. F1 was identified by PCR and Western blot.

Aortas fixed in formalin and embedded in paraffin were sectioned at 5 μm. Hematoxylin-eosin (H-E) or Masson’s trichrome staining was performed using standard procedures. Immunohistochemical staining of PDGF-D was performed using anti–PDGF-D antibody. Antigen retrieval was obtained by heating the tissue slides in 0.01 mol/L citrate buffer, pH 6.0, at 100°C for 5 min. Immunofluorescence staining of Ki67, CD68, α-smooth muscle actin (α-SMA), and fibroblast-specific protein 1 (FSP1) was performed using anti-Ki67 (ab15580; Abcam), anti-CD68 (MCA1957; Bio-Rad), anti–α-SMA (ab21027; Abcam), and anti-FSP1 (ab124805; Abcam) antibodies, and then fluorescence-conjugated secondary antibodies (Alexa Fluor 555, Alexa Fluor 488; Invitrogen). Sections were mounted in Fluorescence Mounting Medium (Dako). Images of H-E and immunohistochemical staining were obtained by Zeiss microscope. Immunofluorescence staining was examined by a laser-scanning confocal microscope (Zeiss). Images were analyzed with ImageJ and Image-Pro Plus software.

For adipocyte culture, stromal vascular fraction cells in the subcutaneous adipose tissue were separated from adipocyte-specific PDGF-D transgenic (PA-Tg) mice and WT mice by a collagenase digestion method. Then stromal vascular fraction to adipocyte differentiation assay was performed by treatment with DMEM containing 10% FBS, 0.5 mmol/L isobutylmethylxanthine, 0.1 μmol/L dexamethasone, 1 μmol/L rosiglitazone, and 1 μg/mL insulin for 2 days, followed by treatment with 1 μg/mL insulin for 8 days. At day 10, PA-Tg–conditioned medium (PA-Tg-CM) and WT adipocyte–conditioned medium (WT-CM) were collected to culture adventitial fibroblasts. Rat adventitial fibroblasts were isolated from the aortas of Sprague-Dawley rats (weight 120–150 g) and cultured in conditioned medium at 37°C in a humidified atmosphere containing 5% CO₂.

Quantitative real-time PCR (QPCR) was performed using the SYBR Premix Ex Taq kits (TaKaRa) in an ABI PRISM 7900HT System (Applied Biosystems). Succinate dehydrogenase complex subunit A was used as a standard reference. Reactions were performed at 95°C for 30 s followed by 40 cycles at 95°C for 5 s, and at 60°C for 30 s. Mouse tissue or cell extracts containing equal amounts of total protein were resolved by SDS-PAGE followed by immunoblot with the immunoblotting antibodies (e.g., PDGF-D, transforming growth factor-β [TGF-β], SMAD2/3). The chemiluminescence was detected using an ECL detection system.

Migration assays were performed using 6.5-mm-diameter and 8.0-μm-pore size Transwells (Costar) coated with 0.5% gelatin. Adventitial fibroblasts were prepared in serum-free medium, and 4 × 10⁴ cells were added to the upper chamber in migration buffer (DMEM containing 0.1% BSA). After 24 h of AngII or PDGF-D stimulation at 37°C, cells were removed from the upper surface of the membranes with a cotton swab, and cells that migrated to the lower surface were fixed with 4% paraformaldehyde for 30 min and then stained with 0.1% crystal violet (C6158; Sigma-Aldrich) for 10 min. Migrated cells were then counted under a microscope. Cell proliferation assays were performed using Cell-Light EdU Apollo594 In Vitro Imaging Kit (Ribobio Co., LTD.). Images were obtained by a laser-scanning confocal microscope (Zeiss).

Adventitial fibroblasts were seeded into six-well plates and incubated in serum-free DMEM for 24 h. The cells were stimulated for indicated stimulations. Interleukin-6 (IL-6), MCP-1, and tumor necrosis factor-α in the supernate, and plasma was quantified using sandwich ELISA kits according to the protocol provided by the manufacturer (Abcam).

Results were expressed as the mean ± SD. Comparisons of experimental groups were analyzed by Student t test (two groups) or one-way ANOVA followed by the post hoc Dunnett test for data with more than two groups (Levene tests for equal variance). The Dunnett T3 test was used as a post hoc test comparison for the analysis of unequal variances (Welch and Brown-Forsythe test). Probability values <0.05 were considered to be statistically significant.

Initially, we found that ob/ob mice manifest a phenotype of AA when subjected to 2 weeks of infusion of AngII (Fig. 1A). The AA incidence in AngII-infused ob/ob mice was 65% (13 of 20) compared with 5% (1 of 20) in WT mice. Among the 13 occurrences of AA in ob/ob mice, 7 were abdominal AA, 5 were thoracoabdominal AA, and 1 was thoracic AA (Fig. 1B). Approximately 30% (6 of 20) of the ob/ob mice died because of aortic rupture, whereas none of the WT mice died (Fig. 1C). Ultrasound analysis showed aortic expansion in ob/ob mice compared with WT mice after AngII infusion (Fig. 1D). Furthermore, AngII infusion resulted in a comparative increase of media thickness in WT and ob/ob mice, whereas adventitia in ob/ob mice was remarkably thickened compared with WT mice after AngII infusion (Fig. 1E), which is accompanied with an increased macrophage infiltration (Fig. 1F). Next, we also demonstrated that AngII caused AA formation in HFD-induced obese mice, including abdominal AA, thoracoabdominal AA and thoracic AA (Supplementary Fig. 1A–C). Accordingly, AngII-infusion resulted in adventitial thickening and macrophage infiltration in HFD mice (Supplementary Fig. 1D and E). These indicate that adventitia activation may be involved in the AA formation during obesity.

To determine whether obesity-induced PVAT dysfunction is involved in the regulation of AA formation and associated with adventitial remodeling in obese mice, we performed gene expression microarray analysis. As indicated in the volcano plots and heatmap, the PVAT cluster showed large overlaps between WT mice, with or without AngII infusion, and the PVAT of ob/ob mice also shared a similar transcriptional profile, with or without AngII treatment. In contrast, there was a great difference in PVAT between WT and ob/ob mice regardless of infusion with saline or AngII (Supplementary Fig. 2A–C). These suggest that obesity, rather than AngII-induced PVAT dysfunction contributes to AA formation in our model. Heatmap and Circos plots showed different expressions in fatty acid metabolism–related, inflammatory response–related, and fibroblast proliferation–related genes (Supplementary Fig. 2C and D). Since AngII-infused ob/ob mice showed significant adventitial hypertrophy and fibroblasts were considered as the main component of the adventitia, we analyzed the transcriptional profile in regulating fibroblast proliferation. Intriguingly, gene pdgfd, which encodes PDGF-D and plays an important role in activating fibroblasts, was highly expressed in the PVAT of ob/ob mice. Next, we examined PDGF-D expression in cells from every layer of the vascular wall, including endothelial cells, smooth muscle cells, adventitial fibroblasts, and adipocytes. The result indicates that PDGF-D is mostly expressed in the adipocyte (Supplementary Fig. 2E). Further, QPCR, immunohistochemistry, and Western blot analysis confirmed the increased expression of PDGF-D in the PVAT of AngII-infused ob/ob mice (Fig. 2A–C) and AngII-infused HFD mice (Fig. 2D–F). According to the phenotype of hypertrophied adventitia in AngII-infused obese mice, we assumed that PDGF-D may be involved in the regulation of adventitia activation and therefore participates in the pathological process of AA formation in obese mice.

To verify the effect of PDGF-D on AA formation in obese mice, we treated AngII-infused obese mice with PDGF-D receptor antagonist CP673451. As shown, CP673451 markedly diminished the incidence of AA and improved the survival rate in ob/ob mice after AngII infusion (Fig. 3A–C). PDGF-D has been taken as a fibrosis inducer in many disease models (24,25). In accordance, the thickness and fibrosis of adventitia were remarkably reduced with CP673451 treatment in AngII-infused ob/ob mice (Fig. 3D and E). Moreover, adventitia proliferation and FSP1-positive fibroblast invasion were also decreased with CP673451 treatment (Fig. 3F and G). CP673451 treatment also decreased CD68-positive macrophage infiltration in the adventitia, as well as plasma IL-6 and MCP-1 concentration in AngII-infused ob/ob mice (Fig. 3H and I). Consistently, CP673451 treatment also reduced the incidence of AA and improved the survival rate in AngII-infused HFD mice (Supplementary Fig. 3A–C), accompanied by decreased adventitia fibrosis and inflammation (Supplementary Fig. 3D–G). Together, these suggest that the inhibition of PDGF-D function ameliorates adventitial remodeling–related AA formation in AngII-infused obese mice.

To establish a direct link between adipocyte-derived PDGF-D and AA formation, we created PA-Tg mice by using adipocyteprotein-2 promoter (Supplementary Fig. 4A). The mRNA level of pdgfd showed no differences in the aorta, heart, kidney, liver, and lung in PA-Tg mice compared with WT littermates (Supplementary Fig. 4B). In contrast, pdgfd mRNA expression increased clearly in the interscapular brown adipose tissue, subcutaneous white adipose tissue, and PVAT of PA-Tg mice (Supplementary Fig. 4C). The increase of PDGF-D protein in PVAT was also verified by Western blot (Supplementary Fig. 4D). Interestingly, AngII infusion caused a 55% (11 of 20) incidence of AA in PA-Tg mice compared with zero (0 of 20) in WT mice (Fig. 4A and B). Among the 11 occurrences of AA, 6 were abdominal AA and the other 5 were thoracoabdominal AA. Fifteen percent (3 of 20) of PA-Tg mice died as a result of aortic rupture after AngII infusion (Fig. 4C). AngII infusion also resulted in a comparative increase of media thickness in WT and PA-Tg mice, whereas the thickness of adventitia showed a greater increase in AngII-infused PA-Tg mice rather than AngII-infused WT mice. In addition, the adventitia of PA-Tg mice was thicker than WT mice without AngII treatment (Fig. 4D). These results confirm that adipocyte-derived PDGF-D promotes AngII-induced AA formation, including a dramatic adventitial remodeling.

Accordingly, PA-Tg mice displayed grossly enlarged fibrotic adventitial areas, and AngII infusion amplified the adventitial fibrosis (Fig. 5A). Ki67 staining showed much more proliferated cells in the adventitia of PA-Tg mice than in WT mice (Fig. 5B). In the aneurismal section, we also detected FSP1-positive fibroblast invasion (Fig. 5C). Of note, we detected a significant CD68-positive macrophage infiltration in the adventitia of PA-Tg mice, which was dramatically aggravated after AngII infusion (Fig. 5D). Consistently, the plasma levels of IL-6 and MCP-1 were significantly increased in PA-Tg mice compared with WT mice, and had a much greater increase after AngII infusion (Fig. 5E). These suggest that PDGF-D–mediated adventitial fibrosis and inflammation are involved in AA formation in PA-Tg mice after AngII infusion.

To determine the effect of adipocyte-derived PDGF-D on adventitial fibroblasts, we cultured adventitial fibroblasts with WT-CM and PA-Tg-CM. PA-Tg-CM had a threefold increase in PDGF-D concentration compared with WT-CM (Fig. 6D). AngII or PA-Tg-CM, respectively, activated adventitial fibroblast proliferation (Fig. 6A), migration (Fig. 6B), and collagen I expression (Fig. 6C). The combined treatment with PA-Tg-CM and AngII further amplified these effects on adventitial fibroblasts, which indicated the synergistic function of adipocyte-derived PDGF-D and AngII. More importantly, CP673451 inhibited PA-Tg-CM–induced adventitial fibroblast activation, whereas this effect of CP673451 was blunted in AngII-stimulated adventitial fibroblasts. Consistent with the proinflammatory effect in vivo, PA-Tg-CM increased IL-6 and MCP-1 expression in adventitial fibroblasts, and this increase was abolished after CP673451 pretreatment (Fig. 6E). Next, we analyzed the effect of recombinant PDGF-D on cultured adventitial fibroblasts. Likewise, the results confirmed the synergistic function of AngII and recombinant PDGF-D on the regulation of adventitial fibroblast proliferation, migration, collagen expression, and IL-6 and MCP-1 expression (Supplementary Fig. 5A–E). These data indicate that PDGF-D activates adventitial fibroblasts in vitro.

Since the TGF-β signaling pathway plays a pivotal role in the regulation of adventitial fibroblast function, as well as AA formation in vivo (26–28), we detected whether PDGF-D stimulated the TGF-β pathway in these mice. First, we found increased TGF-β and downstream collagen I expression in the aortas of PA-Tg mice, and AngII infusion resulted in a further increase (Fig. 7A and B). Consistently, the downstream Smad2/3 phosphorylation was increased in PA-Tg mice, which showed a much higher level after AngII infusion (Fig. 7B). Similarly, recombinant PDGF-D and AngII had a synergistic effect on the TGF-β/Smad signaling pathway in cultured adventitial fibroblasts (Supplementary Fig. 6A). Accordingly, the inhibition of PDGF-D function by CP673451 treatment attenuated the activation of the TGF-β/Smad signaling pathway in AngII-infused ob/ob mice and PA-Tg mice (Supplementary Fig. 6B and C). These suggest that the TGF-β/Smad pathway activation in adventitial fibroblasts is involved in PDGF-D–mediated AA formation during obesity.

It is well known that blood pressure elevation contributes to AA formation. Herein we showed that blood pressure had no further elevation in ob/ob mice, HFD mice, and PA-Tg mice after AngII infusion. Meanwhile, CP673451 treatment did not regulate blood pressure in these mice (Supplementary Fig. 7A–C). Therefore, PDGF-D–mediated AA formation was independent of blood pressure regulation. We also detected plasma level of PDGF-D in mice. The results indicated that circulating levels of PDGF-D increased in ob/ob, HFD, and PA-Tg mice, but it is not influenced by AngII or CP673451 treatment (Supplementary Fig. 7D–F). Together, our data illustrate that PVAT-derived PDGF-D contributes to AA formation via adventitial remodeling during obesity (Supplementary Fig. 8).

This study mainly reveals that obesity-related PVAT dysfunction contributes to AngII-induced AA formation by secreting PDGF-D, which is identified by transcriptome analysis of PVAT in ob/ob mice. Pharmacological blockade of PDGF-D function successfully reduced AngII-induced AA in obese mice. More importantly, we used PA-Tg mice to confirm the direct role of PDGF-D in AA formation via mediating adventitial fibrosis and inflammation.

Obesity is one of the major causes of morbidity and mortality worldwide, and it is a significant public health burden affecting human beings (1,29,30). The effect of obesity on cardiovascular disease mostly depends on the dysfunction of adipose tissue, especially PVAT (31,32). Herein, we showed that the dysfunction of PVAT contributes to AngII-induced AA in obese mice, which is mediated by PDGF-D–induced adventitial fibrosis and inflammation. PDGF-D is a recently identified member of the PDGF family and participates in the regulation of cardiovascular diseases (33). The PDGF protein family is a potent stimulator for cell proliferation and chemotaxis and is documented to play a major role in cell-cell communication for normal development and pathogenesis (34). Among these proteins, PDGF-D has been well documented to regulate organ fibrosis in different animal models (35). In the cardiovascular system, heart-specific overexpression of PDGF-D induces vascular remodeling, resulting in cardiac fibrosis, dilated cardiomyopathy, and cardiac failure (24). Herein, in addition to promoting adventitial fibrosis, we provided definitive evidence that PDGF-D accelerates adventitial inflammation in obese mice and contributes to AA formation after AngII infusion. Although the mechanism of the differential expression of PDGF-D in lean control and obese mice is unclear, we assumed that the differential expression is a result of the different types of adipocytes (36,37). Chronic inflammation is one of the prime reasons for obesity-related adipose tissue dysfunction, where macrophage activation–mediated inflammatory factor release has been attributed to a central function for fat cell disorders (21,38). Thus, the inhibition of PDGF-D function might be a possible intervention to prevent obesity-related metabolic syndrome.

It is increasingly being accepted that adventitial inflammation plays a key role in the pathogenesis of AA, including infiltration of inflammatory cells in the adventitia and activation of adventitial fibroblasts (39,40). It has been reported that leukocyte-fibroblast interactions in the adventitia potentiate local monocyte recruitment and activation, resulting in AA and aortic dissection (19). Even more, local neutrophil recruitment and activation in adventitia facilitate aortic expansion and lead to aortic rupture eventually (18). Herein we are the first to show a direct link between perivascular adipocytes and adventitial fibroblasts in AA formation. In general, PVAT plays a proinflammatory role in AA development. PVAT-derived proinflammatory factors accelerate the recruitment of macrophages, lymphocytes, and mast cells in the vascular wall (41,42). Although another study (43) revealed that higher thoracic and abdominal aortic dimensions are associated with PVAT volume in a cohort study of 3,001 individuals, supporting the notion that PVAT volume may contribute to aortic remodeling. In our study, a PDGF-D supplement activates primary cultured adventitial fibroblasts and increases the production of IL-6 and MCP-1 in vitro. Accordantly, adipocyte-specific overexpression of PDGF-D in vivo aggravates adventitial fibrosis and inflammation. These indicate a crucial molecular mechanism for adipocyte-derived PDGF-D in adventitial inflammation and AA formation.

Obesity could potentially be an independent risk factor for AA, and meanwhile obesity also has been implicated in the pathogenesis of diabetes. Nevertheless, data from large-scale screenings have shown a paradoxically lower prevalence of AA in patients with diabetes, even those who have been reported as having a negative risk factor for human abdominal AA formation (44). Recently, a report (45) showed that vascular cell division autoantigen 1 (CDA1) plays a role in diabetes to reduce susceptibility to aneurysm. It appears that CDA1 reduces aneurysm severity, but not the onset of aneurysm in this model. A CDA1-mediated TGF-β/Smad pathway may limit the growth of the aneurysm via increasing accumulation of extracellular matrix. On the other hand, our study provides an explanation of how the PDGF-D–induced TGF-β/Smad pathway contributes to the onset of aneurysms via promoting vascular fibrotic response and inflammation during obesity. This explanation is in accordance with the results of a previous study (16) showing that AngII induces abdominal AA in diet-induced obese mice as well as genetic obese mice. It has shown that body weight, but not insulin sensitivity, was a significant predictor of abdominal AA formation in the mouse model (16). Consistently, our results provided experimental evidence for obesity-related AA formation, and a molecular mechanism study proposed that obese fat-derived PDGF-D plays a pivotal role in AA formation. Although epidemic study and basic research have claimed the correlation between obesity and AA, the paradox of obesity and diabetes in AA formation still needs further investigation.

In conclusion, this study shows that obesity-induced PDGF-D in PVAT contributes to AA formation. We illustrate the molecular mechanism that adipocyte-derived PDGF-D promotes adventitial fibrosis and inflammation, which contribute to AA formation during obesity.

Acknowledgments. The authors thank Professor Jiqiu Wang (Department of Endocrinology and Metabolism, Ruijin Hospital, Shanghai, People’s Republic of China) for providing HFD mice.

Funding. This work was supported by the National Natural Science Foundation of China (grants 91539202, 81570221, 81770495, and 91739303), the Shanghai Municipal Commission of Health and Family Planning (grants 2017YQ076 and 201540222), the Shanghai Sailing Program (grant 17YF1415900), and the China Postdoctoral Science Foundation (grant 2017M621504).

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. Z.-B.Z. and C.-C.R. performed the AngII infusion animal study and wrote the manuscript. J.-R.L. and L.X. performed animal tissue collection and analysis. X.-H.C. carried out ultrasound imaging. Y.-N.D. and M.-X.F. analyzed survival curve data. L.-R.K. and D.-L.Z. performed the primary cell culture. P.-J.G. designed and initiated the experiments. P.-J.G. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.