Melanocortin-4 receptor in macrophages attenuated angiotensin II-induced abdominal aortic aneurysm in mice

Obesity is recognized as an independent risk factor for abdominal aortic aneurysm (AAA). While mutations in the melanocortin-4 receptor (MC4R) gene is the most common cause of obesity caused by mutations in a single gene, the link between MC4R function and vascular disease has still remained unclear. Here, by using melanocortin-4 receptor (MC4R) deficient mice, we confirmed MC4R deficiency promotes AAA and atherosclerosis. We demonstrated the contribution of two novel factors towards vascular vulnerability in this model: leptin signaling in vascular smooth muscle cells (VSMCs) and loss of MC4R signaling in macrophages. Leptin was shown to promote vascular vulnerability via PI3K-dependent upregulation of Spp1 expression in VSMC. Additionally, Ang II-induced AAA incidence was significantly reduced when MC4R gene expression was myeloid cell-specifically rescued in MC4R deficient (MC4RTB/TB) mice. Ex vivo analysis showed a suppression in NF-κB activity in bone marrow-derived macrophages from LysM(+);MC4RTB/TB mice compared to LysM(−);MC4RTB/TB mice, which exaggerates with endogenous MC4R ligand treatment; α-MSH. These results suggest that MC4R signaling in macrophages attenuates AAA by inhibiting NF-κB activity and subsequent vascular inflammation.


MC4R deficiency leads to vascular vulnerability, and promotes Ang II-induced AAA in mice
To create MC4R-defecient mice, we used a model wherein loxP-flanked transcriptional blocking (TB) cassette is inserted upstream of the Mc4r ATG site, resulting in the disruption of Mc4r expression (Fig. 1A) 13,14 .Western diet (WD)-fed MC4R TB/TB mice exhibited significantly higher bodyweight (Fig. 1B) and sBP (Fig. 1C) in comparison to WD-fed MC4R +/+ mice.While WD-fed MC4R TB/TB mice did not show any presence of AAA, histochemistry analysis revealed an increase in the intima-media area and elastin break number in the aorta of them, suggesting vascular vulnerability (Fig. 1D-F).Moreover, several inflammatory-related genes were upregulated in the aorta of WD-fed MC4R TB/TB mice (Fig. 1G).
Further, to determine whether these findings in MC4R TB/TB mice contribute towards AAA, we first treated WD-fed MC4R TB/TB and MC4R +/+ mice with Ang II (1000 ng/kg/min-usual dose to induce AAA in mice) for 4 weeks 15,16 .However, this dose resulted in a mortality rate of more than 80% in WD-fed MC4R TB/TB , primarily due to the AAA rupture within the first 2 weeks (data not shown), suggesting that WD-fed MC4R TB/TB mice were markedly susceptible for AAA.Therefore, to induce AAA, we chose 500 ng/kg/min of Ang II instead of 1000 ng/kg/min.
While higher bodyweight and sBP were observed in MC4R TB/TB compared to the MC4R +/+ mice (Supplementary Fig. 1A,B), higher elastin break numbers and AAA incidence (Supplementary Fig. 1C,H,I,K) along with a lower survival rate (Fig. 1J) were found in MC4R TB/TB mice in comparison to MC4R +/+ mice fed a WD.An increased AAA incidence and diameter were observed in MC4R TB/TB mice compared to MC4R +/+ mice, even after matching of the sBP levels in MC4R TB/TB and MC4R +/+ mice using hydralazine (Supplementary Fig. 1D-F).Additionally, compared to MC4R +/+ mice, MC4R TB/TB mice showed increased inflammatory-related gene expression (Fig. 1L), as well as the positive F4/80 immunostaining area (Supplementary Fig. 1G,H) in the aorta.

MC4R deficiency promotes Ang II-induced AAA formation via leptin-dependent and -independent mechanisms
Next, to dissect the molecular mechanism by which MC4R signaling is involved in the development of vascular diseases, we determined to focus on Ang II-induced AAA model because this model is considered to take shorter time period to conduct than atherosclerosis model with the background of ApoE −/− ;MC4R TB/TB .In the arcuate nucleus of the hypothalamus, MC4R plays a pivotal role in mediating appetite suppression via the effect of leptin 17 :Consistently, MC4R deficiency resulted in hyperleptinemia in the peripheral blood, possibly via negative feedback (Fig. 3A).We also confirmed that leptin and several proinflammatory genes are significantly upregulated in abdominal perivascular adipose tissue (PVAT) of WD-fed MC4R TB/TB mice compared to that of WD-fed MC4R +/+ mice (Fig. 3B).Since leptin has been reported to stimulate vascular inflammation, oxidative stress, and vascular smooth muscle hypertrophy that may contribute to hypertension, vascular injury, and AAA 18 , we investigated whether leptin contributes to vascular vulnerability as observed in MC4R TB/TB and ApoE −/− ;MC4R TB/TB mice by comparing the three groups of mice including leptin-deficient ob/ob mouse; ob/ ob;MC4R TB/TB , MC4R TB/TB , and ob/ob mice.
Among the three mice fed a standard diet 19 , MC4R TB/TB mice exhibited a slight decrease in bodyweight (Fig. 3C) and higher sBP (Fig. 3D).Moreover, MC4R TB/TB mice showed the highest triglyceride levels (Supplementary Fig. 3B) as well as the highest glucose and serum insulin levels (Supplementary Fig. 3C,D).After treatment with Ang II (500 ng/kg/min), a significant decline in the incidence and size of AAA (Fig. 3E-G), and an increase in the survival rate of ob/ob;MC4R TB/TB mice were observed compared to MC4R TB/TB (Fig. 3E-H).These in vivo findings indicate towards the contribution of leptin in promoting Ang II-induced AAA.In contrast, compared to these in ob/ob mice, AAA incidence and size were found to be higher in ob/ob;MC4R TB/TB , indicating towards a possibility of MC4R inhibiting Ang II-induced AAA in leptin-independent manner.To confirm leptin-dependent and -independent contributions, we determined to do following experiments.

The vascular vulnerability of MC4R TB/TB mice is mediated by an osteopontin receptor CD44
First, we elucidated a leptin-dependent mechanism by which MC4R deficiency promotes vascular vulnerability.In non-alcoholic steatohepatitis (NASH) fibrosis, it has been reported that leptin promotes fibrosis via upregulation of an extracellular matrix glycoprotein, osteopontin (OPN) in hepatic stellate cells (HSCs) 20 .Leptin is also known to upregulate OPN gene (Spp1) expression in vascular smooth muscle cells (VSMCs) 21 .Additionally, OPN promotes Ang II-accelerated atherosclerosis and aneurysm in mice 22 .
In primary culture of VSMCs from WT mice, leptin upregulated Spp1 expression, with its effect abolishing by the phosphoinositide 3-kinase inhibitor LY294002 (Fig. 4A) as previously confirmed in HSCs 20 .Immunostaining of the aorta revealed the stronger OPN expression in the vascular walls of MC4R TB/TB mice than observed for the MC4R +/+ mice (Fig. 4B,C).Moreover, OPN-expressing cells colocalized with α-smooth muscle actin-expressing cells (Fig. 4D).Consistent with the observation, plasma leptin concentration was significantly higher in MC4R TB/TB mice than in MC4R +/+ mice and correlated with plasma OPN concentration in MC4R +/+ and MC4R TB/TB mice (Supplementary Fig. 4A,B).Further, To inhibit OPN action in vivo, we intraperitoneally administered neutralizing antibody against CD44, a cell-surface adhesion receptor of OPN 23 , into Ang II-infused MC4R TB/TB mice.The neutralizing antibody significantly inhibited the Ang II-induced AAA in MC4R TB/TB mice independent of the sBP (Fig. 4E-H).

MC4R gene reconstitution in myeloid cells suppresses Ang II-induced AAA in MC4R TB/TB mice
MC4R is reported to predominantly express in the brain and plays an important role in the regulation of appetite in humans and rodents 6,17 .However, MC4R is recently reported to express in enteroendocrine L cells and regulate peptide YY and glucagon-like peptide 1 release in mice 24,25 .This finding points toward a notion that MC4R expressed in peripheral cells may directly plays an inhibitory role during AAA development.
Western blotting of various tissues in WT mice indicated that MC4R was expressed in the RAW 264.7 macrophage cell line and bone marrow (BM)-derived macrophages, as well as in the brain, liver, and small intestine (Supplementary Fig. 5A).Therefore, to examine the role of MC4R in macrophages, we created myeloid cellspecific MC4R gene reconstitution mice by crossing MC4R TB/TB and LysozymeM (LysM)-Cre expression mice (Supplementary Fig. 5B).We successfully confirmed the MC4R gene expression in BM-derived macrophages from LysM (+);MC4R TB/TB mice (Fig. 5A).On stimulation with α-MSH, an internal ligand of MC4R, while a significant increase in the intracellular cyclic AMP levels in BM-derived macrophages from LysM (+);MC4R TB/TB mice was observed, it did not happen in BM-derived macrophages from LysM (−);MC4R TB/TB mice (Fig. 5B), suggesting a functional expression of MC4R in macrophages.MC4R was colocalized with macrophage marker CD68-expressing cells in the aorta in ApoE −/− mice (Fig. 5C).In vivo, LysM (+);MC4R TB/TB mice exhibited reduction in Ang II-induced AAA incidence and diameter compared to LysM (−);MC4R TB/TB mice (Fig. 5D-F), which was independent of sBP (Supplementary Fig. 5C).
Further, a microarray analysis in BM-derived macrophages from LysM (+);MC4R TB/TB and LysM (−);MC4R TB/TB mice revealed nuclear factor-κB (NF-κB) signaling as a significantly ranked pathway (Fig. 5G).LPS-induced NF-κB activity was also significantly suppressed in BM macrophages from LysM (+);MC4R TB/TB mice (Fig. 5H).These findings indicate that MC4R in macrophages has a significant contribution in suppressing Ang II-induced AAA, probably via the inhibition of NF-κB activity.In macrophages obtained from LysM (+);MC4R TB/TB mice, LPS-induced protein kinase A, IκB, and p65 phosphorylation were significantly attenuated by the α-MSH pretreatment (Fig. 6A, Supplementary Fig. 6A).However, this effect was not observed in macrophages from LysM (−);MC4R TB/TB mice.In ApoE −/− mice, intraperitoneal α-MSH treatment significantly reduced the incidence and diameter of Ang II-induced AAA (Fig. 6B-D), which was accompanied with reduced expression of pro-inflammatory genes in the aorta (Fig. 6E).In contrast, the incidence or diameter of Ang II-induced AAA in MC4R TB/TB mice was not affected by α-MSH treatment (Fig. 6E-H), which was consistent with our ex vivo observation (Fig. 6A).These results demonstrate the possibility that α-MSH exerts its effects through the activation of MC4R signaling and suppression of NF-κB activity and its downstream pro-inflammatory genes expression in macrophages.

Significant correlation of plasma α-MSH concentration in diabetic subjects with atherosclerosis and MC4R gene expression levels in monocytes
As we had already collected the monocytes and plasma samples from human subjects with and without type 2 diabetes mellitus (T2DM) in our previous study 26 , we also investigated the clinical significance of the MC4R gene expression in monocytes and plasma α-MSH concentration in humans (Table 1).We observed a significant negative correlation between the MC4R gene expression in monocytes and the plasma α-MSH concentration both in subjects without and with T2DM (Fig. 6I,J).Furthermore, among subjects with T2DM, the plasma α-MSH concentration was higher in those with carotid plaques than in those lacking it (Supplementary Fig. 6B) and correlated positively with brachial-ankle pulse-wave velocity (Supplementary Fig. 6C).These findings raise the potential that in diabetic subjects with atherosclerosis, plasma α-MSH concentration is elevated to compensate for MC4R signaling in monocytes.Moreover, in subjects with thoracic aortic aneurysm and/or AAA (Table 2), macrophages-like cells obtained from lesions in aneurysm by laser capture microdissection showed lower MC4R gene expression than those obtained from non-lesions in aorta (Supplementary Fig. 7A,B).

Discussion
In this study, we demonstrated that loss of MC4R exacerbates vascular diseases, such as AAA and atherosclerosis by using Mc4r-deficient mice.Furthermore, we confirmed MC4R in macrophages attenuates AAA by suppressing NF-κB activity and subsequent vascular inflammation.
In this study, we first confirmed the occurrence of a leptin-dependent mechanism explaining the high AAA occurrence in WD-fed MC4R deficient mice with Ang II infusion.Locally applied leptin in periaortic area is known to augment the medial MMP-9 synthesis and aortic aneurysm size, implying that PVAT derived leptin may act as a factor promoting AAA development induced by obesity 4 , however, the exact mechanism remains unknown.We demonstrated that leptin exerts its effect through PI3K-dependent upregulation of the Spp1 gene expression in VSMCs.When OPN receptors in VSMCs were neutralized with the anti-CD44 antibody, it significantly suppressed the incidence and size of Ang II-induced AAA, suggesting that leptin, elevated by MC4R deficiency, promotes vascular vulnerability via OPN induction in VSMCs.However, whether PVAT-derived or circulating leptin regulates this need further investigation.Consistent with our findings, a previous study reported a significant increase in OPN expression levels in the VSMCs and medial thickening in the OPN-Tg mice than in non-Tg mice 27 .Moreover, their zymographic analysis confirmed the increased matrix metalloproteinases-2 or -9 in the OPN-Tg aorta, leading to vascular vulnerability.It is also worth noting that metformin downregulates OPN expression in aorta and inhibits Ang II-induced AAA formation 28 .Further studies are needed to determine whether inhibiting OPN action, especially in hyperleptinemic subjects, can be promising strategy to prevent AAA formation.
Importantly, the protective role of MC4R signaling in vascular damage was highlighted when a significant difference in AAA size and survival rate was observed between ob/ob;MC4R TB/TB and ob/ob mice.Additionally, we could successfully demonstrate that myeloid cell-specific rescue of the MC4R gene expression significantly suppressed the Ang II-induced AAA incidence in mice.To the best of our knowledge, this is the first report to suggest that MC4R is functionally expressed in monocytes and macrophages and possibly involved in protecting from AAA in vivo.While further investigation is still needed to validate, we could detect MC4R protein expression on vascular lesion-infiltrating macrophages in ApoE KO mice.In this study, we also confirmed the MC4R's function in BM-derived macrophages by showing an increase in the intracellular cAMP upon α-MSH treatment, which leads to suppression of LPS-induced phosphorylation of the p65 and NF-kB activity.In 2014, MC4R was reported to be expressed in enteroendocrine L cells and regulated peptide YY and glucagon-like peptide 1 (GLP-1) release in mice 24 .Moreover, MC4R was localized in L cells and regulated GLP-1 and peptide www.nature.com/scientificreports/YY secretion from ex vivo human intestine.Additionally, fasting peptide YY levels and oral glucose-induced GLP-1 secretion were reduced in humans carrying a total loss-of-function MC4R mutation 25 .Although, a study reported that MC4R expression in L cells is marginal and may not affect GLP-1 secretion in mice 29 , it does not rule out the possibility of MC4R being functionally expressed in monocytes and macrophages and contributing significantly to protect against AAA.Further research is required to confirm its detailed functional expression in macrophages and determine precise crosstalk between MC4R and NF-κB signaling pathways.
In addition to the finding of MC4R's involvement in attenuating AAA in mice, we found that α-MSH treatment could also inhibit Ang II-induced AAA incidence, which was accompanied by the reduction in inflammatory-related-genes expressions in the aortic lesion.Importantly, these beneficial effects of α-MSH were observed in ApoE −/− but not in MC4R TB/TB mice.Although the possibility is low, this may be attributed to the α-MSH-MC4R signaling in central nervous system and further research is needed to confirm whether MC4R signaling particularly in macrophages is responsible α-MSH's beneficial impact on cardiovascular phenotype.Moreover, α-MSH is known to exhibit anti-atherosclerotic effects by binding to MC1R in the endothelial cells or macrophages 30,31 .However, here, exogenous α-MSH administration did not affect Ang II-induced AAA formation in MC4R TB/TB mice, suggesting that α-MSH inhibited Ang II-induced AAA formation through MC4R.www.nature.com/scientificreports/There are several limitations in this study.First, since ligands other than OPN can also bind with CD44, we can't deny the possibility that the results of anti-neutralizing antibody treatment against CD44 were affected by inhibition of other ligand's binding with CD44.Second, since other pathways such as cell cycle regulation and cholesterol metabolism were ranked higher than NF-κB signaling in Fig. 5G, it is possible that these pathways contributed to inhibition of Ang II-induced AAA development in LysM (+);MC4R TB/TB mice.Since It is reported that surplus cholesterol efflux from macrophage is antiatherogenic 32 , the link of cholesterol metabolism and MC4R in macrophage should be addressed in further research.Third, hydralazine is known to stimulate sympathetic nerve system in addition to antihypertensive effects.Therefore, it is still possible that this effect might affect the AAA incidence in hydralazine-treated MC4R TB/TB mice.Forth, we only used male mice in this study, these observations should not be extrapolated to the females as they may vary with the sex of animal.Lastly, the sole use of the tail-cuff method to monitor blood pressure can be considered as another drawback, as telemetry is usually recommended for measuring subtle BP changes or sympathetic tone.However, because neither slight BP changes nor sympathetic tone are the main parameter being examined in this research, we think tail-cuff measurement is sufficient for this investigation.Despite these limitations, our study demonstrates the functional relevance of MC4R in macrophages.The synthetic agonist of MC4R, setmelanotide, was recently approved by the U.S. Food and Drug Administration to be used for chronic weight management in genetic obesity including leptin receptor deficiency 19 .Therefore, to explore the clinical potential of MC4R, further research investigating its actions in peripheral tissues/cells such as vascular and macrophages is needed.Correlation between the plasma α-MSH concentration and MC4R gene expression levels in cultured monocytes isolated from the peripheral blood of (I) healthy controls (n = 50) and (J) patients with T2DM (n = 33).*P < 0.05, **P < 0.01, ***P < 0.001.

Table 1 .
Characteristics of control subjects and patients with type 2 diabetes.The data are expressed as mean ± SEM or median and interquartile range.The data were compared using Student's t-test or the Mann-Whitney U test.T2DM type 2 diabetes, BMI body mass index, SBP systolic blood pressure, DBP diastolic blood pressure, FPG fasting blood glucose, HbA1c hemoglobin A1c, IRI immunoreactive insulin, HOMA-IR homeostasis model assessment of insulin resistance, TG triglycerides, HDL-C high-density lipoprotein cholesterol, LDL-C low-density lipoprotein cholesterol, AST aspartate transaminase, ALT alanine transaminase, γ-GT γ-glutamyl transferase, eGFR estimated glomerular filtration rate, hsCRP high-sensitivity C reactive protein, ABI ankle-brachial index, baPWV brachial-ankle pulse-wave velocity, IMT intima-media thickness, N/A not available.