Objective: Myeloperoxidase (MPO) is a well described heme-enzyme predominantly expressed in neutrophils and monocytes, being involved in inflammatory reactions by producing hypochlorous acid (HOCl) during oxidative burst. There is evidence for an increase in pro-inflammatory signaling in obesity-related pathological processes such as atherosclerosis and abdominal aortic aneurysm. Aim of this study was to evaluate endothelial function and the role of MPO as an inflammatory surrogate in obesity.

Methods and Results: Plasma, peritoneal white adipose tissue (AT), abdominal aorta and aortic perivascular adipose tissue (PVAT) were harvested from leptin-deficient obese mice (ob/ob) and wild type mice (WT) 10-12 weeks of age.  In addition, we analyzed MPO plasma levels in patients undergoing bariatric surgery. MPO levels were measured using MPO ELISA (Hycult Biotech). Isolated abdominal aorta was stimulated with acetylcholine (ACh) or nitroglycerine (NTG) in organ bath experiments to assess endothelial-dependent and smooth muscle cell (SMC)-mediated vascular relaxation, respectively. In ob/ob mice, plasma MPO levels were significantly elevated as compared to WT mice (797,6 ng/ml vs. 447,9 ng/ml). Moreover, a significant increase in MPO concentration was found in peritoneal white fat tissue of ob/ob mice (1,51 ng/mg vs. 5,25 ng/mg). In the abdominal aortic wall, no difference in MPO level was shown between both genotypes, whereas MPO concentration was notably higher in abdominal PVAT of ob/ob- as compared to WT mice (51,58 ng/mg vs. 32,32 ng/ml). In patients undergoing bariatric surgery with a weight loss of nearly 20%, we could show a significant decrease in plasma MPO levels three months after surgery compared to controls (n=46; 271,39 ± 93,19 pmol/l MPO vs. 229,32 ± 77,08 pmol/l MPO, p=0,03). Organ bath experiments revealed an attenuation of ACh-induced vascular relaxation in the isolated aorta of ob/ob mice, indicating an endothelial dysfunction. Interestingly, no significant difference was seen in SMC-mediated vascular relaxation between WT and ob/ob mice after NTG stimulation. In addition, 3T3 fibroblasts were differentiated to adipocytes in vitro by application of insulin, dexamethasone and rosiglitazone. Treatment with MPO and H₂O₂ induced inflammatory activation of adipocytes as measured by Il-6 cytokine mRNA expression after 24 hours. Furthermore, stimulation with MPO led to a significantly decreased expression of brown adipose tissue (BAT) marker P2RX5, indicating a phenotype switch towards a white adipose tissue (WAT)-like phenotype of adipocytes.

Conclusion:  MPO is elevated in plasma, white adipose tissue and PVAT of  ob/ob mice as well as in plasma from obese patients and show a significant decrease associated with weight loss after bariatric surgery. Our results indicate that MPO might contribute to a phenotype switch from brown-AT to white-AT like adipocytes in PVAT, mediating a pro-inflammatory status with impaired endothelial function in obesity-induced vascular inflammation and dysfunction.