Perirenal Fat Promotes Renal Arterial Endothelial Dysfunction in Obese Swine through Tumor Necrosis Factor-a - Beyond the Abstract
Dysfunction of the renal arterial endothelium predicts the development of renal damage in several animal models.Vascular endothelial dysfunction can be substantially altered by adjacent adipose tissue, especially perivascular fat.9 Ectopic fat in obesity is characterized by elevated expression of TNF-a, which contributes to endothelial dysfunction in resistance arteries embedded in epididymal fat. Kidney derived TNF-a also promotes angiotensin II induced hypertension.11 Yet to our knowledge the potential paracrine effect of perirenal fat TNF-a on endothelial function of the adjacent renal arteries remains unknown. Therefore, this study was designed to test the hypothesis that perirenal fat in pigs with ObM directly impairs endothelial function of the renal artery and this effect is partly mediated by TNF-a.
Animal Protocols
This study was approved by the Mayo Clinic institutional animal care and use committee. A total of 14, 3-month-old female domestic pigs were randomized to 2 groups, including 7 pigs with ObM and 7 lean controls. Lean pigs were fed regular chow while ObM pigs started a high fat/ high fructose diet (Purina Test Diet® 5B4L), which contains 17% kcal protein, 20% kcal complex carbohydrates, 20% kcal fructose and 43% kcal fat (lard, and hydrogenated soybean and coconut oils).
At 16 weeks of the diet the pigs were anesthetized with 0.5 gm intramuscular ketamine and xylazine. Anesthesia was maintained with intravenous ketamine (0.2 mg/kg per minute) and xylazine (0.03 mg/kg per minute). Arterial blood pressure was monitored using an intra-arterial catheter and an intravenous glucose tolerance test was performed. Hemodynamics, function and visceral fat volumes in each kidney were then assessed by MDCT. Renal venous and arterial blood was collected as were urine samples. The animals were allowed a 3-day recovery period and were subsequently sacrificed by a lethal intravenous dose of pentobarbital (Fatal-Plus®, 100 mg/kg).
The kidneys and perirenal fat were removed via a retroperitoneal incision and immediately dissected. Sections were frozen in liquid nitrogen and maintained at e80C. The renal arteries and a small piece of perirenal fat were collected immediately for ex vivo organ bath experiments.
Multidetector Computerized Tomography
At 16 weeks of the diet MDCT was performed to assess kidney volume, RBF and GFR as described previously. Briefly 160 consecutive scans were performed following a central venous injection of iopamidol (0.5 ml/kg 2 seconds). The procedure was repeated after a 15-minute interval and toward the end of 10-minute intra-aortic infusion of Ach (5 mg/kg minute) into a tracker catheter placed above the renal arteries to test endothelium dependent microvascular reactivity in vivo. MDCT images were reconstructed and displayed with ANALYZE software. RBF and GFR were measured from tissue time attenuation curves obtained in regions of interest selected from the aorta, renal cortex and medulla. Renal vascular resistance was calculated by dividing mean arterial pressure by RBF. Intra-abdominal adipose tissue was measured and expressed as volume and fraction, and perirenal fat volume is shown as the ratio to the volume of the corresponding kidney as described previously.
Organ Bath Experiments
Renal arteries were dissected and cut into 5 mm long segments in cold Krebs solution containing 119 mmol/l NaCl, 25 mmol/l NaHCO3, 4.7 mmol/l KCl, 1.2 mmol/l KH2PO4, 1.2 mmol/l MgSO4, 2.5 mmol/l CaCl2 and 11.1
mmol/l glucose, pH 7.4. The renal artery rings were mounted between 2 hooks attached to an isometric force transducer with continuous recording of tension. They were immersed in organ chambers containing 10 ml Krebs solution gassed with 95% O2 and 5% CO2 at 37C. Resting tension was adjusted to 2 gm and the bath solution was replaced every 15 minutes. After a 60-minute equilibra- tion period the viability of each ring was tested by expo- sure to high Kþ Krebs solution (60 mmol/l KCl). The rings were allowed to rest in the organ bath for 30 minutes before any drugs were administered.
The integrity of the endothelium was assessed by relaxation to 10e6 mol/l Ach. Incremental doses of Phe (10e9 to 10e5 mol/l) were used to test the contractile response. To measure the relaxation response a sustained contraction was induced by adding 3 mmol/l Phe. Re- laxations caused by incremental doses of Ach (10e9 to 10e4 mol/l) and SNP (10e9 to 10e4 mol/l) were recorded.
A group of rings was incubated for 30 minutes in 10 ml Krebs solution in the absence or presence of fresh peri- renal fat (100 gm) obtained from allogeneic ObM or lean pigs with or without co-incubation with neutralizing anti- TNF-a antibodies (20 mg/ml), the superoxide radical scavenger Tempol (1 mmol/l) or the TXA2 synthase in- hibitor dazoxiben (1 mmol/l). Vasorelaxant responses are expressed as the percent reduction in tone induced by Phe. Vasoconstriction responses are expressed as the percent contraction in tone induced by 60 mmol/l KCl.
Blood and Urine Assays
Before MDCT an intravenous glucose tolerance test was performed. HOMA-IR (fasting plasma glucose X fasting plasma insulin/22.5) served as the index of insulin resistance. Urine was collected by bladder puncture and blood samples were obtained from the inferior vena cava and the renal artery. Urine microalbumin (Arbor Assays®), serum creatinine and lipid profile (Hoffman-La Roche, Basel, Switzerland) were assessed, including tri- glycerides, total cholesterol, LDL-C and HDL-C. Renal venous and arterial plasma levels of inflammatory cytokines were measured by Luminex® assays, including TNF-a, IL-6, monocyte chemoattractant protein-1 and interferon-g. Cross-kidney gradients of renal venous and arterial levels of these cytokines were calculated and net release from the kidney was calculated as (venous e arterial levels) X RBF.
Tissue Inflammation
MØ infiltration in perirenal fat was assessed by immunofluorescence staining of the MØ marker CD68 (1:50, Abcam). The subpopulations were assessed by coexpression of inducible nitric oxide synthase (1:50) for proinflammatory M1-MØ or arginase-1 (1:50, Abcam) for reparative M2-MØ. Positive and double positive cells were counted. TNF-a and IL-6 immunoreactivity in perirenal fat tissue was examined using specific antibodies (1:50, Abcam). mRNA expression was determined by quantitative polymerase chain reaction. TNF-a and IL-6 immunoreactivity was calculated using a computer aided image analysis program. On each slide immunostaining was quantified in 15 to 20 fields and expressed as the percent of staining of the total surface area. The results of all fields were averaged.
Oxidative Stress
In situ production of superoxide anion in perirenal fat tissues was evaluated by fluorescence microscopy after DHE staining. The ratio of positively stained areas (red) to nuclear stained areas (blue) was calculated.
Statistical Analysis
Results are shown as the mean SE. Statistical analysis was performed using JMP, version 8.0. Comparisons between groups were made with the unpaired t-test and comparisons within groups were made with the paired t-test and repeated measures ANOVA. Statistical significance for all tests was considered at p <0.05.
Animal Characteristics
After 16 weeks of the diet the pigs in the ObM group had higher body weight (p <0.01), systolic blood pressure (p <0.01), HOMA-IR index (p <0.05) and cholesterol levels (p <0.01, fig. 1, A to D). Bilateral perirenal fat tissues reflected by the ratio of perirenal fat volume to kidney volume were markedly higher in the ObM group than in the lean group (each p <0.01, fig. 1, E and F). Additionally the intra-abdominal fat volume fraction of ObM pigs was greater than that of lean pigs (p <0.01, fig. 1, G). Collectively ObM pigs showed typical characteristics of obesity-metabolic syndrome with a cluster of metabolic disorders including overweight, visceral adiposity, hypertension, insulin resistance and dyslipidemia.
Renal Hemodynamics and Function
Baseline and Ach induced RBF were significantly higher in the ObM group than in the lean group (each p <0.01, fig. 1, H). Compared with baseline Ach increased RBF similarly in ObM pigs (34%, p <0.05) and lean pigs (35%, p <0.01). Renal vascular resistance was significantly lower in ObM pigs than in lean pigs before and after Ach injection (each p <0.05, fig. 1, I ). It decreased in both groups in response to Ach. GFR was markedly increased (p <0.01, fig. 1, J ) while plasma creatinine was decreased (p <0.05, fig. 1, K), probably secondary to glomerular hyperfiltration.14 The urinary protein level was unchanged in the ObM group compared to the lean group (data not shown).
Renal Arterial Endothelial Function Ex Vivo
Ach induced, endothelium dependent relaxation of the renal artery ex vivo was significantly impaired in ObM pigs compared to lean pigs (fig. 2, A). SNP induced endothelium independent relaxation and Phe induced vasoconstriction were similar in the 2 groups (fig. 2, B and C). Ach induced relaxation of renal arteries from lean pigs was unaffected by 30-minute incubation with perirenal fat from lean pigs (fig. 2, D). In contrast it was significantly attenuated after incubation with perirenal fat from ObM pigs (fig. 2, D). Similarly Ach induced relaxation of the ObM renal artery was not influenced by incubation with perirenal fat from lean pigs but it was impaired by incubation with perirenal fat from ObM pigs (fig. 2, E). In contrast SNP induced relaxation of lean and ObM renal arteries was not affected by perirenal fat from lean or ObM pigs (fig. 2, F and G). Importantly co-incubation of ObM rings with neutralizing anti-TNF-a antibodies but not with the antioxidant Tempol or the TXA2 synthase inhibitor dazoxiben normalized endothelial function of the renal artery, which had been impaired by ObM perirenal fat (fig. 2, H).
Perirenal Fat Inflammation and Oxidative Stress
The infiltration of pro-inflammatory M1-MØ (p <0.01) but not of anti-inflammatory M2-MØ increased in ObM perirenal fat compared with lean perirenal fat (fig. 3, A to C). Immunoreactivity and mRNA expression of TNF-a but not of IL-6 were several fold higher in ObM perirenal fat (fig. 3, D to H). In contrast neither the renal arterial or the renal venous blood level, or kidney release of theinflammatory markers TNF-a, IL-6, monocyte chemoattractant protein-1 and interferon-g was increased in the ObM group (see table). DHE fluorescent staining was increased in perirenal fat from the ObM group compared with the lean group (p <0.01, fig. 4, A and B), suggesting increased in situ superoxide production.
This study shows that ObM pigs have expansive perirenal fat with increased M1-MØ infiltration, TNF-a expression and oxidative stress compared with lean pigs. Moreover perirenal fat of ObM but not of lean pigs directly impairs renal arterial endothelial function, which can be improved by neutralizing TNF-a. These observations suggest that perirenal fat exerts paracrine effects on the renal circulation in obesity and implicate fat produced TNF-a in these effects.
Our previous studies have demonstrated that swine ObM induces renal hyperperfusion and glomerular hyperfiltration. They may result from increased cardiac output and hyperinsulinemia, and have been viewed as early pathogenic events in the development of CKD. This study shows that despite the increased RBF and glomerular hyperfiltration in ObM pigs endothelial function of the renal arteries was significantly impaired. Endothelial dysfunction is a predictor of renal vascular and parenchymal diseases. However, the mechanism underlying the ObM associated renal artery endothelial dysfunction remained largely unknown.
Adipose tissue is no longer considered a passive energy storage tissue but rather an active paracrine and endocrine tissue. Ectopic fat has a critical role in physiological and pathophysiological regulation of its adjacent organ. In particular the epicardial and coronary perivascular fat is linked to the development of cardiovascular disorders. The adipose tissues surrounding the kidney, including perirenal and renal sinus fat, have received increased attention during recent years. The amount of perirenal fat correlates well with central obesity, poor blood pressure control and development of CKD. However, to our knowledge the causal relationship of perirenal fat with renal vascular and parenchymal disorders is yet to be established.
Adipose tissue imparts protective and deleterious effects on the vascular function of adjacent arteries. In lean healthy subjects perivascular fat exerts an anticontractile effect on adjacent arteries, which is lost in obesity, metabolic syndrome and diabetes. Coronary perivascular adipose tissue from obese Ossabaw pigs in fact potentiates contractions of the coronary vascular smooth muscle.
Similarly our study shows that the perirenal fat of ObM but not of lean pigs impairs endothelial function of the renal artery while endothelium independent relaxation remains intact, suggesting preserved vascular smooth muscle cell responses. The differential effect of ObM fat may be due to fat inflammation and increased oxidative stress elicited by metabolic derangements. Indeed compared with lean pigs we found that perirenal fat from ObM pigs increased pro-inflammatory M1-MØ infiltration, enhanced TNF-a protein and mRNA expression (suggesting local production), and elevated in situsuperoxide production, indicating a local inflammatory and oxidative state.
As observed in many fat depots macrophages infiltrating perirenal fat early during evolution of obesity likely release TNF-a, which in turn causes inflammation and oxidative stress, which can each induce endothelial dysfunction. Notably neutralization of TNF-a attenuated perirenal fat induced endothelial dysfunction of the renal artery ex vivo whereas neither the antioxidant Tempol nor a TXA2 synthase inhibitor did so. We interpret these results to suggest that in our model inflammation had a greater role in mediating fat induced endothelial dysfunction than did superoxide anion or the vasoconstrictor TXA2. Indeed TNF-a exerts pleiotropic effects on the vasculature, including downregulating endothelial nitric oxide synthase expression and limiting nitric oxide bioavailability.
Several previous studies demonstrated that intravascular administration of the TNF-a neutralizing antibody infliximab reduced blood pressure and ameliorated endothelial function in patients with rheumatoid arthritis. The renal arterial blood levels of inflammatory markers were not increased in ObM pigs, arguing against endocrine effects on endothelial function of the renal artery. Similarly the unchanged renal venous blood levels or release of these inflammatory markers in ObM pigs compared to lean pigs suggests that kidney production is unlikely to be the source of the TNF-a which impairs the function of the renal artery. These observations support the contention that perirenal fat derived TNF-a may affect the renal artery in a paracrine manner.
Interestingly RBF and renal vascular resistance responses to Ach in vivo, which reflect microvascular endothelial function, were relatively preserved in ObM kidneys and might initially partly compensate for the impaired reactivity of the large renal arteries. Notably renal vascular resistance is regulated chiefly by intrarenal microvessels and, thus, it may compensate for abnormal renal arterial vasoreactivity. Nonetheless blunted vasorelaxation represents a variety of endothelial functions that might be impaired in ObM, such as antiinflammatory and anti-thrombogenic activities, which may still impact the renal artery locally. A similar dissociation between coronary epicardial and microvascular endothelial function can be observed in patients with early atherosclerosis. However, microvascular dysfunction may well decrease in more advanced or prolonged ObM.
The major limitation of this study is the challenge of demonstrating the direct effect of perirenal fat on the renal artery in vivo. While perirenal fat may damage renal vascular endothelial function through TNF-a, the precise paracrine mechanism also warrants further studies. The mediators of Ach induced dilatation were not further identified due to the limited renal arterial samples so that additional mechanisms may contribute to endothelial dysfunction in this model. In addition, given that adipocyte size increases in obesity, the same mass of perirenal fat used to treat the renal arteries ex vivo might have contained different tissue proportions of adipocytes, vessels and interstitial cells in lean vs ObM pigs. Furthermore, our pigs were exposed to a 16-week diet. A longer duration in humans may accelerate the deterioration of renal function.
Under ObM conditions perirenal fat shows increased volume, inflammation and oxidative stress, and causes endothelial dysfunction of renal arteries, which is likely mediated in paracrine fashion by TNF-a. Blockage of TNF-a may prove to be an important strategy to combat ObM related renal endothelial dysfunction.
Written By: Shuangtao Ma, Xiang-Yang Zhu, Alfonso Eirin, John R. Woollard, Kyra L. Jordan, Hui Tang, Amir Lerman and Lilach O. Lerman*
From the Divisions of Nephrology and Hypertension (SM, XYZ, AE, JRW, KLJ, HT) and Cardiovascular Diseases (AL), Mayo Clinic (LOL), Rochester, Minnesota, and Department of Cardiology, Chengdu Military General Hospital (SM), Chengdu, Sichuan, People’s Republic of China
Read the Abstract
Reference:
1. Despres JP, Lemieux I, Bergeron J et al: Abdominal obesity and the metabolic syndrome: contribution to global cardiometabolic risk. Arterioscler Thromb Vasc Biol 2008; 28: 1039.
2. Rosenquist KJ, Pedley A, Massaro JM et al: Visceral and subcutaneous fat quality and car- diometabolic risk. JACC Cardiovasc Imaging 2013; 6: 762.
3. Caglar V, Songur A, Acar M et al: Volumetric evaluation of fat in the renal sinus in normal subjects using stereological method on computed tomography images and its relation- ship with body composition. Folia Morphol (Warsz) 2014; 73: 302.
4. Lamacchia O, Nicastro V, Camarchio D et al: Para- and perirenal fat thickness is an inde- pendent predictor of chronic kidney disease, increased renal resistance index and hyper- uricaemia in type-2 diabetic patients. Nephrol Dial Transplant 2011; 26: 892.
5. Chughtai HL, Morgan TM, Rocco M et al: Renal sinus fat and poor blood pressure control in middle-aged and elderly individuals at risk for cardiovascular events. Hypertension 2010; 56: 901.
6. Sun X, Han F, Miao W et al: Sonographic eval- uation of para- and perirenal fat thickness is an independent predictor of early kidney damage in obese patients. Int Urol Nephrol 2013; 45: 1589.
7. Ochodnicky P, de Zeeuw D, Henning RH et al: Endothelial function predicts the development of renal damage after combined nephrectomy and myocardial infarction. J Am Soc Nephrol 2006; 17: S49.
8. Wang D, Iversen J, Wilcox CS et al: Endothelial dysfunction and reduced nitric oxide in resis- tance arteries in autosomal-dominant polycystic kidney disease. Kidney Int 2003; 64: 1381.
9. Lim S and Meigs JB: Links between ectopic fat and vascular disease in humans. Arterioscler Thromb Vasc Biol 2014; 34: 1820.
10. Donato AJ, Henson GD, Morgan RG et al: TNF- alpha impairs endothelial function in adipose tissue resistance arteries of mice with diet- induced obesity. Am J Physiol Heart Circ Phys- iol 2012; 303: H672.
11. Zhang J, Patel MB, Griffiths R et al: Tumor necrosis factor-alpha produced in the kidney contributes to angiotensin II-dependent hyper- tension. Hypertension 2014; 64: 1275.
12. Daghini E, Primak AN, Chade AR et al: Assess- ment of renal hemodynamics and function in pigs with 64-section multidetector CT: com- parison with electron-beam CT. Radiology 2007; 243: 405.
13. Li Z, Woollard JR, Wang S et al: Increased glomerular filtration rate in early metabolic syndrome is associated with renal adiposity and microvascular proliferation. Am J Physiol Renal Physiol 2011; 301: F1078.
14. Hjelmesaeth J, Roislien J, Nordstrand N et al: Low serum creatinine is associated with type 2 diabetes in morbidly obese women and men: a cross-sectional study. BMC Endocr Disord 2010; 10: 6.
15. Zhang X, Li ZL, Woollard JR et al: Obesity- metabolic derangement preserves hemody- namics but promotes intrarenal adiposity and macrophage infiltration in swine renovas- cular disease. Am J Physiol Renal Physiol 2013; 305: F265.
16. Tomaszewski M, Charchar FJ, Maric C et al: Glomerular hyperfiltration: a new marker of metabolic risk. Kidney Int 2007; 71: 816.
17. Moody WE, Edwards NC, Madhani M et al: Endothelial dysfunction and cardiovascular disease in early-stage chronic kidney dis- ease: cause or association? Atherosclerosis 2012; 223: 86.
18. Wu-Wong JR: Endothelial dysfunction and chronic kidney disease: treatment options. Curr Opin Investig Drugs 2008; 9: 970.
19. Fliser D, Wiecek A, Suleymanlar G et al: The dysfunctional endothelium in CKD and in car- diovascular disease: mapping the origin(s) of cardiovascular problems in CKD and of kidney disease in cardiovascular conditions for a research agenda. Kidney Int Suppl 2011; 1: 6.
20. Shulman GI: Ectopic fat in insulin resistance, dyslipidemia, and cardiometabolic disease. N Engl J Med 2014; 371: 1131.
21. Noblet JN, Owen MK, Goodwill AG et al: Lean and obese coronary perivascular adipose tissue impairs vasodilation via differential inhibition of vascular smooth muscle Kþ channels. Arte- rioscler Thromb Vasc Biol 2015; 35: 1393.
22. Owen MK, Witzmann FA, McKenney ML et al: Perivascular adipose tissue potentiates contrac- tion of coronary vascular smooth muscle: influ- ence of obesity. Circulation 2013; 128: 9.
23. Kawasaki S, Aoki K, Hasegawa O et al: Sono- graphic evaluation of visceral fat by measuring para- and perirenal fat. J Clin Ultrasound 2008; 36: 129.
24. Van de Voorde J, Boydens C, Pauwels B et al: Perivascular adipose tissue, inflammation and vascular dysfunction in obesity. Curr Vasc Phar- macol 2014; 12: 403.
25. Omar A, Chatterjee TK, Tang Y et al: Proin- flammatory phenotype of perivascular adipocytes. Arterioscler Thromb Vasc Biol 2014; 34: 1631.
26. Greenberg AS and Obin MS: Obesity and the role of adipose tissue in inflammation and meta- bolism. Am J Clin Nutr, suppl., 2006; 83: 461S.
27. Yoshida S, Takeuchi T, Kotani T et al: Infliximab, a TNF-alpha inhibitor, reduces 24-h ambulatory blood pressure in rheumatoid arthritis patients. J Hum Hypertens 2014; 28: 165.
28. Hurlimann D, Forster A, Noll G et al: Anti-tumor necrosis factor-alpha treatment improves endo- thelial function in patients with rheumatoid arthritis. Circulation 2002; 106: 2184.
29. Lavi S, Prasad A, Yang EH et al: Smoking is associated with epicardial coronary endothe- lial dysfunction and elevated white blood cell count in patients with chest pain and early coronary artery disease. Circulation 2007; 115: 2621.
30. Pawar AS, Zhu XY, Eirin A et al: Adipose tissue remodeling in a novel domestic porcine model of diet-induced obesity. Obesity (Silver Spring) 2015; 23: 399.