Article pubs.acs.org/JAFC
Cite This: J. Agric. Food Chem. 2018, 66, 637−644
Neuroprotective Effect of Hydroxytyrosol in Experimental Diabetic Retinopathy: Relationship with Cardiovascular Biomarkers José Antonio González-Correa,† María Dolores Rodríguez-Pérez,† Lucía Márquez-Estrada,† Juan Antonio López-Villodres,† José Julio Reyes,† Guillermo Rodriguez-Gutierrez,‡ Juan Fernández-Bolaños,‡ and José Pedro De La Cruz*,† †
Departmento de Farmacología, Facultad de Medicina, Instituto de Investigación Biomédica (IBIMA), Universidad de Málaga, 29016 Málaga, Spain ‡ Consejo Superior de Investigaciones Científicas (CSIC), Instituto de la Grasa, Ctra. Utrera Km 1, Campus Universitario Pablo de Olavide, Edificio 46, 41013 Sevilla, Spain ABSTRACT: The aim of the study was to test the neuroprotective effect of hydroxytyrosol (HT) on experimental diabetic retinopathy. Animals were divided in four groups: (1) control nondiabetic rats, (2) streptozotocin-diabetic rats (DR), (3) DR treated with 1 mg/kg/day p.o. HT, and (4) DR treated with 5 mg/kg/day p.o. HT. Treatment with HT was started 7 days before inducing diabetes and was maintained for 2 months. In the DR group, total area occupied by extracellular matrix was increased, area occupied by retinal cells was decreased; both returned to near-control values in DR rats treated with HT. The number of retinal ganglion cells in DR was significantly lower (44%) than in the control group, and this decrease was smaller after HT treatment (34% and 9.1%). Linear regression analysis showed that prostacyclin, platelet aggregation, peroxynitrites, and the dose of 5 mg/kg/day HT significantly influenced retinal ganglion cell count. In conclusion, HT exerted a neuroprotective effect on diabetic retinopathy, and this effect correlated significantly with changes in some cardiovascular biomarkers. KEYWORDS: hydroxytyrosol, diabetic retinopathy, neuroprotection, cardiovascular biomarkers
■
INTRODUCTION Diabetes mellitus is one of the main risk factors for cardiovascular disease, and is associated with increased incidences of cardiac, cerebral, and peripheral artery events.1 As a direct complication of hyperglycemia, small blood vessel alterations can lead to diabetic microangiopathy.2 In the pathophysiology of microangiopathy, a number of factors in addition to sustained hyperglycemia affect the development of intracellular endothelial oxidative stress as the main biochemical event.3,4 The consequence is an alteration in vascular function resulting in tissue damage, mainly of the ischemic type.5 In addition, chronic hyperglycemia also directly affects nervous tissues, leading to neurotoxicity in tissues composed of nerve cells, i.e., the central nervous system, peripheral nerves, and retina.6 In light of the predominant role of oxidative stress in the process of vascular and neuronal damage, antioxidant agents could slow or prevent the appearance and progression of these tissue alterations in people with diabetes. The extensive PREDIMED study has shown that a Mediterranean-type diet enriched with virgin olive oil may favor the prevention of cardiac, cerebral and peripheral cardiovascular diseases, metabolic syndrome, neurodegenerative processes,7 and more recently, some types of cancer.8 Preclinical and clinical studies have identified several factors that may be responsible for the beneficial effects of antioxidants. The most relevant factor may be increased micronutrient intake relative to other types of nutrients, particularly flavonoids and polyphenols.9 Hydroxytyrosol (the main polyphenol compound in virgin olive oil) has been shown to decrease a number of © 2017 American Chemical Society
biochemical and/or functional factors recognized as key participants in the development of cardiovascular and neuronal disease in experimental diabetes mellitus, such as platelet aggregation, low density lipoprotein (LDL) oxidation, oxidative stress, and brain inflammatory mediators, among others.10,11 The present study was designed to test the possible effect of hydroxytyrosol on the retina (the nervous tissue directly involved in diabetic retinopathy) in an animal model of diabetes. The main aim of this study was to analyze the effect of the administration of hydroxytyrosol, in an experimental rat model of type 1 diabetes mellitus, on the morphological characteristics of retinal tissue. As a secondary objective, we aimed to evaluate the possible influence of some recognized biomarkers that influence the appearance of diabetic angiopathy, and which have been shown to be important in this experimental model after the oral administration of hydroxytyrosol.
■
MATERIALS AND METHODS
Materials. Thromboxane B2 (TxB2), interleukin-1ß (IL-1ß), 3nitrotyrosine, and 6-keto-prostaglandin F1α (6-keto-PGF1α) enzyme immunoassay kits were from GE Healthcare UK (Little Chalfont, Buckinghamshire, UK). The oxidized LDL cholesterol (oxLDL) enzyme immunoassay kit was from USCN Life Science, Inc. (Bionova ́ S.L., Madrid, Spain). The nitrite/nitrate ELISA kit was from Cientifica Cayman Chemical (Ann Arbor, MI, USA). Collagen was obtained Received: Revised: Accepted: Published: 637
November 2, 2017 December 20, 2017 December 26, 2017 December 27, 2017 DOI: 10.1021/acs.jafc.7b05063 J. Agric. Food Chem. 2018, 66, 637−644
Article
Journal of Agricultural and Food Chemistry from Menarini Diagnóstica (Barcelona, Spain). All other reagents were from Sigma Chemical Corp. (St. Louis, MO, USA). Hydroxytyrosol (Figure 1) was isolated by hydrothermal treatment of the liquid phase obtained from alperujo (a byproduct of the two-
During the follow-up period, diabetic animals were treated with 4 IU every 2 days s.c. of a soluble long-acting basal insulin analogue (Levemir) to reduce mortality due to high blood glucose levels. Control animals received the same volume of isotonic saline solution s.c. All rats were anesthetized with pentobarbital sodium (40 mg/kg i.p.), and were then decapitated with a guillotine. Platelet Aggregometry. Platelet aggregation capacity in whole blood was tested at 37 °C with the electrical impedance method (Chrono-Log 540 aggregometer, Chrono-Log Corp., Haverton, PA, USA). Collagen (10 μg/mL) was used as the aggregation inducing agent. Maximum aggregation intensity (Imax) was determined as the maximum resistance between the two poles of the electrode obtained 10 min after the agonist was added. Platelet Thromboxane B2. Blood samples (0.5 mL) were induced with 10 μg/mL of collagen for 10 min, then blood samples were centrifuged at 10 000 g for 5 min, and the supernatants were frozen at −80 °C until TxB2 production was quantified with an enzyme immunoassay. Vascular 6-Keto-prostaglandin F1α. The aortic segment was cut into two parts and incubated at 37 °C in buffer containing (mM): 100 NaCl, 4 KCl, 25 NaHCO3, 2.1 Na2SO4, 20 sodium citrate, 2.7 glucose and 50 Tris (pH 8.3). Segments were placed in 500 μL fresh buffer, and 10 μL calcium ionophore A23187 (final concentration 50 μM) was added. Thirty minutes later the samples were dried and weighed, and the supernatant was frozen at −80 °C until the assay. The production 6-keto-PGF1α (stable metabolite of prostacyclin) was quantified with an enzyme immunoassay. Plasma Lipid Peroxidation. Thiobarbituric acid reactive substances (TBARS) were measured as an index of plasma lipid peroxide concentration. Samples of plasma were incubated with 500 μL 0.5% thiobarbituric acid in 20% trichloroacetic acid. The samples were shaken and incubated at 100 °C for 15 min, then centrifuged at 2000 g for 15 min at 4 °C. Absorbance of the resulting supernatant was determined spectrophotometrically at 532 nm (FluoStar, BMG Labtechnologies, Offenburg, Germany). Blank samples were prepared in an identical manner except that they were incubated at 4 °C in order to avoid TBARS production. Plasma 3-Nitrotyrosine and Oxidized Low-Density Lipoprotein Cholesterol. Plasma levels of 3-nitrotyrosine were measured as an indirect index of peroxynitrite production. Oxidized LDL was measured as one of the most important biomarkers in the first stages of cardiovascular disease. Plasma was obtained and frozen at −80 °C until the assay. The production of 3-nitrotyrosine and oxLDL was quantified with an enzyme immunoassay. Serum IL-1ß. Two milliliters of native blood was incubated at 37 °C for 30 min, then serum was obtained by centrifugation at 2500 g for 15 min at 4 °C, and frozen at −80 °C until the assay. The production of IL-1ß was quantified with an appropriate enzyme immunoassay. Retina Morphometric Analysis. After the eyeballs were removed and cleaned with saline solution, they were fixed in 10% formaldehyde solution for 48 h with gentle agitation at room temperature. This was followed by fixation and inclusion in paraffin according to a conventional protocol, and sectioning with an HM 325 rotary microtome (Leica Biosystems, Nussloch GmbH, Germany). Sections were cut at a thickness of 7 μm starting from the posterior area at the level of the optic nerve. Morphometric studies were done in 40× images. After dewaxing and hematoxylin-eosin staining, quantitative studies of histological sections were done in a simple-blind manner by two independent observers to record the following parameters:14 (1) Thickness between the inner and outer limiting membrane, equivalent to the thickness of the retina; (2) Outer nuclear layer (ONL) thickness, i.e., the thickness of the outer nuclear or granular layer; (3) Outer plexiform layer (OPL) thickness; (4) Internal nuclear layer (INL) thickness, i.e., the thickness of the internal nuclear or granular layer; (5) Inner plexiform layer (IPL) thickness. Ganglion layer cells were counted in 40× micrographs after staining, in a total of 10 sections per rat at a known distance from the first section. The results are reported as cell counts per 100-μm segment of retinal length to standardize the data reporting across all samples.
Figure 1. Chemical structure of hydroxytyrosol. phase olive oil separation system) at 160 °C for 60 min.12 The liquid was extracted by chromatography fractionation in two steps, with a final yield of 99.6% purity referred to dry matter, according to the process described by Fernández-Bolaños et al.13 For high-performance liquid chromatography, standard tyrosol compound was obtained from Fluka (Buchs, Switzerland) and hydroxytyrosol from Extrasynthese (Lyon Nord, Geney, France). The phenols were quantified using a Hewlett-Packard 1100 liquid chromatography system with an ultraviolet−visible light detector. A Mediterranea Sea C18 analytical column (250 × 4.6 mm i.d.; particle size = 5 μm) (Teknokroma, Barcelona, Spain) was used at room temperature. The system was equipped with Rheodyne injection valves (20 μL loop). The mobile phases were 0.01% trichloroacetic acid in water and acetonitrile, with the following gradient during a total run time of 55 min: 95% initially, 75% at 30 min, 50% at 45 min, 0% at 47 min, 75% at 50 min, and 95% at 52 min until the run was complete. Quantification was carried out by peak integration at 280 nm wavelength with reference to calibrations obtained with external standards. Study Design. The animals were 2-month-old adult male Wistar rats (body weight 200−250 g). All rats were used in accordance with current Spanish legislation for animal care, use, and housing (EDL 2013/80847, BOE-A-2013-6271). The recommendations in the Guide for the Care and Use of Laboratory Animals (NIH publication No. 8623, revised 1985) were followed, as well the Spanish Law on the Protection of Animals where applicable. The study protocol was approved by the University of Malaga Ethics Committee for the Use of Animals (Reference number 10/08/00006/10/10, 2016-0032). Animals (15 rats per group) were divided in four groups: (1) control nondiabetic rats treated with saline, (2) control diabetic rats (DR) treated with saline, (3) DR rats treated with 1 mg/kg/day p.o. hydroxytyrosol, and (4) DR rats treated with 5 mg/kg/day p.o. hydroxytyrosol. These doses were chosen on the basis of previous results with hydroxytyrosol to analyze some of the biomarkers quantified in this study.10,11 Hydroxytyrosol was given once per day for 7 days before diabetes was induced, and then daily until the end of the diabetic period (2 months) via an endogastric cannula at 10:00 a.m. Experimental diabetes was induced with a single intravenous injection of streptozotocin (50 mg/kg). Blood glucose concentration was measured by placing a Glucocard Memory II glucosimeter (Menarini, SA, Barcelona, Spain) in contact with blood from the saphenous vein. Animals were assumed to have diabetes if blood glucose was higher than 200 mg/dL for 2 consecutive days. Rats in the nondiabetic control group received a single intravenous injection of isotonic saline solution, and blood glucose was measured in the same way as in diabetic animals. 638
DOI: 10.1021/acs.jafc.7b05063 J. Agric. Food Chem. 2018, 66, 637−644
Article
Journal of Agricultural and Food Chemistry
Table 1. Mean Values (± SEM) of Body Weight, Blood Glucose, Maximum Intensity of Platelet Aggregation (Imax), Serum Thromboxane B2 (Metabolite of Thromboxane A2), Aortic 6-Keto Prostaglandin F1α (Metabolite of Prostacyclin, 6-KetoPGF1α), Plasma Thiobarbituric Acid Reactive Substances (Lipid Peroxidation, TBARS), Serum Oxidized Low-Density Lipoprotein of Cholesterol (oxLDL), 3-Nitrotyrosine (Peroxynitrites) and Interleukin-1ß (IL-1ß), and Correlation Coefficients (a) and Bilateral Significance (b) between the Number of Retinal Ganglion Cells and the Quantified Vascular Biomarkers, in Nondiabetic Rats (NDR) and Diabetic Rats without Treatment (DR) or Treated with Hydroxytyrosol (HT) (1 and 5 mg/kg/ day p.o.) after 2 Months of Follow-upa body weight (g) blood glucose (mg/dL) Imax (ohms) thromboxane B2 (pg/mL) 6-keto-PGF1α (pg/mg aorta) TBARS (nmol/mg protein) oxLDL (ng/mL) 3-nitrotyrosine (pg/mL) IL-1ß (pg/mL) a
NDR
DR
HT 1 mg/kg/day
HT 5 mg/kg/day
Pearson coefficient (a)
bilateral significance (b)
405 ± 8 96.9 ± 2.8 12.8 ± 1.0 47.3 ± 3.8 1.0 ± 0.1 0.31 ± 0.02 8.2 ± 0.5 0.56 ± 0.03 1.2 ± 0.1
389 ± 10 462 ± 16b 24.5 ± 1.9c 81.6 ± 5.9c 0.5 ± 0.04b 0.70 ± 0.05b 17.6 ± 1.9b 5.9 ± 0.7c 2.5 ± 0.3b
392 ± 12 477 ± 10b 19.2 ± 1.4f 64.7 ± 5.1f 0.6 ± 0.04 0.37 ± 0.04f 11.0 ± 1.4f 2.8 ± 0.3f 2.2 ± 0.4
399 ± 17 480 ± 19b 14.2 ± 1.6f 59.8 ± 6.0f 0.9 ± 0.06d 0.10 ± 0.01f 9.6 ± 1.0f 0.8 ± 0.07f 1.5 ± 0.3e
0.024 −0.158 −0.868 −0.672 +0.965 −0.803 −0.817 −0.835 −0.798
0.256 0.100 0.0001 0.003 0.0001 0.0001 0.0001 0.0001 0.005
N = 15 rats per group. bp < 0.01. cp < 0.0001 with respect to NDR. dp < 0.05. ep < 0.001. fp < 0.0001, with respect to DR.
Statistical Analysis. Data were analyzed with the SPSS statistical package (v. 23.0, licensed to the Central Computer Service of the University of Málaga). Student’s t test was used for unrelated samples, and one-way ANOVA was used with Bonferroni correction. In order to identify relationships between ganglion cell counts as the main outcome variable (to represent retinal damage) and different biochemical (predictor) variables, we calculated Pearson’s correlation coefficients for the associations. The relative risk of the main biochemical parameters that significantly influenced ganglion cell counts was calculated with multiple linear regression analysis (forward method), which allowed us to derive the formula that predicted ganglion cell count in the control and experimental groups (with or without hydroxytyrosol treatment). In all cases, statistical significance was assumed at a P value of
