The correlation between blood oxidative stress and sialic acid content in diabetic patients with nephropathy, hypertension, and hyperlipidemia

Abstract

This clinical study was designed to find out the correlation between oxidative stress and sialic acid (SA) content of plasma and RBCs in patients with type 2 diabetes. We evaluated SA concentration and oxidative stress biomarkers in healthy subjects and diabetic patients with and without complications in a cross-sectional survey. Significant changes in oxidative stress biomarkers and RBC-SA were revealed in the diabetic patients compared to those in the healthy group. Plasma SA significantly increased with an increase in lipid peroxidation of RBCs (LPO-RBC) (P < 0.001) in the diabetic patients without complication. RBC-SA significantly decreased with an elevation in LPO-RBC (P < 0.001) in all the diabetic patients and those with nephropathy. There was no significant correlation between plasma and RBC-SA and other oxidative stress biomarkers in the diabetic subjects. In multiple logistic regression analysis, RBC-SA was independently related to LPO-RBC in all the diabetic patients and those with nephropathy. We conclude that the induction of LPO-RBC in diabetic patients and those with nephropathy may influence the SA decomposition of RBC membrane, thereby altering its functions and transporter activities. Therefore, LPO-RBC and SA levels in RBCs can be used for prediction of diabetic nephropathy, and further studies to evaluate other factors contributing to desialylation of RBC membrane are justified.

Introduction

Type 2 diabetes (T2D) is a metabolic disorder in endocrine system which is the consequence of insufficiency in insulin secretion, action, or both [1]. The prevalence rate of T2D as the most common type of diabetes is increasing over time throughout the world, especially in developing countries [2–4]. The number of diabetic patients in Iran is estimated to be over 9 million cases by 2030 [5]. Common biomarkers for diagnosis of diabetic people are to measure fasting blood glucose levels and oral glucose tolerance test, which are not good predictors for diabetic complication. Other proposed biomarkers such as glycosylated hemoglobin, fructosamine, and glycated albumin have limitations such as moderate sensitivity and specificity [6, 7]. Therefore, it is important to find out novel biomarkers accurately identifying those at risk for diabetic complications.

Extensive evidence in both experimental and clinical studies shows that oxidative stress plays an important role in the pathogenesis of diabetes through an increase in the production of reactive oxygen species (ROS) and decline in the antioxidant defense [8–10]. Induction of oxidative stress can contribute to the development of diabetic complications including neuropathy, nephropathy, retinopathy, etc. Free radicals in diabetes are formed by glucose oxidation, resulting in subsequent non-enzymatic glycation of proteins and oxidative degradation of glycosylated proteins [11, 12]. Free radicals in diabetes are formed by glucose oxidation, resulting in subsequent non-enzymatic glycation of proteins and oxidative degradation of glycosylated proteins [13]. ROS constantly created as a result of induction of oxidative stress through physiological metabolic processes and pathological conditions are very reactive molecules with unconjugated electrons in their outer orbital that react with various macromolecules such as lipids, carbohydrates, and proteins [14, 15].

One of the compounds targeted for oxidative stress is sialic acid (SA) [16, 17], a terminal component of non-reducing end of carbohydrate chains of glycoproteins and glycolipids [18]. SA is an essential compound in all cell membranes and plays an important role in maintaining the structure, permeability, and integrity of the cell membrane [19]. SA also affects cellular adhesiveness, antigenicity, transport process, action of some hormones, and catalytic properties of enzymes [20, 21]. It should be noted that high levels of SA in plasma have been reported in inflammatory processes, alcohol consumption, diabetes, renal diseases, and cancer [22, 23]. It has been suggested that changes in sialylation are associated with oxidative stress induced by several disorders including diabetes [24, 25]. Although the role of SA in diabetic conditions has been established, the mechanisms altering the blood SA levels in diabetic patients have remained unclear.

The present study was designed to evaluate the correlation between oxidative stress parameters and blood SA and their association with the development of diabetic complications, including nephropathy, retinopathy, neuropathy, and cardiovascular complications. To achieve these goals, the correlation among oxidative stress biomarkers including glutathione level, activity of superoxide dismutase (SOD) enzyme, lipid peroxidation (LPO), and SA content was evaluated in plasma and RBCs of diabetic patients with or without complications.

Materials and methods

Participants

This cross-sectional study was conducted on diabetic patients from 2016 to 2018 after obtaining the informed consent. Considering α = 0.5 and β = 0.20, the mean and standard deviation (X ± SD) of SA were estimated 10.3 ± 3.2 and 24 ± 1.8 in diabetic and healthy subjects, respectively [26]. The sample size calculated on the basis of the following two mean comparison formula was at least 15 people in each group:

$$n=\frac{2{\left(z_{1-\frac{\alpha }{2}}+z_{1-\beta }\right)}^2{\delta }^2}{d^2}.$$

The total number of patients enrolled to this study consisted of 161 subjects including 146 diabetic patients aged between 22 and 80 years which were recruited from Diabetes Centers in Afzalipour and Bahonar hospitals, Kerman, Iran and 15 healthy people in the same age range which were selected from the referents to Kerman Blood Transfusion Organization. The subjects were selected using convenience sampling. The inclusion criteria of the diabetic group were the presence of T2D alone or in combination with diabetic complications subjected to clinical examination to confirm the diagnosis of complications based on clinical features and blood biochemical tests. All diabetic patients with nephropathy had chronic renal failure with estimated Glomerular Filtration Rate (eGFR) lower than 15 ml/min/1.73 m² and were undergoing dialysis. The exclusion criteria for both groups were having a history of addiction, smoking, and alcohol consumption.

Blood sampling and preparation

Venous blood samples (5 ml) were taken from the participants in the morning after 8–10 h fasting using EDTA-containing tubes. Plasma were isolated after centrifugation at 3000g for 15 min and the precipitated RBCs were washed and a part of them were used for preparation of RBCs lysates. Another part of RBCs were mixed by 10 volume of 5 mM phosphate buffer pH 7.4 and kept on ice bath 30 min for isolation of RBCs membranes. The RBC membranes were then centrifuged at 20,000g for 40 min, and after removal of supernatant, the precipitants were washed and centrifuged again under the same condition. Finally, the RBCs’ membranes were resuspended in 5 mM phosphate, pH 7.4. All samples were stored at − 80 °C for subsequent tests.

Measurement of glutathione content

According to Ellman colorimetric method [27], 750 ml of EDTA and 750 μL of TCA 10% were added to 1.5 ml of hemolysate and centrifuged at 3500g for 35 min. Then, 2.5 ml of Tris buffer was added to the supernatant. In addition, after addition of 0.5 ml 5,5-dithionitrobenzoic acid (DTNB, 10 mM in 0.1 M phosphate buffer, pH 8), it was incubated in the dark for 15 min at room temperature. The absorbance was read at 412 nm and total thiol content was expressed as µMol/mg protein according to the standard curve of glutathione.

Measurement of SOD enzyme activity

The rate of pyrogallol auto-oxidation was used to measure SOD activity [28]. To carry out this experiment, 610 μl of phosphate buffer was mixed with 90 μl of pyrogallol (2 mM pyrogallol in Tris–HCl buffer, pH 8.2), and the absorbance was determined kinetically at 420 nm alone and after the addition of 50 μl of sample. The amount of SOD needed for 50% inhibition of the pyrogallol autoxidation was considered as one unit of SOD activity and expressed as U/mg protein.

Measurement of LPO

To perform this experiment, malondialdehyde (MDA) level was measured by Draper and Hadley method with a slight modification [29]. The samples were deproteinized by trichloroacetic acid (TCA, 20%) and centrifuged at 4100g for 15 min. One ml of sulfuric acid 0.05% and 800 μl of thiobarbituric acid (TBA, 0.2%) solution were then added to the precipitant, and the tubes were placed in a boiling water bath for 30 min. Afterwards, 800 μl n-butanol was added to each well to extract the colored complex of MDA-TBA, and the absorbance was read by spectrophotometry at 532 nm. Using the MDA standard curve, the concentration of MDA was obtained in each sample and expressed as µMol MDA/ mg protein.

Measurement of SA content

The SA content of RBC membrane (RBC-SA) and that of plasma (PSA) were measured according to Griess method [30]. Briefly, 100 µl of 0.04 M periodic acid was mixed with 500 μl diluted sample and cooled in an ice bath for 30 min at 0 °C. Then, 1.25 ml of resorcinol working solution (5 ml of 6.0% resorcinol solution, 0.12 ml of copper sulfate solution 0.1 M and 19.87 ml of distilled water, brought to a final volume of 50 ml with 10 M HCl) was added to the samples, mixed, and heated at 98 °C for 5 min. The tubes were cooled in the ice bath for approximately 2 min, 3.25 ml of n-butanol was added, and the tubes were placed in a water bath at 37 °C for at 3 min. The absorbance was measured at 625 nm, and according to the standard curve of N-(1-naphthyl) ethylenediamine, the SA value of each sample was expressed as mMol/mg protein in RBC membrane and µMol/l in plasma.

Statistical analyses

SPSS 16 was used to analyze the data. The descriptive results were shown as the mean ± SD and frequency (%). The normality of the data was examined by one-sample Kolmogorov–Smirnov test. The data were analyzed using one-way ANOVA, and the post hoc test was performed using Tukey multiple comparison. The correlation between different parameters was examined using Pearson’s correlation coefficient. To measure factors associated with the level of PSA and RBC-SA, stepwise linear regression was used. Receiver operator characteristic (ROC) curve was plotted to assess whether PSA, RBC-SA, and oxidative stress biomarkers can be used as the biomarkers in the diagnosis of progression of diabetes complications. To determine the best cut point, Youden index was calculated for each biomarker with a significant area under the ROC curve above 0.5 [31]. P value less than 0.05 was considered significant.

Results

The demographic characteristics of the diabetic and healthy subjects are shown in Table 1. The average age of the participants was not significantly different among the groups. The duration of diabetes in the diabetic patients with nephropathy (18.0 ± 12.6) was significantly (P < 0.001) higher than that in the diabetic group without complications (9.0 ± 6.3).

Table 1.

Demographic characteristic of diabetic patients and healthy subjects

Variables	Healthy group (N:15)	Diabetic patients (N:146)
		Without complication	Nephropathy	Hypertension	Hyperlipidemia
Sex, male/female (%)	9/6 (60/40)	11/9 (35.5/64.5)	10/20 (33.3/66.6)	12/43 (36.4/63.6)	10/31 (28.6/71.4)
Age, year	46.54 ± 24.23	53.9 ± 11	63 ± 10.7	58.9 ± 9.5	58.2 ± 9.7
Duration of diabetes, year	–	10.81 ± 9.11	18 ± 12.6	12.1 ± 7.7	10.4 ± 6.2
Fasting blood glucose, mg/dl	97.3 ± 12.5	168.7 ± 62.9	141.7 ± 40	142.5 ± 54.1	142.5 ± 45.3

Data are means ± SD for quantitative or frequency (%) for qualitative data

As shown in Table 2, the RBC glutathione decreased significantly (P < 0.05) in the diabetic patients with nephropathy compared to that in the diabetic subjects without complication. There was no significant difference in the RBC glutathione among the other groups. The activity of SOD in all the diabetic patients with and without complications was significantly lower (P < 0.001) than that of the healthy subjects. The diabetic subjects with nephropathy showed a significant reduction in the SOD activity compared to the diabetic patients without complications (P < 0.001). All the diabetic subjects showed a significant increase (P < 0.001) in the lipid peroxidation of RBC in comparison with the healthy group but not with the diabetic patients without complications. Lipid peroxidation level was not significantly different among the diabetic patients with various complications. In addition, RBC-SA and PSA in the diabetic patients with all the complications were, respectively, lower and higher than those in the healthy group (P < 0.001).

Table 2.

Oxidative stress and plasma and RBC-sialic acid in the healthy and the diabetic subjects

Variables	Healthy group	Diabetic patients (N:146)
		Without complication	Nephropathy	Hypertension	Hyperlipidemia
GSH, µmol/mg pro	741.64 ± 146.50	812.51 ± 405.1	573.53 ± 296.50†	805.40 ± 434.23	874.60 ± 482.93
SOD, U/mg pro	4.30 ± 0.36	1.92 ± 0.22***	0.84 ± 0.53*** †††	1.81 ± 0.61***	2.00 ± 0.40***
LPO, µmol/mg pro	1.20 ± 0.41	1.56 ± 0.31***	1.74 ± 0.34***	1.63 ± 0.32***	1.62 ± 0.20***
PSA, µmol/l	149.9 ± 47.30	175.00 ± 68.89*	174.13 ± 23.20*	176.00 ± 54.82 *	160.80 ± 48.61*
RBC-SA, mMol/mg pro	5.84 ± 1.43	4.20 ± 2.14*	4.00 ± 2.51*	4.14 ± 2.00*	3.93 ± 2.10*

Data are means ± SD

GSH glutathione, LPO lipid peroxidation, PSA plasma sialic acid, RBC-SA red blood cell-sialic acid, SOD superoxide dismutase

*P < 0.05 and ***P < 0.001 by ANOVA in comparison with control, ^†P < 0.05 and ^†††P < 0.001 by ANOVA in comparison with DM

Table 3 shows a significant positive correlation between PSA and LPO in the all diabetic subjects (r: − 0.297, P: 0.018). On the other hand, a significant negative correlation was shown between RBC-LPO and the level of RBC-SA in the diabetic patients (r: − 0.172, P: 0.038) and those with nephropathy (r: − 0.343, P: 0.036). The decrease in the activity of SOD and the glutathione level of RBCs did not show any correlation with PSA and RBC-SA.

Table 3.

Correlation between plasma and red blood cell sialic acid with oxidative stress biomarkers and duration of diabetes in the diabetic groups

Variables	Groups	PSA		RBC-SA
		P	r	P	r
GSH	DM without complication	0.76	0.05	0.82	−0.04
	DM + nephropathy	0.22	0.23	0.75	0.06
	DM + hypertension	0.74	− 0.04	0.38	− 0.12
	DM + hyperlipidemia	0.81	− 0.03	0.38	− 0.12
	All diabetic subjects	0.62	− 0.04	0.61	− 0.04
SOD	DM without complication	0.29	0.19	0.68	0.08
	DM + nephropathy	0.15	− 0.27	0.86	0.03
	DM + hypertension	0.77	− 0.03	0.93	0.01
	DM + hyperlipidemia	0.45	0.1	0.55	− 0.08
	All diabetic subjects	0.89	− 0.01	0.85	0.02
LPO	DM without complication	0.02	0.3	0.3	− 0.19
	DM + nephropathy	0.66	0.08	0.03	− 0.34
	DM + hypertension	0.12	− 0.18	0.33	− 0.12
	DM + hyperlipidemia	0.74	0.05	0.8	− 0.03
	All diabetic subjects	0.14	− 0.12	0.04	− 0.17

Value (r) indicates correlation coefficient by Pearson’s correlation analysis. P indicates level of significance. P < 0.05 is considered statistically significant

DM diabetes mellitus, GSH glutathione, LPO lipid peroxidation, PSA plasma sialic acid, RBC-SA red blood cell-sialic acid, SOD superoxide dismutase

Stepwise multiple regression coefficients are shown in Table 4 for PSA, RBC-SA, and the oxidative stress variables for all the diabetic subjects. In this model, RBC-LPO was significantly and negatively correlated with RBC-SA as the dependent variable in the diabetic patients and in the diabetic patients with nephropathy.

Table 4.

Stepwise multiple linear regression coefficients and statistical significance for plasma sialic acid (PSA), red blood cell-bound sialic acid (RBC-SA) with other variables

Variables	Groups	PSA		RBC-SA
		Regression coefficient	P	Regression coefficient	P
GSH	DM without complication	0.03	0.87	− 0.04	0.82
	DM + nephropathy	0.16	0.38	0.08	0.66
	DM + hypertension	− 0.05	0.65	0.02	0.85
	DM + hyperlipidemia	− 0.05	0.7	− 0.12	0.38
	All diabetic subjects	− 0.04	0.62	− 0.03	0.68
SOD	DM without complication	0.12	0.48	0.08	0.68
	DM + nephropathy	− 0.17	0.38	0.06	0.74
	DM + hypertension	− 0.15	0.19	0.08	0.51
	DM + hyperlipidemia	0.08	0.53	− 0.08	0.55
	All diabetic subjects	− 0.01	0.9	0.02	0.84
LPO	DM without complication	− 0.24	0.15	− 0.19	0.29
	DM + nephropathy	0.04	0.82	− 0.37	0.04
	DM + hypertension	− 0.12	0.3	− 0.17	0.14
	DM + hyperlipidemia	0.01	0.93	− 0.03	0.8
	All diabetic subjects	− 0.12	0.14	− 0.19	0.02

P indicates level of significance. P < 0.05 is considered statistically significant

GSH glutathione, PSA plasma sialic acid, RBC-SA red blood cell-sialic acid, SOD superoxide dismutase, LPO lipid peroxidation

The diagnostic accuracy of PSA, RBC-SA, and the oxidative stress biomarkers was measured by ROC plot. As shown in Fig. 1 and Table 5, RBC-LPO at the best cut point of 1.45 µmol/mg protein, Youden index of 0.57, and area under the curve of 0.80 (P < 0.00) could differentiate the diabetic and non-diabetic subjects with 76% sensitivity and 80% specificity. The optimal RBC-LPO (AUC 0.68, P < 0.02), differentiating the group of diabetic patients from those with nephropathy, was 1.73 µmol/mg protein with the sensitivity of 67%, specificity of 75%, Youden index of 0.42, and AUC of 0.68.

Fig. 1.

Receiver-operating characteristic curve (ROC) of sensitivity plotted against 1-specificity of various biomarkers including RBC-sialic acid (RBC-SA) and plasma sialic acid (PSA), glutathione, superoxide dismutase (SOD), and lipid peroxidation (LPO) in a diabetic subjects, b diabetic patients with hyperlipidemia, c diabetic patients with hypertension, d diabetic patient with nephropathy

Table 5.

Area under the curve (AUC) if the receiver-operating characteristic curve (ROC) for various biomarkers including RBC-sialic acid (RBC-SA), glutathione (GSH), superoxide dismutase (SOD), lipid peroxidation (LPO), plasma sialic acid (PSA) in the healthy, and diabetic groups

Groups	Variables	AUC ± SD	P
Diabetic patients against healthy group	RBC-SA	0.50 ± 0.07	0.99
	GSH	0.49 ± 0.05	0.91
	SOD	0.00 ± 0.00	0.00
	LPO	0.80 ± 0.08	0.00
	PSA	0.60 ± 0.08	0.23
Diabetic patients with hypertension against diabetic group	RBC-SA	0.49 ± 0.06	0.85
	GSH	0.50 ± 0.06	0.94
	SOD	0.51 ± 0.06	0.90
	LPO	0.58 ± 0.06	0.18
	PSA	0.54 ± 0.06	0.51
Diabetic patients with Hyperlipidemia against diabetic group	RBC-SA	0.46 ± 0.06	0.55
	GSH	0.54 ± 0.07	0.52
	SOD	0.60 ± 0.06	0.14
	LPO	0.56 ± 0.07	0.32
	PSA	0.44 ± 0.06	0.33
Diabetic patients with Nephropathy against diabetic group	RBC-SA	0.46 ± 0.08	0.61
	GSH	0.32 ± 0.07	0.02
	SOD	0.06 ± 0.03	0.00
	LPO	0.68 ± 0.07	0.02
	PSA	0.60 ± 0.07	0.10

P indicates that level of significance P < 0.05 is considered statistically significant

Discussion

The findings of the current study showed oxidative stress induction along with PSA elevation and RBC-SA reduction in the diabetic patients. A novel finding of our manuscript was the significant correlation between SA content of RBCs membrane and peroxidation of RBCs membrane lipids in the diabetic patients with chronic renal failure. We did not find any correlation between PSA and the occurrence of the diabetic complications and alteration in the oxidative stress biomarkers. Furthermore, the result of ROC curve in the present study indicated that oxidative stress biomarkers and sialic acid content of plasma and RBCs were not good biomarkers in differentiation of diabetic patients from those with nephropathy, hypertension, and hyperlipidemia, but the LPO as an end product of RBCs’ oxidative damage was more convincing that other biomarkers. Risk of chronic renal failure in diabetic patients is about 30% and develops decades after the onset of diabetes [32]. Krolewski et al. revealed that in majority of diabetic patients, the onset of progressive loss in renal function happens suddenly when patients have normal renal function and normal urinary albumin excretion [33]. Genetic and environmental factors, inflammation, and induction of oxidative damages are contributed in the development of renal failure [34]. Dysregulation of oxidative and mitochondrial enzymes leading to accumulation of free radicals and oxidative damage have a pivotal role in the acceleration of eGFR decline and the development of renal failure [35]. The inflammatory cytokines and oxidative stress biomarkers might be used for prediction of renal failure in diabetic patients [36].

Oxidative stress induction contributed to the development of diabetic complications through production of mitochondrial ROS [37–39]. In addition, it has been shown that inhibition of diabetes-induced oxidative stress can be an effective strategy in control and prevention of the onset and progression of diabetic complications [40, 41]. It has also been reported that total serum SA and PSA are in association with micro- and macrovascular complications including nephropathy, retinopathy, and hypertension in diabetic patients [42–46]. In consistence with the finding of the current study, the previous studies also demonstrated a negative correlation between SA content of RBC membrane and elevation of LPO products like MDA in diabetes [47, 48]. Elevation of oxidative stress during aging process in human has a positive and negative correlation, respectively, with PSA and RBC-SA [30]. Since RBC membrane undergoes various modifications through induction of oxidative stress in blood circulation, the negative correlation of RBC-SA with LPO may be due in part to detachment and degradation of RBC-SA from membrane as a result of the oxidative damage [30]. RBC-SA is thought to favor negative charge of RBC membrane and its biophysical properties which affect RBC interactions with the other blood cells and vascular wall [49]. It should be mentioned that desialylation of carbohydrate domain of RBC membrane is in association with development of macro and microvascular complications of diabetes by elevating RBC aggregation [50, 51].

The results of this study indicate that one of the mechanisms which can affect desialylation of RBC membrane is induction of oxidative damage of membrane lipids during development of diabetic nephropathy. Considering the potential of SA and oxidative stress biomarkers in diagnosis of diabetic conditions, further studies should be performed to evaluate other factors contributing to desialylation of RBCs’ membrane including Na⁺ and K⁺ cation permeability, intracellular concentration of Ca²⁺, enzymatic desialylation, and cytoskeletal changes in diabetic patients. In addition, also it is required to develop studies in diabetic patients with different stages of nephropathy for a better evaluation of the involvement of oxidative stress biomarkers in pathology of diabetic renal failure.

Compliance with ethical standards

Conflict of interest

The authors declare that there is no conflict of interest associated with this manuscript.

Human rights statement

All procedures followed were in accordance with the ethical standards of the responsible committee on human experimentation (the research ethics committee of Kerman University of Medical Sciences, approval date. 2016.10.17, approval no. IR.79.KMU.REC.1395.273) and with the Helsinki Declaration of 1964 and later versions.

Informed consent

Informed consent was obtained from the participants before they were included in the study.