Oxidative stress during erythropoietin hyporesponsiveness anemia at end stage renal disease: Molecular and biochemical studies

Graphical abstract

Abstract

Inflammation and oxidative stress are two faces of one coin in end stage renal disease patients (ESRD) on maintenance hemodialysis. Their interconnection induces anemia complicated with erythropoietin hyporesponsiveness. The biochemical bases behind the resistance to erythropoietin therapy with frequent hemoglobinemia, oxidative stress and iron status have not been fully understood. Here two equal groups (40 patients each) of responders and non-responders to recombinant human erythropoietin therapy (higher than 300 IU/kg/wk of epoetin) were investigated. Hematological and biochemical analyses of collected blood and serum samples were performed along with serum electrophoretic protein footprinting. The leukocytic DNA fragmentation was used to evaluate the degree of oxidative insult. The good responders showed lower erythrocyte malondialdehyde (E-MDA) level and less DNA fragmentation of circulating leukocytes than poor responders with elevated hemoglobin, albumin, A/G ratio, total iron, and ferritin levels. Contrariwise, lower erythrocyte superoxide dismutase (E-SOD) and catalase activities in EPO poor responder group were noticed. Neither other serum constituents nor electrophoretic protein pattern showed any difference between the two groups. There were higher levels of inflammatory markers, interleukin-6 (IL6) and C-reactive protein (CRP) in EPO poor responder than good responder. The negative correlations between Hb and both IL6 and CRP levels in the present data remotely indicate a positive correlation between inflammatory markers and severity of anemia. A direct correlation between Hb and antioxidant enzymes (E-SOD and catalase) was noticed, while inverse correlation with E-MDA was recorded. The study proved that oral supplementation of vitamin C to ESRD patients might mitigate the previously elevated serum MDA level in these patients.

Introduction

Anemia is a risk factor for progression of chronic kidney disease (CKD) to end stage renal disease (ESRD) [1]. The degree of anemia in CKD patients tends to be parallel with altered kidney function manifested by considerable patients’ variability [2]. Insufficient erythropoietin (EPO) production is the primary cause of renal anemia during ESRD, due to partially or completely depleted or injured specialized peritubular cells responsible for its production [3]. Anemia was secondary promoted by other factors, including active blood loss, haemo-globinopathies, aluminum overload, hypothyroidism [4]. Impaired erythropoiesis also contributes to anemia as a result of poor response to EPO with reduced proliferative activity of erythroid precursors in bone marrow and erythrophagocytosis [5]. Further oxidative damage of RBCs membrane in chronic hemodialysis (HD) patients that decreases erythrocytes life span could exaggerate renal anemia [6]. The condition is consequently associated with a decreased quality of life [7] and cardiovascular complications [8]. That requires intense hospitalization [9]. Such anemia should be corrected by erythropoiesis stimulating agents (ESAs) therapy that slows down the progression of CKD [1]. Although the majority of CKD patients respond adequately to ESAs, 10% of these patients showed marked resistance to recombinant human erythropoietin (rhEPO) therapy [10]. Resistance to ESAs has been associated with an increased risk of cardiovascular events in CKD patients [11] with increased mortality and morbidity rates [12].

Oxidative stress is a constituent of the inflammatory mechanisms that contributes to anemia of ESRD patients. The degree of oxidative stress is closely correlated with the inflammatory status of ESRD patients undergoing HD [13]. The mechanism includes the depletion of redox capacity with membrane structural deformity and shortened life span of erythrocytes. This consequently elevates the production of hepcidin; a hormone that inhibits both intestinal absorption of iron and mobilization of iron stores by binding to ferroportin on the cell membrane with diminished expression of iron-transport protein transferrin and induction of erythropoietin resistance [14].

A close association between high levels of inflammatory markers and ESAs resistance in CKD patients has been reported [15]. Elevated circulating interleukin 6 (IL6) –as one of inflammatory cytokines intimately correlated with poor response to EPO treatment in ESRD on HD. These inflammatory cytokines can impair bone marrow function and significantly alter iron metabolism. This increased state of pro-inflammatory cytokine activity in CKD adversely restrains erythroid progenitor cell production that advances to hyporesponsiveness to ESAs and poor treatment outcomes [16]. Another gold standard as a micro-inflammatory marker in HD is the C-reactive protein (CRP) that predicts mortality after adjustment for other risk factors. Its level significantly increases comparable with other acute-phase proteins, making it a convenient clinical evaluator [17]. Furthermore, Costa et al., [18] reported that CKD patients on HD present with high levels of inflammatory markers, namely CRP, IL6, tumor necrosis factor-alpha (TNF-α), and interferon-γ and with lower serum levels of albumin.

Maintenance of balanced redox state is an important modulator in immune system homeostasis [19]. Imbalanced cellular redox state evokes cellular free radicals flooding with elevated inflammatory mediators that amplify the cascade of vicious cycle of generation of reactive oxygen species (ROS). Either chronic or acute production of free radicals contributes to lipid peroxidation, protein denaturation, and deoxyribonucleic acid remodeling [20]. Long periods of HD treatment are linked to DNA damage due to increased oxidative stress [21].

Vitamin C (ascorbic acid), on the other hand, is a potent water-soluble antioxidant in biological fluids by scavenging ROS (O₂⁻ and OH⁻) and reactive nitrogen species by forming semi-dehydroascorbic acid that mitigates oxidative damage of important biomolecules. It is an effective antioxidant against lipid peroxidation. Vitamin C deficiency in CKD patients on HD may be secondary to dietary restriction of fresh fruits and vegetables, to avoid hyperkalemia and loss of the vitamin when receiving dialysis [22]. Therefore, supplementation of ascorbic acid is essential, because the need for vitamin C increases in HD patients [23].

The current study aimed to better clarify the mechanisms of resistance to rhEPO therapy and the influence of inflammatory cytokines on erythropoietin production, and understand the interplay of the multiple factors involved in the pathogenesis of the anemia of chronic disease. Moreover, studying biochemical changes is associated with hyporesponsiveness to rhEPO therapy in HD patients with particular interest on oxidative status in the form of cell membrane deterioration and accelerated apoptosis in the form of DNA-fragmentation, as well as the disturbances in antioxidant enzymatic activity of superoxide dismutase and catalase on erythropoietin response. The study also addressed the antioxidant role of Vitamin C in alleviation of the hazard potentially induced by elevated ROS during progression of CKD.

Patients and methods

Patients and study design

A group of 80 ESRD patients (30 males and 50 females) undergo regular HD; their ages ranged from 39 to 55 years; duration of HD 7.59 ± 2.3 years, treated with rhEPO, was selected from more than 170 HD surveyed outpatients at nephrology clinic – Maadi Armed Forces Hospital.

All patients were routinely dialyzed three times a week, 4 h per session, using high flux polysulfone capillary dialyzers (Fresenius® Medical Care, Bad Homburg, Germany) and bicarbonate dialysate (Na⁺: 103 mmol/L; K⁺: 2.0 mmol/L; Ca²⁺: 1.75 mmol/L; Mg²⁺: 0.5 mmol/L; Cl⁻: 109.5 mmol/L; HCO₃⁻: 35 mmol/L). The blood flow rate ranged from 80 to 200 mL/min, according to body weight; dialysis flow was 500 mL/min with heparin anticoagulant. Mean dialysis dose Kt/V was 1.89. All patients were supported with L-Carnitine® and B-complex® supplementation after each session of HD.

The ESRD patients included 40 poor responders (Hb < 11 g/dL and rhEPO dose > 300 IU/kg/week) and 40 good responders to rhEPO therapy (Hb > 11 g/dL and rhEPO dose < 300 IU/kg/week). Classification of the patients into poor or good responder was performed in accordance with the European Best Practice Guidelines [24], which defines resistance to rhEPO as a failure to achieve target hemoglobin levels (between 11 and 12 g/dL) with maintained doses of rhEPO higher than 300 IU/kg/week of epoetin (Eprex®).

For studying the effect of antioxidant therapy on dialysis patients, another group of 20 ESRD patients was included. This group has been divided into 10 ESRD patients on HD orally supplemented with 500 mg vitamin C, twice daily for one month, according to a previous recommendation of Deicher and Horl [23], and 10 ESRD on HD patients served as control (without vitamin C supplementation).

Patients with recent blood transfusion, autoimmune disease, malignancy, hematological disorders, parathormone level >250 pg/mL and acute or chronic infection before the beginning of the study, as well as patients who were on supplementation with vitamin C and/or E during the 3 months before the beginning of the study were excluded.

The written patients’ approval consents were taken and the study followed the required Ethics Committee obligations stated by Cairo University scientific research protocol for handling of non-invasive samples (blood samples) from human subjects.

Blood sampling

Blood samples were taken twice from the HD patients, immediately before (pre-HD) and after (post-HD) dialysis sessions from arteriovenous fistulas, in vacutainer tubes with anticoagulants (ethylene diamine tetra-acetic acid (EDTA), sodium citrate, and lithium heparin) and without anticoagulant for obtaining serum samples.

Biochemical analysis

Serum samples were obtained after centrifugation (3000 rpm for 10 min), and subjected to the measurement of biochemical parameters including kidney function tests (blood urea, serum creatinine, and uric acid), liver function tests (total bilirubin, alanine aminotransferase (ALT), aspartate aminotransferase (AST), and alkaline phosphatase (ALP)), protein profile (total proteins, albumin, globulin, and A/G ratio), lipids (total cholesterol level, triglycerides), blood glucose level, total iron, calcium, and phosphorus by Hitachi 917 (Hitachi Corp., Roche-Diagnostic®, Mannheim, Germany), automatic clinical chemistry analyzer using routine laboratory techniques and available commercial kits (Roche-Diagnostic®).

Blood electrolytes

Na⁺ and K⁺ were determined in serum samples by a direct ion selective electrode method using Audicom 9101 (AC9101), an electrolyte analyzer (Horiba Medical®, Audicom Medical Instrument Co., Ltd. Jiangsu, China).

Osmolality test

Citrated plasma samples were used for determination of plasma osmolality by measuring the freezing point depression, using automatic cryoscopic osmometer (Osmomat 030, Gonotec®, Berlin, Germany).

Complete blood count (CBC)

EDTA-coated tubes were used for the CBC measurements, including hemoglobin, hematocrit, leukocytes, mean cell volume and platelets using automatic cell counter (Sysmex® XP300, Hamburg, Germany).

Determination of serum ferritin level

An electrochemiluminescence immunoassay “ECLIA” technique was performed for detection of ferritin level in serum samples by using Elecsys 2010 immunoassay analyser (Roche-Diagnostic®, Mannheim, Germany). Two monoclonal mouse antibodies – M-4.184 and M-3.170 (available kit from Roche-Diagnostic®) were used to form the sandwich complex in the assay.

Inflammatory markers

T-cell and monocyte function were assessed by measuring proinflammatory cytokine secretion from the mononuclear cells – IL-6 in serum, using commercially available enzyme linked immunosorbent assay (ELISA) kits (Biosource, Diagnostic Corporation, USA). The plates were read at 450 nm on a computerized automated VersaMax micro-plate ELISA reader (Molecular Devices Inc., Sunnyvale, CA, USA).

CRP level assay

A quantitative determination of CRP level (mg/L) in serum samples was assayed by a turbidometric immunoassay in which a serum sample is mixed with latex beads coated with anti-human CRP antibodies forming an insoluble aggregate, using Indiko auto analyzer (Indiko™ Plus; Thermo-scientific, North America).

Determination of erythrocyte malondialdehyde (E-MDA) in hemolysate

For preparation of erythrocyte hemolysate, EDTA blood sample was centrifuged at 4000 rpm for 15 min and the plasma was removed. The recovered erythrocytes were successively washed with saline solution and lysed at room temperature incubation in hypotonic double distilled water containing 5 mL/L Triton X 100. This was followed by vigorous vortex mixing. The membrane free hemolysate was obtained by centrifugation at 10,000 rpm for 5 min. Estimation of E-MDA was carried out according to the procedure described by Albro et al. [25]. Two and half mL of 10% Trichloracetic acid (TCA) was added to 0.5 mL hemolysate in a centrifuge tube, mixed well and kept for 15 min in boiling water bath. The tubes were cooled under tap water prior to addition of 1 mL of distilled water. After good mixing, tubes were centrifuged at 4000 rpm for 10 min. Two mL of filtered supernatant was mixed with 1 mL of thiobarbituric acid (TBA) and the mixture was placed in boiling water bath for 20 min. The optical density of the pre-cooled mixture was measured by spectrophotometer (Model 752N, Anjing Everich Medicare Co., Ltd. Jiangsu, China) at 532 nm against TBA blank. E-MDA concentration was expressed as μmol/g Hb.

Determination of erythrocyte superoxide dismutase activity (E-SOD)

The enzymatic activity of E-SOD in erythrocyte hemolysate was assessed according to the method of Marklund and Marklund [26]. Addition of 5 μL hemolysate to 25 μL pyrogallol (24 mmol/L prepared in 10 mmol HCl) and the final volume were adjusted to 3 mL using Tris HCl buffer (0.1 M, pH 7.8). The change in absorbance was recorded by spectrophotometer (Model 752N, Anjing Everich Medicare Co., Ltd. Jiangsu, China) at 420 nm for 3 min. E-SOD activity was expressed as U/mg Hb.

Preparation of purified leukocytes for ex vivo estimation of catalase activity (CAT)

The leukocytes-rich buffy coat was separated from heparinized blood samples by standard techniques using Ficoll-Hypaque gradient density (density 1077 g/L), centrifugation at 1000 rpm for 30 min at 20 °C (Pharmacia LKB, Uppsala, Sweden). Cells (peripheral blood mononuclear cells) were washed in Hanks balanced salt solution (HBSS; Life Technologies BRL, Life Technologies Ltd, Paisley, UK) (600g, 10 min, and 4 °C). Two additional washes with HBSS were performed (200g, 10 min, and 4 °C). As an indirect reflection of real immune status an ex vivo leukocytic catalase activity was then measured as a function of the rate of oxygen release using Clark oxygen electrode unit (Rank Brothers, Cambridge, UK) connected with chart recorder (Kipp and Zonen BD112 medical desk-top dual channel flatbed chart recorder, Holland). Ten μg protein from previously purified leukocytes was used to estimate their catalase activity. Taking normal subjects as control, the catalase activity was arbitrary measured by the developing slope formed by the oxygen release in unit time (rate of oxygen release from hydrogen peroxide via catalase action). The hydrogen peroxide substrate adequately added in access (>> above 25 mM H₂O₂ which is its catalase Km value) to drive the reaction into zero order kinetics. The zero order kinetics is a state at which the rate of catalytic conversion is mainly dependent on the enzyme activity under investigation. The oxygen concentration baseline was previously estimated by previous consumption of buffer ambient O₂ via equilibration of buffer in the measurement unit with 0.5 nmol of sodium hydrosulfite (Sigma–Aldrich).

DNA fragmentation assay as an oxidative stress marker

EDTA blood samples were used for isolation of DNA. The assay was basically performed after salting out extraction following the method described by Aljanabi and Martinez [27]. Briefly the proteins and other cellular contaminants were salted out from saturated 5 M NaCl salt solution. The DNA was then precipitated by ice-cold absolute isopropanol. The purified genomic DNA was resolved by agarose gel electrophoresis (1.5% agarose in TAE buffer with 5 V/cm migration voltage) using horizontal minigel electrophoresis unit (BIO-Rad laboratories Inc., CA, USA). The ethidium bromide pre-stained resolved DNA was visualized by UV detector (Bench top Visible/UV transilluminator, Thermo-scientific®, North America). The intensity of DNA was measured via Gel Pro Analyzer free Software using 100 bp DNA ladder marker (Vivantis Technologies®, Selangor, Malaysia).

Protein foot printing

Protein pattern analysis of serum samples was performed according to the procedure of Laemmli [29], using denatured sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS–PAGE). Prestained wide range molecular weight marker (BIO-RAD Laboratories, Hercules, CA, USA) was used.

Antioxidant effect of vitamin C

The plasma vitamin C level was determined colorimetrically, following the method of Kyaw [30] using available kit (Biodiagnostic®, Giza, Egypt). The method was based on mixing of 2, 6 dichlorophenol indophenol (DCPIP) with plasma samples in acidic medium. DCPIP dye reduced to a colorless leuco base while ascorbate is oxidized to dehydroascorbate. This redox reaction was measured by spectrophotometer (Model 752N, Anjing Everich Medicare Co., Ltd., Jiangsu, China) and ascorbate concentration was expressed as mg/L.

Determination of lipid peroxidation product in serum

A colorimetric method for quantitative analysis of lipid peroxide in serum samples was used according to the procedure described by Satoh [28], depending on TBA dissolved in sodium sulfate solution in order to avoid the interference of sialic acid. TBA reacts with MDA in acidic medium at temperature of 95 °C for 30 min to form thiobarbituric acid reactive product. By adding n-butyl alcohol, the resulting chromogen was extracted and the absorbance of the organic phase was measured spectrophotometrically at 530 nm (Model 752N, Anjing Everich Medicare Co., Ltd. Jiangsu, China). Serum MDA concentration was expressed as nmol/L, using standard curve.

Statistical analysis

Statistical analysis was performed using Statistical Package for Social Science (SPSS; SPSS version 16.0 for Microsoft Windows, Inc. Chicago, IL, USA). Power analysis of several biomarkers has been determined in patients with CKD, and resulted in N = 40 with power = 0.8 and alpha = 0.05 (data not shown). Independent t-test was used to compare the results between groups and paired t-test was used for comparison between pre and post dialysis within group. All values are expressed as mean ± SE. A P-value less than 0.05 was considered statistically significant, except serum ferritin level, which is non-normally distributed variable. Thus, the P-value is based on nonparametric test of Mann–Whitney U Test. Spearman’s rank correlations were performed to explore relationships among the blood variables.

Results

The rhEPO therapy in HD group showed non-significant difference in most of biochemical parameters measured e.g. urea, creatinine, uric acid, total bilirubin level, as well as ALP and AST activity between EPO poor responders and good responders Table 1. A significant elevation was observed in ALT activity in good responder patients when compared with poor responder ones but still, however, within the reference ranges. Both serum albumin and A/G ratio (Table 1 and Fig. 1) were significantly elevated in EPO good responders than poor responders group (3.99 ± 0.076 vs. 3.77 ± 0.07) and (1.43 ± 0.08 vs. 1.16 ± 0.05), respectively. In respect of circulating bioenergetic reserve parameters neither blood glucose nor did lipids (cholesterol and triglycerides) show any significant variation between groups. More importantly there was non-significant difference noticed between both groups in plasma osmolality, electrolytes, as well as calcium and phosphorus levels.

Table 1.

Biochemical parameters of EPO poor responder and good responder ESRD patients.

Parameters	Poor responders	Good responders	P-value
Urea (mg/dL)	140.13 ± 5.91	143.70 ± 5.84	0.669
Creatinine (mg/dL)	10.28 ± 0.41	11.22 ± 0.53	0.172
Uric acid (mg/dL)	6.87 ± 0.24	7.01 ± 0.32	0.735
ALP (U/L)	139.96 ± 23.44	154.54 ± 21.50	0.739
AST (U/L)	17.83 ± 1.45	22.24 ± 3.46	0.250
ALT (U/L)	15.58 ± 2.05	24.58 ± 3.53	0.034
Total bilirubin (mg/dL)	0.48 ± 0.02	0.50 ± 0.03	0.646
Total proteins (g/dL)	7.11 ± 0.11	6.97 ± 0.18	0.531
Albumin (g/dL)	3.77 ± 0.07	3.99 ± 0.07	0.048
Globulin (g/dL)	3.33 ± 0.10	2.98 ± 0.17	0.092
A/G ratio	1.16 ± 0.05	1.43 ± 0.08	0.010
Cholesterol (mg/dL)	176.25 ± 8.13	191.17 ± 12.70	0.329
Triglycerides (mg/dL)	151.67 ± 13.74	213.83 ± 25.81	0.101
Blood glucose (mg/dL)	94.95 ± 3.94	99.45 ± 4.58	0.461
Total iron (μg/dL)	50.91 ± 5.12	84.29 ± 7.12	<0.001
Calcium (mg/dL)	8.70 ± 0.21	8.36 ± 0.15	0.213
Phosphorus (mg/dL)	5.41 ± 0.18	5.24 ± 0.24	0.584
Sodium (mmol/L)	137.70 ± 1.29	140.50 ± 1.31	0.138
Potassium (mmol/L)	5.83 ± 0.10	5.75 ± 0.13	0.608
Plasma osmolality (mos moL/kg)	232.64 ± 21.28	247.5 ± 16.02	0.582
Hb (g/dL)	8.96 ± 0.244	12.18 ± 0.11	<0.001
Ferritin (ng/mL)	409.81⁎ (56–727.2)	672.25⁎ (414–1220)	0.020

All values represented as Mean ± S.E (N = 40). NS = no significant differences at P > 0.05. S = significant difference between groups at P < 0.05. HS = highly significant difference between groups at P < 0.001.

^⁎Nonparametric values (50–75th).

Fig. 1.

Pre HD levels of protein profile.

Logically, serum iron level recorded highly significant elevation in EPO good responders than poor responders (84.29 ± 7.12 vs. 50.91 ± 5.12; P < 0.001) as presented in Table 1 with highly significant decline in Hb levels (8.96 ± 0.24 vs. 12.18 ± 0.11) with P-value < 0.001. Also, lower serum ferritin levels (ng/mL) 409.81 (56–727.2) were noticed in EPO poor responder group.

Table 2 and Fig. 2a displayed the significant increase in serum IL6 levels (ng/mL) in ESRD (uremic) patients above the normal ranges (<9 ng/mL) in poor responders to EPO than good responders pre HD, while post HD level showed non-significant rise in poor responder patients when compared with good responders. These data also declared a slight non-significant elevation in IL6 levels after HD than pre HD levels in poor responder group (24.06 ± 3.01 vs. 19.01 ± 2.08), while a significant elevation was recorded in good responders after HD as compared with pre HD levels (18.38 ± 2.20 vs. 11.45 ± 1.13). A significant elevation of CRP levels was observed in EPO poor responder than good responder ESRD patients (Table 2 and Fig. 2b) before and after dialysis (14.83 ± 1.33 vs. 4.20 ± 0.61 and 30.00 ± 2.17 vs. 16.75 ± 1.54, respectively).

Table 2.

Pre and post HD levels of inflammatory marker (IL6 and CRP) in serum of EPO poor responder and good responder ESRD patients.

	Poor responders	Good responders	P-value
IL6 (ng/ml)
Pre HD	19.01 ± 2.08	11.45 ± 1.13	0.039
Post HD	24.06 ± 3.01	18.38 ± 2.20	0.301
P-value	0.387	0.049
CRP (mg/l)
Pre HD	14.83 ± 1.33	4.20 ± 0.61	0.020
Post HD	30.00 ± 2.17	16.75 ± 1.54	0.049
P-value	0.048	0.014

All values are represented as Mean ± S.E (N = 40). NS = no significant differences at P > 0.05. S = significant difference between groups at P < 0.05.

Fig. 2a.

IL6 levels.

Fig. 2b.

CRP levels.

In Table 3 a highly significant rise was observed in E-MDA levels (μmoL/g Hb) in EPO poor responders than in good responders (42.60 ± 2.19 vs. 31.7 ± 0.96; P < 0.001), while significant decline in SOD activity (U/g Hb) was recorded in EPO poor responders than in good responder group (1.62 ± 0.25 vs. 2.62 ± 0.15; P < 0.05). In the same way, Fig. 3 displayed falling of catalase activity in leukocytes of both ESRD groups on HD when compared with control subjects.

Table 3.

Pre HD levels of E-MDA and E-SOD of EPO poor and good responder patients.

Parameters	Poor responders	Good responders	P-value
E-MDA (μmol/g Hb)	42.60 ± 2.19	31.72 ± 0.96	0.001
E-SOD (U/g Hb)	1.62 ± 0.25	2.62 ± 0.15	0.004

All values are represented as Mean ± S.E (N = 40). S = significant difference between groups at P < 0.05. HS = highly significant difference between groups at P < 0.001.

Fig. 3.

Catalase activity.

Table 4 recorded a significantly direct correlation between Hb and each of iron, ferritin, albumin levels, and E-SOD activity (0.655^∗∗, 0.327^∗, 0.355^∗∗, 0.678^∗∗), respectively, while inverse correlation was noticed between Hb and each of oxidative stress (E-MDA) and inflammatory markers (IL6 and CRP) (−0.468^∗, −0.297^∗, and −0.376^∗∗), respectively. In the same way, serum iron also negatively correlated with IL6 and CRP (−0.307^∗, −0.336^∗), respectively, along with positively correlated with E-SOD and ferritin (0.564^∗∗, 0.562^∗∗), respectively.

Table 4.

Interrelations between parameters of EPO poor responder and good responder ESRD patients.

Parameters	Correlation with Hb	P-value	Parameters	Inter correlations	P-value
Iron	0.655⁎⁎	<0.001	CRP & IL6	0.331⁎	0.022
Ferritin	0.327⁎	0.023	Iron & ferritin	0.562⁎⁎	<0.001
IL6	−0.297⁎	0.041	Iron & IL6	−0.307⁎	0.034
CRP	−0.376⁎⁎	0.009	Iron & CRP	−0.336⁎	0.020
Albumin	0.355⁎⁎	0.013	Iron & E-SOD	0.564⁎⁎	0.010
E-MDA	−0.468⁎	0.037	E-MDA & E-SOD	−0.394	0.096
E-SOD	0.678⁎⁎	0.001

^⁎Correlation is significant at the 0.05 level (2-tailed), N = 40.

^⁎⁎Correlation is significant at the 0.01 level (2-tailed), N = 40.

Fig. 4 represented agarose gel analysis of DNA fragmentation extracted from plasma and leukocytes of EPO poor and good responder ESRD patients pre and post HD. DNA fragmentation was clearer in poor responder patients especially post HD (as in lane 3), as DNA fragments showed dense lower smearing (more fragmentation) when compared with good responder group post HD (lane 6) observed with less dense lower smearing (less fragmentation). The appearance of upper smearing in lanes 3 and 6 after dialysis that disappeared from lanes 7 and 8 prior to it might confirm the stressful condition induced by dialysis.

Fig. 4.

Electrophoretic pattern of DNA fragmentation assay. Agarose gel electrophoresis of DNA fragmentation. Lane (M): 100 Bp ladder. Lane (1): DNA extracted from leukocytes of control subject showed intact DNA. Lane (2): DNA extracted from plasma of control subject. Lane (3): DNA extracted from leukocytes of EPO poor responders post HD. Lane (4): DNA extracted from plasma of EPO good responders’ pre HD. Lane (5): DNA extracted from plasma of EPO poor responders pre HD. Lane (6): DNA extracted from leukocytes of EPO good responders post HD. Lane (7): DNA extracted from leukocytes of EPO poor responders pre HD. Lane (8): DNA extracted from leukocytes of EPO good responders pre HD. Lane (9): DNA extracted from plasma of EPO poor responders post HD. Lane (10): DNA extracted from plasma of EPO good responders post HD.

Protein foot printing of serum from control subject and ESRD patients (poor and good responders to EPO) displayed in Fig. 5 showed minimal variations in protein expression between the studied groups. There is no obvious difference in protein banding between poor responders and good responders except slight increase in concentration of the heavy molecular weight proteins (>130 kDa) in poor responders that is more obvious in lane 8.

Fig. 5.

Protein foot printing of serum from control and ESRD patients (poor and good responders to EPO). Lane (M): Marker (7 μl). Lanes (1): serum from control subject. Lane (2, 4, 6, 8): serum from EPO poor responder HD patient. Lane (3, 5, 7, 9): serum from EPO good responder HD patient.

The data displayed in Table 5 recorded a significant decrease in plasma vitamin C level (mg/L) after HD in both vitamin C supplemented and unsupplemented groups (P < 0.001; % of change: −62% and −75%), respectively when compared with pre HD levels. However, vitamin C supplemented patients (Fig. 6a) showed a significant higher vitamin C values than unsupplemented groups before and after HD session (53.278 ± 4.825 vs. 30.94 ± 2.186) and (20.288 ± 3.688 vs. 7.782 ± 1.549), respectively. Table 5 and Fig. 6b also recorded a highly significant increase in serum MDA level (nmol/L) after HD in non-vitamin C supplemented ESRD group than pre HD level (P < 0.001, % of change: 52%). Studying the effect of vitamin C supplementation on lipid peroxidation product in serum, a significantly lowering effect of vitamin C supplementation was pronounced in serum MDA level of vitamin C supplemented patients after HD when compared with un-supplemented group (5.7 ± 0.377 vs. 8.38 ± 0.687), as no significant difference was noticed between both groups before HD. In respect of serum iron levels, vitamin C supplemented group recorded lower values than unsupplemented group, and thus reaching significant difference in post HD levels (107.6 ± 8.60 vs. 166.7 ± 15.18).

Table 5.

Pre and post HD levels of plasma vitamin C, serum MDA, and serum total iron of vitamin C supplemented and unsupplemented ESRD patients.

	Vitamin C unsupplemented	Vitamin C supplemented	P-value
Plasma vitamin C (mg/L)
Pre HD	30.94 ± 2.18	53.27 ± 4.82	<0.001
Post HD	7.78 ± 1.54	20.28 ± 3.68	0.005
% of change	−75	−62
P-value	<0.001	<0.001
Serum MDA (nmol/L)
Pre HD	3.99 ± 0.29	4.94 ± 0.51	0.122
Post HD	8.38 ± 0.68	5.7 ± 0.37	0.002
% of change	52	13
P-value	<0.001	0.264
Total iron (μg/dL)
Pre HD	113.2 ± 10.63	89.11 ± 9.79	0.112
Post HD	166.7 ± 15.18	107.6 ± 8.60	0.003
% of change	32	17
P-value	0.009	0.189

All values are represented as Mean ± S.E. (N = 10). NS = no significant differences at P > 0.05. S = significant difference at P < 0.05. HS = highly significant difference at P < 0.001.

Fig. 6a.

Plasma Vitamin C levels in Vitamin C supplemented and unsupplemented HD patients.

Fig. 6b.

Serum MDA levels in Vitamin C supplemented and unsupplemented HD patients.

Discussion

Alterations of the immune system in ESRD constitute a complex issue. Here the results confirm the rise in IL-6 levels-as pro-inflammatory cytokines-above the normal levels found in HD patients. Accumulation of proinflammatory cytokines as a consequence of decreased renal elimination and/or increased generation (hypercytokinemia) is an intimate feature of uremia following induction by uremic toxins and oxidative stress [31]. On the other hand, uremia is associated with immunosuppression due to the influence of the uremic milieu and a variety of associated disorders exerted on immune-competent cells. Pecoits-Filho et al. [32] also suggested that the level of proinflammatory cytokines in HD patients is higher than healthy control as ESRD potentiates hypercytokinemia involving IL-6.

Many studies have linked a high level of proinflammatory cytokines with poor outcome in renal patients. Eschbach [33] revealed that chronic immune activation in ESRD patients possibly caused by either contact of immune cells with dialysis membranes or from repeated infection. These patients frequently present deviated homeostasis of body iron with typical anemia observed in chronic disease. Del Vecchio et al. [34] also reported that cytokine-induced inflammation suppresses bone marrow erythropoiesis in HD patients with possible aggravation of anemia.

Surprisingly, the level of IL6 showed serum persistent elevation even after hemodialysis regardless of EPO therapy. This agreed with Massey and McPherson [35] finding that suggested a diverse elimination kinetic behavior of IL-6 cytokines despite its relative low molecular weight (26.5 kDa). Tarakc et al. [36] stated that HD session does not change serum levels of IL-1, IL-6, and TNF-α, underlining importance of the structural characteristics of the molecules. Girndt et al. [37] also reported unchanged serum IL-6, during HD concurrent with increased clearance or membrane adsorption of these cytokines. The relatively short half-lives (3–7 min) with potential rapid binding of plasma cytokines cell surface receptors might imply stable plasma concentration achieved by continuous high production rate of IL-6.

The effect of immune dysfunction on the erythropoietin response in ESRD patients on maintenance HD was displayed in the present study by a significant rise in both IL-6 and CRP levels in EPO poor responder than good responder patients pre HD, with clear negative correlation between Hb concentrations with both IL-6 level and CRP. The same observations were previously noticed by Carrero et al. [38], who attributed the higher serum levels of IL-6 that reflects CRP concentration with lower Hb level in HD patients to quench the EPO therapy response. Furthermore, the inverse correlation between CRP and Hb indirectly correlates CRP with anemia in patients kept on HD [39]. The recorded inverse correlation between CRP and mean cell volume, mean corpuscular hemoglobin, and serum iron in our investigation might support the notion of Costa et al. [40] that chronic inflammation was intensified in EPO non-responder patients with potential macrophages iron trapping and consequent serum iron reduction.

In respect of iron status, a significant negative correlation between IL-6 and serum iron levels parallel with positive correlation between the later and ferritin levels explained the role of IL-6 proinflammatory in induction of EPO resistance anemia and direct inhibitory effect on erythroid progenitor cells with possible disruption of iron metabolism due to hepcidin up-regulation [41]. Hepcidin is a type II acute-phase protein produced in the liver that has been proposed to be the central regulator of iron metabolism. These findings consistent with Nemeth et al. [42], who illustrated that proinflammatory cytokines, mainly IL-6, negatively affect iron metabolism by stimulating the synthesis of hepcidin. Moreover, relation of Hb levels with serum iron and serum ferritin demonstrated the effect of these parameters on the incidence of anemia. These relations were described by Weiss and Goodnough [43] who stated that anemia of chronic disease is immune driven.

Obviously, the highest ferritin level in EPO good responder patients is in agreement with previous study [44].

Serum albumin, on the other hand, offers a predictor of baseline Hb level in patients with chronic HD, since hypoalbuminemia is usually connected with suppressed erythropoiesis [45] with considerably minor response to rhEPO in patients with lower serum albumin levels. Therefore, it is not surprisingly that the circulating albumin displayed a direct correlation with Hb levels. Moreover, serum albumin has an antioxidant activity through inhibition of erythrocyte membrane lipid peroxidation. Since there was a proved correlation between deviated redox homeostasis and altered serum proteome [46], the serum proteome foot-printing was compared among the two groups. Unlike the previous finding [47], there was no major variation seen in electrophoretic protein pattern between the two groups.

Oxidative stress aggravates lipid peroxidation [48]. In accord with our results, Zachara et al. [49] revealed that HD induces oxidative stress with initiation of lipid peroxidation. Focusing on ESAs therapy, the data support previous notion of Ludat et al. [50] that EPO good responder patients recorded lower MDA level than poor one with improved Hb status and potential decrease in free radicals generation.

ESRD patients on HD are usually suffering from the interaction between dialysis membranes and their blood. Sosa et al. [51] proved that this direct contact has to trigger the release of oxygen free radicals and oxidizing agents, such as superoxide anion, hydrogen peroxide, and myeloperoxidase. In turn, these molecules contribute to the oxidation of various bio-polymers e.g. lipids, proteins and nucleic acids. Thus, accelerated demise of circulating erythrocytes, and shorten RBCs life span are the core of next pathological consequences. Therefore, the observed negative correlation between Hb level and lipid peroxidation marker (E-MDA) with positive correlation between Hb and antioxidant activity presented by E-SOD, can be established by this phenomena. Along with the predominant DNA fragmentation – as an indicator of total genomic damage noticed in our study, Stopper et al. [52] demonstrated a clear relationship between renal failure and genomic damage in HD patients. Likewise, Gonzalez et al. [53] indicated that this damage is a sole convenient parameter of increased free radicals production and imbalanced cellular redox homeostasis, which potentially deteriorates metabolism. Superoxide dismutase action is intra-cellularly augmented with catalase and GSH-Pxs in quenching harmful effects induced by ROS [54]. Significant decline in both leukocytic catalase activity and erythrocyte SOD activity with higher MDA levels, was recorded in the study. These results were supported by Farzaneh et al. [55] who suggested that long duration of dialysis initiates decrease in erythrocytic antioxidant activity with subsequent rise in the rate of lipid peroxidation.

Serum MDA level is the global marker of polyunsaturated fatty acid peroxidation with twofold increase above the basal level after dialysis in HD patients [56]. Our study proved the antioxidant effect of vitamin C – as oral supplementation – in reducing the post dialysis serum MDA level (32%) than that in unsupplemented group.

Since vitamin C has a regulatory role in declining serum iron level via facilitating iron absorption and utilization with EPO effect augmentation [57], this investigation explains the virtue of orally-supplemented ascorbic acid as more convenient, less invasive, and shortly administrated way than the widely accepted parenteral route. Thus, we can deduce that the goal of reducing the HD-induced oxidative damage can be achieved by oral vitamin C supplement. In contrary to our finding however, Fumeron et al. [58] recognized that normalization of plasma total vitamin C forms by oral supplementation did not correct the level of previously identified oxidative stress and/or inflammatory markers.

Conclusions

The results have conveyed an excessive linkage between inflammatory status and oxidative imbalance induced by chronic dialysis process in EPO hyporesponsiveness HD patients. In order to improve responsiveness to rhEPO therapy oral supplementation with vitamin C was prescribed and monitored before and after HD. The work proved that antioxidant vitamin C supplementation – in low oral dosing – may enhance antioxidant defense mechanisms with mitigation of oxidative damages and thus reducing the requirement of high rhEPO dosing.

Limitations of the study

There were some limitations to the present study. For example, we did specify the patients who had vitamin C deficiency, and measure their plasma vitamin C levels before the beginning of the study, which burden the ability to generalize the findings. In addition, relatively few number of patients was under study. Thus, it is recommended to carry out studies with larger sample sizes, and longer term of vitamin C supplementation.

Conflict of interest

The authors have declared no conflict of interest.