Increased Oxidation as an Additional Mechanism Underlying Reduced Clot Permeability and Impaired Fibrinolysis in Type 2 Diabetes

Aims. We sought to investigate whether enhanced oxidation contributes to unfavorable fibrin clot properties in patients with diabetes. Methods. We assessed plasma fibrin clot permeation (K s, a measure of the pore size in fibrin networks) and clot lysis time induced by recombinant tissue plasminogen activator (CLT) in 163 consecutive type 2 diabetic patients (92 men and 71 women) aged 65 ± 8.8 years with a mean glycated hemoglobin (HbA1c) of 6.8%. We also measured oxidative stress markers, including nitrotyrosine, the soluble form of receptor for advanced glycation end products (sRAGE), 8-iso-prostaglandin F2α (8-iso-PGF2α), oxidized low-density lipoprotein (oxLDL), and advanced glycation end products (AGE). Results. There were inverse correlations between K s and nitrotyrosine, sRAGE, 8-iso-PGF2α, and oxLDL. CLT showed a positive correlation with oxLDL and nitrotyrosine but not with other oxidation markers. All these associations remained significant for K s after adjustment for fibrinogen, disease duration, and HbA1c (all P < 0.05), while oxLDL was the only independent predictor of CLT. Conclusions. Our study shows that enhanced oxidative stress adversely affects plasma fibrin clot properties in type 2 diabetic patients, regardless of disease duration and glycemia control.


Introduction
There is compelling evidence that type 2 diabetes (T2DM) is associated with a prothrombotic state including increased thrombin generation and platelet hyperactivity as well as endothelial dysfunction [1][2][3]. Higher plasma D-dimer levels in diabetic patients were also observed in several studies [4].
The structure of a fibrin clot, the final step of blood coagulation, has shown to be influenced by both genetic and environmental factors [5]. A number of studies have demonstrated that the structure and function of fibrin clots are unfavorably altered in subjects with T2DM [5,6]. Fibrin clots made from plasma of diabetic subjects have a denser, less porous structure than those formed from control plasma and they are more resistant to fibrinolysis [5]. Similar findings have been reported in studies performed on purified fibrinogen obtained from diabetic patients at various levels of metabolic control [7]. Denser fibrin networks have been shown to be related to worse glycemic control, reflected by elevated glycated hemoglobin (HbA1c). Glycation of fibrinogen may interfere with fibrin polymerization, cross-linking by FXIII, tissue plasminogen activator (tPA), plasminogen binding, and plasminogen to plasmin conversion [6]. However, it has also been demonstrated that, in some diabetic patients as compared to controls, there are no differences in clot permeability despite a marked difference in fibrinogen glycation, suggesting that other potent factors contribute to clot phenotype in this disease [7]. Of note, treatment with insulin has been shown to make fibrin more permeable despite no improvement in glycemia control [8]. Recently, it has been demonstrated that glycation of plasminogen also contributes to impaired fibrinolysis in T2DM [9]. Decreased clot susceptibility to fibrinolysis, together with high levels of plasminogen activator inhibitor-1 (PAI-1), which is typical of T2DM [10] and associated with abdominal obesity [5], may contribute to an increased atherothrombotic risk in this disease. Clinical relevance of resistance to lysis has been reinforced by observations indicating that T2DM predicts the failure of thrombolytic therapy in patients with acute myocardial infarction [11]. Recently, we have demonstrated that, even in patients with good glycemia control, prolonged duration of T2DM is associated with impaired fibrinolysis and formation of more compact fibrin clots [12].
Growing evidence indicates that enhanced oxidative stress is implicated in the pathogenesis of diabetes [13]. Elevated levels of 8-iso-prostaglandin F 2 (8-iso-PGF 2 ), an isoprostane that is produced by the nonenzymatic peroxidation of arachidonic acid in patients with T2DM, are detected in 24-hour urinary samples [14]. Moreover, acute glucose fluctuations were correlated with urinary excretion of 8iso-PGF 2 [14,15]. High serum peroxynitrite levels have been shown in diabetic patients by Al-Nimer et al. [16]. Elevated levels of circulating oxidized low-density lipoprotein (oxLDL), another oxidation marker, have also been demonstrated in patients with T2DM [17,18]. Free radicals are formed disproportionately by glucose oxidation, nonenzymatic glycation of proteins, and the subsequent oxidative degradation of glycated proteins. High levels of free radicals and the simultaneous decline of antioxidant defense can lead to increased lipid peroxidation and insulin resistance [19]. Oxidative modifications in the fibrinogen molecules can also alter the clot structure. Fibrin deposits formed in the presence of ferric ions are remarkably resistant to proteolytic and chemical degradations. Fibrinogen modification can alter the rate of assembly of fibrin monomers into a fibrin clot and the fibrin structure [20,21]. Associations between plasma fibrin clot properties and increased oxidative stress, reflected by plasma levels of 8-iso-PGF 2 , have been reported in patients with acute myocardial infarction [22] and in heavy cigarette smokers [23]. It is unknown whether oxidative stress modifies fibrin clot characteristics in diabetes. Since T2DM markedly alters fibrin clot structure and enhances clot resistance to fibrinolysis that might predispose to cardiovascular thrombotic complications, elucidation of mechanisms underlying these abnormalities is of utmost importance. We hypothesized that increased oxidation contributes to altered plasma fibrin clot properties in T2DM.

Subjects.
We enrolled 163 consecutive white patients with T2DM diagnosed based on the WHO criteria. The exclusion criteria were as follows: signs of acute infection, known cancer, chronic inflammatory disease, liver injury (alanine and aspartate transaminase >1.5 times above the upper limit of the reference range), estimated glomerular filtration rate (eGFR) <30 mL/min, pregnancy, arterial or venous thromboembolic events within the previous 6 months, and current anticoagulant therapy, all the states known to substantially alter blood coagulation or fibrin clot structure [24].
Demographic and clinical data were collected using a standardized questionnaire. Dyslipidemia was defined as total cholesterol (TC) >5.0 mmol/L, low-density lipoprotein (LDL) cholesterol >2.6 mmol/L, and triglycerides (TG) >1.7 mmol/L or ongoing lipid-lowering treatment. Obesity was defined as the body mass index (BMI) >30 kg/m 2 . Coronary artery disease (CAD) was diagnosed based on positive results of cardiac stress test, cardiac catheterization, or coronary computed tomography angiogram. Peripheral artery disease (PAD) was defined based on the ankle-brachial index (ABI) <0.9. Previous myocardial infarction (MI), ischemic stroke, or revascularization was established based on medical records. Diabetic nephropathy was defined as macroalbuminuria (albumin to creatinine ratio [ACR] >34 mg/mmol [300 mg/g]) or microalbuminuria (ACR 3.4 to 34 mg/mmol [30-300 mg/g]) associated with retinopathy. Diabetic neuropathy was diagnosed according to the Toronto Diabetic Neuropathy Expert Group definition [25]. Diabetic retinopathy was diagnosed by an independent ophthalmologist.
The local ethics committee approved the study. All enrolled subjects provided a written informed consent.

Laboratory
Methods. Blood was drawn after an overnight fast from an antecubital vein. Lipid profile, glucose, creatinine, blood cell, and platelet counts were assayed by routine laboratory techniques. To obtain plasma, blood samples were mixed with 3.2% sodium citrate (9 : 1) centrifuged once for 20 min and stored at −80 ∘ C. HbA1c and high sensitivity Creactive protein (CRP) were measured using immunoturbidimetric methods (Roche Diagnostics GmbH, Mannheim, Germany). Fibrinogen was determined with the von Clauss method.
We measured using commercially available enzymelinked immunosorbent assays (ELISA) the following oxidation markers: nitrotyrosine (Millipore, USA), advanced glycation end products (sRAGE) (R&D Systems, Minneapolis, MN, USA), serum 8-iso-PGF 2 (Cayman Chemicals, Ann Arbor, MI), and serum oxLDL (Mercodia AB, Uppsala, Sweden). To assess the advanced glycation end product (AGE) concentration, plasma fluorescence spectra were recorded with Fluoroskan Ascent FL, Labsystems (excitation 355 nm/emission 460 nm). AGE measurements were expressed in arbitral units (U/mL). This method measures a combination of glycation and oxidation products such as pentosidine, cross links, and others. Total serum fluorescence in humans is associated mainly with high molecular mass proteins, particularly albumin [26].

Clot Fibrin Variables.
To perform fibrin clot analysis, venous blood samples were collected (vol/vol, 9 : 1) into 3.2% trisodium citrate. Plasma samples were centrifuged at 2560 g within 20 minutes of collection, immediately frozen, and stored in aliquots at −80 ∘ C until further use. Plasma fibrin clot variables were determined in duplicate by technicians not familiar with the origin of samples (intra-assay and interassay coefficients of variation, <8%), as described with slight modifications [12,27].
Journal of Diabetes Research 3 Fibrin clot permeation was assessed using a pressuredriven system, with calculation of a permeation coefficient ( ), using the following equation: where is the flow rate in time , is the length of a fibrin gel, is the viscosity of liquid (in poise), is the cross-sectional area (in cm 2 ), Δ is a differential pressure (in dyne/cm 2 ), and is percolating time. Lower values indicate reduced permeability. Briefly, 100 L of citrated plasma was mixed with 20 L of a reagent mixture at final concentrations of 10 pmol/L TF (Innovin, Siemens Healthcare Diagnostics, Marburg, Germany), 4 mol/L phospholipids (Rossix, Mölndal, Sweden) and 20 mmol/L CaCl 2 . The mixture was immediately moved to the mechanically scratched tubes. Then the gel was stabilized for 2 hours in a moisture chamber at room temperature. The tubes were connected via silicone tube to a reservoir containing Tris-buffered saline (TBS, 50 mmol/L Tris-HCl and 0.1 mol/L NaCl, pH 7.4), with a pressure drop of 4 cm H 2 O. After washing, flow rates of buffer through the clots were measured.
Clot lysis time (CLT) was measured using a tissue factorinduced clot lysis assay. Briefly, citrated plasma was mixed with calcium chloride at final concentration of 15 mmol/L, human tissue factor (Innovin, Dade Behring, Marburg, Germany) at final concentration of 0.6 pmol/L, phospholipid vesicles (Avanti Polar Lipids, Alabaster, AL) at a final concentration of 12 mmol/L, and recombinant tPA (Boehringer Ingelheim, Germany) at final concentration of 50 ng/mL. All the dilutions were prepared in Tris buffered saline (50 mmol/L Tris-HCl, 0.1 mol/L NaCl, pH 7.4). Plasma represented 50% of the mixture volume. Turbidity was measured at 405 nm at 37 ∘ C. CLT was defined as the time from the midpoint of the baseline clear to maximum turbid transition, to the final plateau phase.

Statistical Analysis.
Statistical analysis was performed by use of MedCalc for Windows, version 13.1.2.0. All the text and tables results are expressed as mean ± standard deviation (SD) or median and interquartile range (IQR) or number and percentage. Continuous variables were compared between groups using -test (variables with normal distribution) or Mann-Whitney test (nonnormally distributed variables). Categorical variables were compared by use of the chi-square test. To determine the relationship between variables the Spearman rank-order or Pearson correlations were analyzed. Finally a multivariate regression model was constructed using all variables, that is, oxidative markers, duration of diabetes, HbA1c, fibrinogen, and other markers (if in univariate regression a value was <0.1) in backward approach to determine which factors had influence after adjustment. A value <0.05 was considered statistically significant.

Results
The final analysis included 163 patients (92 men and 71 women aged 65 ± 8.8 years, range 49-82 years) with a mean duration of diabetes of 5 years and mean HbA1c of 6.8% (Table 1). There were 127 subjects with DM duration ≤10 years and 36 with DM duration above 10 years. Vascular manifestations of T2DM were as follows: diabetic nephropathy = 29 (17.8%), diabetic retinopathy = 24

Fibrin Clot Properties and Its Relationship to Oxidation
Markers. As expected, and CLT investigated in vitro in plasma samples were inversely correlated in diabetic patients ( = −0.5, < 0.0001). We found inverse correlations between and both diabetes duration ( = −0.32, = 0.0001) and HbA1c ( = −0.30, = 0.0001) in the whole group. No similar associations were observed for CLT (Table 3). There were no associations between lipid profile or renal function and fibrin properties (data not shown).
Importantly, there were inverse correlations between and all the 5 oxidation markers, that is, nitrotyrosine, sRAGE, 8-iso-PGF 2 , AGE, and oxLDL (Table 3). There were also positive correlations of CLT with oxLDL and nitrotyrosine but not with other oxidation markers (Table 3).

Discussion
This study is the first to show that enhanced oxidative stress reflected by its circulating markers including nitrotyrosine, sRAGE, 8-iso-PGF 2 , and oxLDL unfavorably affects the fibrin network structure and susceptibility to lysis in T2DM. Known factors contributing to the prothrombotic fibrin clot phenotype observed in patients with diabetes represent increased plasma fibrinogen concentrations and enhanced glycation of the fibrinogen or plasminogen molecules [5][6][7][8][9]28]. Our observations suggest a novel prothrombotic mechanism mediated by oxidative stress that may lead to the formation of denser and resistant to lysis plasma fibrin meshwork in T2DM and thus enhancing unfavorable effects of protein glycation. Moreover, our findings provide additional evidence that oxidation can modify the clot structure. In vitro experiments demonstrated that the extent of oxidation and substances used to its generation strongly affect alterations to fibrin clot properties and these effects range widely from markedly impaired fibrin formation to tendency to produce compact networks [20]. We found that oxidation superimposed on the glycation in diabetes makes fibrin clots less permeable and poorly lysable.
To assess the impact of oxidation in vivo, we have chosen five parameters detectable in circulating blood that have been used in several studies performed in T2DM patients [29,30]. To our knowledge, associations of these oxidation markers with fibrin clot characteristics have not been studied yet. We observed the heterogeneity of the effects of circulating oxidation markers on fibrin variables in our experimental approach, which supports the concept that complex oxidative processes occurring in vivo may alter fibrin clot structure and function in a different manner.
Measurement of 8-iso-PGF 2 is a reliable tool for the identification of subjects with enhanced rates of lipid peroxidation. This prostaglandin is produced by the nonenzymatic peroxidation of arachidonic acid in membrane phospholipids 6 Journal of Diabetes Research   [19,31]. Of special interest are our findings regarding nitrotyrosine, a product of peroxynitrite action, in the plasma of diabetic patients. A direct correlation between postprandial hyperglycemia and the production of nitrotyrosine was observed in T2DM [32]. There is evidence that peroxynitrite is generated in diabetes and might be involved in the development of diabetes complications [33]. We observed that in T2DM patients nitrotyrosine could be a good predictor of . AGEs are modifications of proteins or lipids that become nonenzymatically glycated and oxidized after contact with aldose sugars. Early glycation and oxidation processes result in the formation of Schiff bases and Amadori products. Further glycation of proteins and lipids causes molecular rearrangements that lead to the generation of AGEs [34]. Circulating AGE levels are considered a more accurate biomarker than plasma glucose or HbA1c for the assessment of cumulative glycemic control [35] and a major contributor to accelerate both the microvascular and macrovascular complications in T2DM patients [36].
Activation of RAGE, the receptor for advanced glycation end products by AGEs, causes upregulation of the transcription factor nuclear factor-B and its target genes, which control cellular expression of RAGE, linking RAGE to the inflammatory response [37]. Plasma sRAGE correlates positively with albumin excretion in T2DM patients, which may represent an early marker of diabetic nephropathy [38]. Some observations suggest the active participation of AGEs-RAGE axis in accelerated atherosclerosis and development of diabetic cardiomyopathy via inducing endothelial dysfunction, inflammatory, fibrotic, and proapoptotic reactions in the myocardium [39]. Kajikawa et al. reported that the ratio of AGEs to soluble form of RAGE may be a new chemical biomarker of endothelial function [40].
However, in our study we observed slight differences in circulating AGE levels in diabetic patients who had low and high levels of HbA1c. We have found that sRAGE alters fibrinrelated prothrombotic mechanisms in T2DM.
A marker of the oxidative modification of LDL particles, plasma oxidized LDL (oxLDL), is elevated among individuals with abdominal fat, and it is positively associated with T2DM, fasting glucose, and HbA1c [41]. It has been demonstrated that oxLDL is an independent risk factor of future clinical events in a general population [42]. Importantly, oxLDL was found in the present study as the only independent predictor of prolonged CLT in T2DM, suggesting that oxLDL might directly affect the binding of fibrinolysis proteins to the fibrin fibers and thus impairing plasmin-mediated fibrin degradation.
Previous studies have demonstrated that oxidative stress parameters were elevated in patients with long-standing diabetes [43]. On the other hand, we have shown that prolonged diabetes duration is linked to persistent prothrombotic alterations despite good glycemia control [12].
A role of oxidative stress in maintaining the metabolic memory has also been suggested [44]. The current study expands the current knowledge by indicating that oxidative stress contributes to the fibrin-related aspects of metabolic memory. We showed that hypofibrinolysis and denser fibrin clot networks aggravated by oxidative stress have been associated with glycemia control and also with prolonged duration of diabetes. However, diabetes duration itself is the risk of thrombotic events despite good glycemia control. It is known that glucose control decreases fibrinogen glycation in subjects with T2DM. Glycated fibrinogen correlates well with HbA1c but it does not explain increased blood coagulation in long-term diabetes and good metabolic control of the disease [45]. It is known that oxidative stress is positively correlated with glycemic variability over a daily period. It should be noted that dipeptidyl peptidase-IV (DPP-4) inhibitors can influence oxidative stress and reduce HbA1c and glycemic fluctuations on a daily basis [46]. Pharmacological modulation of oxidative stress in diabetes merits further studies in the context of a prothrombotic state.

Journal of Diabetes Research 7
The study has several limitations. First, the size of the study population was relatively small but representative of real-life patients with T2DM. Second, we determined each variable in a single time point, and some changes in the levels during the follow-up cannot be excluded. Most HbA1c levels did not exceed 10%, so we did not evaluate patients with extremely high HbA1c. It should also be noted that duration of diabetes represents the time from diagnosis, not the actual duration of the disease. Finally, a large prospective study is needed to assess the links between the fibrin phenotype in the context of oxidative stress markers and thromboembolic events in patients with diabetes.

Conclusions
We demonstrated that prothrombotic alterations detectable in plasma fibrin clot characteristics of T2DM patients are associated not only with inadequate glycemia control and disease duration but also with increased oxidation.
Mechanisms behind complex interactions between glycation and oxidation remain to be established.