Human plasma proteome analysis has the potential to ease disease diagnosis and therapeutic monitoring (1,2). Plasma protein biomarkers are useful for the diagnosis and prognosis of numerous diseases, including diabetes. Inability to utilize blood glucose is a hallmark of diabetes and leads to the development of numerous complications, including neuropathy, retinopathy, nephropathy and atherosclerosis (3). Development of these complications is 2.5 times higher in patients with long-term poorly controlled glycaemic levels than controlled levels (4). Furthermore, on the basis of glycated haemoglobin (HbA1c) levels diabetic subjects are categorized as having controlled diabetes (CD) and poorly controlled diabetes (PCD) with HbA1c levels ≤8% and >8%, respectively, of total haemoglobin (5). To understand the molecular mechanisms of pathophysiology of diabetic complications, numerous studies have utilized proteomic approaches, which have been reviewed in great detail (6). Pathology of diabetic complications is associated with increased generation of reactive oxygen species (ROS) resulting in oxidative, glyco-oxidative and carbonyl stress (7). Following engagement with receptor for AGE (RAGE), advanced glycation end-products (AGEs) induce the generation of ROS and activation of the transcription factor NF-κB, causing changes in gene expression (8). AGEs are also known to affect the activity of several plasma proteins. For example, ~50% aspartate aminotransferase enzyme activity is inactivated as a result of glycation (9). Similarly, impaired activity of glycated α-1-antitrypsin was observed in diabetes, thereby leading to protease-antiprotease imbalance (10). Glycated transferrin demonstrated the deterioration of antioxidant capacity in diabetic patients (11). Additionally, glyco-oxidative modification leads to protein aggregation resulting in protein instability. To prevent serious metabolic disturbances caused by the accumulation of glyco-oxidatively modified proteins, these proteins are further degraded by the proteasomal system (12). Glyco-oxidative modification of protein results in the elicitation of autoantibodies against several diabetic plasma proteins. These proteins include albumin, insulin, carbonic anhydrase and heat shock proteins, thereby resulting in their decreased levels in diabetic plasma (13). However, any variation in insulin levels affects the insulin-regulated protein synthesis of several proteins. For example, decreased insulin synthesis and resistance in diabetes affects the gene expression of albumin and fibrinogen (14,15). These factors contribute to differential protein expression. To compensate for the altered protein function and loss, enhanced or altered gene expression may occur, resulting in varying levels of proteins in diabetes.

Previous studies have reported differential expression of various proteins in diabetes, including α-1-antitrypsin, fibrinogen, vitamin D binding protein, complement C3 and apolipoprotein A1 (Apo A1) (16–19). However, it is important to study differential protein expression in PCD to understand the pathophysiology associated with the development of diabetic complications. In the present study, we analyzed differential protein expression in the plasma of CD and PCD subjects using proteomic methods and validating the results by western and dot blot analysis. Analysis of PCD subjects has not previously been performed.

Differential plasma protein expression was studied in CD and PCD clinical plasma samples following analysis of fasting plasma glucose, HbA1c levels, serum creatinine, high density lipoproteins (HDL), triglycerides and cholesterol levels (Table I). Differentially expressed proteins were identified by 2DE and LC-MS^E and are listed in Table II. Densitometric analysis of 2DE gels revealed upregulation of fibrinogen and haptoglobin and downregulation of vitamin D binding protein, α-1-antitrypsin, transthyretin and Apo A1 in CD and PCD compared with ND samples (Fig. 1). However, no significant difference in differentially expressed proteins was detected between CD and PCD plasma with the exception of Apo A1, whose downregulation was more prominent in PCD. The observed downregulation of Apo A1 was validated by western blot analysis (Fig. 1) and dot-blot analysis of each clinical plasma sample from ND, CD and PCD (Fig. 2). Apo A1 downregulation in PCD was considered statistically significant (P<0.05).

Diabetic plasma proteomic studies are important for understanding the molecular mechanisms which lead to complications. Numerous studies have been performed comparing ND and CD plasma samples (6). However, the pathophysiology of diabetic complications is demonstrated more clearly in PCD than CD. The present study was performed to identify differentially expressed proteins in PCD. Although fibrinogen, haptoglobin, VDBP, α-1-antitrypsin and transthyretin were found to be differentially expressed in patients with diabetes, their levels were not significantly different between CD and PCD. However, significant downregulation of Apo A1 was detected in PCD. Downregulation of Apo A1 is associated with the development of diabetic vascular complications mediated by the reverse cholesterol transport system (21). Decreased levels of Apo A1 in PCD may be attributed to numerous reasons, including autoantibodies against Apo A1 (22,23), elevated levels of inflammatory molecules (24) and insulin resistance (25). The Apo B100/Apo A1 ratio predicts cardiovascular risk more accurately than the lipids, lipoproteins and lipid ratios (26) and the ratio increases in PCD compared with CD plasma (27). Furthermore, a negative correlation was observed between serum creatinine levels and Apo A1, which supports the incidence of chronic kidney disease with lower plasma HDL-cholesterol levels (28). Therefore, lower levels of Apo A1 in PCD may be associated with increased risk of cardiovascular and kidney disorders. Thus, downregulation of Apo A1 may serve as an early predictive marker for diabetic complications.