Plasminogen activator inhibitor-1 gene polymorphism as a risk factor for vascular complications in type 2 diabetes mellitus

Abstract

Background

Diabetes mellitus (DM) can lead to microvascular and macrovascular damages through hyperglycemia that is the main cause of diabetic complications. Other factors such as hypertension, obesity, and hyperlipidemia may worsen or accelerate the others. Several studies have revealed definitive genetic predispositions to the development of type 2 diabetes mellitus (T2DM) and development of vascular complications. This study aimed to address the association between plasminogen activator inhibitor-1 (PAI-1) gene polymorphism and T2DM, and if this gene polymorphism may have a possible role in the development of vascular complications in T2DM. This study is a case control; it included 200 patients with T2DM, 117 patients had no vascular complications, and 83 had previous vascular complications (VCs). One hundred eighty volunteer blood donors were selected as a healthy control group. All patients and controls were subjected to clinical examination, and laboratory investigations included lipid profile, fasting and 2 h blood glucose, complete blood cell count, d-dimer, PAI-1, thrombin activatable fibrinolysis inhibitor (TAFI), and detection of PAI-1 gene polymorphism by real-time polymerase chain reaction (PCR).

Results

The most prevalent genotype of PAI-1 gene polymorphism in all studied groups, including controls, was 4G/5G with the highest allele frequency as 4G. The 4G/5G and 4G/4G genotypes were associated with increased risk of DM development as compared to 5G/5G genotype. The 4G/5G and 4G/4G genotypes also had a highly significant increased risk of VCs among diabetic patients, as compared to 5G/5G. The 4G allele also was highly associated with DM with VCs. The d-dimer TAFI, PAI-1 showed the highest levels in 4G/5G genotype followed by 4G/4G genotype. The lowest level was expressed in 5G/5G genotype in diabetic patients with and without VCs. The univariable analysis showed that genotypes 4G/5G and 4G/4G were potentially risk factors for development of VCs with T2DM patients.

Conclusion

This study concludes that the PAI-1 4G/5G polymorphism may be associated with T2DM and may be considered as a risk factor for development of thrombotic events. It may also help in selection and dosing of patients being treated with anticoagulant and fibrinolytic agents. Further large-scale studies are recommended to assess the possible role of environmental factors and gene interactions in the development of T2DM vascular risks.

Background

Type 2 diabetes mellitus (T2DM) is a common metabolic disease in developing countries, in which combined insulin resistance and beta-cell impairment lead to hyperglycemia [1]. The long-term complications of diabetes mellitus (DM) are the major causes of morbidity, mortality, and high healthcare costs [2]. Vascular complications of DM represent a leading health problem worldwide [3]. Most of the complications caused by hyperglycemia involve damage to small vessels which leads to neuropathy, nephropathy and retinopathy, and large blood vessels, as in cardiovascular diseases. Risk factors such as hypertension, dyslipidemia, and obesity can also increase the risk of type 2 DM. Insulin resistance and hyperglycemia are associated with low-grade inflammation as well as chronic enhancement of oxidative stress, triggering endothelial dysfunction and promoting atherogenesis [4].

Plasminogen activator inhibitor-1 (PAI-1) is a 50-kDa glycoprotein that belongs to the serine protease inhibitor (serpins) family. PAI-1 is the main regulator of the endogenous fibrinolytic system [5]. PAI-1 is composed of three beta sheets and nine alpha helices. PAI-1 can bind to the somatomedin B domain, interact with the proteasome, and interfere with cell adhesion to the extracellular matrix [6]. PAI-1 is expressed and secreted in a variety of tissues, including the liver and spleen. The synthesis of PAI-1 is regulated by insulin, very-low-density lipoprotein (VLDL), low-density lipoprotein (LDL), and glucose [7].

PAI-1 is the main physiological inhibitor of tissue-type plasminogen activator (tPA) in the fibrinolytic system, which produces active plasmin from plasminogen that then cleaves fibrin. Impaired fibrinolytic function induced by increased PAI-1 expression is commonly observed in patients with thrombotic disease [8].

The PAI-1 gene is located on human chromosome 7q21.3-22, spans 12.3 kb and contains nine exons and eight introns [9]. The genetic expression and polymorphisms of PAI-1 are still incompletely understood. The main polymorphism of PAI-1 consists of a common single insertion/deletion of a guanine (G) base at position 675 in the promoter region of PAI-1 and can cause changes in the rate of gene transcription [10]. This polymorphism produces two alleles that contain either four or five sequential guanosines (4G and 5G) that differ in their regulation of the concentration of PAl-1 [11].

Individuals who are homozygous for the 4G allele (4G/4G) have higher levels of gene transcription and a higher PAI-1 plasma concentration than those who are homozygous for 5G (5G/5G) and therefore possibly have an increased risk for intravascular thrombosis [12]. Individuals who are heterozygous (4G/5G) have intermediate levels of PAI-1. The 4G allele produces up to six times more messenger ribonucleic acid (mRNA) than the 5G and is associated with increased PAI-1 activity [13]. An increase in the plasma concentration of PAI-1 is, therefore, associated with increased thrombotic events [14], recurrent myocardial infarction (MI) in young patients [15], and ischemic events in individuals with pre-existing atherosclerosis [16].

The role of the PAI-1 polymorphism as a risk factor for thrombotic events in many other pathological conditions has been disused in previous studies. However, the relationship between the PAI-1 gene polymorphism and the development of vascular complications in T2DM is still a matter of debate. Therefore, the present study investigated the possible association between the PAI-1 gene polymorphism and the development of T2DM in Egyptian patients and identified the possible relationship of this gene polymorphism with vascular complications in these patients.

Methods

Our study included 200 patients with T2 DM (123 males and 77 females, with ages ranging from 37–68 years). The patients were from the Internal Medicine Departments of Menoufia, Al Zahraa and Helwan University Hospitals. Diagnosis of diabetes was performed according to American Diabetes Association (ADA) 2018, with either fasting blood glucose greater than 126 mg/dl, 2–h post-prandial blood sugar greater than 200 mg/dl during oral glucose tolerance test (OGTT), or hemoglobin A1C (HbA1C) greater than 6.5%. We also included patients who received an anti-diabetic treatment.

Hypertension was defined as systolic blood pressure (SBP) greater than 130 mmHg or diastolic blood pressure (DBP) greater than 80 according to American Heart Association (AHA) 2017 or patients receiving an anti-hypertensive treatment.

Inclusion criteria

All 200 included patients had T2DM, and 117 of them had T2DM without any vascular complications with a median age of 48 years. The remaining 83 patients had previous vascular complications with a median age of 49 years (17 patients had a proven myocardial infarction (MI), 15 had cerebral strokes (CSs), 17 had proven diabetic nephropathy (DN), 20 had coronary artery disease (CAD), and 12 had proven deep venous thrombosis (DVT)).

Exclusion criteria

Non-diabetic patients or patients with type 1 diabetes mellitus (T1DM), cardiogenic shock, risk factors of vascular complications due to causes other than T2DM, or acute renal failure due to causes other than DM; patients who were overt obese; patients who refused to participate in the study; patients who were very debilitated; and patients above 70 years of age.

One hundred eighty-one blood donor volunteers were selected as the healthy control group (137 males and 44 females, with ages ranging from 32 to 58 and a median age of 46 years).

Written consent was obtained from all participants, and the study was approved by the ethical committee of National Liver Institute-Menoufia University. All procedures performed in this work were carried out in accordance with the 1964 Helsinki declaration and its later amendment.

All patients and controls were subjected to the following:

• I. Complete history and clinical examination.

• II. Laboratory investigations:

  • Blood samples were collected after overnight fasting from patients and controls into plain tubes for determination of their lipid profile [total cholesterol (TC), triglycerides (TG), low-density lipoprotein (LDL), high-density lipoprotein (HDL)] and fasting and postprandial blood glucose by an auto analyser Integra-800 (Roche-Diagnostics, Mannheim, Germany).

  • Ethylene-diamine-tetraacetic acid (EDTA) blood samples were used for complete blood cell counts and were measured by a Sysmix XT-1800 Haematology automatic cell counter (Germany).

  • Plasma samples were isolated from EDTA samples and stored at − 80 °C until used for detection of d-dimers by latex agglutination turbidimetry with a normal range of < 0.5 μg/ml.

  • A sodium citrate sample was used to determine the PAI-1 antigen using a commercial kit from Asserachrom PAI-1(Stago Diagnostic-France).

  • Human thrombin activatable fibrinolysis inhibitor (TAFI) was detected in EDTA plasma samples by a Sandwich enzyme-linked immunosorbent assay (ELISA) Kit (Catalog No: MBS2500580, MyBio-source). Briefly, standards or samples were added to the wells of a micro ELISA plate that was pre-coated with an antibody specific to human TAFI, followed by incubation with a biotinylated detection antibody specific for TAFI conjugated to avidin-horseradish peroxidase (HRP). After washing, the substrate solution was added to each well. Then, a blue color was developed, and the reaction was terminated by the addition of stop solution. The color turned yellow. The optical density (OD) was spectrophotometrically measured at a wavelength of 450 nm ± 2 nm. The OD value was proportional to the concentration of TAFI in the sample. The final concentration of TAFI in the samples was obtained from the standard curve. The detection range was 5–320 ng/ml, with a coefficient of variation of < 10%.

  • Detection of PAI-1 genotyping by real-time PCR:

  • 1. DNA extraction:

    Genomic DNA was extracted from whole EDTA blood samples using the Invitrogen DNA Blood Mini Kit according to the manufacturer’s instructions. Briefly, genomic wash buffers 1 and 2 were placed in micro-centrifuge tubes. Then, 200 μl of whole blood was added, followed by the addition of 20 μl proteinase k and 20 μl RNase to each sample. The mixture was vortexed and incubated at room temperature for 2 min. After the addition of 200 μl buffer, vortexing was performed, and the mixture was incubated at 55 °C for 10 min to promote protein digestion. Two hundred microliters of ethanol (96–100%) was added to the sample and vortexed for 5 s. The mixture was applied to a Pure Link mini spin column and centrifuged at 8000 rpm for 1 min and filtered. Five hundred microlitres of buffer was added to the DNA pellet and centrifuged at full speed (13,000 rpm) for 3 min, followed by the addition of 100 μl Pure Link elution buffer, incubation at room temperature for 1 min, and centrifugation at maximum speed for 1 min. The extracted DNA was stored at − 20 °C until used for PAI-1genotyping by real-time polymerase chain reaction (PCR).

  • 2. TaqMan genotyping assay:

    The PAI-1 polymorphism was genotyped by real-time PCR fluorescence detection on a Rotor Gene Real Time PCR System (QIAGEN, GmbH-Germany) using fluorescent-labelled probes. TaqMan probes are sequence-specific oligonucleotides containing a fluorophore and a quencher. TaqMan minor groove binder (MGB) probes additionally include a MGB moiety, which helps to increase allelic discrimination using two probes that only differ by one nucleotide.

    PCR included two TaqMan probes; one was specific for the 4G allele and the other was specific for the 5G allele. These probes specifically annealed to the target region between the two primers. The fluorescent dye VIC was used for the homozygous 4G allele, and the fluorescence dye FAM was used for the homozygous 5G allele; the fluorescence signals for both dyes were heterozygous for both alleles (4G/5G). During PCR, DNA polymerase causes primer extension, cleaves the probe at the 5′ end, and separates the fluorophore dye from the quencher dye. The probe then perfectly hybridizes to the target DNA. The fluorescence signal results from this cleavage were monitored by a real-time PCR detection system. The increase in the fluorescence signal was proportional to the amount of the specific released fluorophore, indicating which alleles are present in the sample.

    PCR was performed according to the Duggan et al.’s [17] protocol: 95 °C for 10 min, 95 °C for 15 s and 60 °C for 1 min and then repeating steps of 95 °C for 15 s and 60 °C for 1 min for 40 cycles (for amplification); the post-read step was performed when the PCR was completed. The results of the allelic discrimination run were plotted by the real-time PCR instrument software on a scatter plot for the 4G allele versus the 5G allele, and each well of the rotor was represented as an individual point on the plot. The allelic discrimination plate read was analyzed, and the allele types were documented and verified.

Statistical methods

The results were statistically analysed using the statistical package of social sciences (SPSS 22.0, IBM/SPSS, Inc., and Chicago, IL). Normally distributed variables are expressed as the mean and standard deviation (mean ± SD), while non-normally distributed variables are expressed as the median and interquartile range (IQR). The categorical test results are expressed as the frequency and percentage. For comparing continuous variables, ANOVA was used when normality and homogeneity assumptions were met; if not, then on-parametric equivalent Kruskal-Wallis or Mann-Whitney tests were applied. The Chi-square test was used to compare categorical variables or, alternatively, Fisher’s exact test was used when the Chi-square assumptions were violated.

The association of the genotypes and clinical data with vascular complications in DM patients was explained by odds ratios (ORs) and 95% confidence intervals (95% CI). After univariable analysis was used to identify the potential risk factors for vascular complications, multivariable logistic regression analysis was conducted to estimate the adjusted ORs and 95% CIs of the independently associated risk factors. A P value of 0.1 was set for the variables included in the multivariable model, and highly correlated variables were excluded to avoid multicolinearity between entered variables. Statistical significance was set at a P value < 0.05.

Results

The baseline demographics and clinical characteristics of the studied groups were shown in Table 1. The majority of the studied groups were men (75.7% of the healthy controls, 71.8% of the DM without VCs patients’ and 73.5% of DM with VCs patients). A history of smoking was recorded in 24.8% of diabetic patients without VCs, 28.9% in diabetic patients with VCs, and 18.2% of the healthy controls. Dyslipidemia was observed more frequently in diabetic group with VCs. The comparison of between patients groups and the control showed a matched age and gender with no statistically significant differences regarding smoking, diabetes disease duration, or hypertension.

Table 1 Comparison between control, DM without vascular complications, and DM with vascular complications groups regarding demographic and clinical data

The mean levels of body mass index (BMI), SBP, DBP, fasting blood sugar (FBS), 2-hr PBS, hemoglobin A1C (HbA1c %), total cholesterol, and LDL among the studied groups were significantly higher in diabetic patients (with or without VCs) compared to control group, but were not significantly different between both patient groups. Triglycerides level was significantly higher in diabetic patients with VCs compared to control group. HDL level was significantly lower in diabetic patients without VCs compared to control group.

There were significant difference in the level of PAI-1, TAFI, and d-dimer between studied groups, with the highest levels observed in DM with VCs more than DM without VCs and the lowest levels observed in the control group. PAI-1 median levels were (59, 43, and 32 respectively), while TAFI median levels were (89, 61, and 6.5 respectively) and d-dimer median levels were 2.4, 1.5, and 0.7 respectively).

In Table 2, the most prevalent genotype among the diabetic patients with vascular complications (VCs) was 4G/5G (56.6%), followed by 4G/4G (33.7%), and finally 5G/5G (9.6%). The allelic frequencies of the 4G and 5G alleles were 62.0% and 38% respectively. In diabetic patients without VCs, it was 41.0% with 4G/5G, 32.2% with 4G/4G, and 24.8% with 5G/5G comprising 54.7% with 4G allele and 45.3% with 5G allele. In the control group, 38.1% of the subjects had the 4G/5G genotype, 32.6% had the 4G/4G genotype, and 29.3% had the 5G/5G genotype. In this group, there were 51.7% with 4G allele and 48.3% with 5G allele. The genotype distribution among the three studied groups revealed a highly significant difference (P = 0.007), but there was no statistical difference regarding the allelic frequencies among the studied groups (P = 0.084). On adding the genotypes 4G/4G to 4G/5G in the genetic dominant model of 4G allele (4G/4G + 4G/5G), the distribution among the studied groups revealed more significant difference (P = 0.002) while it revealed non-significant difference (P = 0.957) in the genetic recessive model (4G/4G).

Table 2 Comparisons of genotype distribution and allele frequencies of PAI-1 gene polymorphism between healthy control, DM without vascular complications, and DM with vascular complications groups

The analysis of genotype distribution, allelic frequencies, dominant, and recessive models of PAI-1 polymorphism in solely DM (without VCs) and control groups indicated that 4G/5G and 4G/4G genotypes were significantly associated (P < 0.05) with increased risk of DM as compared to 5G/5G genotype (OR = 1.27, 95% CI 0.71–2.28; OR = 1.24, 95% CI 0.68–2.27, respectively). Similar results were obtained for 4G allele compared to 5G allele (OR = 1.13, 95% CI 0.81–1.57), dominant model [4G/4G + 4G/5G vs. 5G/5G, (OR = 1.26, 95% CI 0.74–2.13)], and recessive model [4G/4G vs. 5G/5G + 4G/5G, (OR = 1.07 95% CI 0.66–1.76)].

Additionally, in DM with VCs and control group, both 4G/5G and 4G/4G genotypes had a highly significant increased risk of VCs associated DM as compared to 5G/5G (OR = 4.51, 95% CI 1.91–10.36, and P < 0.001; OR = 3.14, 95% CI 1.32–7.50, and P = 0.008, respectively). The subjects with the 4G allele also were associated significantly with increased risk of VCs of DM as compared with those carrying the 5G allele (OR = 1.53, 95% CI 1.05–2.23, and P = 0.026). The 4G allele also showed a highly significant association with the risk of VCs of DM in dominant model (OR = 3.88, 95% CI 1.75–8.61, and P < 0.001) but different results were found in recessive model that showed non-significant association (OR = 1.05, 95% CI 0.61–1.83, and P = 0.855) (Table 2).

In the same way, considering both diabetic groups (with and without VCs), the 4G/5G genotype had a highly significant increased risk of VCs among diabetic patients, as compared to 5G/5G (OR = 3.55, 95% CI 1.47–8.56, and P = 0.004), whereas patients with 4G/4G genotype hardly had a significant risk (OR = 2.54, 95% CI 1.01–6.37, and P = 0.044). Patients carrying the 4G allele missed the significant association (OR = 1.35, 95% CI 0.90–2.03, and P = 0.143). However, the 4G allele in dominant model was associated highly significant with the increased risk of VCs (OR = 3.09, 95% CI 1.33–7.16, and P = 0.007) but the association was not significant in the recessive model (OR = 0.98, 95% CI 0.54–1.78), and P = 0.94.

Figure 1 showed that the most prevalent genotype among the diabetic patients with vascular complications (VCs) was 4G/5G (56.6%), followed by 4G/4G (33.7%), and finally 5G/5G (9.6%). In diabetic patients without VCs, it was 41.0% with 4G/5G, 32.2% with 4G/4G, and 24.8% with 5G/5G. In the control group, there were 38.1% of the subjects had the 4G/5G genotype, 32.6% had the 4G/4G genotype, and 29.3% had the 5G/5G genotype.

Fig. 1

Genotype distribution of PAI-1 gene polymorphism in the studied goups

Figure 2 showed that among the diabetic patients with vascular complications (VCs) allelic frequencies of the 4G and 5G alleles were 62.0% and 38.%, respectively and the dominant model (4G/5G + 4G/4G) is more prevalent 90.4 % compared with the recessive model 4G/4G; 33.7%. In diabetic patients without VCs, comprising 54.7% with 4G allele and 45.3% with 5G allele and the dominant model (4G/5G + 4G/4G) is more prevalent 75.2 % compared with the recessive model 4G/4G; 34.2%. In the control group, there were 51.7% with 4G allele and 48.3% with 5G allele and the dominant model (4G/5G + 4G/4G) is more prevalent 77.7 % compared with the recessive model 4G/4G; 32.6%.

Fig. 2

Allelic frequencies, dominant and recessive genotypes of PAI-1 gene polymorphism in the studid groups

Table 3 showed comparison of clinical and biochemical data of the diabetes without vascular complication patient group and the genotypes of PAI-1. There were significant differences between the genotypes of PAI-1 regarding the number of smokers and BMI of the patients with the 4G/5G genotype having the highest prevalence of smokers and highest BMI in the patients. In addition, significant differences were detected between the PAI-1 genotypes regarding the levels of SBP, DBP, PAI-1, TAFI, d-dimer, dyslipidemia, FBS, 2Hrs BS, HbA1c %, total cholesterol, and LDL with the highest levels being expressed in 4G/5G genotype followed by 4G/4G genotype and the lowest level being expressed in 5G/5G genotype. Meanwhile, the level of triglycerides was the highest in 4G/4G genotype followed by 4G/5G genotype with the lowest level being expressed in 5G/5G genotype.

Table 3 Comparison between PAI-1 genotypes regarding demographic and clinical data in diabetes without vascular complication group

In contrast, no significant differences were observed between the PAI-1 genotypes regarding age, gender, the diabetes duration nor the HDL level.

Table 4 showed comparison of clinical and biochemical data of the diabetes with vascular complication patients groups and the genotypes of PAI-1. There were significant differences between the PAI-1 genotypes regarding the levels of PAI-1, TAFI, d-dimer, total cholesterol, and triglycerides with the highest levels being expressed in 4G/5G genotype followed by 4G/4G genotype and the lowest level being expressed in 5G/5G genotype.

Table 4 Comparison between the genotypes of PAI-1 regarding demographic and clinical data in diabetes with vascular complication group

There was no significant difference between the genotypes of PAI-1 genotypes regarding age, gender, number of smokers, hypertension, diabetes duration, dyslipidemia, BMI, 2-Hrs PBS, HbA1c %, LDL nor the HDL levels.

In Table 5, univariable analysis was used to identify the potential risk factors associated with VCs among diabetic patients. The genotypes 4G/5G and 4G/4G were significantly associated risk factors meaning that carrying one 4G allele in dominant model (4G/4G + 4G/5G) had a significant risk (P = 0.009) for VCs (OR = 3.09, 95% CI 1.33–7.16). Additional risk factors significantly (P < 0.05) associated with VCs included PAI-1 (OR = 1.06, 95% CI 1.04–1.08), TAFI (OR = 1.04, 95% CI 1.03–1.05), d-dimer (OR = 1.58, 95% CI 1.20–2.07), FBS (OR = 1.01, 95% CI 1.00–1.03), 2hr BS (OR = 1.01, 95% CI 1.00–1.03), triglycerides (OR = 1.02, 95% CI 1.00–1.03), and HDL(OR = 1.16, 95% CI 1.09–1.24).

Table 5 Potential risk factors independently associated with vascular complications in diabetic patients

Many of these factors were highly correlated to each other and to gene polymorphism (data not shown) so the most relevant variable was selected to enter the multivariable logistic regression model to avoid multicolinearity. Other conventional risk factors and confounders did not associate significantly (P ≥ 0.05) with VCs, those included age, sex, smoking, hypertension, dyslipidemia, LDL, and duration of diabetes (Table 5). The P value at 0.1 was set for variable inclusion in multivariable model. Five variables, PAI-1, FBS, total cholesterol, triglycerides, and HDL, were included into the model where PAI-1 (OR = 1.06, 95% CI 1.03–1.09, and P < 0.001) and HDL OR = 1.18, 95% CI 1.11–1.26 and P < 0.001) were the only variables to be independent risk factors significantly associated with VCs in diabetic patients of the study population.

Discussion

The major causes of disability and death in patients with diabetes mellitus are vascular diseases, particularly atherosclerosis. Diabetes mellitus substantially increases the risk of developing coronary, cerebrovascular, and peripheral arterial diseases. A better understanding of the mechanisms leading to vascular dysfunction may help develop new strategies to reduce cardiovascular morbidity and mortality in patients with diabetes [18].

The PAI-1 gene, also known as serpin E1, is located on human chromosome 7q21.3-q22, spans 12.3 kb, and contains nine exons [19]. PAI-1 is a fast-acting fibrinolytic inhibitor. Increased plasma levels of PAI-1 are associated with increased incidences of thrombophilia and osteonecrosis [20]. The PAI-1 4G/5G polymorphism is associated with high levels of PAI-1, which is induced by the suppression of fibrinolysis by inhibition of the Plasminogen activator and promotion of thrombosis [21, 22].

In our study, there was an association between high levels of TAFI and D-dimers and other risk factors involved in vascular complications in T2DM. Those risk factors included smoking (in 28.9% of diabetic patients with VCs), dyslipidemia, and triglycerides (the level of which was significantly higher in diabetic patients with VCs). These findings agree with the study of Nilsson et al. [23] 2008, who stated that the etiology of cardiovascular disease is multi-factorial but strongly involves genetic and environmental factors. An increase in PAI-1 in vulnerable atherosclerotic plaques associated with an increased inflammatory response might provide the necessary conditions for an atherothrombotic event [23].

Concerning the lipid profile, there were no statistically significant differences in the levels of HDL, LDL, total cholesterol, or triglycerides between diabetic patients without VCs and diabetic patients with VCs. These results were in agreement with those of Wijesuriya et al. [24] and Elnaggar et al. [25]. The HbA1c levels among the studied groups were significantly higher in both diabetic patients groups compared to the control group but were not significantly different between both patient groups. This result agreed with Elnaggar et al. [25], Rahimiet al. [26], and Mtiraoui et al. [27]. A study conducted by Eroglu et al. [28], however, found that the mean value of HbA1c was significantly higher in patients with diabetic nephropathy.

In the present study, there were significant differences in the plasma level of PAI-1 among the studied groups. The highest levels were observed in DM patients with VCs followed by DM patients without VCs. This result was in agreement with the study conducted by Madan et al. [29], who reported that PAI-1 was significantly increased in type 2 patients with microvascular complications, such as diabetic nephropathy. By contrast, Elnaggar et al. [25] reported no statistically significant association between the PAI-1 level and diabetic nephropathy [25].

In our study, the most prevalent genotype among all of the studied groups was 4G/5G (38.1% of controls, 41% of DM patients without VCs, and 56.6% of DM patients with VCs). The allelic frequency of the 4G allele was the highest (51.7% of controls, 54.7% of DM patients without VCs, and 62% of DM patients with VCs). In agreement with our results, Salas et al. [30] found that in the majority of populations around the world, the 4G allele appears at a greater frequency than the 5G allele [30].

Analyses of the genotype distribution, allelic frequencies, and dominant and recessive models of the PAI-1 polymorphism in the solely DM (without VCs) and control groups indicated that the 4G/5G and 4G/4G genotypes were significantly associated with an increased risk of DM development compared to the 5G/5G genotype. Similar results were obtained for the 4G allele compared to the 5G allele, dominant model (4G/4G+4G/5G vs. 5G/5G), and recessive model (4G/4G vs. 5G/5G + 4G/5G). Similarly, Zhao and Huang [19] found that the PAI-1 4G/5G polymorphism was significantly associated with type 2 DM risk and that circulating PAI-1 levels could predict the development of type 2 DM [19].

In the same way, considering both diabetic groups (with and without VCs), the 4G/5G genotype was found to be a highly significant risk factor for VCs among diabetic patients compared to the 5G/5Ggenotype. Patients with the 4G/4G genotype did not have a significant risk of developing type2 DM. However, the 4G allele in the dominant model was highly significantly associated with an increased risk of VCs, but the association was not significant in the recessive model. Similar to our results, Abdel Rasol et al. [31] reported that the 4G/4G genotype contributed to the genetic susceptibility to diabetic retinopathy. A high level of PAI-1 was also independently associated with an increased risk of retinopathy among Egyptians. Additionally, in agreement with our findings, Salas et al. [30] showed that the 4G allele was a risk factor for myocardial infarction in diabetic patients.

In contrast to our findings, a meta-analysis performed by Kuanfeng et al. [32] showed that the PAI-1 (4G/5G) polymorphism might not be a risk factor for DM, diabetic nephropathy, diabetic retinopathy, or diabetic coronary artery disease (CAD).

In the current study, diabetic patients without vascular complications with the PAI-1 4G/5G genotype had the highest prevalence of smoking, high BMI, high SBP and DBP, dyslipidemia, FBS, 2Hrs PBS, HbA1c %, total cholesterol, and LDL. Additionally, the highest levels of d-dimer, TAFI, and PAI-1 were found in patients with the 4G/5G genotype, followed by those with the 4G/4G genotype, and the lowest levels were expressed in those with the 5G/5G genotype. In diabetic patients with vascular complications, nearly the same results were reported, indicating that these risk factors are closely associated with the occurrence of thrombotic complications accompanying DM. The most commonly studied functional variant of the PAI-1 gene is the guanine deletion polymorphism at position 675 relative to the transcription start site (rs1799889). The 4G/5G polymorphism is located in the PAI-1 gene promoter region. The PAI-1 -675 4G allele has higher transcriptional activity than the PAI-1 -675 5G allele, and homozygous 675 4G is associated with higher plasma PAI-1 levels [9], which may explain our findings. In partial agreement with our finding, Salas et al. [30] found that 4G/4G homozygous subjects had the highest plasma concentrations of PAI-1; the lowest plasma concentrations were observed in subjects with the 5G/5G genotype, and intermediate concentrations were recorded in heterozygote subjects (4G/5G) [30]. The differences between our results and those of other studies may be explained by differences in the distribution of the 4G/5G polymorphism as a determining factor for the plasma concentration of PAI-1.

The PAI-1 4G/5G polymorphism is a DNA sequence variation that plays a key role in regulating PAI-1 gene expression. Studies have shown that the PAI-1 activity of the 4G allele promoter is higher than that of the 5G allele promoter in a cytokine-stimulated state. The PAI-1 4G/5G polymorphism also influences PAI-1 gene transcription similarly in non-stimulated cells [19]. Unlike the 5G allele, which binds a transcription repressor protein, resulting in low PAI-1 expression, the 4G allele does not bind a transcription repressor and those confers high PAI-1 expression [33, 34].

Univariable analysis was used to identify the potential risk factors associated with VCs among diabetic patients. The 4G/5G and 4G/4G genotypes were significantly associated risk factors, meaning that carrying one 4G allele in the dominant model (4G/4G+4G/5G) had a significant risk for VCs. Additional risk factors significantly associated with VCs included PAI-1, TAFI, d-dimer, FBS, 2 hr PBS, triglycerides, and HDL. Other conventional risk factors and confounders did not significantly associate with VCs. The other conventional risk factors and confounders included age, sex, smoking, hypertension, dyslipidemia, LDL, and duration of diabetes. In the multivariable model, five variables, PAI-1, FBS, total cholesterol, triglycerides, and HDL, were included, and PAI-1 and HDL were the only variables that were independent risk factors significantly associated with VCs in diabetic patients in the study population.

Variability of PAI-1 plasma concentrations has been reported in different ethnic groups around the world [25]. In some cases, this variability appears to be governed by the 4G/5G polymorphism, while in others, environmental factors, such as smoking, are involved along with certain components of metabolic syndrome, such as dyslipidemia, obesity, and insulin concentration or the interaction between smoking and this syndrome, which increase the risk of cardiovascular disease [26]. The possibility that PAI-1 plays a role in the development of vascular diseases is supported by the biological characteristics of PAI-1 and the association between high plasma PAI-1 and other vascular disease risk factors [26].

Conclusion

In conclusion, the present study suggests that the PAI-1 polymorphism can be considered a risk factor that influences thrombotic events in T2DM. The PAI-1 polymorphism might be useful in further research to provide new strategies for the prevention and treatment of these complications in their early stages.