The rs2442598 polymorphism in the ANGPT-2 gene is associated with risk for diabetic retinopathy in patients with type 1 diabetes mellitus in a Brazilian population

ABSTRACT Objective: As studies have reported the involvement of angiopoietin-2 (ANGPT-2) in the pathogenesis of diabetic retinopathy (DR), the aim of this study was to investigate the association between the ANGPT-2 rs2442598 polymorphism and DR. Materials and methods: This case-control study comprised 107 patients with type 1 diabetes mellitus (T1DM) and DR (cases) and 129 patients with T1DM without DR (controls) and with ≥ 10 years of DM. The ANGPT-2 rs2442598 (G/A) polymorphism was genotyped by real-time PCR using TaqMan MGB probes. Results: Genotype distributions of this polymorphism were consistent with the Hardy-Weinberg equilibrium. The frequency of the rs2442598 A allele was higher in cases compared to controls (p = 0.011). Moreover, the A/A genotype was more frequent in cases than in controls (p = 0.017) and was associated with risk for DR after adjustments for duration of DM, HbA1c, triglycerides, estimated glomerular filtration rate, and hypertension (odds ratio [OR] = 5.19, 95% confidence interval [CI] 1.21-22.27). This association was maintained under recessive (OR = 4.78, 95% CI 1.14-19.99) and additive (OR = 6.861, 95% CI 1.45-32.38) inheritance models. Conclusion: Our data demonstrated, for the first time, an association between the ANGPT-2 rs2442598 A allele and risk for DR in T1DM patients from southern Brazil. Additional studies are necessary to replicate this association in other populations.


INTRODUCTION
D iabetic retinopathy (DR) is one of the most disabling microvascular complications of diabetes mellitus (DM) (1). According to the International Diabetes Federation, DR is the leading cause of blindness in working-age adults (2). Clinically, DR is characterized by the presence of typical retinal microvascular signs, such as microaneurysms, hemorrhages, cotton-wool spots, hard exudates, and neovascularization (1). The development and progression of DR depend on the complex interaction of clinical risk factors, environmental factors, and genetic factors (1,3). In this context, the identification of new genetic polymorphisms associated with DR can contribute to a better understanding of the risk factors and predisposition to this diabetic complication.
Of particular interest for DR are genes related to angiogenesis and vascular development. Among them, angiopoietins (ANGPT) are a family of glycoproteins mainly expressed in endothelial cells, which play an important role in angiogenesis and in the regulation of vascular remodeling by binding to the endothelial receptor tyrosine kinase-2 (Tie-2) (4,5). Tie-2 is highly expressed in endothelial cells and demonstrates strong kinase activity. It becomes phosphorylated on several cytoplasmic tyrosine residues, resulting in the activation of pathways related to the inhibition of de novo blood vessel growth and vascular hyperpermeability such as PI3/AKT and ERK (5,6). ANGPT-1 is involved in vascular maturation, promotes endothelial cell (EC) survival and migration, inhibits EC apoptosis, limits vascular permeability, and exerts anti-inflammatory effects on EC [review in (5,7)]. In contrast, ANGPT-2 can antagonize the effect of ANGPT-1 on Tie-2 receptors and may destabilize the vasculature [review in (5,7)]. Evidence suggests that ANGPT-2, in conjunction with vascular endothelial growth factor (VEGF), can induce vascular sprouting, while ANGPT-2 in the absence of VEGF is associated with EC apoptosis (7,8).
A number of studies have reported the involvement of ANGPT-2 in DR (5,7,9,10). Hammes and cols. (10) demonstrated ANGPT-2 to be upregulated in the retina of diabetic rats compared to controls. Rangasamy and cols. (9) also showed that ANGPT-2 was increased in the retina of diabetic rats after 8 weeks of DM and in human retinal endothelial cells cultured under high glucose conditions. Single nucleotide polymorphisms (SNPs) in the ANGPT-2 gene may thus be associated with DR susceptibility. rs2442598 is an intron SNP and one of the most studied variants of the ANGPT-2 gene in different diseases, being associated with risk for rheumatoid arthritis (11), psoriasis vulgaris (12), and acute lung injury (13). However, to date, no study has analyzed SNPs in the ANGPT-2 gene in patients with DR. Therefore, we investigated for the first time the association between the rs2442598 SNP in ANGPT-2 and DR in patients with type 1 DM (T1DM).

Participants and clinical and laboratory analyses
This case-control study was performed following STROBE and STREGA guidelines for association studies (14,15). A total of 236 participants, including 107 patients with T1DM and DR (cases) and 129 patients with T1DM but no DR and with more than 10 years of DM diagnosis, were enrolled in this study. All patients were recruited at Hospital de Clínicas de Porto Alegre (Rio Grande do Sul, Brazil) between January 2005 and December 2013.
T1DM was diagnosed according to American Diabetes Association (ADA) guidelines (16). The diagnosis of DR was made by an experienced ophthalmologist using fundoscopy (eye fundus examination) through dilated pupils. DR was classified as "absent DR" (without fundus abnormalities), "nonproliferative DR" (NPDR) (microaneurysms, hemorrhage, and hard exudates), or "proliferative DR" (PDR) (newly formed blood vessels and/or growth of fibrous tissue into the vitreous cavity). DR classification was based on the most severely affected eye, according to the Global Diabetic Retinopathy Group scale (17). The Research Ethics Committee at Hospital de Clínicas de Porto Alegre approved the study protocol, and all participants provided assent and written informed consent prior to inclusion in the study.
A standard questionnaire was applied to collect information regarding age, age at T1DM diagnosis, and drug treatments. All patients were subjected to physical examination and laboratory tests, as described previously (18). Serum and plasma samples were collected after 12 h fasting for laboratory analyses (18). Glucose levels were measured using the glucose oxidase method. Glycated hemoglobin (HbA1c) measurements were performed by different methods and the results were traceable to the Diabetes Control and Complications Trial (DCCT) method by offline calibration or through conversion formulae (19). Creatinine was measured by the Jaffé reaction; total plasma cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides by enzymatic methods, and urinary albumin excretion (UAE) by immunoturbidimetry (Sera-Pak immuno microalbuminuria, Bayer, Tarrytown, NY, USA; mean intra-and inter-assay coefficients of variance of 4.5% and 11%, respectively) (20). The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation (21).

Genotyping
Total DNA was extracted from peripheral blood leukocytes using a standardized salting-out procedure (22). The ANGPT-2 rs2442598 SNP (G/A) was SNP in the ANGPT-2 gene and DR Arch Endocrinol Metab. 2021;65/6 genotyped using TaqMan SNP Genotyping Assay 20x (Assay ID = C__15803341_10, Thermo Fisher Scientific, Foster City, CA, USA). Real-time polymerase chain reactions (PCRs) were performed in 384-well plates, in a 5 µL volume, using 2 ng of DNA, the TaqPath ProAmp Master Mix (Thermo Fisher Scientific) 1x, and the TaqMan SNP Genotyping Assay 1x (Thermo Fisher Scientific). Plates were then placed in a real-time PCR thermal cycler (ViiA7 Real-Time PCR System; Thermo Fisher Scientific) and heated for 10 min at 95 ºC, followed by 50 cycles of 95 ºC for 15 s and 62 ºC for 1 min.

Statistical analysis
Allele frequencies were calculated by gene counting, and departures from the Hardy-Weinberg equilibrium (HWE) were verified using the c 2 test. Allele and genotype frequencies were compared between groups using c 2 tests. Additionally, genotypes were compared between case and control groups under additive, recessive, and dominant inheritance models, categorized as suggested by a previous publication (23). Clinical and laboratory characteristics were compared between groups of patients categorized according to the different SNP genotypes using an unpaired Student's t-test, one-way analysis of variance (ANOVA), or the c 2 test, as appropriate. Variables with normal distribution are shown as means ± standard deviations (SDs) or percentages. Variables with skewed distribution were log-transformed before analyses and are shown as medians (25th-75th percentile values). Multivariate logistic regression analyses were done to evaluate the independent association of the rs2442598 SNP with DR, adjusting for possible confounding factors. Variables with significant associations with DR in the univariate analysis or with relevant biological association with this complication were chosen for inclusion in the multivariate model. Statistical analyses were performed using SPSS 18.0 software (SPSS, Chicago, IL, USA), and p-values < 0.05 were considered significant.
Sample sizes were calculated using the OpenEpi website (www.openepi.com). Since no studies have yet investigated the association between the rs2442598 SNP and DR, we used frequencies published by previous studies that assessed the association of this SNP with other diseases (mean minor allele frequency = 0.30 and odds ratio [OR] = 2.2) (11,13). Therefore, the calculated sample size was estimated as 108 participants per group.

Sample description
The clinical and laboratory characteristics of patients with T1DM with and without DR are shown in Table 1. Men comprised 52.3% of the cases and 53.5% of the control group (p = 0.964). The mean age of participants was higher in cases with DR when compared to controls (41.4 ± 12.1 vs. 34.6 ± 12.2 years, p < 0.001). As expected,   Figure 1). However, the rs2442598 SNP was not associated with DR under the dominant model (p = 0.073), and this result did  Figure 1).

Genotype and allele frequencies
In an exploratory analysis, the clinical and laboratory characteristics of patients with T1DM, regardless of their DR status, were compared after stratification according to the presence of the rs2442598 A/A genotype under the recessive model ( Table 3). The frequency of male participants and mean values of body mass index (BMI), HbA1c, age at T1DM diagnosis, duration of DM, total cholesterol, triglycerides, HDL cholesterol, LDL cholesterol, eGFR, and UAE were similar between patients with A/A or G/G + G/A genotypes (p > 0.050). Interestingly, the presence of hypertension was more frequent in patients carrying the A/A genotype in comparison with patients carrying the G/G + G/A genotypes (76.9% vs. 44.3%, p = 0.045).

DISCUSSION
Our results demonstrated, for the first time, that the A allele of the rs2442598 SNP in the ANGPT-2 gene is associated with risk for DR in patients with T1DM from southern Brazil. Although the exact role of ANGPT-2 in angiogenesis and its involvement in disease mechanisms are not fully understood, there is enough data to link ANGPT-2 to DM and its vascular complications (10,(24)(25)(26).
El-Asrar and cols. (24) evaluated serum ANGPT-2 in children and adolescents with T1DM as a potential marker of vascular complications. Compared with healthy controls, ANGPT-2 levels were higher in patients with T1DM and were positively correlated with fasting blood glucose and HbA1c levels. When comparing patients with T1DM with or without vascular complications, the highest ANGPT-2 levels were observed in those with complications. ANGPT-2 also had a positive correlation with carotid and aortic intima-media thickness (24). In T2DM patients, a study showed that serum ANGPT-2 levels were higher in patients with cardiovascular disease (CVD) compared to those without CVD or healthy controls. Moreover, ANGPT-2 levels were associated with a 2.83-fold higher risk of T2DM and were positively correlated with fasting blood glucose, HbA1c, homeostatic model assessment of insulin resistance (HOMA-IR), triglycerides, and LDL cholesterol (25).
The involvement of ANGPT-2 in PDR has also been well demonstrated. Klaassen and cols. (26) studied vitreous samples from patients with PDR (with and without fibrovascular membrane [FVM] formation) and from non-diabetic patients with vitreous debris to identify proteins associated with neurite outgrowth, angiogenesis, and fibrosis in the formation of FVM in PDR. Among 8 other associated proteins, ANGPT-2 showed the strongest correlation with the degree of fibrosis and presence of FVM in patients with PDR, pointing to an additional role in the development of PDR apart from active angiogenesis (26). Another study demonstrated increased ANGPT-2 levels in patients with DM compared to healthy controls, with the highest ANGPT-2 levels being observed among patients with pre-proliferative or PDR (27). In contrast, Malik and cols. (28) did not find any differences in serum ANGPT-2 levels between patients with diabetes with or without DR. Another evidence of the involvement of ANGPT-2 in DR is related to its effect on pericyte loss in the early phases of DR development. An experimental study conducted by Hammes and cols. (10) demonstrated that Angpt-2 was upregulated more than 30-fold in the retina of diabetic rats when compared to nondiabetic controls and was associated with pericyte loss in DR. When recombinant Angpt-2 was injected into the eyes of normal rats, pericyte loss was observed in a dose-dependent manner (10). Park and cols. (29), studying the molecular mechanisms behind the role of hyperglycemia-induced Angpt-2 in pericyte loss, demonstrated that it could induce pericyte apoptosis via the p53 pathway.
Furthermore, Angpt-2 was associated with astrocyte loss in DR. Astrocytes are neurovascular components of the retina that contribute to vessel integrity and are closely related to the formation of tight junctions in the retina (30). Yun and cols. (30) studied the role of ANGPT-2 in astrocyte loss and vascular leakage in the retina of early diabetic mice. They were able to show that, as Angpt-2 increased, vascular leakage with loss of astrocytes occurred and was induced by apoptosis via the GSK-3β/β-catenin pathway under high glucose conditions. Similarly, Rangasamy and cols. (9) demonstrated the ability of Angpt-2 to alter the retinal endothelial cell barrier in vitro and in vivo. Retinal endothelial cells cultured with increasing concentrations of purified recombinant ANGPT-2 showed a dosedependent increase in monolayer permeability. Moreover, a threefold increase in vascular permeability was observed when nondiabetic rats received a single intraocular injection of recombinant purified Angpt-2 (9). Hence, ANGPT-2 seems to be closely involved with the damage and loss of pericytes and astrocytes that directly participate in the pathogenesis of DR.
This study has a few limitations. First, we cannot rule out the possibility of population stratification bias when analyzing our samples, even though only White participants were studied and both the cases and controls were recruited from the same hospital, thus reducing the risk of false-positive/negative associations due to this bias. Second, this is the first study that demonstrated an association between the rs2442598 SNP in the ANGPT-2 gene and risk for DR, and we did not perform a replication of the observed association in another Brazilian sample. Third, there is a lack of information about how rs2442598 affects ANGPT-2 expression and what is its functional role on DR susceptibility. Therefore, additional genetic studies are needed in order to confirm the association between the rs2442598 SNP in ANGPT-2 and the risk for DR in different ethnicities and populations. Moreover, functional studies are necessary to better understand how rs2442598 affects ANGPT-2 expression and activity.
In conclusion, the present data suggest an association between the A allele of the rs2442598 SNP in ANGPT-2 and susceptibility to DR in patients with T1DM. To our knowledge, this is the first report of a polymorphism study of the ANGPT-2 gene suggesting risk for DR. This finding needs to be further evaluated in other populations, although there is a good rationale for this association to be relevant given the role of ANGPT-2 in the mechanisms involved in DR pathogenesis.