The association between plasma concentration of pigment epithelium-derived factor and diabetic retinopathy

Introduction

Diabetes mellitus (DM) is a metabolic disease that is caused by the deficiency/absence of insulin in the body or the resistance to the effects of insulin, and progresses with disorders in carbohydrate, lipid and protein metabolism as a result of chronic hyperglycemia [1]. The DM incidence gradually increases all over the world. 537 million adults (20–79 years) are living with diabetes. This number is predicted to rise to 643 million by 2030 and 783 million by 2045 [2]. Microvascular and macrovascular complications are the most significant causes of work loss, morbidity, and mortality in diabetic persons. It has been detected that approximately 14 % of health expenditures are provided to these complications of diabetes. Therefore, clarifying pathophysiological mechanisms has drawn attention to preventing these diabetes-induced complications. Recent studies have shown that the leading cause of organ damage in diabetes mellitus is inflammation [3, 4]. The most common microvascular complication of diabetes is diabetic retinopathy (DRP). DRP is the most important cause of blindness and appears in individuals between 24 and 70 years of age who actively work [5].

Pigment epithelium-derived factor (PEDF) is a protein with a weight of 50 kDa identified as a neurotropic factor first detected in human retinal pigment cells [6]. Although PEDF was first seen in the eye, it is known to be synthesized in various organs such as the brain, spinal cord, liver, heart, placenta, bone, pancreas, and prostate [7]. Primary circulatory resources of PEDF include liver and adipose tissue [8]. PEDF is a multifunctional protein with neurotropic, neurotrophic, neuroprotective, anti-tumorigenic, anti-angiogenic, and anti-vascular permeability characteristics. PEDF is one of the most potent endogenous angiogenesis inhibitors [9]. It has been reported that PEDF may bring a new approach to treating vascular complications in diabetic patients thanks to these characteristics [10, 11].

It has been previously shown that the reduction of retinal (retinal capillary endothelial cells and Müller cells) PEDF levels in an animal model with DRP is pathological. It was stated in the study above that the decrease/loss of PEDF levels may increase retinal inflammation and thus contribute to the pathogenesis of DRP [12]. However, the mechanism of action of extracellular PEDF remains unclear. This study aimed to determine any possible association between serum PEDF levels and DRP.

Materials and methods

This cross-sectional case-control study was conducted with ophthalmology and internal medicine clinics. The Helsinki Declaration was adhered to throughout the study. Ethical approval from Ethical Committee for Clinical Researches (2019/67) and informed consent form from all participants were obtained. Totally 82 individuals, including 41 patients with DRP and 21 diabetic patients without DRP who, were referred to our hospital between December 2019 and November 2020 and 20 healthy participants were enrolled into the study.

The complete ophthalmological examinations of the participants were performed by the eye clinic (T.S). Systemic diseases and medication history were investigated. Fundus Fluorescein Angiography (FFA) was performed in all patients with DRP findings in the fundus examination to determine the stage of DRP. DRPs of the patients were divided into 2 groups as non-proliferative DRP (NPDRP) (28 patients) and proliferative DRP (PDRP) (13 patients) depending on FFA findings—blood samples for analyses including serum PEDF levels of all participants collected taken by the internal medicine clinic. All participants’ body mass index (BMI) was calculated by dividing the weight in kg into the square of the height in meter.

Results

A total of 82 volunteers, consisting of sick and healthy volunteers, were included in the study. 41 of them were diabetic patients with DRP, 21 of them were non-DRP diabetic patients, and 20 of them were healthy volunteers. DRP finding was absent in 21 diabetic patients; however, 28 patients had NPDRP, and 13 had PDRP.

Demographic data of the groups, age, gender distribution, body mass index (BMI), and systolic and diastolic blood pressures were similar in all groups (Table 1).

When we examined the biochemical data of the groups, FBG and HbA_1c levels in the DRP and non-DRP groups were significantly higher than those in the control group [for FBG; for p<0.001 and HbA_1c; p<0.001]. All groups had similar Triglyceride, total cholesterol, HDL, and LDL cholesterol values. In addition, serum creatinine values, total bilirubin, indirect bilirubin and direct bilirubin values, and neutrophil-lymphocyte ratio (NLR) were similar in the three groups (Table 2).

There was a significant difference between serum PEDF levels of diabetic patients (1.533 ± 0.233 μg/mL) and (2.163 ± 0.343 μg/mL) healthy participants (p=0.002). (Mann Whitney U test) (variation coefficient %105).

The PEDF level in the DRP subgroup of diabetic patients (n=41) was 1.341 ± 0.182 μg/mL, whereas the PEDF level in the non-DRP subgroup (n=21) was 1.910 ± 0.592 μg/mL (p=0.337) (Mann Whitney U test).

The blood PEDF levels in the healthy individuals were higher than the diabetic group; however, PEDF levels decreased significantly as diabetic retinopathy progressed (2.163 ± 0.343 μg/mL, 1.910 ± 0.592 μg/mL, 1.588 ± 0.247 μg/mL, 0.806 ± 0.127 μg/mL in the control group, non-DRP, NPDRP, PDRP groups, respectively) (p=0.001) (Figure 1) (Table 3).

Figure 1:

PEDF levels according to the severity of the disease. NPDRP, non-proliferative DRP; PDRP, proliferative DRP.

A significant negative correlation was observed between the progression of retinopathy and PEDF levels (Spearman rho correlation coefficient ρ= −0.288; p=0.009). Progression of retinopathy was associated with fasting blood glucose (ρ=0.489; p=0.001) and HbA_1c (ρ=0.613; p=0.001) levels. No significant correlation was detected between total bilirubin, direct bilirubin, indirect bilirubin, neutrophil-lymphocyte ratio, total cholesterol, HDL, LDL, triglyceride, BMI, creatinine levels and progression of retinopathy (p>0.05).

Discussion

The two leading causes that threaten vision in diabetic patients are retinal neo-vascularization and macular edema [13]. The result of the study revealed that the level of PEDF, which is known as one of the most potent inhibitors of vascularization in the serum of diabetic patients, was lower than in the healthy control group. We observed that blood PEDF levels decreased further in diabetic patients as retinopathy progressed.

It is consensually accepted that DRP is an inflammatory disease. The earliest histopathological indicator for DRP is the loss of pericytes. Loss of pericytes causes basal membrane thickening, increased permeability, and formation of microaneurysms. It was reported that proliferative changes associated with neovascularization occur after loss of pericytes. It is known that the increase in vascular permeability is the main reason that threatens vision in DRP. Advanced glycation end-products (AGEs) were shown to induce the expression of intercellular adhesion molecule (ICAM-1) and increase the leukocyte adhesion into endothelial cells. VEGF was detected as the main etiological factor that increases ICAM-1 expression and vascular permeability [14].

PEDF is a protein synthesized from lung, brain, kidney, liver, and adipose tissue [15]. PEDF blocks endothelial cell activation, smooth muscle cell proliferation, and leukocyte adhesion to endothelial cell [16]. It has been shown that PEDF reduces TNF-α expressed by hypoxia in retinal capillary endothelial cells [17]. When Wang et al. administered PEDF-expressing adenovirus to diabetic rats intravenously, they showed that the expression of proinflammatory factors such as TNF-α, MCP-1, ICAM-1, and VEGF was suppressed, and retinal neovascularization was inhibited [18]. Moreover, PEDF is known as one of the most potent inhibitors of angiogenesis [11]. The hypothesis was previously suggested that PEDF would suppress the proliferative inflammatory response against injuries on endothelial cells and may play a protective role against vascular injuries [19].

It was found that diabetic patients, especially those with proliferative retinopathy, had lower PEDF levels in the vitreous and aqueous humor [20, 21]. Sabater et al. showed a positive correlation between PEDF levels and insulin resistance. They detected that the serum PEDF level in diabetic patients was significantly higher than in the control group. They also showed a significant decrease in PEDF levels after weight loss in the diabetic group [22]. Jenkins et al. found plasma PEDF levels higher in diabetic patients when compared to healthy individuals [23]. These two studies reported that this difference in PEDF level between the groups was positively correlated with BMI. It was stated that increased PEDF level was associated with insulin resistance and obesity. It was found in the study above that serum PEDF levels decreased in individuals who lost weight, and PEDF levels increased in those with metabolic syndrome and type 2 diabetes [24]. Akın et al. obtained different results in their study. They reported that PEDF levels increased in diabetic patients receiving metformin treatment and those who lost weight [17]. They stated that the most likely reason for this may be associated with the fact that metformin directly increases PEDF production and/or release from adipose tissue. They also reported that these results may be due to the small number of patients and the shorter follow-up period. One strength of our study was the absence of significant differences in BMIs between groups. We believe the similarity of the BMI factor affecting PEDF between the groups would be a substantial aspect of our study.

Ogata et al. reported that plasma PEDF levels were significantly higher in type 2 diabetic patients, especially in patients with PDRP, compared to the healthy control group. They stated that higher PEDF levels in plasma may be associated with the progression of diabetic retinopathy [25]. However, Hui et al. suggested that the serum PEDF level was significantly higher in type 2 diabetic patients and that this increase in PEDF was related to the violence of diabetic nephropathy. They reported that such increase in PEDF level in patients with nephropathy may be a compensatory system to prevent disease worsening [26]. Similarly, Matsuyama et al. suggested that the increase in plasma PEDF levels is associated with the progression of the disease in both diabetic retinopathy and nephropathy. They remarked in their study that elevated PEDF levels in the blood may show microvascular damage in diabetic patients and may be a predictor of retinopathy and nephropathy progression [27]. However, the BMI difference between the groups was not considered. Higher PEDF levels in diabetic patients in this study may be associated with higher BMI (It is known that one of the primary sources of PEDF is adipose tissue) values. No information was provided about BMI values of the participants. Jenkins et al. announced that elevated PEDF levels in the sera of diabetic patients may be associated with microvascular complications. Jenkins et al. also reported that serum PEDF levels were associated with microvascular complications, poor vascular health, hyperglycemia, adiposity, and inflammation in patients with Type 1 diabetes. There was a significant difference between the groups’ BMI values that were compared [28]. Another study detected an increase in plasma PEDF levels in patients with diabetic microangiopathy (retinopathy or nephropathy). In the analysis above, the authors suggested that the increase in PEDF works like a defense system to inhibit vascular injury in diabetes [28]. Although it was known that BMI affects PEDF levels in previous studies, this was not kept similar when comparing groups. The BMI values of the groups were similar in our study. The possible effect of BMI on PEDF levels was thereby eliminated. Diabetic patients enrolled into our study were those diagnosed with DM for over a decade. We detected that PEDF levels decreased in diabetic patients and as the stage of DRP progressed. We believe that this (as the duration and complications of diabetes increase) may be related to the decrease in blood level due to consumption of PEDF in peripheral tissues.

We have a limited number of patients and sub-group patients. We had a challenging time collecting patients due to the Covid-19 pandemic that occurred during patient collection phase, and we could not increase the number of patients to the level we wanted.

Conclusions

Blood PEDF levels change in DM. Decreased blood PEDF levels may cause emergence and progression of DRP. Agents containing PEDF may be used intraocularly/systemically for therapeutic purposes to prevent vascular complications of diabetes shortly. Further studies with more extensive patient series would guide us.