Changes of Circulating Acylated Ghrelin and Neutrophils Elastase in Diabetic Retinopathy Progression

1. Introduction

Diabetic retinopathy (DR) represents the main cause of blindness, especially in industrialized countries, with a significant impact on health care costs [1]. Although intravitreal injections of anti-VEGF drugs and steroids seem to be effective for the management of both Non-Proliferative (NPDR) and Proliferative (PDR) DR forms [2,3], DR detection is still a challenge since it requires complex and expensive equipment, along with well trained technicians [4]. Therefore, the identification of new systemic biomarkers correlated to the stage of the disease, and involved in its evolution, could be a useful novel tool to predict or early identify NPDR and to slow its progression to PDR [5].

To this regard, ghrelin is a peptide hormone produced by gut that beyond its role as an orexigenic signal [6], mediates anti-apoptotic, autophagic and anti-inflammatory effects at the level of the eye structures, as for example in the epithelial cells of the lens membrane [7], and in the retina [8]. Interestingly, previous preclinical studies reproducing DR in vitro and in vivo have shown that ghrelin reduces retinal angiogenesis, apoptosis and inflammation induced by high glucose through the activation of the ghrelin receptor [9,10]. A receptor known as growth hormone secretin receptor 1a (GHSR-1a) and characterized by its higher affinity to acylated form of ghrelin compared to the des-acylated one [11]. Through this receptor, ghrelin is able to promote the apoptosis of neutrophils [12,13], leukocytes highly involved in inflammation and endothelial damage during DR evolution, contributing to DR complications with the extrusion of neutrophil extracellular traps (NETs) [14]. A process called NETosis, worsened by hyperglycemia [15,16] and exacerbated in DR patients [17,18]. Thus, ghrelin could impact the progression of DR through a cross-talk with neutrophils and the related risk of NETosis process in diabetic patients.

Therefore, the present study aimed to evaluate in humans whether there are putative changes in total levels of serum ghrelin and/or its acylated/des-acylated forms during different phases of DR, and whether they correlate with clinical signs and serum elastase levels, indicative of NET formation [19] and vascular damage [15,19,20].

2. Results

2.1. Characteristics of DR Patients

57 patients (31 males and 26 females) were enrolled in this study and divided in the following four groups, based on ocular evaluations and DR clinical diagnosis:

• Non-diabetic subjects with absence of ocular pathologies (N = 12) - CTRL group;
• Diabetic patients without DR signs (N = 15 ) - Diabetic group;
• Diabetic patients with diagnosis of non-proliferative DR (N = 15) - NPDR group;
• Diabetic patients with diagnosis of proliferative DR (N = 15) - PDR group;

Differences in age or diabetes duration were not observed in diabetic groups. Conversely, PDR patients exhibited significantly higher glycaemic levels compared to Diabetic and NPDR groups (both P < 0.05) (Table 1).

Moreover, neutrophils percentage significantly increased in NPDR patients compared to Diabetic group (P < 0.01). This was even more elevated in PDR patients (P < 0.01 vs NPDR) (Table 1), although the values for all the 4 clinical groups were in the normal range.

2.2. Serum Ghrelin Levels in DR Patients

Serum total ghrelin (TG) was significantly decreased in Diabetic group compared to CTRL subjects (CTRL: 1056+146 pg/mL; Diabetic: 925+86 pg/mL, P < 0.05 vs CTRL). A reduction in serum TG levels, although not significant in comparison with Diabetic, was detected also in NPDR group (NPDR: 861+126 pg/mL, P > 0.05 vs Diabetic), while it was significant in PDR patients compared to Diabetic group (PDR: 807+127 pg/mL, P < 0.05 vs Diabetic) (Figure 1A).

Similarly, serum acylated ghrelin (AG) was significantly decreased in Diabetic group compared to CTRL subjects (CTRL: 275±93 pg/mL; Diabetic: 171 ±31 pg/mL, P < 0.01 vs CTRL). A further significant reduction was evident in NPDR patients compared to Diabetic group (NPDR: 115 ±24 pg/mL, P < 0.05 vs Diabetic), with the lowest serum AG levels detected in PDR sera (PDR: 63±32 pg/mL, P < 0.01 vs NPDR) (Figure 1B). Conversely, serum des-acylated ghrelin (DAG) was not differentially modulated in our clinical setting (CTRL: 800±182 pg/mL; Diabetic: 926±165 pg/mL; NPDR: 917±248 pg/mL; PDR: 990±193 pg/mL) (Figure 1C).

2.3. Serum AG/DAG Ratio and Its Correlation with Retinal Abnormalities in DR Patients

Consequently, AG/DAG ratio was significantly decreased in Diabetic patients (0.23 ± 0.05, P < 0.01 vs CTRL) compared to CTRL group (0.35 ± 0.1). A progressive significant reduction of AG/DAG ratio was evident also in NPDR (0.16 ± 0.04, P < 0.05 vs Diabetic CTRL) and PDR (0.08 ± 0.04, P < 0.01 vs Diabetic and P < 0.05 vs NPDR) patients (Figure 2A).

The decrease of AG/DAG was paralleled by an increase of microhemorrages, microaneurysms and exudates detected in NPDR and PDR patients respectively by retinography . Indeed, CTRL subjects evidenced a normal retinal vasculature without abnormalities and Diabetic patients showed a slight irregularity of the caliber and vessel course without DR classic signs (Figure 2B). Conversely, NPDR patients evidenced multiple microhemorrhages (1- ≥ 20), microaneurysms and exudates in the four retinal quadrants, while PDR group showed retinal neovascularization, and/or extensive fibrovascular proliferation, or vitreous/preretinal hemorrhages. This increase of retinal abnormalities identifying DR progression was significantly inversed correlated with serum AG/DAG ratio (r = -0.83, P < 0.01) (Figure 2C). Therefore, since the modulation of AG/DAG ratio in DR patients was due to serum AG changes, this was considered for further correlations.

2.4. Serum Neutrophil Elastase (NE) Levels in DR Patients and Its Association with Serum AG

Serum AG were inversely associated in DR patients with serum neutrophils percentage (r = -0.74, P < 0.01) (Figure 3A).

As expected, this positively correlated with serum NE levels (0.80, P < 0.01) (Figure 3B). Indeed, serum NE was significantly increased in Diabetic group compared to CTRL subjects (CTRL: 0.85±0.3 ng/mL; Diabetic: 1.6±0.4 ng/mL, P < 0.05 vs CTRL) (Figure 3C). Moreover, NPDR patients showed serum NE levels significantly elevated compared to Diabetic group (NPDR: 2.2±0.4 ng/mL, P < 0.01 vs Diabetic), with the highest value reached in PDR patients (PDR: 2.9±0.6 ng/mL, P < 0.01 vs NPDR) (Figure 3C). A strong positve correlation was observed between serum NE levels with DR progression (r = 0.86, P < 0.01) (Figure 3D), while a significant inverse association was evident between serum NE and AG (r = -0.73, P < 0.01) (Figure 3E).

3. Discussion

Ghrelin, often referred to as the “hunger hormone”, mediates several functions beyond its effects on appetite, food intake, body weight and adiposity [7]. Indeed, once acylated by the ghrelin-O-acyl transferase (GOAT), the 28-aa Ser3 AG binds with high affinity to the G-protein coupled receptor GHSR1-a on central neurons and peripheral cells [21,22,23]. On these, ghrelin acts by increasing energy intake [24,25], improving cardiac functions [26,27], reducing muscle atrophy [28,29], promoting bone mass or formation [30,31] and through this receptor it promotes the apoptosis of neutrophils [12,13], leukocytes highly involved in inflammation and endothelial damage during DR evolution.

Noteworthy, diabetic retinopathy is a complication of diabetes that damages retinal blood vessels [32]. The typical microvascular alterations that characterize this disease arise with a predictable progression and this allows serious damage to vision to be prevented. Indeed, if neglected, diabetic retinopathy can cause severe vision loss or even blindness [33]. In the NPDR early stage, vascular occlusion and dilation occur; subsequently the condition evolves into PDR, with the growth of new blood vessels on the retinal surface, a process known as neovascularization [33]. At the basis of these changes in the retina there is a wide range of mediators, some pro-proliferative, others anti-proliferative. Among these is ghrelin, a peptide hormone whose endogenous levels can be compromised by pathologies such as diabetes, that seems to have anti-proliferative effects. In fact, previous studies have found a reduced circulating ghrelin concentration in type 2 diabetes [34,35,36], both in the early stages and advanced stages with complications [37,38]. In line with this evidence, here it is indirectly confirmed the protective role of ghrelin in DR since diabetic patients have decreased serum levels of total ghrelin compared to non-diabetic patients, especially those with progression from the NPDR to PDR. However, the novelty of the research with respect to previous one resides in the fact that of the two known isoforms of the total ghrelin, acylated and des-acylated, the acylated form follows the reduction trend shown by the total ghrelin. Conversely, serum des-acylated ghrelin was not differentially modulated in the present clinical context, despite it is the main form secreted in physiological conditions, with an AG/DAG ratio of 1:10 [39,40]. An explanation for this may reside in the fact that diabetes impairs the expression of enzymes that rapidly deacylates AG, such as acyl protein thioesterase 1 which normally deacylates AG [41,42]. This in line with a recent study showing that serum AG seems to be related to elevated blood glucose levels in diabetic obese patients, while DAG seems to be involved in excess body fat mass in the same clinical setting [43]. More specifically, the present study evidenced that changes of serum TG in DR patients are due to AG constant decrement in NPDR and PDR patients. Particularly, serum AG levels decreased significantly in NPDR and PDR patients and showed a strong negative correlation with DR evolution Therefore, in line with the novel clinical applications suggested for AG in Alzheimer’s and Parkinson’s Disease [44], lipodystrophy [45], reproductive toxicity in cancer patients [46] and heart failure [47], serum AG could be considered a sensitive peripheral marker of clinical progression of DR.

From the mechanistic point of view, the changes of AG in DR evolution mirrored the changes and behavior of neutrophils, where ghrelin and its receptor are expressed [48,49]. Indeed, here it is demonstrated an increase of neutrophils percentage in DR stages, which negatively correlates with serum AG levels. During DR, neutrophils play a very important role in microangiopathy occurrence and progression, as well as in the endothelial cell wall inflammatory processes [14,50], since they secrete proteolytic enzymes involved in endothelial damage [14]. This latter leads to chronic inflammation which exacerbates microvascular complications, contributing to DR progression [14,51]. Indeed, DR patients are characterized by high serum neutrophil count [52], with a higher neutrophil-lymphocyte ratio [52,53,54]. This correlating with DR severity in clinical settings [55,56,57]. So, to our knowledge this is the first evidence of a possible endogenous crosstalk between AG and neutrophils for the development and progression of DR. On another note, it is well known that acting on these immune cells, ghrelin exerts anti-inflammatory effects by reducing their migration and infiltration [13,58,59,60,61]. Similarly, ghrelin has been shown to promote neutrophil apoptosis [12,60,61], overall favoring the resolution of inflammation in several diseases, such as arthritis, sepsis and respiratory pathologies [13,58,62,63,64].

The progression of diabetes and diabetic retinopathy (DR) has recently been linked to elevated levels of neutrophil elastase (NE) in circulation, a protein found within the granules of neutrophils. Studies have shown that levels of NE are higher in both type-1 and type-2 diabetic patients when compared to individuals without diabetes [15], along with PDR patients compared to NPDR group [17]. Particularly, NE generally reflects the neutrophil count and contributes to DR pathological vascular permeability [15,19,20]. Moreover, NE is one of the constituents of NETs, along with histones, DNA, fibers and several proteins such as NE, myeloperoxidase (MPO), cathepsin G, cathelicidin and proteinase 3 (PR3) [65]. NETs are a sort of physical barrier against pathogens [66,67] organized by apoptotic netrophils also implicated in the pathology associated with a growing number of immune-mediated conditions [68] and in a process called NETosis. NETosis is induced in an organ in response to endogenous danger signals that must be tightly regulated to prevent excessive tissue damage during acute inflammation or chronic inflammatory and autoimmune disease [68]. Among the other, NETs can occlude the vasculature by promoting thrombosis and obstruct important organ areas. Both diabetes and DR progression seem to be characterized by NETs enhanced formation and release [15,17]. In line with this evidence, it is found here a positive correlation between the increase of neutrophils in the two DR stages analyzed and the serum NE levels. These were increased in both groups with higher levels in PDR patients compared to NPDR patients. Serum NE levels were positively correlated with DR progression, by evidencing a higher risk of NETs formation, which strongly impact the occurrence of thrombosis and hemorrhages [69,70], two of the PDR hallmarks [71]. Furthermore, for the first time it is shown here an inverse correlation between serum NE and AG in DR patients. Therefore, it could be hypothesized that the physiological role of AG in diabetes and uncomplicated retinopathy is to promote neutrophil apoptosis, consequently reducing the risk of NET formation and complications of PDR.

In conclusion, the data presented suggest that circulating ghrelin and its acylated form could be potential markers for the progression of diabetic retinopathy since associates itself to reduced circulating neutrophils activation and reduced NETs formation during the progression of DR from NPDR to PDR. However, it's important to acknowledge the limitations of this study, including the small sample size of patients, the absence of consistent follow-up, and the lack of a direct cause-and-effect study on ghrelin. Bearing these constraints in mind, it's still plausible to consider ghrelin monitoring as a potentially valuable approach in the proactive prevention of DR progression.

4. Materials and Methods

4.1. Clinical Design

This study was performed at the Eye Clinic of the University of Campania “Luigi Vanvitelli” (Naples, Italy), with all its procedures adhering to the Declaration of Helsinki and Good Clinical Practice guidelines. The study was approved by the ethic committee of the University of Campania “Luigi Vanvitelli” (protocol number 42 DEC, 30-1-2019 and protocol number 0003239/i, 01-02-2023). The patients were enrolled after a basal examination, needed to obtain the written informed consent and to evaluate the adherence to the following inclusion criteria: I) Age of at least 46 years; II) Duration of diabetes of at least 4 years; III) Clinical diagnosis of NPDR or PDR by medical history and ocular fundus examination; IV) Hypoglycemic therapy with Sodium-Glucose Cotransporter 2 inhibitors [72]. Patients with body mass index higher than 30 kg/m² (obese patients), uncontrolled diabetes, other severe diabetic complications (ketoacidosis/nephropathy), recent or systemic infection and recent cardiovascular diseases were excluded. Furthermore, patients using lipid-lowering agents or immunosuppressive, steroidal and non-steroidal anti-inflammatory drugs (NSAIDs) were not enrolled, as well as patients receiving intravitreal steroids/anti-Vascular Endothelial Growth Factors (VEGFs) , and argon laser coagulation or vitrectomy in the last 6 months.

4.2. DR Diagnosis

DR diagnosis was performed by obtaining the patient’s medical history and by performing ocular fundus examination. DR stages were determined by the number of intraretinal microvascular abnormalities, microaneurysms, haemorrhages and retinal neovascularization, according to the American Academy of Ophthalmology Retina-Vitreous Panel [73]. Retinography analysis was performed by using Topcon Engineering.

4.3. Serum Samples Collection and Analysis of Serum Markers

Sterile dry vacutainer tubes were used to obtain fasting venous blood samples from the enrolled patients. Within 2 hours from sampling, blood was incubated at 20°C for 30 minutes before being centrifuged at 4°C for 15 minutes at 3000 rpm, to have serum samples as supernatants. Their aliquots were stored at -80°C for subsequent analysis of serum markers, performed at the Pharmacology Section of University of Campania “Luigi Vanvitelli”) by Enzyme-Linked Immunosorbent Assays (ELISAs). Serum AG, total ghrelin (TG) and NE levels were assessed by using commercial ELISAs according to the manufacturers’ instructions (respectively, EH2601 Human Acylated Ghrelin ELISA Kit, FineTest – Wuhan, China; EH0355 Human Ghrelin ELISAS Kit, FineTest – Wuhan, China; BMS269 Human PMN-Elastase ELISA Kit, Invitrogen – Waltham, MA, USA). Serum DAG levels were calculated by subtracting AG from TG [74].

4.4. Statistical Analysis

Prism 6.0 (GraphPad, San Diego, CA) software was used for both statistical analysis and graph design. One-way analysis of variance (ANOVA), followed by Tukey multiple comparison post hoc test, was used for the determination of differences between groups. The strength of association between 2 quantitative variables was evaluated by Pearson correlation analysis. A P-value < 0.05 was considered significant.