No Correlation between Serum Progranulin Levels and Severity of Diabetic Retinopathy in Chinese People with Type 2 Diabetes and Normal or Mildly Impaired Renal Function

Background: Progranulin (PGRN), a novel pro-inflammatory adipokine, was reported to be related to the development and progression of diabetic retinopathy(DR). However, recently PGRN was established as a renal function-dependent protein, but Kruskal-Wallis test followed by post hoc analysis. Categorical parameters were analyzed using the χ2 test. PGRN: progranulin, BMI:body mass index, SBP: systolic blood pressure, DBP: diastolic blood pressure, HbA1c: hemoglobin A1c, FPG: fasting plasma glucose, FINS: fasting serum insulin, FCP: fasting C-peptide, HOMA-IR: Homeostasis Model Assessment for Insulin Resistance, TyG: triglycerides and glucose index, BUN: blood urea nitrogen, CRE: creatinine, uric acid, eGFR: estimated glomerular filtration rate, ACR: albuminuria/creatinine ratio, TG: triglyceride, total cholesterol, HDL-C: high-density lipoprotein cholesterol, LDL-C: low-density lipoprotein cholesterol, IL-6: interleukin-6. retinopathy,


Background
Diabetic retinopathy (DR), the most common microvascular complication of type 2 diabetes, is a leading cause of vision impairment and blindness in working-age population globally [1,2]. Among diabetic populations, the estimated prevalence of DR is 34.6%, which brings a substantial worldwide public health burden [1].In the last decade, the clinical management of DR have been dramatically changed by the using of anti-vascular endothelial growth factor (VEGF) drugs and corticosteroids [3,4].
However, because of the limited efficacy and potential adverse effects, a comprehensive understanding of the pathophysiology of DR is urgently needed for the development of therapy.
Recently, increasing evidence points to chronic low-grade inflammation as a crucial contributor to the development and progression of DR [5]. Various pro-inflammatory cytokines and chemokines, such as interleukin-6 (IL-6), tumor necrosis factor-α (TNF-α), and monocyte chemoattractant protein-1 (MCP-1) , are elevated in serum and ocular samples (vitreous and aqueous humor) from patients with DR [6,7,8]. Many other features of inflammation including leukostasis, neutrophil and macrophage infiltration, vascular permeability and tissue edema also have been described in animal models and human patients with DR [5,9]. In addition, interlinked molecular pathways, namely increased formation of advanced glycation end-products, oxidative stress, increased expression of VEGF, neurodegeneration have been established the contribution to the inflammatory response [5,9,10].
Moreover, the aforementioned improvement of DR by corticosteroids also supports that inflammation participates the pathophysiology of DR. However, the exact molecular mechanism is not yet fully understood.
It has been shown that PGRN is an adipokine related to obesity-induced insulin resistance(IR), revealing its metabolic function and pro-inflammatory properties [11,16]. Youn et al. [18] revealed that serum PGRN levels were elevated in patients with type 2 diabetes, and correlated with visceral fat accumulation and macrophage infiltration into adipose tissue. Matsubara et al. [16] identified PGRN as a novel pro-inflammatory adipokine by differential proteome analysis of cellular models of IR. They found that PGRN levels were increased in IR state both in vivo and in vitro, and PGRN induced the expression of IL-6, a pro-inflammatory adipokine.
Further study has shown that the role of PGRN in chronic subclinical inflammation contributes to diabetic retinopathy. Xu et al. [19] showed that serum PGRN was significantly increased in the proliferative diabetic retinopathy (PDR) patients compared with the normal controls, simple DM patients, and background diabetic retinopathy (BDR) patients. But later, serum PGRN was found being affected by a decrease in estimated glomerular filtration rate (eGFR) of diabetic kidney disease (DKD) patients with eGFR < 60 ml/min/1.73m 2 , but not affected in DKD patients with eGFR ≥ 60 ml/min/1.73m 2 [20,21,22]. However, the PDR group in Xu et al.'s research presented reduced eGFR (< 60 ml/min/1.73m 2 ), which may affect their conclusions.
No data to date is available about the association between serum PGRN levels and DR excluding the effect of deteriorating renal function, and the role of PGRN in the pathophysiology of DR is still unknown. Therefore, the primary objective of this study was to investigate the association between serum PGRN levels and DR in patients with type 2 diabetes and eGFR ≥ 60ml/min/1.73m 2 . We also aimed to assess the factors correlated with serum PGRN levels.

Subjects
For this cross-sectional study, 262 patients(162 males/100 females) previously or newly diagnosed type 2 diabetes with eGFR ≥ 60 ml/min/1. The following exclusion criteria were applied: age < 18 years, eGFR < 60 ml/min/1.73m 2 , pregnancy, past history of malignancy, degeneration disease of the nervous system, acute or chronic inflammation, acute infectious disease, acute surgery or trauma, other endocrine diseases affecting glucose and lipid metabolism, taking thiazolidinedione drugs in 2 weeks, and history of drug abuse.
The study was approved by the local ethical committee. Signed informed consents were obtained from all participants in the study.

Methods
Standardized questionnaires were used to obtain age, sex, diabetes duration, hypertension duration, the history for the present and past illness, and medication.
Standardized protocols were used to measure height, body weight, and blood pressure (BP) . Body mass index(BMI) was calculated as weight/height 2 (kg/m 2 ).
Overnight fasting blood samples were collected for the determination of Early-morning urine samples were collected for calculating albuminuria/creatinine ratio (ACR).
Serums were separated from the blood samples and frozen at -80℃ prior to analysis of PGRN and IL-6. Concentrations of serum PGRN and IL-6 were measured by enzyme linked immunosorbent assay (ELISA) using the commercially available Human ELISA kits (mlbio, CHN) according to the manufacturers' instructions. All samples were diluted to 5 times during testing. The assay sensitivity was 0.08μg/L and assay range was 0.8-28μg/L for PGRN, while the assay sensitivity was 0.08 ng/L and assay range was 0.8-20ng/L for IL-6. The specificity of the immunosorbent assay for both PGRN and IL-6 was estimated to be 99.6%.
Fundal examinations were performed on all subjects by a professional ophthalmologist using direct ophthalmoscopy through dilated pupils. Fundus photography was performed on patients with type 2 diabetes, and some confounding cases were finally diagnosed by fundus fluorescein angiography.    Table 2) .

Multiple Stepwise Regression Analysis.
Multivariate stepwise regression analysis was performed to evaluate the independent factors of serum PGRN with the factors identified in the above univariate analysis including eGFR, TG, and LDL-C as independent variables. The analysis demonstrated that only LDL-C (β= 1.030, P<0.05)was independently associated with serum PGRN (Table 3). Similar results were obtained when TyG index instead of TG was included in the model (data not shown).

Discussion
In the present study, we could better investigate the association between serum Even as various research reported previously [20,21], we also observed a negative association between serum PGRN and eGFR. Although, the association was lost in multivariate stepwise regression analysis, which was in accordance with Nicoletto et al.'s research [20]. It seemed that the serum PGRN might be affected by moderate to severe renal insufficiency. However, the mechanism underlying the correlation between PGRN and deteriorating renal function was still unclear. The hypothesis was proposed that renal clearance was an important route of PGRN elimination [20,21,22].
Despite both Richter et al. [21] and Nicoletto et al. [20] did not find a significant correlation between serum and urinary PGRN, it was supposed that PGRN was at least partly degraded during the elimination [21].
In previous research, the association of serum PGRN and DR was regardless of the deteriorating renal function. The development and progression of DR was explained by the increase of serum PGRN, elevated IL-6 and then promoted chronic low-grade inflammation [16]. As we did not observe the significant difference of serum PGRN between NC, SDM, NPDR, and PDR groups, we did not find the remarkable difference of serum IL-6 either, and there was no relationship between serum PGRN and IL-6. So the present study did not allow us to identify serum PGRN as a marker for the occurrence and severity of DR. DR is identified not only as a retinal vasculopathy involving the increase of VEGF [24] but also as a retinal neuron injury [25]. Interestingly, the retina is composed of abundant epithelial cells and nerve cells which may secret PGRN [9]. Furthermore, PGRN has been established as a upstream stimulating factor of VEGF concerned in angiogenesis [14], and a neurotrophic and neuroprotective factor involved in central nervous system [15].
However, no data is currently available about the PGRN in ocular samples, i.e., retinal tissue, vitreous and aqueous humor, of DR in humans or animal models.
Previous studies showed that circulating PGRN was increased in patients with type 2 diabetes when compared to subjects without diabetes and there was a positive correlation between circulating PGRN and FPG, HbA1c, as well as HOMA-IR [18,26].
However, these studies did not take eGFR into account. Notably, after controlling for eGFR, the associations between circulating PGRN and FPG, HOMA-IR, as well as other components of the metabolic syndrome including DBP, waist to hip ratio (WHR), waist to height ratio (WHtR), TC, TG, HDL-C, and LDL-C were all lost in the research of Richter et al [21]. Furthermore, both Xu et al. [19] and Nicoletto et al. [20] did not find a significant difference of serum PGRN between non-diabetic subjects and normoalbuminuric patients without deteriorating eGFR, and there was no relationship between circulating PGRN and FPG, HBA1C, FINS or HOMA-IR. In agreement with these studies, no remarkable correlationship was found between serum PGRN and FPG, HbA1c, FINS, HOMA-IR, TyG index in our study.
Other metabolic disorders have also been reported to associate with PGRN [11].
A positive correlation between circulating PGRN and TC, TG indicated PGRN was involved in dyslipidemia [18,17,27]. In this study, we found that LDL-C was positively and independently associated with serum PGRN. LDL-C was acknowledged as a determining risk factor of carotid intima media thickness (CIMT), a useful surrogate marker for atherosclerosis [28]. Yoo et. al. [27] found that serum PGRN was an independent predictor for increased CIMT in subjects without metabolic syndrome after adjusting for other cardiovascular risk factors. Although the exact explanation for the association between PGRN, LDL-C, and CIMT was still unclear, it suggested that PGRN might have an influence on atherosclerosis. Further studies are needed to elucidate the connection.
There were some limitations in our study. Firstly, the sample sizes of NPDR and PDR groups were small. It is quite possible that nonsignificant associations in multivariate analyses would have become statistically remarkable if larger samples were studied. Secondly, our study was a retrospective study, and some data missing might affect the results. Thirdly, the study was a cross-sectional design, and therefore, causality could not be established.

Conclusion
In conclusion, we demonstrated that serum PGRN levels did not correlate with the occurrence and severity of DR in Chinese patients with type 2 diabetes and eGFR ≥ 60 ml/min/1.73m 2 . Furthermore, serum PGRN was positively and independently associated with LDL-C, which suggested that PGRN might be associated with diabetic macrovascular complications. Prospective studies are needed to confirm our observations and to validate the current findings.

Ethics approval and consent to participate
The study was approved by the Ethics Committee of the Affiliated Wuxi People's Hospital of Nanjing Medical University (number: RYSX0021) , and the acquisition of specimens from all cases was performed with the consent of the patient.

Consent to publish
Not applicable.

Availability of data and materials
The datasets used and analysed during the current study are available from the corresponding author on reasonable request.

Competing interests
The authors declare that they have no competing interests.

Funding
This work was supported by the National Natural Science Foundation of China (grant number 81500630), and the Wuxi science and technology development Foundation (grant number CSE31N1704).

Authors' Contributions
WJW was contributed to the study design. HYC, LX, RFB performed the screening of patients. CJ performed the experiments. HYC, and LJY managed the data.
HYC and LX were responsible for the statistical analyses, interpretation of data and drafting the manuscript. WJW reviewed and edited the manuscript. All authors approved the final manuscript.