The relationship between plasma serglycin levels and the diagnosis of diabetic retinopathy

Abstract Background Diabetic retinopathy (DR), a microvascular complication which is closely related to diabetes, remains the leading cause of vision loss around the world among older adults. Serglycin (SRGN) was known as a hematopoietic cell granule proteoglycan, exerting its function in the formation of mast cell secretory granules and mediates the storage of various compounds in secretory vesicles. The present study illustrates the potential clinical value and experimental mechanisms of SRGN in the DR. Methods Firstly, the mRNA expression and protein expression of SRGN in plasma samples from NPDR, PDR patients, type 2 diabetes mellitus (T2Dm) cases, and healthy controls were measured by qPCR and Western blotting assays, respectively. Then, the potentials of SRGN functioning as a diagnostic indicator in DR were verified by the receiver operating characteristic (ROC) analysis. We established in vitro DR model of human retinal endothelial cells through high‐glucose treatment. The expression of SRGN and its mechanisms of regulating cellular processes were illustrated subsequently. Results Our data revealed that SRGN was dramatically upregulated in NPDR and PDR cases compared with healthy controls and T2DM patients; meanwhile, the expression of SRGN was further increased in the PDR group with regard to the NPDR group. Furthermore, the ROC analysis demonstrated that SRGN could distinguish the DR cases from type 2 diabetes mellitus (T2DM) patients and healthy controls with the area under the curve (AUC) of 0.7680 (95% CI = 0.6780 ~ 0.8576, sensitivity = 47.27%, specificity = 100%, cutoff value = 1.4727) and 0.8753 (95% CI = 0.8261 ~ 0.9244, sensitivity = 69.23%, specificity = 100%, cutoff value = 1.6923), respectively. In vitro high‐glucose treatment showed that the SRGN expressions were dramatically increased. The loss of SRGN could partially counteract the inhibition of HREC proliferation caused by high‐glucose stimulation. Meanwhile, SRGN knockdown could reverse the promotion of HREC apoptosis induced by high glucose as well. Conclusions Consequently, our study implied that SRGN might serve as a promising biomarker with high specificity and sensitivity in the DR diagnosis and progression.


| INTRODUC TI ON
Diabetic retinopathy (DR), a microvascular complication caused by chronic diabetes mellitus (DM), is a significant threat to vision in adults, which may even lead to blindness and retinal detachment. 1 The clinical symptoms of DR include fibrosis, capillary obstruction, neovascularization, and increased vascular permeability. 2 According to the lesion's severity, DR could be classified as proliferative DR and non-proliferative DR. 3 Epidemiological studies showed that there were more than 285 million DM patients worldwide. 4 It was estimated that about 92.4 million adults in China have DM; 43% of them developed DR. 4 DR was broadly classified into two types: non-proliferative diabetic retinopathy (NPDR) and proliferative diabetic retinopathy (PDR). 5 The NPDR stage was mainly characterized by damage to the microvasculature such as massive apoptosis of retinal pericytes and endothelial cells, causing local ischemia, ischemia in the retina, increased vascular permeability, capillary dilatation, and microaneurysm formation. 6 When neovascularization broke through the inner boundary membrane of the retina to form pathological neovascularization, it was called the PDR stage, which was also the main period of blindness from DR. 7 Meanwhile, almost all type 1 diabetes cases and more than 60% of type 2 diabetes (T2DM) cases within two decades of onset had DR complications. 4,8 More importantly, only one-third of T2DM patients were diagnosed with DR symptoms at admission. 8 With the lapse of time, when fragile new blood vessels formed on the surface of the retina, the DR process had already progressed to an advanced stage, causing severe visual impairment and even blindness. 9 Therefore, given the constant increase in DR caused by T2DM, clinical strategies should be developed to identify and treat this complication as early as possible.
To date, the examination of glycosylated hemoglobin (HbA1c), fructosamine, and glycosylated albumin index has been widely used in the clinical diagnosis of the DR level. 10 However, the specificity, sensitivity, and accuracy of these indicators are limited, and the expenses related to diagnosis constitute a significant challenge. 11 Therefore, finding novel and accurate factors related to the DR pathogenesis and contributing to clinical diagnosis and treatment is a severe challenge.
Recently, several genes have been identified as having a potential diagnostic and prognostic value in DM and related complications such as diabetic nephropathy and diabetic cardiomyopathy. [12][13][14] In the realm of DR, various genes were also found to be potential candidates for DR diagnosis and progression. For instance, Li et al found that angiogenesis growth factors VEGFC, ANGPT1, ANGPT2, and EFNB2 were upregulated in DR patients, presenting their potentials as biomarkers for DR diagnosis and treatment. 15 Another report demonstrated that netrin-1 and netrin-4 were overexpressed in the DR group, functioning as promising therapeutic agents for DR. 16 Serglycin (SRGN) was a hematopoietic granulosa protein-polysaccharide composed of a relatively small (*17 kDa) core protein. 17 Studies showed that SRGN was mainly expressed in normal hematopoietic cells, endothelial cells, uterine decidua, and embryonic stem cells. [18][19][20][21][22] Meanwhile, SRGN was also involved in forming mast cell secretory granules and the localization of granular proteins and mediating the storage and secretion of various proteases, chemokines, and cytokines. 23,24 Furthermore, novel evidence suggested that SRGN may address an essential role in cancer development. For instance, elevated SRGN expression levels were deemed an unfavorable indicator in primary nasopharyngeal carcinoma. 25 Furthermore, abnormal expression of SRGN was associated with tumor progression and metastasis in breast cancer and non-small cell lung cancer. 26,27 According to a previous study, SRGN was abundantly represented in DN patients' fibrovascular membranes, implying its potential value as a parameter for predicting DR occurrence and development. 28 However, the clinical and cellular mechanisms of SRGN remain not fully understood. Consequently, the present study aimed to explore whether SRGN might be a candidate for DR's diagnosis and treatment.

| qPCR analysis
The expressions of SRGN in plasma samples and cells were measured by qPCR analysis. In brief, total RNA was extracted

| Western blot assay
Total proteins were isolated from plasma samples and cells using

| Flow cytometric and cell apoptosis assay
After transfection, cells were seeded at a 6-well plate with a  SPSS 21.0 (SPSS Inc, Shanghai, China) and GraphPad Prism 6.0 software were used to analyze the data. All the data obtained in the present study were conducted at least in triplicate. Differences among three groups or more groups were compared and analyzed using one-way ANOVA followed by Tukey's test. ROC curve analysis was conducted to verify the diagnostic value of SRGN for discriminating DR cases from healthy controls. Differences were considered statistically significant if the p-value was <0.05.

| Clinical features
The clinical parameters of the DR patients, T2DM patients, and healthy controls are presented in Table 1. As shown in Table 1, there were significant differences in fasting blood glucose (FBG), HbA1c, and SRGN levels among the three groups (p < 0.001).
However, there were no differences in age, gender, body mass index (BMI), diastolic blood pressure (DBP), systolic blood pres-

| Comparison of SRGN levels in plasma samples of NPDR cases, PDR cases, T2DM cases, and healthy controls
The mRNA and protein expressions of SRGN in plasma samples were measured by qPCR and Western blot assays. As shown in Figure 1,

| The diagnostic potential of SRGN for DR
According to ROC curve analyses, the diagnostic sensitivity and specificity of SRGN for DR were measured. As shown in Figure 2

| Comparison of SRGN levels in high-glucosetreated cells
To identify the mechanisms of SRGN at the cellular level, we further constructed a high-glucose-treated cell model to address further experiments. After cell culture, we found a significant increase in SRGN expression in high-glucose-treated HRECs compared with the control group, which contains average glucose concentration ( Figure 3).

| Knockdown of SRGN significantly increased proliferation in high-glucose-treated HRECs
After transfection and treated with high glucose, cell proliferation was measured by CCK-8 assay. As shown in Figure 5, after highglucose treatment, HREC proliferation was significantly inhibited; however, the inhibition could be partially reversed by transfection with siRNA1-SRGN or siRNA2-SRGN.

| Knockdown of SRGN significantly decreased apoptosis in high-glucose-treated HRECs
After transfection and treated with high glucose, cell apoptosis was measured by flow cytometry analysis. As shown in Figure 6, after high-glucose treatment, HREC apoptosis was significantly facilitated; however, the promotion could be partially counteracted by transfection with siRNA1-SRGN or siRNA2-SRGN.

| D ISCUSS I ON
DR is a multifactorial chronic diabetic eye complication that can eventually lead to microvascular disease, retinal dysfunction, and  32 Xu et al demonstrated that the overexpression of SRGN increased colorectal cancer cell F I G U R E 5 The knockdown of SRGN significantly promoted HREC proliferation after high-glucose treatment. A, Cell proliferation after treated with high glucose and transfection with siRNA1-SRGN. B, Cell proliferation after treated with high glucose and transfection with siRNA2-SRGN. Control, untreated HRECs with 5.5 mmol/L glucose F I G U R E 6 The knockdown of SRGN significantly inhibited HREC apoptosis after high-glucose treatment. A, The cell apoptosis after treated with high glucose and transfection with siRNA1-SRGN. B, The cell apoptosis after treated with high glucose and transfection with siRNA2-SRGN. * p < 0.05, HG + siRNA1(2)-SRGN vs HG + siRNA-NC group; ** p < 0.01, HG vs control group; control, untreated HRECs with 5.5 mmol/L glucose migration and invasion and was associated with poor prognosis. 33 Another study reported that SRGN might be implicated in diabetic tubulointerstitial injury confirmed by multiple-microarray analysis, which may provide new insights for DN's diagnosis and therapeutics 34 . However, whether SRGN could function as a therapeutic and diagnostic biomarker in DR is not fully illustrated.
This study found that SRGN expression was dramatically increased in plasma samples from DR cases compared with T2DM or healthy participants. Meanwhile, according to the international criteria for DR, SRGN was further elevated in the PDR group compared with the NPDR group. In addition to the SRGN level, biomarkers such as FBG level and HbA1c level showed statistically significant differences among the three groups. This finding was consistent with the previously reported HbA1c level associated with the severity of DR, and the FBG level was also associated with DR's occurrence. 35 We further used ROC analysis to analyze the diagnostic value of SRGN for DR. As expected, SRGN can differentiate between DR and T2DM patients, and healthy populations with the AUC results from all are higher than 0.8. Therefore, the use of SRGN as a biomarker for DR diagnosis had important clinical significance.
Concerning the in vitro experiments, we established a high-glucose-treated HREC model to mimic the intracellular environment of DR. From the results, we found upregulation of SRGN in highglucose-treated HRECs. Then, the knockdown of SRGN could reverse the inhibition of HRECs proliferation caused by high glucose.
Meanwhile, blocking of SRGN could also decrease the apoptosis induced by high-glucose treatment as well.
It is worth noting that this study still has the following limitations. First, we recruited 130 DR patients, 55 T2DM patients, and 46 healthy controls, with relatively small sample size. Simultaneously, participants were all from Ningbo Eye Hospital (Ningbo, China), and the source of samples may be biased. So, in the future, we need to test this in other populations. We completely excluded the effects of unknown confounders, such as clinical testing methods, disease stage, and drug use. Finally, the mechanism by which SRGN affects DR progression at the molecular level needs to be further investigated.
To sum up, in this study, we found that SRGN was significantly upregulated in the serum of DR patients, and after high-glucose treatment, SRGN promoted the proliferation of human retinal pigment epithelial cells and inhibited the modulation of SRGN. At the same time, SRGN levels were able to screen DR patients from T2DM and healthy populations. Therefore, SRGN may be a potential diagnostic and therapeutic target for DR.

ACK N OWLED G M ENTS
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.