Glucose-regulated protein 78 in the aqueous humor in diabetic macular edema patients.

In this study, we explored the presence and elevation of glucose-regulated protein 78 (GRP78) in aqueous humor of patients with diabetic macular edema (DME).After comparing DME patients with the controls, we analyzed GRP78 and vascular endothelial growth factor (VEGF) levels in DME patients. We examined factors associated with GRP78 levels in DME patients.GRP78 was detected in aqueous humor with elevated levels in DME patients. Stepwise backward regression analysis showed that GRP78 levels were associated with the VEGF levels and the duration of diabetes (P < .001 and P = .002, respectively). However, no statistical significance was observed between GRP78 levels and the decrease in CST following 3 monthly anti-VEGF treatments in univariate regression analysis (P = .695).We showed that GRP78 is elevated in DME patients. In addition, there is a correlation between GRP78 and VEGF levels in aqueous humor. However, GRP78 levels were not associated with the responsiveness of anti-VEGF in DME patients.

Introduction

The endoplasmic reticulum (ER) is a eukaryotic cell organelle, which is responsible for the metabolism and synthesis of substances. The rough ER, through binding with ribosomes, synthesizes, and processes proteins. In various physiological and pathological conditions, an accumulation of unfolded or misfolded proteins in the ER is associated with cell survival.[ An abnormal accumulation of these proteins, known as ER stress, can lead to apoptosis or autophagy.[ Under ER stress, ER chaperone proteins are activated to facilitate folding and assembly of proteins, or to export abnormal proteins for degradation.[

Among the ER chaperone proteins, glucose-regulated protein 78 (GRP78), a member of the heat-shock protein 70 family, is a widely used ER stress marker protein.[ GRP78 has multiple functions, being involved in proper protein folding and assembly, misfolded protein degradation, calcium homeostasis, and control of the activation of transmembrane ER stress sensor proteins.[ Most studies with GRP78 have been conducted in poorly perfused solid tumors that induced ER stress because of altered glucose metabolism and microenvironmental factors, such as glucose deprivation, hypoxia, and acidosis.[ The induction of GRP78 by ER stress leads not only to an increase in GRP78 in the ER but also to relocalization from the ER to the cell surface, cytoplasm, mitochondria, and nucleus.[ Some studies have even reported that exocrine pancreatic cells and oviduct epithelial cells secrete GRP78.[ Although studies on glaucoma have reported elevated GRP78 in trabecular meshwork cells and tissues,[ there have been no reports about the secreted form of GRP78 in the eye. Based on studies showing that diabetic retinopathy also occurs under ER stress conditions,[ we designed this study to investigate the presence of GRP78 in the aqueous humor of patients with diabetic macular edema (DME).

Methods

Study subjects

This study characterized GRP78 levels in the aqueous humor, and analyzed the association between GRP78 and vascular endothelial growth factor (VEGF) levels in the aqueous humor, of naïve DME patients with type II diabetes mellitus (DM). The study adhered to the tenets of the Declaration of Helsinki, and the protocol was approved by the Institutional Review/Ethics Board of the Catholic University of Korea. All methods were performed in accordance with the relevant guidelines and regulations by the protocol. All participants gave written informed consent to the use of their clinical records.

We enrolled eyes with a central subfield thickness (CST) > 300 μm due to DME; eyes with cataract without DM were the controls. The aqueous humors of 103 naïve DME patients with type II DM and 14 cataract patients were analyzed. Exclusion criteria included eyes with glaucoma, retinal degeneration, and macular edema due to other causes, such as epiretinal membrane, medication, or vitreomacular traction. Eyes with concurrent diseases, such as retinal vascular occlusion, as well as eyes with a history of ocular trauma, uveitis, or prior intraocular surgery that could influence the enzyme levels of the aqueous humor, were also excluded.

We measured glycated hemoglobin levels, and all patients underwent a full ophthalmic examination that included measurement of the best-corrected visual acuity (BCVA), a dilated fundus examination, and classification of the eye according to the Early Treatment of Diabetic Retinopathy Study as mild nonproliferative diabetic retinopathy (NPDR), moderate NPDR, severe NPDR, or proliferative diabetic retinopathy. Macular thickness was measured using optical coherence tomography (OCT; Cirrus High-Definition OCT; Carl Zeiss Meditec, Dublin, CA). Axial length was measured using an IOL Master instrument (Carl Zeiss Meditec). We evaluated changes in the CST and BCVA at the 1 month visit after 3 consecutive monthly intravitreal bevacizumab (IVB) injections.[

Quantitation of GRP78 in the aqueous humor

The volume of the collected aqueous humor was at least 85 μL. Levels of GRP78 protein were determined using 1:1 diluted aqueous humor samples and a human GRP78 enzyme-linked immunosorbent assay (ELISA) kit (Enzo Life Sciences, Lausen, Switzerland) with a detection range of 1.4 to 4500 ng/mL. The VEGF levels were determined using 1:2 diluted samples and a human VEGF ELISA kit (DuoSet ELISA; R&D Systems, Minneapolis, MN) with a detection range of 31.20 to 2000 pg/mL. The assays were performed according to the manufacturers’ instructions.

Statistical analyses

The study group patients were compared with the control group using the nonparametric Mann–Whitney U test and the chi-squared test. Univariate and stepwise backward regression analyses were used to identify associations of GRP78 levels and VEGF levels with age, axial length, BCVA (logMAR), and CST at the initial visit, and changes in BCVA and CST after 3 consecutive IVBs. Statistical analyses were performed using SPSS for Windows statistical software (ver. 20.0; SPSS Inc., Chicago, IL) and R software (ver. 3.2.3 [2015-2-0, Platform: x86_64-redhat-linux-gnu, R Core Team (2015)]. R is a language and environment for statistical computing (R Foundation for Statistical Computing, Vienna, Austria; https://www.R-project.org/).

Results

The average age of the cataract patients was 66.29 ± 11.98 years, and that of the DME patients was 55.41 ± 8.54 years (P < .001). There were 51 males and 52 females in the study group and 7 males and 7 females in the control group. There was no significant difference in axial length between the study and control groups. GRP78 and VEGF levels were significantly elevated in the DME group (P < .001 and P = .001, respectively; Table 1).

Table 1

Demographics and baseline clinical characteristics of the study participants.

In univariate analyses, GRP78 levels correlated with VEGF levels and the duration of DM (P < .001 and P = .005, respectively; Table 2; Fig. 1A and B) Stepwise backward regression analysis also showed a statistically significant relationship between GRP78 levels and VEGF levels and duration of DM (P < .001and P = .002, respectively; Table 2). VEGF levels correlated with GRP78 levels in stepwise backward regression analysis (P < .001; Table 3). However, in a further regression analysis with the same parameters, initial CST was not associated with GRP78 or VEGF levels (P = .956 and P = .984, respectively); the initial BCVA was also not associated with the level of either of these proteins (P = .597 and P = .883, respectively).

Table 2

Variables associated with level of GRP78 in aqueous humor of DME patients in univariate and stepwise backward regression analyses.

Figure 1

(A) Scatter plot showing the correlation between GRP78 and vascular endothelial growth factor (VEGF) levels in the aqueous humor. (B) Scatter plot showing the correlation between GRP78 levels and duration of diabetes mellitus in univariate regression analyses. Both correlations were statistically significant (P < .001 and P = .005, respectively). GRP78 = glucose-regulated protein of 78 kDa, VEGF = vascular endothelial growth factor

Table 3

Variables associated with level of VEGF in aqueous humor of DME patients in univariate and stepwise backward regression analyses.

Discussion

ER stress occurs during inflammation, ischemia, and tumor conditions, and is associated with cell survival.[ The early pathogenesis of DME is usually characterized by damage to the inner blood-retinal barrier due to hypoxia and inflammation.[ We therefore designed this study to explore the possible correlation between ER stress and DME, by determining the level of GRP78 in the aqueous humor of DME patients.

GRP78 has several functions in the ER, including acting as a chaperone under non-stress conditions.[ Under excessive ER stress or pathological conditions, GRP78 can translocate to the cell surface or be secreted.[ GRP78 at the cell surface controls cellular proliferation, apoptosis, and/or immune functions.[ The secreted form of GRP78 is not as well known as other forms. All chaperone proteins have a lysine-aspartate-glutamate-leucine (KDEL) motif at the C terminus, functioning as an ER retention signal. When KDEL receptors are saturated or defective, GRP78 is thought to be secreted.[ Although the secreted form of GRP78 has been mainly studied in cancer research, some studies have shown its presence in the oviduct or pancreatic exocrine samples.[ Recent studies have even reported that blood GRP78 levels are associated with metabolism in the polycystic ovary, obesity and physical activity.[ Secreted GRP78 may therefore be involved in biological processes in both pathological and physiological conditions.[ We showed that GRP78 was also present in the aqueous humor, and that it was elevated in DME patients.

An association between ocular disease and ER stress has been reported in several studies. Studies of glaucoma patients reported that GRP78 levels were elevated in trabecular meshwork cells and tissues.[ Other studies reported that the unfolded protein response occurred in retinal vascular diseases, and that VEGF levels correlated with ER stress.[ However, because investigation with human retinal tissue is very difficult, most studies involved in vitro samples or animal models. We therefore characterized ER stress in the aqueous humor, which reflects the retinal status and is easily obtained during ocular surgery.[ Our multivariate regression results showed that GRP78 levels in the aqueous humor are positively correlated with VEGF levels and the duration of diabetes significantly (Table 2). Similarly, aqueous VEGF levels were associated with GRP 78 levels (Table 3). Some other studies have reported that GRP78 levels were correlated with the degree of neovascularization, as well as with elevated VEGF levels.[ These results all together suggest that ER stress and angiogenesis be associated.

There were some limitations to this study. First, although the mean GRP78 level in the aqueous humor was significantly higher in the study group, the sample size was relatively small to investigate the functional and anatomical relationships with GRP78 levels. Second, there was insufficient sample volume to determine the possible presence of inflammatory cytokines, such as interleukins and adhesion molecules, or of other growth factors that act as biomarkers for DME.[ In future studies, we plan to characterize vitreous samples using other techniques, such as the Luminex x MAP suspension array technology (Luminex Corp., Austin, TX). Third, we measured only GRP78 levels in the aqueous humor, and did not determine which cells or tissues are involved in GRP78 secretion. As mentioned previously, human ocular tissue is difficult to obtain, so our future studies will involve animal models or in vitro experiments. Fourth, breakdown of blood retinal or aqueous barrier could lead to elevation of overall protein concentration in aqueous humor. Therefore, normalization of GRP78 concentration to total protein concentration in aqueous humor needs, and this normalized data should be compared with GRP78 concentration from blood samples.

In conclusion, this study is the first to reveal increased GRP78 levels in DME patients. And the GRP 78 levels in aqueous had positive correlations with the aqueous VEGF levels and duration of diabetes. However, this protein levels showed no association with the degree of responsiveness to anti-VEGF therapy. Future studies should aim to explore the role and origin of this ER chaperone in the aqueous humor.

Acknowledgment

The authors wish to acknowledge the support of Center for Integrative Rheumatoid Transcriptomics and Dynamics in College of Medicine, Catholic University of Korea and financial support of Aju Pharm from Catholic Medical Center Research Foundation in the program year of 2016.

Author contributions

Conceptualization: Jin-woo Kwon.

Data curation: Jin-woo Kwon, Ilyoung Jung.

Formal analysis: Jin-woo Kwon.

Funding acquisition: Donghyun Jee.

Investigation: Jin-woo Kwon.

Methodology: Jin-woo Kwon.

Resources: Jin-woo Kwon.

Software: Jin-woo Kwon.

Supervision: Donghyun Jee.

Writing – original draft: Jin-woo Kwon.

Jin-woo Kwon orcid: 0000-0003-2093-4284.