Upregulation of Retinal Neuronal MCP-1 in the Rodent Model of Diabetic Retinopathy and Its Function In Vitro

Abstract

Purpose.: Toevaluate the expression of monocyte chemoattractant protein-1 (MCP-1) in the rodent model of diabetic retinopathy (DR) and to study the stimulation of microglial activation by retinal neuronal MCP-1 in vitro.

Methods.: Diabetes mellitus was induced by streptozotocin (STZ) injection. The expression of MCP-1 was determined using immunohistochemical methods, Western blotting and RT-PCR analyses. Retinal neurons and microglia were separated and co-cultured in a Transwell apparatus. The levels of soluble MCP-1 that were produced after stimulation of retinal neurons by adding advanced glycation end products (AGEs) to the medium were measured by ELISA. The degree of microglial activation was measured by testing microglial migration and the level of soluble TNF-α in the medium by ELISA. The ability of neuronal MCP-1 to stimulate microglia activation was examined by pre-exposing the retinal neurons to AGEs and an MCP-1 antibody or to AGEs and SiRNA specific to MCP-1.

Results.: A marked increase in the expression of MCP-1 was detected 4 weeks after STZ injection, and the expression was consistently upregulated at 3 and 5 months in the rodent DR model. Stimulation with AGEs significantly increased the expression of MCP-1 in retinal neurons, which activated microglial cells, including increased microglial migration and upregulated secretion of TNF-α. Retinal neurons that were pre-exposed to AGEs and an MCP-1 antibody or MCP-1 knockdown displayed greatly reduced microglial migration and TNF-α secretion.

Conclusions.: Upregulation of MCP-1 began during the early stage of DR and increased with the development of the disease. Retinal neurons are the main source of MCP-1, and they play an important role in retinal microglial activation, which may be an important link in the pathogenesis of DM.

Introduction

Diabetic retinopathy (DR) is a chronic complication of diabetes that is characterized by degeneration of neurons and glial activation accompanied by diffuse vascular abnormalities. ¹ Whereas the mechanisms of vascular and neuronal pathology in DR are not yet fully understood, increasing evidence indicates that inflammation plays a pivotal role in the pathogenesis of DR. ^2–4 The increased expression of adhesion molecules and proinflammatory cytokines, as well as microglial activation in the retina or vitreous, all demonstrate that local inflammation may represent the central pathway leading to DR. 

Microglia are resident immunocompetent and phagocytic cells in the central nervous system (CNS). Activated microglia not only act as scavengers but also serve as rapid sensors of neuronal damage that are responsible for tissue repair and neural regeneration. ^5,6 Microglial activation has been shown to be a major histopathological change in DR. ³ Cytokines released by activated microglia regulate the influx of inflammatory cells to the damaged area, causing vascular breakdown and releasing cytotoxins that kill retinal neurons. Activated microglia release TNF-α, which has the potential to induce apoptosis, fibroblast proliferation, nuclear factor-kappaB activation, and cell adhesion molecule activation. ^7–9  

In recent years, accumulating in vivo and in vitro evidence has shown that microglial activity is prominent after neuronal damage, suggesting that neurons play an important role in activating microglia. ^10–13 In response to injury, neurons release chemokines that act on microglial cell receptors to induce migration and activation. ^14,15 The retina is a highly organized structure in the CNS, and many studies show that activated microglia are involved in a variety of retinal neurodegenerative and inflammatory diseases, such as DR, proliferative eye diseases, light damage, glaucoma, damage from laser photocoagulation, and age-related macular degeneration. ^16–21 Nevertheless, what mediates microglial activation and the origin of the chemostimulants in the retina remain unknown. 

Studies have shown that the chemokine monocyte chemoattractant protein-1 (MCP-1) is upregulated in models of endotoxin-induced uveitis, retinal neovascularization, and in retinal degeneration in mice. ^22–26 In these retinopathies, MCP-1 has been shown to activate or attract microglia; however, no research concerning the expression or the source of MCP-1 in DR or its function in retinal microglial interaction has been reported. 

For this reason, in this study, a rodent model of DR was used to test the expression of MCP-1 in the retina. Retinal neurons and microglia were separated and cultured in a Transwell apparatus in which the retinal neurons and microglial cells shared the same medium but could not form direct cell-cell interactions. Retinal neurons were stimulated with advanced glycation end products (AGEs) to mimic the diabetic microenvironment. We subsequently investigated the expression of MCP-1 in the retinal neurons. The function of neuronal MCP-1 in the neuron–microglia interaction was studied using an MCP-1 antibody or small interfering RNA (siRNA) specific to MCP-1. 

Materials and Methods

Animals

Diabetes mellitus (DM) was induced in 48 Sprague-Dawley rats weighing approximately 160 to 200 g by intraperitoneal injection of streptozotocin (STZ, 60 mg/kg; Sigma, St. Louis, MO) in 10 mmol/L sodium citrate buffer (pH 4.5). Control animals received the buffer alone. Animals with blood glucose levels higher than 16.7 mmol/L 3 days after receiving STZ were considered to be diabetic. Blood glucose levels were measured weekly. All experiments were conducted according to the statement from the Association for Research in Vision and Ophthalmology (ARVO) on the Use of Animals in Ophthalmology and Vision Research. At different time points (i.e., 0, 1, 3, and 5 months after STZ injection), control and model rats were killed, and their eyes were collected, under anesthesia. For the immunohistochemistry assays, the eyes were immediately enucleated and embedded in optimum cutting temperature (OCT) compound. Cryostat sections (6-μm thick) were obtained using a freezing microtome. The retinas were removed on ice, frozen in liquid nitrogen, and kept at −80°C until use in the Western blot and RT-PCR assays. 

Immunohistochemistry Assay

The methods of the immunohistochemistry assay have been described in detail previously. ²⁷ Sections were incubated with a primary MCP-1 antibody (1:100; eBioscience, San Diego, CA) at 4°C overnight and were further incubated with a biotin–avidin complex containing peroxidase (Vector Laboratories, Burlingame, CA). 

RT-PCR

The method of RT-PCR has been described in detail previously. ²⁷ Total mRNA was extracted from the retinas using the TRIzol (Invitrogen, Carlsbad, CA) method according to the manufacturer's instructions. The sequence of the primer used for rat MCP-1 was CCA GAA ACC AGC CAA CTC TC; TTC CTT ATT GGG GTC AGC AC. The primer used for rat glyceraldehyde-3-phosphate-dehydrogenase (GAPDH) was TGC CAC TCA GAA GAC TGT GG; TTC AGC TCT GGG ATG ACC TT. A semiquantitative analysis was performed by contrasting the band densities that were measured using Quantity One 1-D Analysis software (Bio-Rad, Hercules, CA) and normalized to each internal control GAPDH. 

Western Blot Analysis

The method used for Western blot analysis has been described in detail previously. ²⁷ Total proteins were extracted from the retinas with Protein Extraction Reagent (PIERCE, Rockford, IL) according to the manufacturer's instructions. Equal amounts of each protein extract (50 μg) were incubated with the primary MCP-1 antibody (1:200; eBioscience) at 4°C overnight. After washing, the membranes were incubated with a peroxidase-conjugated secondary antibody and visualized with a chemiluminescence detection system (Immobilon P; Millipore, Billerica, MA). Images were obtained using a densitometer (Bio-Rad), and analyzed quantitatively with Multi-Analyst Macintosh software (Bio-Rad). The band densities were normalized relative to the level of β-actin in each sample, which was detected as an internal control. 

Primary Retinal Microglia and Neuronal Culture

The primary microglial culture was performed according to the protocol of Roque, and the primary retina neural cell culture was performed based on Zhou's published protocol with minor modifications. ^28,29 In brief, retinas were collected and digested with 0.125% trypsin for 20 minutes at 37°C. The trypsin was subsequently inactivated with Dulbecco's modified Eagle's medium (DMEM)/F-12 (Invitrogen) containing 10% fetal bovine serum (FBS) (Invitrogen). Subsequently, the tissue was passed through 200-μm filters. Then, the filtered cells were resuspended in DMEM/F-12 culture medium containing 10% FBS, 100 U/mL penicillin, and 100 mg/mL streptomycin. Next, the cells were seeded into 75-cm² tissue culture flasks (Corning, Oneonta, NY) at a density of 1 × 10⁶ cells/cm². The cells were kept in a humidified atmosphere of 5% CO₂ and 95% air. The culture medium was changed at 24 hours and twice weekly thereafter. After 2 weeks, the microglia were harvested by shaking the flasks at 200 rpm for 1 hour, then used for Transwell culture and the following experiments. For the primary retina neuronal culture, the process of cell digestion, filtration, and centrifugation was the same as described above, the resuspended cells were seeded in 6-well or 24-well Boyden chambers (Corning) that were precoated with poly-D-lysine for the following Transwell culture and experiments. On the second day after seeding, Cytosine β-D-arabinofuranoside (10 μM; Sigma) was added to the cultures to suppress the proliferation of glial cells. The culture medium was changed at 24 hours and twice weekly thereafter. 

MTT Cell Viability Assay

Seven-day-old primary cultured retina neurons (1 × 10⁵ cells/well) and freshly isolated retina microglial cells (4 × 10⁴ cells/well) plated in 96-well plates (Corning) were used for the 3-(4,5-Dimethyl-2-thiazolyl)-2,5-diph-enyl-2H tetrazolium bromide (MTT) test. After 4 hours, AGE (#2221-10; BioVision, Milpitas, CA) was added at different concentrations (0, 100, 250, 500, 750, 1000, or 1500 μg/mL), and the cells were cultured for 24 hours. Next, 15 μL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT) solution (5 mg/mL in PBS) was added to each well, and the plates were incubated for 4 hours at 37°C. The reaction was terminated by adding an extraction solution (100 μL/well) that consisted of 20% (wt/vol) sodium dodecyl sulfate N,N-dimethylformamide to lyse the cells and dissolve the crystals. The plates were incubated overnight at 37°C. The optical density was measured at 570 nm in a dual-beam microtiter plate reader (Molecular Devices, Sunnyvale, CA) with 630 nm as the reference.This procedure provided dose–response results, which determined that the optimal concentration for the next experiment was 750 μg/mL. 

MCP-1 siRNA Preparation and Transfections

MCP-1 siRNA was purchased from Ambion (Austin, TX) and had the following sequences: sense 5′-GGCUAAAAGUAGUGAAUGAtt-3′, antisense 5′-UCAUUCACUACUUUUAGCCtt-3′, and negative control siRNA (sense 5′-CCUACGCCACCAAUUUCGUtt-3′, antisense 5′-ACGAAAUUGGUGGCG UAGGtt-3′). MCP-1 siRNA was complexed with Lipofectamine 2000 (Invitrogen, Karlsruhe, Germany) in 6-well plates in which 7-day-old primary purified retinal neuronal cells were incubated according to the manufacturer's instructions. Then, 2 μL Lipofectamine 2000 was diluted in 50 μL neurobasal medium without supplements and combined with 0.01 to 0.20 μg siRNA after 5-minute incubation at room temperature. The transfection was continued for 24 hours at room temperature. The knockdown of MCP-1 in neuronal cells was determined by Western blot analysis. The process of Western bolt was the same as described above. 

Transwell Culture and Treatment

Freshly isolated microglia were resuspended and plated at 2.5 × 10⁵ cells per well on polycarbonate Transwell inserts (4.67 cm²) with a 0.4-μm pore size (Corning). After 4 hours, these inserts were placed on top of 6-well plates containing the 7-day-old primary cultures of retinal neurons described above (nearly 5 × 10⁶ cells per well) to set up a retinal neuron-microglia Transwell system. Next, the experiments were processed in 4 groups with different treatments. In Group 1, the cells were cultured with culture medium alone for 24 hours and used as a control. In Group 2, the cells were treated with AGE (750 μg/mL) in the culture medium for 24 hours. In Group 3, to determine the role of MCP-1 produced by the retinal neuron, primary cultured retinal neurons were pre-incubated with AGE (750 μg/mL) and MCP-1 blocking peptide (15 μg/mL) for 8 hours, then co-cultured with the isolated microglia in the Transwell apparatus for another 24 hours. In Group 4, retinal neurons were pre-exposed to AGEs (750 μg/mL) and MCP-1 siRNA for 24 hours, then co-cultured with the previously described isolated microglia in the Transwell apparatus for another 24 hours. 

Cell Immunofluorescence

For immunocytochemistry analysis, the 12-mm glass coverslips (BD Pharmingen, San Diego, CA) were put into both the upper and bottom chamber of the retinal neuron-microglia Transwell culture system before the cell seeding. After stimulation, the coverslips were taken out and the cells were rinsed twice with PBS and fixed for 15 minutes in 4% paraformaldehyde at room temperature. The retinal neurons were permeabilized with 0.3% Triton-X100 in PBS for 20 minutes at room temperature and were incubated overnight with specific primary antibodies against MCP-1 (1:50; eBioscience). Microglial cells were directly incubated with goat anti-CD11b (1:100; Serotec, Oxford, UK) and rabbit anti-TNF-α (1:50; eBioscience) antibodies overnight at 4°C without permeabilization. Subsequently, the cells were incubated with FITC- or PE-conjugated secondary antibodies (1:50; Chemicon, Temecula, CA) in PBS for 1 hour at room temperature. Finally, the cells were counterstained with 4′,6-diamidino-2-phenylindole, mounted in glycerol, and examined by confocal microscopy (Fluoview1000; Olympus, Tokyo, Japan). The specificity of the antibodies was confirmed by replacing each primary antibody with a nonspecific protein of the same isotype. Quantitative analysis of the number of positive microglial cells was carried out by counting eight microscopic fields in the control cultures and treated cultures. 

In Vitro Chemotaxis Assay

The in vitro migration of microglia was assessed in 24-well Boyden chambers (Corning) containing polycarbonate membranes (8 μm pore size) that were precoated with fibronectin (BD Bioscience, Bedford, MA). Seven-day-old purified retinal neurons (nearly 2.5 × 10⁵/well), as described above, were cultured in the bottom wells of the Boyden chambers and served as a chemoattractant. Samples from the primary cultures of microglial cells (nearly 5 × 10⁴ in 200 μL) were added to the upper chamber. For the following procedures, the cells were divided into four groups as described above. After incubation, the filter was removed from the apparatus, and the cells that did not migrate and stayed on the upper membrane surface were removed with a cotton swab. The migrated cells on the bottom surface of the membrane were fixed in 4% paraformaldehyde for 10 minutes and were stained using 0.5% crystal violet. The number of migrated cells was counted in six randomly chosen fields under ×200 magnification and averaged. 

ELISA

The concentrations of MCP-1 and TNF-α that were released in the Transwell culture were tested with ELISA kits (PIERCE). Briefly, samples from the four groups described above were incubated in 96-well plates coated with MCP-1 or TNF-α antiserum for 2 hours. The samples were treated with enzyme working reagent for 30 minutes and with TMB One-Step substrate reagent (DAKO, Carpinteria, CA) for 30 minutes in the dark, and the reaction plates were read within 30 minutes in an ELISA plate reader (Molecular Devices, Eugene, OR) at 450 nm, with 620 nm serving as the reference. The detection limit was 1 pg/mL. 

Statistical Analysis

All experiments were performed at least three times. Quantitative data are presented as the mean ± SE and were analyzed with one-way ANOVA or Student's t-test. A P value less than 0.05 was considered statistically significant. 

Results

Expression of MCP-1 in the Retinas of Rats with Diabetes Mellitus

In control rats, relatively few MCP-1–positive cells appeared in the granular cell layer (GCL) of the retina. However, markedly increased expression of MCP-1 was detected 4 weeks after STZ injection. With the development of diabetes, the expression of MCP-1 was increased at 3 and 5 months after injection (Fig. 1). 

Figure 1. 

 

In control animals, MCP-1 (A) was detectable in the GCL of the retina. Increased expression of MCP-1 (B) can be observed at 1 month postinjection. Consecutively increased expression of MCP-1 was evident at 3 and 5 months after injection (C, D). The positive staining of MCP-1 was distributed mainly around vessel walls, as well as in the cell bodies and synapses of the RGCs. (MCP-1 uses DAB staining [brown], other organizations use the hematoxylin counterstain [blue]; original magnification ×200). Arrows represent MCP-1 positive cells.

Corresponding to the immunohistochemical results, the basic protein production and mRNA level of MCP-1 were detectable in the retinas of control rats. STZ injection led to increased MCP-1 mRNA expression and protein production at 1 month postinjection. The increased expression of MCP-1 became aggravated with the development of diabetes at 3 and 5 months after injection (Fig. 2). 

Figure 2. 

 

STZ injection led to increased expression of MCP-1 mRNA (A) and protein (B). (*P < 0.05; **P < 0.01)

Retinal Cell Culture

After 7 days of culture, the purity of the retinal neurons was more than 90%, as assessed by flow cytometry with Alexa Fluor 488 rat anti-Thy-1.1 (CD90) antibody (#202506; BioLegend, San Diego, CA), which is a specific label for retinal ganglion cells (RGCs, the main neuron in retina neuronal culture) (Fig. 3A). The purity of freshly separated microglial cells was more than 98% (Fig. 3B). 

Figure 3. 

 

After 7 days of retinal neuron culture, neurons represented more than 90% of the cells present (A). The purity of retina microglia separated by shaking the mixed primary retina culture was more than 98% (B).

MCP-1 Knockdown in Primary Cultured Retinal Neurons

To determine the role of MCP-1 produced by the retinal neuron, we transfected primary cultured retinal neurons with siRNA specific to MCP-1. We achieved downregulation of MCP-1 as monitored by Western blot analysis, with a maximum knockdown of approximately 75% at 24 hours (Fig. 4). 

Figure 4. 

 

Primary cultured retinal neurons were treated with specific siRNA to knockdown the expression of MCP-1. Representative MCP-1 knockdown was identified with Western blot analysis with a maximum downregulation of approximately 75%. Results are statistically significant (*P < 0.05 Student's t-test).

AGEs Upregulate the Expression of MCP-1 in Retinal Neurons

A Transwell culture system was designed in which retinal neurons and microglial cells shared the same medium but had no direct cell-cell interaction. In this system, interactions between neurons and microglia were mediated only by their release of soluble factors. The effect of AGEs on the expression of MCP-1 in the RGCs was analyzed with a cell immunocytochemistry assay, and the concentration of soluble MCP-1 in the culture medium was measured with an ELISA kit (PIERCE). 

MCP-1–positive neurons were detectable in the control medium (1.524 ± 0.582 cells/mm²) (Fig. 5A), and the number of MCP-1–positive cells increased significantly following treatment with 750 μg/mL of AGE (8.524 ± 1.362 cells/mm²) (Fig. 5B), and the number of MCP-1–positive cells decreased significantly following the treatment with 750 μg/mL of AGE accompanied with siRNA specific to MCP-1 (4.643 ± 1.587 cells/mm²) (Fig. 5C). 

Figure 5. 

 

The basal level of expression of MCP-1 was observed in the control medium (A). The number of MCP-1–positive cells increased markedly after AGE treatment (B). The number of MCP-1–positive cells increased moderately after AGE treatment accompanied with siRNA specific to MCP-1 (C).

Soluble MCP-1 was detected in the unstimulated medium (12.36 ± 1.83 pg/mL); however, exposure to AGEs led to upregulation of MCP-1 release in the retinal neuron–microglia Transwell culture system (73.26 ± 9.26 pg/mL); exposure to AGEs accompanied with MCP-1 knockdown led to downregulation of MCP-1 release in the retinal neuron–microglia Transwell culture system (34.96 ± 4.35 pg/mL) (Fig. 6). 

Figure 6. 

 

The concentration of soluble MCP-1 in the culture medium was measured using an ELISA kit. Soluble MCP-1 was detected in the unstimulated medium (12.36 ± 1.83 pg/mL). AGE stimulation resulted in upregulation of MCP-1 release (**P < 0.05). Exposure to AGE accompanied with MCP-1 siRNA led to downregulation of MCP-1 release under AGE stimulation (*P < 0.05).

Retinal Neurons Activated Microglia following AGE Exposure through MCP-1 Expression

Our data show that when cultured with AGEs, retinal neurons had a marked effect on retinal microglial activation in the Transwell culture system. The medium from retinal neurons stimulated with AGE induced increased retinal microglial migration compared with the control medium (36.24 ± 4.26 vs. 4.28 ± 0.86) (Fig. 7) and upregulated the number of CD11b and TNF-α double-stained cells (10.32 ± 6.26 vs. 1.28 ± 0.48) (Fig. 8). TNF-α was detectable in the control medium (20.46 ± 3.69 pg/mL), but the level increased by a factor of 20 (420.21 ± 50.11 pg/mL) following stimulation with AGEs (Fig. 9). This result shows that after exposure to AGEs, retinal neurons activate microglia in the neuron–microglia Transwell culture system. 

Figure 7. 

 

AGE-induced increase in retinal microglial migration (B) compared with the control medium (A, **P < 0.01). Microglia migration was greatly inhibited when pre-incubated with AGE and an MCP-1 antibody (C, *P < 0.05). Microglia migration was greatly inhibited when pre-incubated with AGE and MCP-1 knockdown (D, *P < 0.05).

Figure 8. 

 

Purified microglia were stained with FITC-CD11b (A1, B1, green). When stimulated with AGE, the expression of TNF-α up-regulated labelled with PE (A2, B2, red). CD11b and TNF-α double-stained cells were detectable in the control medium (A3). The number of double-stained cells (activated microglia) increased markedly after AGE treatment (B3).

Figure 9. 

 

The level of TNF-α secretion is detectable in the culture medium, and the expression of TNF-α increased significantly with the stimulation of AGEs (**P < 0.01). When pre-incubated with AGE and an MCP-1 antibody, the level of TNF-α secretion by the microglia was greatly decreased (*P < 0.05). When pre-incubated with AGE and MCP-1 knockdown, the level of TNF-α secretion by the microglia was also greatly decreased (*P < 0.05).

To determine the role of MCP-1 released by retinal neurons in neuron-microglia interactions, retinal neurons were pre-incubated with AGE and an MCP-1 antibody to neutralize the endogenous MCP-1 secreted by the retinal neurons in response to AGE. At the same time, MCP-1 knockdown was designed to downregulate the neuronal MCP-1 secretion in response to AGE. Our data showed that not only neutralizing neuronal MCP-1 but also MCP-1 knockdown can greatly inhibit the effect of retinal neurons on microglial activation. Microglial migration was further inhibited (Fig. 7), and the level of TNF-α production was also greatly decreased (Fig. 9). 

Discussion

MCP-1(CCL2) is a member of the C-C chemokine family, which regulates migration and infiltration of monocytes/macrophages. ³⁰ This protein has powerful activating and chemotactic effects on mononuclear phagocytes and is a well-known pro-inflammatory cytokine that plays a central role in the development of inflammation in the CNS. ^31–33  

In this study, we show that with the development of diabetes mellitus induced by STZ injection, the increased expression of MCP-1 was aggravated at 3 and 5 months. In our rodent diabetic model, the positive MCP-1 staining was mainly distributed around vessel walls, as well as in the cell bodies and synapses of the RGCs. With the development of disease, the positive staining of MCP-1 in the synapses of the RGC extended to the inner nuclear layer of the retina. These findings implied that MCP-1 may play a role in the pathologic mechanism of DR. To testify our hypothesis, we designed a series of experiments in vitro. 

AGEs modify the structures of intracellular and extracellular proteins, thereby directly impairing their functions. Although the effect of AGEs on cells in vitro cannot completely represent tissue damage in diabetic disease, AGEs can be used to mimic the microenvironment of DR in vitro. ^34–36 From our in vitro data, we showed that retinal neurons are the main source of MCP-1 release following exposure to AGEs. In combination with our in vivo data, we presumed that the RGCs may be the main source of increased MCP-1 expression in a rodent model of DR. 

In certain pathological processes, activation of microglia occurs rapidly after injury of the CNS, which forms the first line of defense against CNS injury through their capacity to migrate, proliferate, secrete inflammatory factors, and phagocytose damaged cells and debris. ^37,38 During these processes, neuronal chemokines have been indicated as promising candidates for neuron–microglia signaling. ^39,40 Many in vitro studies support the important role of chemokines in the migration and neurotoxic activity of microglia on CNS injury and neuroinflammation. ^41,42  

In retinal degeneration mice, Zeng and colleagues ²⁶ showed that damaged photoreceptors (another kind of neuron in retina) induced microglial activation and migration, which play a major role in retinal degeneration. In this process, the expression of chemokines, including MCP-1, precede the onset and peak of microglial migration, suggesting that chemokines play a role in microglial activation in response to neuronal dysfunction. The authors hypothesized that the chemokines were produced by the diseased photoreceptors, but this notion was not confirmed. In the rodent diabetic model, retinal microglia are activated early during the onset of diabetes, and the activated microglia appear primarily in the GCL, ³ but no reports have shown which chemoattractant activated the microglia in the retina. 

Although astrocytes and microglia are the primary source of chemokines, there is evidence that neurons also express and secrete chemokines. ^43–45 In our Transwell system, retinal neurons were stimulated with AGE, and the simultaneous addition of an anti–MCP-1 antibody or MCP-1 knockdown confirmed the role of neuronal MCP-1 in microglial activation by retinal neurons. Neutralizing neuronal MCP-1 with specific antibody or downregulating MCP-1 with SiRNA can greatly inhibit the effect of retinal neurons on microglia activation (Figs. 7–9). We demonstrate for the first time that retinal neurons release MCP-1 as a result of AGE stimulation, and retinal neuronal MCP-1 plays a role in retinal microglial activation and migration in vitro. Based on our in vitro and in vivo study, we hypothesize that the pathological increase in the expression of MCP-1 originates from diseased RGCs in STZ-induced DR, which activate and recruit retinal microglia to the GCL of the retina. The activated microglia accelerate RGC death by secreting cytotoxic substances, such as TNF-α, as well as effectively phagocytose damaged cells and debris. With the development of RGC dysfunction, a pathological increase in the expression of MCP-1 was boosted, followed by strong microglial activation. Therefore, a cycle developed that resulted in severe diabetic neuronal degeneration in the advanced stage. 

There is convincing evidence to suggest that neuropathy is a feature of DR because ganglion cell death was observed in the diabetic retina. ⁴⁶ Our results strongly support the hypothesis that the neuropathy in DR may be attributed to the pathological expression of chemokines by themselves. This research also provides insights into the role of neurons in neuron–microglia activation, thereby indicating a neuronal contribution to chemokine signaling. Understanding this bridge between retinal neurons and microglia may contribute to the development of new therapeutic targets for DR mediated by neuronal chemokines. 

In summary, upregulation of MCP-1 appeared in the early stage of DR and increased with the development of disease. Retinal neurons are the main source of MCP-1 release and play an important role in retinal microglia activation, which may be an important link in DM pathogenesis. 

Acknowledgments

The authors thank Xiaoxin Li, PhD, for advice and assistance in the development of the animal model. 

Footnotes

 Supported by the China Natural Science Foundation (30700919) and the China Postdoctoral Science Foundation (20070410028).

Footnotes

 Disclosure: N. Dong, None; X. Li, None; L. Xiao, None; W. Yu, None; B. Wang, None; L. Chu, None