Article Text
PDF
Abstract
Background The complement activation molecule C5a has been found in the eye and is implicated in the pathogenesis of ocular inflammatory diseases. In this study, the authors sought to investigate C5a's effects on human retinal pigment epithelial (RPE) cells and peripheral blood mononuclear cells (PBMCs), and on the interaction between RPE cells and PBMCs.
Methods Arising retinal pigment epithelia cell line-19 and PBMCs isolated from healthy donors were used in this study. Western blot, real-time PCR and cell surface receptor staining were used to detect C5a receptor expression. Real-time PCR was used to detect cytokine mRNA expression. A thiazolyl blue tetrazolium bromide assay was used to detect cell viability. Cells were stained with Annexin V and 7-aminoactinomycin D for an apoptosis assay. Cell proliferation was measured using a tritiated thymidine incorporation assay.
Results C5a receptors were present on RPE cells, and receptor expression was increased by pro-inflammatory cytokines. C5a suppressed RPE cells' production of transforming growth factor β2, an important immunosuppressive agent in the eye. In addition, the viability of RPE cells was decreased in the presence of C5a, and this effect was not due to apoptosis. C5a increased proliferation of PBMCs and upregulated their production of pro-inflammatory cytokines. Finally, C5a decreased RPE cells' ability to suppress immune cell proliferation.
Conclusion The results provide a direct link between complement activation and intraocular inflammation. This line of information may help to understand the mechanism of the pathogenesis of intraocular inflammatory diseases. Moreover, the authors show that a close, reciprocal interaction between the innate immune system and the adaptive immune system may be involved in the development of such diseases.
- Complement activation molecule C5a
- retinal pigment epithelial (RPE)
- intraocular inflammation
- immunology
- inflammation
Statistics from Altmetric.com
- Complement activation molecule C5a
- retinal pigment epithelial (RPE)
- intraocular inflammation
- immunology
- inflammation
Introduction
Activation of the complement system is a double-edged sword: while it can provide protection against infections, it can also damage self-tissue by promoting inflammatory reactions. In a normal eye, the complement system is continuously activated at low levels, and intraocular complement regulatory proteins keep this spontaneous activation under check. However, complement dysregulation can occur, and an imbalance between complement activation and complement inhibition is implicated in the development and progression of several immune-mediated ocular conditions such as age-related macular degeneration (AMD) and uveitis.1
A critical event in complement activation is the cleavage of complement component C5 into C5a and C5b. The C5a fragment is an inflammatory mediator that induces cell migration, cell adhesion and cytokine release.2 Several studies have suggested that C5a plays a role in the pathogenesis of ocular inflammatory diseases. Nozaki et al showed that C5a is present in drusen of patients with AMD, and that it induces vascular endothelial growth factor expression promoting choroidal neovascularisation.3 Copland et al demonstrated that anti-C5 therapy reduces the disease severity in experimental autoimmune uveoretinitis, an animal model of posterior uveitis.4 C5a stimulates cells via interaction with C5a receptors (C5aR), which belong to a family of G-protein coupled receptors with seven transmembrane segments.2 5 C5a receptors are expressed on immune cells including macrophages, neutrophils, and T cells. Recently, Fukuoka et al have found that retinal pigment epithelial (RPE) cells also possess C5a receptors.6 7
RPE cells are important ocular resident cells critical for vision. Under normal conditions, RPE cells downregulate the intraocular immune response.8–12 However, under pro-inflammatory conditions, RPE cells can become activated and further contribute to inflammation through a number of mechanisms.13–17 Complement fragment C5a is one of the potent activators of RPE cells. Fukuoka et al found that C5a caused RPE cells to increase their expression of inflammatory cytokines including interleukin-1β (IL-1β), interleukin-6 (IL-6), monocyte chemotactic protein-1 (MCP-1), granulocyte-macrophage colony-stimulating factor (GM-CSF) and interleukin-8 (IL-8).6 7
In this study, we sought to elucidate the role of the complement system in intraocular inflammation by understanding C5a's effect on RPE cells, on peripheral blood mononuclear cells (PBMCs), and on the interaction between RPE cells and PBMCs.
Materials and methods
Reagents, cell lines and antibodies
All cell lines were cultured at 37°C in 5% CO2. Arising retinal pigment epithelia cell line-19 (ARPE-19) cells (ATCC, Manassas, Virginia) were cultured in complete Dulbecco's Modified Eagle's Medium supplemented with 10% fetal bovine serum, 2 mM l-glutamine (Invitrogen, Carlsbad, California) and 1× penicillin. Only cells younger than passage 20 were used. ARPE-19 cells were seeded as 30–40% confluent. Treatment was applied on the second day when cells were 50% confluent. PBMCs were isolated from normal healthy donors (NIH blood bank) using Ficoll gradient centrifugation. PBMCs were cultured in Roswell Park Memorial Institute (RPMI) medium supplemented with 10% inactivated human AB serum, 2 mM l-glutamine, and 1× penicillin (RPMI complete medium). Mouse antihuman CD3 and anti-CD28 antibody (eBioscience, San Diego, California) were used specifically to stimulate T cells. C5a was purchased from R&D Systems (Minneapolis, Minnesota). C5aR antagonist was a kind gift from Dr Anthony Adamis. Antihuman C5aR, Biotin rat anti-mouse IgG and APC Streptavidin were purchased from BD Pharmingen (Franklin Lakes, New Jersey). C5aR primary antibody for western blot was purchased from Santa Cruz Biotechnology (Santa Cruz, California). Antiglyceraldehyde-3-phosphate dehydrogenase (GAPDH) antibody was from Abcam (Cambridge, Massachusetts).
Sodium dodecyl sulfate PAGE and western blotting
A total of one million ARPE-19 cells were lysed in 100 μl of lysis buffer (50 mM Tris-Cl, 1% Triton X-100, 100 mM NaCl, 2 mM EDTA, 50 mM NaF, 50 mM glycerol-phosphate, 1 mM NaVO4 and 1× protease inhibitor cocktail (Roche Molecular Biochemicals, Indianapolis, Indiana)). Samples were prepared by adding an equal amount of 2× sodium dodecyl sulfate protein loading buffer and boiled at 95°C for 5 min to achieve complete cell lysis. Immunoblotting was performed according to standard protocols.
Real-time PCR (RT-PCR)
Total RNA was extracted from confluent monolayers of ARPE-19 (RNeasy Kit; Qiagen, Valencia, California). Total RNA (100 ng) was reverse-transcribed to cDNA (ReactionReady First Strand cDNA Synthesis Kit, SABiosciences, Frederick, Maryland). RT-PCR runs were performed in triplicate using a 96-well format PCR array and an ABI 7500 real-time PCR unit (Applied Biosystems, Foster City, California). Human GAPDH was used for internal standards. The results were expressed as the n-fold expression of C5aR and transforming growth factor (TGF)-β2, normalised to that of GAPDH. Primers for TGF-β2, C5aR and GAPDH genes were purchased from SABiosciences.
Three-step cell-surface C5a receptor staining
ARPE-19 cells were first incubated with purified antihuman C5aR at a final concentration of 10 μg/ml for 20 min at 4°C. A control mouse IgG was used as background staining. They were then incubated with Biotin rat antimouse IgG (10 μg/ml) for 20 min at 4°C, and lastly with APC Streptavidin (4 μg/ml) for 15 min at room temperature. Cells were washed with PBS between each step and subsequently fixed with 1% paraformaldehyde. Data were acquired using a FACSCalibur flow cytometer (Becton Dickinson, San Jose, California) and analysed with FlowJo software (TreeStar, San Jose, California).
Thiazolyl blue tetrazolium bromide assay
Ten thousand ARPE-19 cells per well were seeded in 96-well plates with or without C5a. After 48 h, thiazolyl blue tetrazolium bromide (MTT) was added to the medium at a final concentration of 0.5 mg/ml for 4 h to allow MTT to be metabolised. Media in wells were discarded, and cells were resuspended in formazan (MTT metabolic product) in 200 μl of dimethl sulfoxide (DMSO) per well. Five minutes after adding DMSO, wells were read at an optical density of 540 nm.
Apoptosis assay
Apoptotic cells were detected by staining cells with both annexin-V-FITC and Via-probe (7-amino-actinomycin D; 7-AAD) according to the manufacturer's instructions (BD Biosciences). Briefly, 2.5×105 cells were incubated with saturating concentrations of annexin V-FITC and 7-AAD for 15 min at room temperature and immediately analysed by flow cytometry as described above.
Measurement of cytokines
Cell supernatants were generated from culturing PBMCs alone or co-culturing PBMCs with ARPE-19. Complete RPMI medium was used for both PBMCs cultured alone and co-cultured with ARPE-19. Cell supernatants were collected after 5 days of incubation, aliquoted immediately and kept in a −80°C freezer. One aliquot was shipped overnight on dry ice to Aushon Biosystems (Billerica, MA) for multiplexing ELISA cytokine array assay. For a different aliquot, an ELISA kit (R&D Systems) was used to detect the tumour necrosis factor (TNF)-α level.
RPE-PBMC co-culture and proliferation assay
ARPE-19 cells were seeded first and allowed to grow to 80% confluent. PBMCs were then added for co-culture. Cells were pulsed with (3H) thymidine for 12–16 h after a 5-day culture and then harvested, and uptake of (3H) was measured with a β counter in units of counts per minute.
Statistical analysis
All statistical analyses were performed using the independent Student t test. In the figures, ** indicates statistically significant p value <0.01. Error bars represent standard deviations.
Results
C5a receptors were present on ARPE-19 cells, and their expression was increased by pro-inflammatory cytokines
We wanted to confirm the presence of C5a receptors (C5aR) on the ARPE-19 cell line as well as investigate how different conditions may modulate the expression level of C5aR. We treated separate groups of ARPE-19 cells with C5a, H2O2, pro-inflammatory cytokines (IL-1β, TNF-α and IFN-γ) and anti-inflammatory cytokines (IL-4, IL-10, TGF-β2) and compared these results with the control. We chose the pro-inflammatory cytokine mixture because the three cytokines we used are found to be elevated in intraocular inflammatory diseases, and this may better mimic the in vivo intraocular inflammation.18 19 Western blot showed C5aR expression in ARPE-19 cells, confirming the presence of C5aR on ARPE-19 cells (figure 1A). We then performed a time-course real-time PCR to examine levels of C5aR mRNA expression following treatment with pro-inflammatory cytokines and anti-inflammatory cytokines. After stimulation with pro-inflammatory cytokines only, the C5aR mRNA level increased, starting at 3 h, reached a peak at 6 h and returned to the prestimulation level at 24 h (figure 1B). After stimulation with anti-inflammatory cytokines only, the C5aR mRNA level did not change compared with the control, indicating that the increase in mRNA expression was specific to pro-inflammatory cytokines (figure 1B). We wanted to confirm our observation by staining for cell surface C5aR and analysing with flow cytometry. Cell-surface C5aR expression was increased 24 h after pro-inflammatory cytokines treatment (figure 1C). This upregulation was transient, and the C5aR level returned to the pretreatment level 48 h poststimulation (figure 1C).
Upregulation of C5a receptor expression on arising retinal pigment epithelial cell line-19 (ARPE-19) cells mediated by pro-inflammatory cytokines. These cells were treated with or without C5a (50 ng/ml), H2O2 (1 mM), pro-inflammatory cytokines (10 ng/ml interleukin (IL)-1β, 10 ng/ml tumour necrosis factor-α, 25 ng/ml IFN-γ), and anti-inflammatory cytokines (10 ng/ml IL-4, 10 ng/ml IL-10, and 20 ng/ml transforming growth factor-β2). (A) After 24 h of treatment, cells were trypsinised and processed for western blot analysis for indicated proteins. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control. Two separate experiments were performed, and the figure shows representative data. Densitometry graph is also shown. (B) At 3 h, 6 h, 10 h and 24 h post-treatment, RNA was collected from cells, and real-time PCR was performed. Results are expressed as fold change in C5aR mRNA expression compared with the control (no treatment). Open bars represent cells treated with anti-inflammatory cytokines only, while solid bars represent cells treated with pro-inflammatory cytokines only. Each experiment was performed in triplicate, and the results shown are average data from three separate experiments. Error bars represent±SD. **Statistically significant p<0.01. (C) At 24 h and 48 h post-treatment, cells were collected for cell-surface C5a receptor staining. The IgG peak indicates background staining. Three separate experiments were performed, and the figure shows representative data. Mean fluorescence intensity (MFI) values are also listed.
C5a suppressed ARPE-19 cells' production of TGF-β2
TGF-β2 is an important factor that contributes to immune suppression in the eye and is produced by RPE cells.10 20 21 We wanted to examine whether C5a could affect RPE cells' production of TGF-β2. A time-course RT-PCR was performed to investigate the production of TGF-β2 mRNA following C5a stimulation. As shown in figure 2A, TGF-β2 levels decreased following C5a stimulation, and the decrease was most dramatic at 6 h (p<0.01). We then tested whether the TGF-β2 mRNA expression resulted in a difference in protein production. ELISA results showed that C5a significantly reduced TGF-β2 secretion from ARPE-19 cells (p<0.01, figure 2B).
Downregulation of transforming growth factor (TGF)-β2 from arising retinal pigment epithelial cell line-19 cells mediated by C5a. (A) Cells were treated with or without C5a (50 ng/ml). At 3 h, 6 h, 10 h and 24 h post-treatment, RNA was collected, and real-time PCR was performed. Results are expressed as fold change in TGF-β2 mRNA expression compared with the control (no treatment). Each experiment was performed in triplicate, and the results shown are average data from three separate experiments. Error bars represent±SD. (B) Cells were treated with or without C5a (50 ng/ml) for 2 days. Cell supernatants were collected for ELISA. Results shown are average data from three separate experiments. Error bars represent±SD. **Indicates statistically significant p<0.01.
C5a decreased ARPE-19 cells' mitochondrial activity, and this effect was not due to cell apoptosis
We were interested in examining what effect C5a may have on RPE cells' mitochondrial activity. ARPE-19 cells were treated with C5a, and the MTT assay, which measures the amount of mitochondrial reductase activity, was performed. Compared with control, C5a significantly decreased ARPE-19 cells' mitochondrial activity (p<0.01, figure 3A). A decrease in mitochondrial reductase activity may be due to cell death, or an overall decrease in cell metabolism in the absence of cell death. Therefore, we investigated the cause by staining ARPE-19 cells with annexin V, an apoptosis marker, and 7-aminoactinomycin D (7AAD), a cell-viability marker. Compared with the control, ARPE-19 cells treated with C5a stained comparably with annexin V and 7AAD, suggesting that the decrease in cell mitochondrial activity was not due to cell apoptosis (figure 3B).
Suppression of arising retinal pigment epithelial cell line-19 cell viability mediated by C5a. (A) Cells were treated with or without C5a (50 ng/ml) for 48 h. A thiazolyl blue tetrazolium bromide assay was performed, and absorbance of solubilised formazan (thiazolyl blue tetrazolium bromide metabolic product) was read at 540 nm. Each experiment was performed in triplicate, and the results shown are average data from three separate experiments. Error bars represent±SD. (B) Cells were treated with or without C5a (50 ng/ml) for 48 h and stained with Annexin V and 7-aminoactinomycin D (7AAD). Three separate experiments were performed. The figure shows representative data, and the quadrant statistics of flow cytometry is also shown. LL, lower left; LR, lower right; UL, upper left; UR, upper right. ** Indicates statistically significant p<0.01.
C5a increased PBMCs' production of pro-inflammatory cytokines
We measured the secretion of various cytokines by PBMCs after C5a treatment. Figure 4 shows that PBMCs' production of IFN-γ, IL-1β, IL-6 and TNF-α increased following stimulation with both C5a and anti-CD3/anti-CD28 antibodies, compared with groups that only received anti-CD3/anti-CD28 treatment. The secretion of IL-8, IL-10, IL-12p70, IL-17 and IL-23 was also measured, but no difference was found between the groups receiving C5a treatment and the groups without C5a treatment (data not shown).
Upregulation of pro-inflammatory cytokines from peripheral blood mononuclear cells (PBMCs) mediated by C5a. Cell supernatants were collected after 5 days of incubation. Cytokine concentrations of interferon (IFN)-γ, tumour necrosis factor (TNF)-α, interleukin (IL)-1β and IL-6 were determined by a multiplexing ELISA cytokine array assay. Two separate experiments were performed, and the figure shows representative data.
C5a increased proliferation of PBMCs and decreased RPE cells' ability to suppress immune cell proliferation and secretion of TNF-α
We wanted to investigate whether C5a could increase immune cell proliferation. As shown in figure 5A, groups that did not receive stimulation by anti-CD3 and anti-CD28 antibodies proliferated very little. PBMCs that were treated only with anti-CD3 and anti-CD28 antibodies proliferated dramatically, while those treated with C5a in addition to the antibodies proliferated even more (figure 5A). Adding C5a antagonist abrogated the additional stimulatory effect of C5a (figure 5A). RPE cells have been shown to potently suppress T cell response to T cell receptor stimulation.10–12 We were curious as to whether the presence of C5a could decrease this local immunosuppressive function of RPE cells. As expected, when PBMCs were co-cultured with RPE cells, and then stimulated with the same antibodies, RPE cells significantly suppressed the level of proliferation (figure 5A). However, when C5a was added, proliferation once again increased, but not quite back to the original level when RPE cells were not present (figure 5A). We then tested whether C5a affected RPE cells' ability to suppress immune cell secretion of TNF-α. As shown in figure 5B, PBMCs co-cultured with RPE cells secreted a significantly lower level of TNF-α. However, when C5a was added, TNF-α secretion showed an increasing trend, although it was not statistically significant (p=0.08).
Increased proliferation of peripheral blood mononuclear cells (PBMCs) and decreased retinal pigment epithelial (RPE) cells' ability to suppress PBMC proliferation and tumor necrosis factor (TNF)-α production mediated by C5a. (A) arising retinal pigment epithelia cell line-19 cells (ARPE-19 cells) were seeded first in 96-well plates and cultured in complete Dulbecco Modified Eagle Medium. After 2 days, PBMCs were isolated from healthy donors using Ficoll gradient and seeded in wells with or without ARPE-19 cells. At this time, the culture medium was changed abruptly to complete Roswell Park Memorial Institute. Cells were treated with or without C5a (50 ng/ml), C5a antagonist (2.5 μM), and anti-CD3 (2 μg/ml) and anti-CD28 (2 μg/ml) antibodies that specifically stimulate T cells. After 5 days, cells were pulsed with (3H) thymidine for 12–16 h. Cells were harvested, and uptake of (3H) thymidine was measured with a beta counter in units of counts per minute. Results are expressed as stimulation index, which is the fold change compared with control PBMCs (no treatment and no antibody stimulation). Two sets of data are presented side by side. Error bars represent±SD. (B) Cell supernatants were collected after 5 days of co-culturing PBMCs with irradiated (9000 rad) ARPE-19 cells. Cytokine concentration of TNF-α was determined by ELISA assay. The results shown are average data from two separate experiments. Error bars represent±SD.
Discussion
Dysregulation of the complement system has been linked to intraocular inflammatory diseases such as uveitis and AMD. In experimental autoimmune uveoretinitis, complement depletion by cobra venom factor has been shown to decrease disease severity.1 Several reports have found that a single nucleotide polymorphism of complement factor H (CFH), an important complement inhibitor, confers a genetic risk of developing AMD.22–25 This CFH single nucleotide polymorphism most likely causes a functional impairment in inhibiting the alternative complement pathway, thus leading to excessive complement activation.26 There are also reports indicating that in AMD, the complement system is activated both locally and systemically. C3d, Ba, C3a, C5a and factor D, which all reflect systemic complement activation, are significantly elevated in the circulation of patients with AMD.27 Studies have found that giant drusen, found in the eyes of patients with AMD, contain complement proteins C3 and C5, C3a and C5a, CFH, and membrane attack complex component C5b-9, suggesting that the complement pathway has been locally activated in the human retina.28–30 In this study, we examined C5a's effects on RPE cells and PBMCs. Our results demonstrate that C5a decreased the immune privilege function of RPE cells by suppressing TGF-β2 expression and inhibiting cell growth. In addition, C5a promoted lymphocyte proliferation and the production of pro-inflammatory cytokines. These results suggest that C5a helps to produce a pro-inflammatory intraocular environment which may contribute to local inflammation.
Excessive inflammation in the eye is detrimental and could lead to vision loss. Normally, a healthy eye is under a state of immune suppression, termed the downregulatory intraocular environment (DIE).21 TGF-β2 is an important factor that contributes to DIE and is produced by RPE cells.10 20 21 In our study, we found that C5a decreased RPE cells' production of TGF-β2, compromising DIE (figure 2). Furthermore, we also found that C5a can stimulate and increase T cell proliferation (figure 5A). Moreover, the presence of C5a decreased RPE cells' ability to suppress T cell proliferation, even though this effect is not dramatic (figure 5A). Our observations suggest that C5a establishes a favourable environment for T cells and could dramatically increase the number of these cells present in the eye during an inflammatory event. A subsequent functional study revealed that these T cells increased their production of pro-inflammatory cytokines such as IFN-γ, IL-1β, IL-6 and TNF-α in the presence of C5a (figures 4, 5B). Interestingly, these same pro-inflammatory cytokines significantly upregulated C5a receptor expression on RPE cells, increasing the sensitivity of RPE cells to the effects of C5a (figure 1B,C).
The origin of C5a in ocular tissues has been speculated. Sohn et al found that the complement system is continuously activated (through the alternative pathway) at a low level in the normal rat eye, and intraocular complement regulatory proteins in the intraocular fluid as well as on cell membranes tightly regulate this spontaneous complement activation.31 Nozaki et al examined laser injury-induced complement in wild-type mice and found that C5a and C3a were localised to the RPE cells and their immediate vicinity before the infiltration of neutrophils and macrophages, suggesting that resident cells produce complement anaphylatoxins.3 Strainic et al found that the cognate interaction between antigen presenting cells and T cells caused both types of cells to generate C5a and C3a, suggesting that infiltrating immune cells during ocular inflammation may also be a source of C5a.32 The concentration of C5a in the eye likely varies among individuals and is also highly dependent on the immune status of the eye. Our decision to use 50 ng/ml was based on published articles.3 6 Fukuoka and Medof showed that the expression of IL-8 by ARPE-19 cells responded in a dose-dependent manner to C5a stimulation with maximal effect at 100 ng/ml.6
Currently, ocular injection with antivascular endothelial growth factor is used to treat neovascular AMD; no efficient treatment is available for the atrophic form.33 The results of our study suggest that new therapeutic strategies can be developed to disrupt the innate-adaptive circuitry. One of the approaches may be to target the complement activation molecule C5a, which may benefit in controlling ocular inflammation.
In summary, our results emphasise the importance of reciprocal interaction between the innate and the adaptive immune system in the development of intraocular inflammatory diseases: C5a favours T cell proliferation and the secretion of pro-inflammatory cytokines, which in turn, increase the number of C5a receptors on RPE cells, rendering these cells more sensitive to the detrimental effects of C5a. C5a inhibits RPE cell growth and suppresses RPE cells' ability to produce TGF-β2, further compromising the DIE and promoting inflammation. Understanding C5a's role in the pathogenesis of AMD may lead to the development of new therapeutic agents.
Acknowledgments
We thank A Adamis for kindly providing the C5aR antagonist.
References
Footnotes
Funding This research was supported by the Intramural Research Program of NIH, National Eye Institute.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.