Extracellular Nucleotides and Interleukin-8 Production by ARPE Cells: Potential Role of Danger Signals in Blood–Retinal Barrier Activation

Abstract

purpose. RPE cell activation is an important feature of autoimmune uveitis. This investigation focused on whether extracellular nucleotides could contribute to this activation, and the effects of ATPγS, UTP, and UDP on the production of IL-8 by RPE cells was studied in relation to their expression of functional P2Y receptors.

methods. ARPE-19 cells were cultured with ATPγS, UTP, UDP, and TNF. IL-8 gene transcription and protein production were measured by semiquantitative RT-PCR and ELISA. Western blot analysis and RT-PCR were used to investigate ERK 1/2 activation and P2Y expression. Changes in intracellular calcium and cAMP concentration were analyzed by spectrofluorometry and radioimmunoassay.

results. Stimulation of ARPE-19 cells with ATPγS, UTP, and UDP induced IL-8 gene transcription and protein secretion. TNFα induction of IL-8 secretion was also increased by ATPγS, UTP, and UDP. Nucleotide induction of IL-8 production was blocked by PD98059, and all nucleotides stimulated ERK 1/2 phosphorylation. P2Y₂ and P2Y₆ mRNAs were detected in ARPE-19 cells. All tested nucleotides induced a pulse of intracellular calcium.

conclusions. ATPγS, UTP, and UDP stimulate both basal and TNFα-induced IL-8 secretion in RPE cells through an ERK 1/2-dependent pathway. The results suggest that those effects are mediated by P2Y₂ and P2Y₆ receptors.

Retinal pigment epithelial (RPE) cells constitute the most external part of the retina. In physiological conditions, RPE cells are nondividing cells that form the external blood–retinal barrier (BRB) between the neural retina and the choroidal circulation. However, during autoimmune uveitis, RPE cells are activated and participate in the recruitment and stimulation of inflammatory cells into the eye. ¹ Hence, it has been described, in vitro, that RPE cells can produce different chemokines such as interleukin (IL)-8. ² IL-8 attracts and activates neutrophils and its role in uveitis is highlighted by the fact that endotoxin-induced uveitis is partially inhibited by anti-IL-8 antibody treatment. ³ Moreover, experimental autoimmune uveitis is abrogated when mice are depleted of neutrophils. ⁴ Accordingly, understanding the pathways involved in IL-8 production by RPE cells is a keystone for the future management of this blinding condition. 

It has been extensively described that cytokines such as tumor necrosis factor (TNF)-α and interleukin (IL)-1β can strongly activate RPE cells to secrete various inflammatory mediators, including IL-8. ² However, it is likely that other stimuli play a role in triggering RPE cell activation during uveitis. Of interest, the role of “danger signals” in the development of an inflammatory response has recently emerged. The danger theory postulates that the immune system recognizes substances that cause danger to the host. The danger signals can arise from pathogens but also from stressed or necrotic cells, and it is now accepted that inflammatory cells respond to molecules normally found inside the cells. Among those, it has been shown that nucleotides are involved in the regulation of immune responses. ^{5 6}  

Nucleotides are released in extracellular fluids through cell lysis, exocytosis of nucleotide-concentrating granules, vesicular trafficking, or membrane transport proteins. ⁷ Extracellular nucleotides play autocrine and paracrine roles by activating two families of receptors: P2X receptors, which are ATP-gated ion channels, and P2Y receptors, which are coupled to G proteins. ⁷ P2Y₂ receptors have been identified in RPE cells, and their activation has been shown to strongly stimulate fluid reabsorption and retinal reattachment in a rat model of retinal detachment. ^{8 9} More recently, Tovell and Sanderson ¹⁰ have demonstrated the presence of P2Y₁ and P2Y₆, extending the potential role of purine signaling in RPE function. 

In this study, we demonstrated that the P2Y₂ and P2Y₆ receptors ligands ATPγS, UTP, and UDP stimulate IL-8 production by human RPE cells through activation of mitogen-associated kinase. 

Methods

Cell Culture and Reagents

ARPE-19 (American retinal pigment epithelium type 19) is a spontaneously arising human RPE cell line obtained from the American Type Culture Collection (ATCC, Manassas, VA). These cells were cultured in a 1:1 mixture of DMEM (Dulbecco’s modified Eagle’s medium; Invitrogen, Paisley, UK) and Ham’s F12 with 2.5 mM l-glutamine (Invitrogen), supplemented with 10% fetal bovine serum (FBS), 100 IU/mL penicillin, and 100 μg/mL streptomycin in a 5% CO₂ and 95% humidity incubator at 37°C. To avoid rapid hydrolysis of ATP, we used the hydrolysis-resistant ATP analogue adenosine 5′-O-(3-thiotriphosphate) (ATPγS). ATPγS, uridine triphosphate (UTP), and uridine diphosphate (UDP) were purchased from Sigma-Aldrich (St. Louis, MO), 8-(p-sulfophenyl) theophylline (8-p-SPT) from Research Biochemicals International (Natick, MA), TNF-α from Biosource (Nivelles, Belgium), and PD98059 from Cell Signaling (Leiden, The Netherlands). 

Real-Time RT-PCR Analysis of IL-8

ARPE-19 cells were seeded at 5 × 10⁵ cells/well in 24-well plates (2.5 × 10⁵ cells/cm²). One day later extracellular nucleotides were added to the cultures for 3 hours. RNA was extracted (Trizol reagents; Life Technologies, The Netherlands; and RNeasy kit; Qiagen, Westburg, The Netherlands). Real-time RT-PCR was performed as previously described. ¹¹ Briefly, diluted cDNA was analyzed with 2× SYBR green PCR master mix (Applied Biosystems [ABI], Lennik, Belgium) on a PCR system (7500 Fast Real-Time and 7500 Fast software; ABI), according to the manufacturer’s protocol. Gene-specific primers were designed according to sequences covering the conserved peptide sequence region and in interexonic gene sequences: human IL-8 forward: 5′-CTTCCTGATTTCTGCAGCTCTGT-3′; human IL-8 reverse: 5′-GGTGGAAAGGTTTGGAGTATGTCTT-3′; human β-actin forward: 5′-AGAAAATCTGGCACCACACC-3′; and human β-actin reverse: 5′-GGGGTGTTGAAGGTCTCAAA-3′. The relative mRNA amount in each sample was calculated based on its threshold cycle in comparison to the threshold cycle of the human β-actin housekeeping gene. Real-time PCR were conducted in triplicate in two independent experiments, and the ensuing mean value was calculated. The results were calculated as follows (2^{(Ct of IL-8 − Ct of β-actin housekeeping gene)}) and expressed in arbitrary units. Results are expressed as a stimulated/control ratio. 

IL-8 Measurement

To measure IL-8 production, 5 × 10⁵ ARPE-19 cells were seeded with complete medium in 24-well plates (2.5 × 10⁵cells/cm²). One day later, medium was replaced by medium without FBS, and after 4 hours the cells were stimulated with the agonist for the indicated time. The medium was then collected and IL-8 secretion quantified by a specific ELISA (Biosource). 

RT-PCR Detection of P2Y Messengers

We designed specific primers for P2Y₁, P2Y₂, P2Y₄, and P2Y₆ receptors that were synthesized by Eurogentec (Seraing, Belgium). Total RNAs were isolated (RNeasy kit; Qiagen) and reverse transcribed by using random hexamers (Superscript II Reverse Transcriptase; Invitrogen). PCR experiments were performed with Taq polymerase (Invitrogen) according to the manufacturer’s instructions. PCR amplification conditions were 94°C, 4 minutes for 1 cycle; 94°C, 45 seconds, 50°C, 30 seconds, 72°C, 1 minute 30 seconds for 30 cycles; 72°C, 6 minutes for 1 cycle. 

Calcium Measurements

After 2 days of culture, ARPE-19 cells were washed in HBSS without phenol red (Hanks’ Balanced Salt Solution; Invitrogen) containing 0.1% BSA and resuspended at 10⁷cells/mL. Cells were then incubated with 10 μg/mL Fura-2 (Invitrogen-Molecular Probes, Leiden, The Netherlands), 0.1 mg/mL pluronic acid F127 (Sigma-Aldrich) and 2.5 μg/mL sulfinpyrazole (Sigma-Aldrich) for 30 minutes at 37°C in darkness. The cells were washed and suspended in HBSS at 1 × 10⁶/mL before fluorometric analysis (LS50B; Perkin Elmer, Boston, MA). Calcium transients were determined with the 340/380-wavelength excitation ratio at an emission wavelength of 505 nm. 

Western Blot Analysis

Assays were performed on cells cultured in 10% FBS medium or after 4 hours of serum deprivation. Ten minutes after stimulation, the cells were washed with PBS at 4°C before being lysed in a buffer containing Tris-HCl 50 mM (pH 7.5), 100 mM NaCl, 1% Brij, 200 mM NaF, 5 mM Na₄P₂O₇, 4 mM Na₃VO₄, 2 mM pefablock_, 10 μg/mL aprotinin, and 10 μg/mL leupeptin. Protein concentration was determined by Minamide and Bamburg’s method. ¹² Equal amounts of proteins (75 μg) were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to a nitrocellulose membrane. Nonspecific binding was blocked by using 5% fat-free milk powder. Immunodetection was performed by incubating the membrane with specific primary antibodies (phospho-ERK 1/2, total ERK 1/2; Cell Signaling), for 2 hours at room temperature. Blots were then washed and incubated for 1 hour with a secondary horseradish peroxidase–conjugated antibody before being developed using a chemoluminescent detection kit (ECL system; Perkin Elmer). 

Statistical Analysis

Results were analyzed with commercial software (Instat; GraphPad, San Diego, CA) using a two-tailed unpaired Student’s t-test. 

Results

Effects of ATPγS, UTP, and UDP on IL-8 Production by ARPE-19 Cells

To investigate whether nucleotides could act as danger signals in RPE cells, we measured the effects of ATPγS, UTP, and UDP on IL-8 production by ARPE-19 cells. We chose ATP and UTP, since P2Y₂ expression has been described in RPE cells, and UDP, as this molecule is known to regulate IL-8 production in different cell types. ^{8 13} ARPE-19 cells were first stimulated with 100 μM of ATPγS, UTP, or UDP for 4 hours and the level of IL-8 gene transcription analyzed by semiquantitative RT-PCR. Figure 1Ashows that ATPγS, UTP, and UDP induce a significant increase in IL-8 mRNA. We then evaluated whether the protein level was also raised. To measure IL-8 production, 5 × 10⁵ ARPE-19 cells were seeded with complete medium in 24-well plates (2.5 × 10⁵ cells/cm²). One day later, the medium was replaced by medium without FCS, and after 4 hours, the cells were stimulated with the agonist for the indicated time. The medium was then collected, and IL-8 secretion was quantified by a specific ELISA (Biosource). As for gene expression, all tested nucleotides increased IL-8 production (Fig. 1B) . Given that culture conditions can influence the properties of ARPE-19 cells, we also tested the effects of extracellular nucleotides on IL-8 production after a longer term culture protocol recently described. ^{14 15} ARPE-19 cells (2.5 × 10⁵cells/well, six-well plate or 2.63 × 10⁴cells/cm²) were seeded for 3 days in 10% FBS-containing medium and then were serum starved for 24 hours in medium with 0.1% FBS. The cells were then stimulated with nucleotides in 0.5% FBS-containing medium for another 24 hours. We found that in those conditions all tested nucleotides also induce IL-8 secretion. Hence, IL-8 level was 2423 ± 336 pg/mL in control subjects, 6339 ± 571 pg/mL in ATPγS-stimulated cells, 5823 ± 293 pg/mL in UTP-stimulated cells, and 6185 ± 569 pg/mL in UDP-stimulated cells (P < 0.01 for each nucleotide compared with the control). Since TNFα plays a major role in uveitis development and is a strong activator of RPE cells, we decided to analyze also the effects of ATPγS, UTP, or UDP on the secretion of IL-8 after stimulation of ARPE-19 cells by TNFα. ¹⁶ ARPE-19 cells were activated by TNFα in the presence or absence of ATPγS, UTP, or UDP during 24 hours and IL-8 production measured by ELISA. As shown in Figure 2 , TNFα stimulation induced a considerable upregulation of IL-8 secretion by ARPE-19 cells. Addition of ATPγS, UTP, or UDP to the culture significantly increased the effect of TNFα. 

ATPγS, UTP, and UDP Stimulation of IL-8 Production through ERK1/2

It has been demonstrated that IL-8 secretion by ARPE-19 cells is regulated by ERK 1/2, and P2Y receptors are known to regulate this pathway. ^{2 17} To analyze the possible involvement of MAPK in the induction of IL-8 secretion by nucleotides, we first investigated whether ERK 1/2 was activated by nucleotide stimulation. ARPE-19 cells were treated with 100 μM ATPγS, UTP, or UDP for 5 and 30 minutes and the level of phosphorylated ERK 1/2 (which represents the active form of the protein) assessed by Western blot. We found that all tested nucleotides induced, after 5 minutes, a strong phosphorylation of ERK 1/2 which was sustained after 30 minutes (Fig. 3A) . We further determined whether this activation of ERK 1/2 was required for nucleotide stimulation of IL-8 production. ARPE-19 cells were preincubated for 1 hour with PD98059 (25 μM). The cells were thereafter treated with ATPγS, UTP, or UDP for 24 hours. ELISA assays showed that PD98059 significantly inhibits IL-8 release induced by nucleotide stimulation of ARPE-19 cells (Fig. 3B) . 

Expression of UDP-Sensitive P2Y₆ Purinergic Receptors on ARPE-19

P2Y₂ receptor expression has already been described on RPE cells. ⁸ ATP and UTP are known to be equipotent in activating those receptors. It is thus likely that the effects of ATPγS and UTP on IL-8 production and ERK 1/2 activation are mediated through P2Y₂ in ARPE-19 cells. However, the similar effects of UDP on IL-8 production and ERK 1/2 activation in ARPE-19 cells suggest a coexpression of other P2Y receptors. Accordingly, we looked at the expression of P2Y₆ and P2Y₄ receptors that were UDP- and UTP-sensitive by RT-PCR, and found the expression of P2Y₆ but not P2Y₄ receptors on ARPE-19 cells (Fig. 4) . In addition, we found the expression of P2Y₁, recently described by Tovell and Sanderson in primary RPE cells (Ref. ¹⁰ and data not shown). We then investigated whether UDP could activate the expected signal transduction linked to P2Y₆ receptor expression, a release of calcium from reticuloendoplasmic stores. We thus analyzed the calcium flux in ARPE-19 cells stimulated by UDP. We found that all tested nucleotides induced a pulse of intracellular calcium (Fig. 5) . Responses were reduced but persisted when the extracellular medium was depleted in calcium by EGTA (data not shown). No variation of intracellular cAMP level was found after ATPγS, UTP, or UDP stimulation (data not shown). 

Discussion

RPE cells form the external part of the BRB and normally protect the eye from the potential damage of inflammatory cells. However, this immune privilege is broken during autoimmune uveitis. In this situation, RPE cells are activated and secrete chemokines such as IL-8. Proinflammatory cytokines have been described as the major stimuli for this secretion, yet it is likely that there are many other contributing factors as well. In this work, we have postulated that extracellular nucleotides participate in the activation of RPE cells by acting as danger signals. Our data demonstrated that ATPγS, UTP, and UDP stimulate both basal and TNFα-induced IL-8 secretion in RPE cells through an ERK 1/2-dependent pathway. In addition, our results suggest that these effects were mediated by P2Y₂ and P2Y₆ receptors. 

IL-8 attracts and activates neutrophils. ² The role of these cells in autoimmune uveitis has been recently emphasized by Su et al., ⁴ who showed that experimental autoimmune uveitis induction is abrogated in neutrophil-depleted animals. In vitro, the main stimuli of IL-8 secretion by RPE cells are TNFα and IL-1β but stimulation of Toll-like receptor-3 and -9 has also been found to induce IL-8 secretion by RPE cells. ^{2 18 19} Of interest, Toll-like receptors recognize specific pathogen-associated patterns and belong to the so-called “danger signals.” Similarly, since nucleotides are normally found inside the cells and are released during tissue damage, extracellular nucleotides are also considered as danger signals. Our findings are thus in accordance with the fact that RPE cells act as sentinels that can be alerted when the eye is in danger. IL-8 secretion by RPE cells being basal, nucleotides release may help to recruit inflammatory cells from the choroid to the retina. ²⁰ This reaction is clearly advantageous during infection, for clearing the retina from pathogens. However, dysregulation of this response can also expose the eye to abnormal IL-8 production and in theory be involved in the physiopathology of inflammatory retinal disease. In this context, it is interesting to note that an inappropriate response to danger signals has been associated in vivo with the development of autoimmune diseases. ²¹  

Together with p38 and JNK, ERK1/2 belongs to the mitogen-activated protein kinase (MAPK) family. ERK 1/2 controls many cellular processes, including cytokine production by different cell types. ²² Herein, we have shown that the specific MAPK kinase inhibitor PD98059 inhibits IL-8 secretion induced by ATPγS, UTP, and UDP in RPE cells and that all tested nucleotides induce a strong activation of ERK 1/2. A similar involvement of MAPK in the effect of glycated human serum albumin on IL-8 production by RPE cells has been demonstrated by Bian et al. ²³ ERK 1/2 was also implicated in the secretion of IL-8 induced in RPE cells by coculturing them with monocytes. ²⁴ Altogether, this suggests that ERK 1/2 plays a central role in the regulation of IL-8 secretion by RPE cells. 

Cellular responses to extracellular nucleotides are mediated through P2 receptors which include P2X and P2Y receptors. P2X receptors are ATP-gated ion channels whereas P2Y are G protein-coupled receptors. P2 receptors display a very complex pharmacology. Much evidence suggests that the effect of ATPγS and UTP on IL-8 secretion by ARPE-19 cells is mediated through P2Y₂ receptors. First, we found transcripts for this receptor, confirming the work by Sullivan et al., ⁸ which has shown the presence of functional P2Y₂ receptors on RPE cells, but not for the UTP-sensitive P2Y₄ receptor. Second, ATP and UTP are equipotent agonists of P2Y₂ receptors, and indeed we have observed that both nucleotides have similar effects on IL-8 secretion and ERK 1/2 activation in ARPE-19 cells. Third, as expected for a P2Y₂ stimulation, ATPγS and UTP induce a calcium release from endoplasmic stores, and no modulation of cAMP. Of interest, it was also shown that UDP had effects similar to those of ATPγS and UTP on IL-8 production and ERK 1/2 stimulation. UDP’s selective binding of the P2Y₆ receptor suggests that RPE cells express it. Accordingly, after the recent work of Tovell and Sanderson, ¹⁰ we observed the expression of P2Y₆ messengers in ARPE-19 cells and calcium release from endoplasmic stores after UDP stimulation. Because gene expression differences between primary RPE and ARPE-19 cells have been demonstrated, studies performed on ARPE-19 cells should be interpreted with caution. ^{14 25} In their work, Tovell and Sanderson ¹⁰ found no differences in term of P2Y₆ expression between native and cultured RPE cells. Moreover, animal studies have found P2Y₆ expression on RPE cells in vivo. ²⁶ Altogether, our work and those data suggest that functional P2Y₆ receptors are expressed in RPE cells. 

In conclusion, we found that ATPγS, UTP, and UDP activate IL-8 production by RPE cells through a MAPK-dependent pathway. The emerging role of danger signals in tissue activation suggests that extracellular nucleotides can participate in BRB breakdown during retinal inflammation. 

³  Contributed equally to the work and therefore should be considered equivalent authors.

 Supported by the Fonds de la Recherche en Ophtalmologie (FRO), the FRIA, and l’ASBL Vésale.

 Submitted for publication February 21, 2008; revised June 22 and July 29, 2008; accepted January 15, 2009.

 Disclosure: L. Judice M. Relvas, None; C. Bouffioux, None; B. Marcet, None; D. Communi, None; M. Makhoul, None; M. Horckmans, None; D. Blero, None; C. Bruyns, None; L. Caspers, None; J.-M. Boeynaems, None; F. Willermain, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

ATPγS, UTP, and UDP induced IL-8 production by ARPE-19 cells. (A) ARPE-19 cells were treated for 3 hours with 100 μM ATPγS, UTP, or UDP. RNA was then extracted and semiquantitative RT-PCR performed with primers specific for IL-8. Results are expressed as the increase (x-fold) of IL-8 mRNA expression over β-actin. (B) ARPE-19 cells were treated during 24 hours with or without 100 μM extracellular nucleotides. Supernatants were collected, and IL-8 was measured by ELISA. Results are expressed as induction ± SEM (x-fold) of IL-8 secretion versus the control. The data are representative of results in 2 (A) and 10 (B) representative experiments performed in triplicate and conducted independently. * P < 0.05, ** P < 0.01.

Figure 2.

 

ATPγS, UTP, and UDP upregulated TNFα induction of IL-8 secretion. ARPE-19 cells were treated for 24 hours with or without extracellular nucleotides (100 μM) and TNFα (4 ng/mL). Supernatants were collected and IL-8 was measured by ELISA. The results are expressed as the induction ± SEM (x-fold) of IL-8 secretion versus the control. The data are representative of three experiments performed in triplicate and conducted independently. ** P < 0.01.

Figure 3.

 

ATPγS, UTP, and UDP stimulate IL-8 production through ERK 1/2. (A) ARPE-19 cells were stimulated with ATPγS, UTP, or UDP (100 μM) during 5 and 30 minutes and phospho-ERK 1/2, which represents the active form of the protein, detected by Western blot analysis using a specific antibody. Total ERK 1/2 was used as the control. (B) ARPE-19 cells were preincubated with PD98059 (25 μM) for 1 hour and then treated during 24 hours with 100 μM extracellular nucleotides. Supernatants were collected, and IL-8 was measured by ELISA. The results are expressed as the induction ± SEM (x-fold) of IL-8 secretion versus the control. Data are representative of results in three experiments performed in triplicate and conducted independently. ** P < 0.01.

Figure 4.

 

Expression of P2Y receptors subtypes in ARPE-19 cells. RNA was extracted from confluent ARPE-19 cell culture. RT-PCR analysis (30 cycles) was performed using primers for P2Y₂, P2Y₄, and P2Y₆ receptors. PCR products were 439 bp for P2Y2 and 423 bp for P2Y6. Data are representative of two independent experiments. +RT, samples with reverse transcriptase; −RT, samples without reverse transcriptase.

Figure 5.

 

ATPγS, UTP, and UDP induce calcium release in ARPE-19 cells. After 2 days of culture, ARPE-19 cells were loaded with the calcium indicator Fura-2 and then directly stimulated with the indicated nucleotides. (A) ATPγS, (B) UTP, and (C) UDP. Calcium transients were determined with the 340/380 wavelength excitation ratio at an emission wavelength of 505 nm. Arrow (a) corresponds to nucleotide (100 μM) stimulation and arrow (b) corresponds to digitonin (0.5 mM) stimulation showing the maximum sustained increase in intracellular calcium. The results are representative of those in three experiments conducted independently.