Inhibition of Human Retinal Pigment Epithelial Cell Attachment, Spreading, and Migration by Alkylphosphocholines

Abstract

purpose. To investigate the effect of alkylphosphocholines (APCs) on human retinal pigment epithelium (RPE) attachment, spreading, migration, and microfilament assembly in vitro.

methods. Cultured RPE cells of five human donors were treated with one of four APCs (C18:1-PC, C20:1-PC, C21:1-PC, or C22:1-PC) in the presence of fetal calf serum. Cell viability was tested by the trypan blue exclusion assay. Attachment was assessed after a 2-hour incubation of RPE cells on coated 96-well-plates and subsequent MTT testing. Cellular spreading is characterized by cytoplasmic halo formation and was quantified by counting four separate fields of RPE cells allowed to spread on coated 24-well plates for 4 hours. Migration was measured by a modification of the Boyden chamber method in microchemotaxis chambers with polycarbonated filters. Microfilament assembly was assessed by immunofluorescence analysis after incubation with rhodamine-phalloidin.

results. All four APCs inhibited RPE cell attachment by more than 70% of their IC₅₀ (C18:1-PC: 30 μM; C20:1-PC: 10 μM; C21:1-PC: 10 μM; and C22:1-PC: 10 μM). Also, APCs inhibited RPE cell spreading by more than 80% and migration by more than 90% at similar concentrations. Trypan blue staining revealed a toxicity within control limits within the concentration interval tested. Microfilament organization was significantly disturbed after incubation of RPE cells with one of the four APCs close to its IC₅₀.

conclusions. APCs inhibit RPE cell attachment and spreading in vitro at nontoxic concentrations. As a possible mechanism of action, APCs disturb microfilament assembly, since they are known to interfere with protein kinase C (PKC) function. This could represent a novel method of preventing even early stages of proliferative vitreoretinal diseases like proliferative vitreoretinopathy (PVR).

Proliferative vitreoretinopathy (PVR) is a major cause of persistent reduction in visual acuity after retinal detachment or severe ocular trauma. ^{1 2 3} Clinically, PVR is characterized by the dispersion and proliferation of retinal pigment epithelial cells in the vitreous and the subsequent formation of membranes on both retinal surfaces. ^{4 5} The contraction of these epiretinal and subretinal membranes results in tractional retinal detachment or redetachment. 

Retinal pigment epithelial (RPE) cells seem to play a major role in PVR development (e.g., in retinal membrane formation and as a contractile force resulting in complex retinal redetachment). ^{6 7} The cascade of cellular events leading to PVR is cell mediated and stimulated by serum and/or a variety of growth factors. ^{8 9} Initial events in this cellular cascade are attachment to the retinal surface, spreading along the surface, and migration with subsequent formation of cellular membranes. ¹⁰ Later cellular events present as proliferation and cell-mediated contraction of retinal membranes resulting in redetachment of the retina. 

Pharmacologic prevention of PVR membrane formation has been based chiefly on antiproliferative and anti-inflammatory agents. ^{11 12 13 14} Although many of these substances lack efficacy and may even have toxic side effects, some of them seem to reduce PVR recurrence. ¹¹ However, no pharmacological adjuvant has been established as part of a routine treatment for PVR, as recent reviews of the current literature show. ^{15 16}  

In previous in vitro studies, we have shown that alkylphosphocholines (APCs) are effective inhibitors of human RPE cell proliferation and RPE-mediated collagen contraction at nontoxic concentrations. ¹⁷ APCs are a new class of pharmacological agents in routine clinical use with known antitumoral (Miltex; Zentaris GmbH, Frankfurt, Germany) and antiparasitic (Impavido; Zentaris GmbH) properties. ^{18 19} The mode of action of APCs is mediated by the cell membrane and involves the inhibition of protein kinase C in human RPE cells ¹⁷ as well as in other cellular systems in vitro. ^{20 21} The activation of PKC seems to be an important feature, not only in the late phase of PVR, as in RPE-mediated membrane contraction, ²² but also during the early cellular events of PVR development, as in cellular migration, ²³ adhesion, and spreading. ²⁴ In addition, PKC inhibitors have been implicated in the regulation of microfilament network organization. ²⁵ Thus, a selective inhibition of this pathway by APCs could offer a new perspective on prevention of PVR. 

The purpose of this study was to test whether APCs inhibit early cellular events in PVR pathogenesis, such as attachment, spreading and migration of human RPE cells in vitro and to assess microfilament assembly in the presence of serum. 

Materials and Methods

Alkylphosphocholines

The APCs oleyl-phosphocholine (C18:1-PC), (Z)-10-eicosenyl-phosphocholine (C20:1-PC), (Z)-12-heneicosenyl-phosphocholine (C21:1-PC), and erucyl-phosphocholine (C22:1-PC) were synthesized and kindly provided by Hansjoerg Eibl, PhD, (Göttingen, Germany), as described previously. ¹⁷  

All substances were dissolved in ethanol and stored at 4°C. Independent dilution series in ethanol were used to obtain final concentrations of APCs in equal volumes of ethanol. Equal volumes of ethanol without addition of APCs served as the control in all experiments performed. 

Human RPE Cell Culture

RPE cells were isolated from five human donors, age 20 to 71 years, within 4 to 14 hours after death as described previously ²⁶ and obtained from the Eye Bank of the Ludwig-Maximilians University Munich. Methods for securing human tissue were humane, included proper consent and approval, complied with the Declaration of Helsinki, and were approved by the local ethics committee. 

In brief, whole eyes were rinsed with 0.9% NaCl solution, immersed in 5% polyvinyl pyrrolidone iodine, and rinsed again with the sodium-chloride solution. After removal of the anterior segment, the posterior pole of each donor eye was examined with a binocular stereomicroscope to exclude any gross retinal disease. The neuroretina was then carefully peeled away from the RPE layer with a fine forceps. The eye cup was rinsed with Ca²⁺- and Mg²⁺-free Hank’s balanced salt solution and treated with 0.25% trypsin (Invitrogen-Gibco, Karlsruhe, Germany) for 1 hour at 37°C. After aspiration of trypsin, the eye cup was filled with Dulbecco’s modified Eagle’s medium (DMEM; Biochrom, Berlin, Germany) supplemented with 20% fetal calf serum (FCS; Biochrom). The medium was then gently aspirated with a pipette, to detach RPE into the medium without damaging Bruch’s membrane. 

The cell suspension was transferred to a 50-mL flask (Falcon; Corning-Costar, Wiesbaden, Germany) containing 20 mL of DMEM supplemented with 20% FCS and maintained at 37°C and 5% carbon dioxide. All cell culture dishes used were uncoated. Epithelial origin was confirmed by immunohistochemical staining for cytokeratin using a pancytokeratin antibody (Sigma-Aldrich, Deisenhofen, Germany) as described previously (data not shown). ²⁷ The cells were tested and found free of contaminating macrophages (anti-CD11, Sigma-Aldrich) and endothelial cells (anti-von Willebrand factor; Sigma-Aldrich; data not shown). 

Several primary cell cultures were grown to confluence. Subcultures between the third and sixth passages (passages 3–6) were maintained with DMEM containing 10% FCS at 37°C and 5% carbon dioxide. In the present study, RPE cells of five human donors were used when cells were actively proliferating. 

Cell Attachment Assay

Ninety-six well plates (Nunc, Wiesbaden, Germany) were coated with 70 μL/well fibronectin (50 μg/mL in PBS [pH 7.4]; Sigma-Aldrich) for 16 hours at 4°C. Unspecific binding was blocked by 2 mg/mL ovalbumin (Sigma-Aldrich) in PBS for 1 hour at 37°C as described previously. ²⁸  

RPE cells (passages 3–6) were seeded in 96-well plates (Nunc) at a density of 1 × 10⁶ cells per well in 1 mL DMEM, 10% FCS. APCs were added in concentrations spanning the 50% inhibitory concentration (IC₅₀) interval between 0.01 and 50 μM, as determined in preliminary assays. ¹⁷ The IC₅₀ is defined as the concentration of a drug that produces a 50% reduction in cell number at nontoxic concentrations. After 2 hours, RPE cells were then carefully washed three times with PBS using an automated plate washer (Molecular Devices, Garching, Germany). 

After the washing step, the number of attached RPE cells was determined by the tetrazolium dye-reduction assay [MTT (3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyl tetrazolium bromide]; Sigma-Aldrich). The MTT test was performed as described by Mosmann, ²⁹ with some modifications. In brief, 200 μL/well MTT solution (1.5 mL MTT stock [2 mg/mL in PBS] plus 28.5 mL DMEM) was added. RPE cells were incubated at 37°C for 1 hour and carefully observed to exclude crystal formation outside of cells. Formazan crystals formed were then dissolved by the addition of dimethyl sulfoxide (DMSO; 200 μL/well). The final concentration of DMSO in the cell culture medium was found to have no antiproliferative effect on RPE cells. Absorption was measured by a scanning multiwell spectrophotometer at 550 nm (Molecular Probes, Göttingen, Germany). The number of attached living cells correlated with the absorbance (optical density, OD) measured by the MTT at 550 nm. Results from the wells were expressed as the mean percentage of the control (control OD 0.6 at 550 nm assigned as 100%). Experiments were performed in triplicate and repeated four times. RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control. 

Cell Spreading Assay

For analysis of cell spreading, 24-well plates (Nunc) were coated with 70 μL/well fibronectin (50 μg/mL in PBS, pH 7,4; Sigma-Aldrich) for 16 hours at 4°C. Unspecific binding was blocked by 2 mg/mL ovalbumin (Sigma-Aldrich) in PBS for 1 hour at 37°C, as described previously. ¹⁰ Human RPE cells (passages 3–6) were seeded at a density of 5 × 10⁴ cells per well on the fibronectin-coated wells in 1 mL DMEM and 10% FCS. APCs were added in concentrations spanning the 50% inhibitory RPE cells, as determined previously. ¹⁷ RPE cells were allowed to spread for 4 hours at 37°C and 5% CO₂. After they were washed with PBS three times, cells were fixed with 4% paraformaldehyde in PBS, and then stained with Coomassie brilliant blue for 3 to 4 minutes and rinsed again with PBS. For quantification of cell spreading, four separate fields were photo-documented with a phase contrast microscope (Leica, Heidelberg, Germany). Cell spreading was characterized by the formation of a clearly defined cytoplasmic halo around the cell nucleus. RPE cells demonstrating this phenomenon were counted and the percentage of the total number of RPE cells per field was calculated. Results from the wells were expressed as the mean percentage of the control. Experiments were performed in triplicate and repeated four times. RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control. 

Cell Migration Assay

Migration was determined by a modification of the Boyden’s chamber method using microchemotaxis chambers (NeuroProbe, Gaithersburg, MD) and polycarbonate filters (Nucleopore, Karlsruhe, Germany) with a pore size of 8.0 μm. ³⁰ The filters were coated with fibronectin and placed between the chambers. The lower half of the Boyden chamber was filled with 125 μL DMEM containing 20 ng/mL human recombinant platelet-derived growth factor (PDGF-BB; PeproTech, London, UK). Human RPE cells (passages 3–6) were then suspended at a density of 5 × 10⁵ cells/mL in DMEM and 0.5% FCS and placed in the upper half of the Boyden chamber (500 μL). APCs were added in concentrations spanning the 50% inhibitory RPE cells, as determined previously. ¹⁷ The cells were incubated 6 hours at 37°C in 5% CO₂. After the upper side of the filter was scraped with a cotton tip to eliminate RPE cells that had not migrated through it, the filter was removed and fixed in methanol before staining with hematoxylin. Cell number in five fields chosen randomly was determined at a magnification of ×200 using a phase contrast microscope (Leica). Experiments were performed in triplicate and repeated four times. RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control. 

Immunofluorescence Analysis of Microfilaments

Microfilament organization of human RPE cells (passages 3–6) was assessed by a modification of the immunofluorescence analysis protocol with rhodamine-phalloidin (AlexaFluor 488 phalloidin A 12379; Molecular Probes, Inc.), as described elsewhere. ²⁵ Cells were seeded in chamber slides (2 × 10⁴ cells/well in DMEM and 10% FCS). After incubation for 4 hours at 37°C in 5% CO₂, the medium was aspirated and adherent cells were fixed with 3.7% paraformaldehyde in PBS for 10 minutes at 20°C. After they were washed with PBS (pH 7.4), the cells were permeabilized with 0.1% Triton X-100 for 3 minutes and blocked with 1% BSA (bovine serum albumin) in PBS. Cells were then incubated with rhodamine-phalloidin (200 U/mL) and Hoechst stain (1 μg/mL) for 20 minutes in the dark. PBS was used as the base of all solutions and intervening rinses, and incubations were performed at room temperature. After mounting (Kaiser’s Glyceringelatine; Merck, Darmstadt, Germany), slides were examined under an epifluorescence microscope with appropriate excitation and emission filters under 40× magnification. Pictures of observed fields were recorded digitally. Experiments were repeated three times. RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control. 

Statistical Evaluation

Statistical analysis was performed (SPSS V 11.0; SPSS Inc., Chicago, IL). All results are expressed as the mean ± SEM as indicated. For determination of the significance of differences an ANOVA was performed. Differences with P < 0.01 were considered statistically significant. 

Results

Inhibition of RPE Cell Attachment by APCs

All APCs investigated inhibited attachment of human retinal pigment epithelium cells on fibronectin-coated cell culture dishes in a dose-dependent manner in vitro (Fig. 1) . At the same concentration, however, APCs with a longer carbon chain of more than C20 were more effective than C18:1-PC. This finding correlates well with the difference in the IC₅₀ concentration determined previously ¹⁷ (Table1). Consequently, the APC with the highest IC₅₀ of approximately 30 μM, C18:1-PC, has the lowest efficacy in the inhibition of RPE cell attachment compared with those of C20:1-PC, C21:1-PC, and C22:1-PC with an IC₅₀ of only approximately 10 μM. 

Oleyl-phosphocholine (C18:1-PC) displayed an inhibition of RPE cell attachment at concentrations between 0.1 and 50 μM (Fig. 1) . Cells exposed to C18:1-PC at concentrations below 0.1 μM were not significantly compromised in their ability to attach compared with the control. Close to its IC₅₀ of 26.5 μM ± 1.3 (Table 1) , C18:1-PC inhibited attachment of more than 70% of treated RPE cells. At 10 μM, 50% of RPE cells did not attach. 

All APCs with a carbon chain length of more than C20, eicosenyl-phosphocholine (C20:1-PC) heneicosenyl-phosphocholine (C21:1-PC) and erucyl-phosphocholine (C22:1-PC), inhibited RPE cell attachment at concentrations between 0.01 and 30 μM (Fig. 1) . Close to their IC₅₀ of approximately 10 μM, respectively, more than 70% of RPE cells were unable to attach was was shown for C18:1-PC at its IC₅₀. However, a 50% inhibition of cell attachment was achieved for all APCs with a carbon chain length of more than 20 at a concentration of only 1 μM. 

Inhibition by APCs of RPE Cell Spreading

All four APCs applied caused a significant concentration-dependent inhibition of RPE cell spreading on fibronectin-coated cell culture dishes (Fig. 2) . The data suggest a difference in efficacy of the APCs in parallel to that demonstrated for RPE cell attachment. Again, close to their IC₅₀, all four APCs were equally effective, achieving an inhibition of RPE spreading in more than 80% of cells. The concentration necessary to induce a spreading inhibition in 50% of cells was even more than 10 times higher for C18:1-PC (10 μM) than for all other APCs applied (0.1–1 μM). This corresponds well with the findings for RPE cell attachment (Table 1) . 

Inhibition of RPE Cell Migration by APCs

RPE cell migration was inhibited by all four APCs applied in a dose-dependent manner (Fig. 3) . Close to their IC₅₀, respectively, each APC inhibited migration in more than 90% of RPE cells (Table 1) . APCs with a longer carbon chain of more than C20 were again far more effective at equal concentrations than was C18:1-PC. The concentration achieving a 50% inhibition of RPE cell migration was 0.1 to 1 μM for C20:1-PC, C21:1-PC, and C22:1-PC compared with 10 μM for C18:1-PC. 

Disturbance of Microfilament Dynamics by APCs

F-actin assembly in human RPE cells was significantly disturbed after incubation with each one of the four APCs applied close to its IC₅₀ (Fig. 4c 4d 4e 4f 4g 4h 4i 4j ; C18:1-PC, 30 μM: Figs. 4c 4d ; C20:1-PC, 10 μM: Figs. 4e 4f ; C21:1-PC, 10 μM: Figs. 4g 4h ; C22:1-PC, 10 μM: Figs. 4i 4j ). Control cells displayed a well organized actin skeleton with fibers extending throughout the cytoplasm into the cell membrane (Figs. 4a 4b) . 

Discussion

On the cellular level, the development of PVR can be described as a cascade of specific cellular events within the retina and on both retinal surfaces involving the vitreous. Initial cellular events in early PVR can be defined as attachment of RPE cells on the retinal surfaces followed by cytoplasmic spreading and cellular migration along these surfaces. Later stages in PVR pathogenesis are characterized by cellular proliferation and membrane formation with subsequent cell-mediated membrane contraction and tractional redetachment of the retina. In a previous in vitro study, we could demonstrate that APCs are able to inhibit the late stages in PVR development associated with RPE proliferation and cell-mediated membrane contraction. ¹⁷ In this study, APCs were found to be effective inhibitors of early cellular events in PVR pathogenesis like RPE cell attachment, spreading, and migration at nontoxic concentrations. 

We are aware of the limitations of our results by using fibronectin alone as a substrate for the investigation of key cellular events in early PVR. Alternatively, we and others ^{31 32} demonstrated that RPE cells show a clear preference for fibronectin over other provisional extracellular matrix (ECM) components such as collagen type I and laminin in early PVR. ³³ Therefore, we believe that the inhibition of RPE cell attachment by APCs on fibronectin is a good example for the potential of these substances. Even if APCs were insufficient in their inhibition of these key cellular events on other provisional ECM substrates, their effect on fibronectin would interfere with early PVR development significantly. 

Our data suggest that the effect of APCs on the key cellular events in early PVR is due to an interference of these substances with the microfilament assembly in human RPE cells. The inhibition of PKC activity by APCs ¹⁷ seems to be linked with the microfilament assembly in these cells resulting in a decreased ability of the RPE to attach, spread and migrate. ^{23 25} Thus, the inhibition of the initial stages in PVR development could prevent a disease progression to the more advanced stages commonly associated with a severe reduction in visual acuity. 

In the present in vitro study, we have shown that APCs can inhibit even early cellular events in PVR development at concentrations below the IC₅₀ ¹⁷ in a dose-dependent manner in the presence of serum. The dispersion and accumulation of serum components in the vitreous cavity is known to be a crucial event for in vivo PVR development due to a breakdown of the blood–retinal barrier. ^{34 35} Serum has been identified as a strong stimulator of RPE proliferation ³⁶ and contains fibrogenic (basic fibroblast growth factor; bFGF) and inflammatory (interleukin 6; IL-6) growth factors. These factors have recently been found to be concentrated in the retro-oil fluid after complicated retinal detachment surgery prompting the development of PVR after surgery. ³⁷ A new therapeutic approach for the treatment of PVR should therefore be able to act in the presence of these growth factors. 

Other pharmacologic agents currently in discussion for PVR treatment have not been effective in the presence of serum and, moreover, are not able to influence all the cellular events mentioned. For example, the most promising combination of 5-fluorouracil (5-FU) and low-molecular-weight heparin (LMWH) has been shown to inhibit RPE proliferation only. ³⁸ Clinically, there was no statistically significant difference in visual outcome of treated versus untreated patients. ³⁴ Daunomycin, another candidate for PVR treatment, inhibits cellular proliferation and migration, but the effective dose seems to be close to the toxic one which has prevented a routine clinical application so far. ^{39 40} Also, a combination therapy of dexamethasone and heparin has not shown statistically significant results regarding PVR recurrence with at the same time an increased incidence of postoperative vitreous hemorrhage in the treated group. ⁴¹ Because of a lack of efficacy and side effects of these substances, none of them is currently in routine clinical use for the prevention and/or treatment of PVR. 

In summary, our results indicate that APCs are able to inhibit not only the late stages of PVR development like RPE cell proliferation and cell-mediated membrane-contraction but also the early cellular events like RPE cell attachment, spreading and migration in a dose-dependent manner at nontoxic concentrations. As a mechanism of action, APCs seem to interfere with microfilament assembly in human RPE cells by inhibition of PKC. The first intraocular in vivo toxicity study of erucylphosphocholine in a rat model displayed no retinal toxicity even at the 10-fold IC₅₀ when administered intravitreally. ⁴² Therefore, APCs may become a new treatment option for prevention of PVR. 

 Supported by Deutsche Forschungsgemeinschaft (DFG) Grant WE 2577/2-1.

 Submitted for publication May 25, 2005; revised August 31, 2005; accepted November 23, 2005.

 Disclosure: K.H. Eibl, None; D. Kook, None; S. Priglinger, None; C. Haritoglou, None; A. Yu, None; A. Kampik, None; U. Welge-Lussen, 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.

 

Inhibition of RPE cell attachment by APCs measured by a colorimetric test (MTT). RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control. Results was expressed as the mean percentage of the control (control OD 0.6 at 550 nm was deemed 100%). Data are the mean results of four experiments, each performed in triplicate. Error bars, SEM.

Table 1.

 

 

Summary of the Inhibitory Effects of APCs on Early Cellular Events of PVR Development In Vitro

Table 1.

 

Summary of the Inhibitory Effects of APCs on Early Cellular Events of PVR Development In Vitro

APC	IC50 Concentration (μM ± CI)	Cell Attachment (% ± SEM)	Cell Spreading (% ± SEM)	Cell Migration (% ± SEM)
C18:1-PC	26.5 ± 1.3	19.0 ± 2.2	17.3 ± 2.2	9.0 ± 2.0
C20:1-PC	11.0 ± 0.7	15.0 ± 2.2	19.0 ± 0.6	4.0 ± 1.0
C21:1-PC	9.0 ± 0.9	20.7 ± 2.4	19.0 ± 2.2	4.0 ± 1.0
C22:1-PC	8.5 ± 2.3	24.3 ± 2.1	20.7 ± 0.5	4.0 ± 1.0

 Effect of the 50% inhibitory concentration (IC₅₀)¹⁷(in μM ± 95% CI) on RPE cell attachment, spreading, and migration for each APC, respectively (% control ± SEM).

Figure 2.

 

Human RPE cell spreading inhibition by APCs (a). Data are the mean results of four experiments each performed in triplicate. Error bars, SEM. Phase-contrast microscope photographs of control (b; arrows: halos) and treated human RPE cells (10 μM C22:1-PC; c; arrows: cells without halos) RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control. Magnification, ×100.

Figure 3.

 

Human RPE cell migration inhibition by APCs. Migration was assessed by a modification of the Boyden’s chamber method. Number of cells in five fields chosen randomly was assessed at a magnification of ×200 with a phase-contrast microscope. Experiments were performed in triplicate and repeated four times. RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control.

Figure 4.

 

Microfilament organization of human RPE cells (passages 3–6) was assessed by immunofluorescence analysis with rhodamine-phalloidin. Cells were seeded in chamber slides, and APCs were added close to their IC₅₀(Table 1) . After incubation for 4 hours, adherent cells were fixed, washed, and incubated with rhodamine-phalloidin and Hoechst stain. Slides were examined under an epifluorescence microscope with appropriate excitation and emission filters. Images of observed fields were recorded digitally. Control cells displayed an elaborate network of precisely organized F-actin filaments (a, b; arrows). The F-actin filament architecture becomes disorganized with massive disruption of filament bundles after exposure to APCs in nontoxic concentrations (c–j; arrows). Experiments were repeated three times. RPE cells of the same passage number incubated with equal volumes of ethanol without addition of APCs served as the control (a, b: control; c, d: C18:1-PC, 30 μM; e, f: C20:1-PC, 10 μM; g, h: C21:1-PC, 10 μM; i, j: C22:1-PC, 10 μM).

 

The authors thank Katja Obholzer for expert technical assistance.