Human CMV (HCMV) retinitis frequently leads to blindness in iatrogenically immunosuppressed patients and in the end stage of AIDS. Despite the general proinflammatory potential of HCMV, virus infection is associated with a rather mild cellular inflammatory response in the retina. To investigate this phenomenon, the influence of HCMV (strains AD169 or Hi91) infection on C-X-C chemokine secretion, ICAM-1 expression, and neutrophil recruitment in cultured human retinal pigment epithelial (RPE) cells was studied. Supernatants from infected cultures contained enhanced levels of IL-8 and melanoma growth-stimulating activity/Gro α and induced neutrophil chemotaxis compared with supernatants from uninfected RPE cells. Despite HCMV-induced ICAM-1 expression on RPE cells, binding of activated neutrophils to HCMV-infected RPE cells and subsequent transepithelial penetration were significantly reduced. Reduced neutrophil adhesion to infected RPE cells correlated with HCMV-induced up-regulation of constitutive Fas ligand (FasL) expression. Functional blocking of FasL on RPE cells with the neutralizing mAbs NOK-1 and NOK-2 or of the Fas receptor on neutrophils with mAbB-D29 prevented the HCMV-induced impairment of neutrophil/RPE interactions. Fas-FasL-dependent impairment of neutrophil binding had occurred by 10 min after neutrophil/RPE coculture without apoptotic signs. Neutrophil apoptosis was first detected after 4 h. Treatment of neutrophils with a specific inhibitor of caspase-8 suppressed apoptosis, whereas it did not prevent impaired neutrophil binding to infected RPE. The current results suggest a novel role for FasL in the RPE regulation of neutrophil binding. This may be an important feature of virus escape mechanisms and for sustaining the immune-privileged character of the retina during HCMV ocular infection.

Disseminated human CMV (HCMV)3 infection is a life-threatening infection, e.g., in AIDS patients. A significant number of these patients develop HCMV retinitis (1, 2). The pathological features of HCMV retinitis include transmission of virus from retinal capillaries and necrosis of the retinal layers, causing retinal detachment and blindness in untreated patients (3, 4). Infection of the retinal pigment epithelium (RPE) as an important part of the blood-retina barrier is suspected to be crucial in the development of inflammatory disease in the retina (1, 3, 5, 6).

Because HCMV infection is known to modulate the immunological status of its host cells by altering the expression of cellular genes coding for proinflammatory proteins, it is possible that HCMV-infected RPE cells support retinal inflammation. Moreover, HCMV infection of other cell types induced the up-regulation of several binding molecules, such as ICAM-1 or LFA-3 (7, 8), C-X-C chemokines IL-8 and melanoma growth-stimulating activity (MGSA)/Gro α (9, 10, 11, 12), and C-C chemokines such as RANTES (13). These changes in the expression of cellular proinflammatory molecules may result in infiltration of the infected tissue with leukocytes and thus lead to HCMV-associated inflammatory effects.

Despite these generally accepted proinflammatory properties, HCMV infection mostly elicits a rather mild cellular inflammatory response in the retina (14), which is usually characterized by a sparse infiltrate of mononuclear cells in both naturally and iatrogenically immunosuppressed patients. Lymphocytes are the predominant cell type, and approximately 22–50% of patients with AIDS and HCMV retinitis have foci of neutrophilic infiltrates in retinal tissue at autopsy (5, 14). A possible explanation for the weak inflammatory responses might be the severe immunodeficiency in most patients with HCMV retinitis. Alternatively, the eye has an immune-privileged status (15, 16, 17) that may be sustained by various features such as local secretion of immunosuppressive factors or expression of Fas ligand (FasL), as found in different parts of the eye, including the retina (17, 18, 19, 20). However, it is unclear whether HCMV-infected RPE cells evoke neutrophil immune responses or cause local immune deviation. As a result, the present study investigated neutrophil attraction and adhesion in a cell culture model using mock-infected and HCMV-infected human RPE cell lines.

HUVECs and human foreskin fibroblasts (HFF) were cultured as described previously (21). RPE cells were isolated from freshly enucleated bulbi for corneal transplantation of three donor eyes; the tenets of the Declaration of Helsinki were followed. RPE isolation and culture were performed as described previously with some modifications (22). Briefly, the corneoscleral disc was first removed, followed by the lens and vitreous. The residual eye cup was sectioned with a longitudinal incision toward the optic nerve. Repeated rinsing with Ca2+ and Mg2+ Dulbecco’s PBS allowed prompt separation of the remaining vitreous and neural retina from the layer of RPE and permitted detachment of the choroid from the sclera. The RPE cells adhering to Bruch’s membrane on the choroidal sheets obtained were washed with PBS and treated with 0.25% trypsin-EDTA solution. Detached cells were resuspended in IMDM, supplemented with 20% FBS, and transferred to 25-cm2 flasks. The homogeneity of cultured RPE cells was confirmed by positive immunostaining with mAb to cytokeratins (Pan) and to cellular retinaldehyde binding protein (mAbs were donated by J. Saari, Department of Ophthalomology, University of Washington, School of Medicine, Seattle, WA) (23). The cell cultures used in this study were designated RPE-I, RPE-II, and RPE-III. Cells were routinely tested for mycoplasma and were not used in the experiments later than passage 3.

Neutrophils were isolated from the venous blood of healthy adult volunteers by centrifugation on granulocyte separation medium (Polymorphprep, Nycodens) and immediately used for the experiments. The purity of the neutrophils was >95%, and the viability was >99% as determined by trypan blue exclusion.

The HCMV laboratory strain AD169 was obtained from American Type Culture Collection (Manassas, VA). Strain Hi91 was isolated from urine of an AIDS patient with HCMV retinitis (24). Virus stocks were prepared in HFF grown in MEM with 4% FBS. The respective titers were determined by plaque titration in HFF cells as described previously (21). Mock-infected inocula were prepared in an identical fashion, except that cell monolayers were not infected with HCMV.

Inactivation of virus was achieved by exposure of virus solution to UV light (220 V, 12 W) for 15 min (9). Samples of irradiated virus were then used to infect RPE cell cultures. UV-irradiated samples were free of infectious virus as demonstrated by plaque titration (not shown).

Filtered virus inocula were prepared by filtering virus stocks through a Microsep microconcentrator with a cut-off at 300,000 m.w. (Filtron Technology, Northborough, MA) at 3,000 × g for 1 h at 4°C. The filtrate collected from the bottom of the filter apparatus was added to RPE cell cultures. The filtrate samples were free of infectious virus as demonstrated by plaque titration.

Confluent cultures of RPE cells were incubated with HCMV at a multiplicity of infection of 10. After incubation for 1 h, which was required for virus adsorption, cells were washed with PBS and incubated in maintenance medium containing 4% FBS. As described in detail previously (21), cells producing HCMV-specific Ags were detected 24 and 72 h postinfection (p.i.) by immunoperoxidase staining using mAbs directed against 72-kDa immediate early Ag (IEA; DuPont, Bad Homburg, Germany) and 67 kDa late Ag (LA), respectively. For control purposes an irrelevant Ab directed against HSV glycoprotein B was used.

Ganciclovir (GCV; Hoffman-La Roche, Grenzach-Wyhlen, Germany) was prepared in distilled water. ISIS 2922, a phosphorothioate oligonucleotide that is complementary to HCMV IE mRNA and the noncomplementary control oligonucleotide ISIS 26062 as well as the FITC-conjugated variants were provided by ISIS Pharmaceuticals (Carlsbad, CA). Both ISIS 2922 and ISIS 26062 were dissolved in PBS at a concentration of 10 mM, and aliquots were stored at −20°C until use. To enhance oligonucleotide uptake, ISIS 2922 or ISIS 26062 was complexed to cationic liposomes (DOTAP, Roche, Mannheim, Germany) immediately before virus infectivity assay. The mixing of oligonucleotides with DOTAP was performed according to the manufacturer’s instructions. To visualize the uptake and confirm the stability of ISIS 2922 and ISIS 26062 during the experiments, the same procedure was conducted with FITC-conjugated oligonucleotides (25). For the antiviral assays cell monolayers were pretreated with GCV, ISIS 2922, or ISIS 26062 overnight in MEM containing 0.2% FBS and then were washed three times with PBS before infection with HCMV. After 1 h of incubation, virus was removed, and fresh medium containing GCV, ISIS 2922, or ISIS 26062 was added. The number of cells producing viral Ags was determined 24 and 72 h after infection by immunoperoxidase staining. IEA were detected using the mAb Mab810 (Chemicon, Hofheim, Germany), which binds to a shared epitope at the amino termini of the IE55, IE86, and IE72 proteins. LA were detected using the mAb directed against 67-kDa LA (DuPont, Bad Homburg, Germany).

Amounts of IL-8 and MGSA/Gro α in culture supernatants were assessed by quantitative sandwich enzyme immunoassays (ELISA; R&D Systems, Wiesbaden Nordenstadt, Germany) according to the manufacturer’s instructions. Supernatants harvested from mock- or HCMV-infected RPE cultures at various times p.i. were stored at −80°C until measurement. As positive controls, standards provided by the manufacturer were used. Fresh culture medium was used as a negative control. OD was determined with a microplate reader set at 450 nm.

To investigate the expression of Fas, FasL, or ICAM-1, 5 × 105 cells were fixed for 10 min in 4% buffered formaldehyde. After washing the cells twice in PBS containing 0.5% Tween 20, cells were incubated for 30 min with mAbs against FasL (clones NOK-1 and NOK-2; PharMingen, Heidelberg, Germany) or ICAM-1 (clone BBIG-I1; R&D Systems). Cell pellets were washed twice and incubated with FITC-conjugated goat anti-mouse IgG (Becton Dickinson, Heidelberg, Germany) for 30 min. For control purposes, cells were stained with an irrelevant primary Ab (isotype) or without a primary Ab to determine unspecific and background fluorescence, respectively. Instrument settings of the flow cytometer (FACScan, Becton Dickinson) were adjusted to obtain background mean fluorescence in the histogram mode between 1 and 10 on the logarithmic scale. Data were analyzed using CellQuest software. All experiments were repeated at least three times.

Total RNA was isolated from tissue samples of neurosensory retina (mock- or AD169-infected) using TRIzol according to the manufacturer’s instructions (Life Technologies, Gaithersburg, MD). RNA was reverse transcribed using random hexamer priming. One microgram of total RNA was denatured at 70°C for 10 min and chilled on ice. The denatured RNA was then coincubated with 2.5 μM random hexamer oligonucleotides, 1 μM of each dNTP, 5 mM MgCl2, 1 μl of RNase inhibitor (Roche), and 1 μl of Moloney murine leukemia virus reverse transcriptase (Life Technologies) in 1× PCR buffer II (Perkin-Elmer, Norwalk, CT) for 1 h at 37°C. The RT was inactivated for 5 min at 95°C before amplification. FasL primers used were: FasL sense, 5′-ATG CAG CAG CCC TTC AAT TAC-3′ (position 86–106); and FasL antisense, 5′-GCC TCT GGA ATG GGA AGA CAC C-3′ (position 559–580) (26). The sequence of the GAPDH primers used as the control were 5′-TGG GGA AGG TGA AGG TCG GA-3′ (position 61–81) and 5′-GAA GGG GTC ATT GAT GGC AA-3′ (position 151–171) (24). PCR amplification of the cDNA was conducted by adding 0.5 μg of Taq DNA polymerase (Roche). PCR amplification of FasL fragment was performed using 30 cycles in a DNA thermocycler with denaturation for 1 min at 94°C, annealing for 1 min at 60°C, and extension for 1 min at 72°C, whereas conditions for amplification of GAPDH fragment were denaturation for 1 min at 94°C, annealing for 1 min at 52°C, and extension for 1.5 min at 72°C in a Perkin-Elmer Thermocycler. PCR products were resolved alongside DNA marker on an agarose gel, stained with ethidium bromide, and photographed. The photographs were further analyzed by scanning densitometry using the E.A.S.Y. RH system (HeroLab, Wiesloch, Germany), and the ratio of FasL/GAPDH band intensity was calculated. To ascertain that FasL transcripts were specifically amplified, sequence analysis of PCR products was performed. Amplified sequences fully matched FasL nucleotide sequences (results not shown).

Supernatant from mock- or HCMV-infected RPE was harvested 4, 24, or 72 h after infection and immediately assessed for neutrophil chemotactic activity in a transendothelial migration assay. HUVECs were seeded onto 6.5-mm diameter Transwell filters (Becton Dickinson, Mountain View, CA) with a pore size of 3 μm. The formation of confluent monolayers was confirmed the following day by microscopic examination. The Transwell filters with the endothelial cell monolayers were then washed and placed in six-well plates (Becton Dickinson). Wells contained 1) fresh medium (negative control), 2) medium plus FMLP (Sigma; 1 × 10−7 M) as a positive control, 3) supernatants from mock-infected RPE cells, and 4) supernatants from HCMV-infected RPE cells. Neutrophils (0.5 × 106) were added to each upper chamber, and the wells were incubated for 60 min in a humidified atmosphere at 37°C. Subsequently, Transwell filters were removed, and neutrophils in the lower chamber were counted microscopically, including neutrophils that were attached to the inverse side of the filter. The latter were obtained by swabbing off the cells with a cotton wool tip. The integrity of the endothelial monolayer was microscopically confirmed before and after termination of the experiments. In some experiments the functional relevance of IL-8 and MGSA/Gro α was determined. Therefore, 20 μg/ml neutralizing Abs against IL-8, MGSA/Gro α, or irrelevant IgG isotype (all from R&D Systems) were added to the lower chamber and incubated for 30 min at 37°C before addition of Transwell filters to the wells. The Abs were present throughout the transmigration assays.

RPE cells were transferred to round coverslips, treated with 3-aminopropyl-triethoxy-silan (2%; Sigma, Munich, Germany) and placed into six-well multiplates. When confluence was reached, 0.5 × 106 neutrophils/well were added to the RPE monolayer for 10, 30, or 60 min. Neutrophils were stimulated with 1 × 10−6 M PMA (Sigma) to allow activation of the LFA-1 adhesion ligand, which is necessary for binding via ICAM-1 receptor according to Dustin and Springer (27). Nonadherent neutrophils were washed off using warm (37°C) IMDM. The remaining cells were fixed with 1% glutaraldehyde (Merck, Darmstadt, Germany). Adherent neutrophils were counted in five different fields (5 × 0.25 mm2) using a phase contrast microscope (×20 objective).The mean cellular adhesion rate was obtained by calculating the mean of five countings.

The ability of adherent neutrophils to penetrate under the RPE cell monolayer (transepithelial migration) was studied by means of a reflection interference contrast microscope (Leitz, Wetzlar, Germany) with a Ploem apparatus (×100 oil immersion objective). This method has been described in detail previously (28). Images were visualized and amplified using a Proxitronic CCD camera (Proxitronic, Bensheim, Germany), and the number of penetrated neutrophils was quantified by the image-analyzing system ARGUS 20 (Hamamatsu, Hersching, Germany). To optimize the signal/noise ratio, online background subtraction and averaging of eight images were performed by using the image-processing system QUANTIMET Q 520 (Cambridge Instruments, Bensheim, Germany). Cells were counted in five different fields, and the mean penetration rate was obtained by calculating the mean of five countings.

RPE cells were pretreated for 60 min at 37°C with mAbs against FasL molecules (clones NOK-1 and NOK-2) or with mAb against ICAM-1 (clone BBIG-I1). Alternatively human neutrophils were treated for 60 min with mAb against CD95 (CD95 antagonist; clone B-D29) purchased from Laboserv (Staufenberg, Germany). Irrelevant Abs of the same isotype were used as controls. All Abs were used at a concentration of 5 μg/ml.

Apoptosis of neutrophils incubated with mock- or HCMV-infected RPE cells was determined using the Apoptosis Detection Kit (R&D Systems). The basis of this kit is the ability of living cells to expel propidium iodide (PI) and the inability to bind annexin V. Neutrophils (0.5 × 106/ml) were added to RPE cells for various time periods. Subsequently, neutrophils were washed off and incubated with 0.25 μg/ml FITC-conjugated annexin V and 10 μl of PI and analyzed by means of flow cytometry. Three subpopulations were identified: viable cells (FITC/PI), apoptotic cells (FITC+/PI), and necrotic cells (FITC+/PI+). In addition, the extent of apoptotic neutrophils incubated with mock- or HCMV-infected RPE cells was determined by measurement of cytoplasmic histone DNA fragments (mono- and oligonucleosomes). Measurements were performed by photometric enzyme immunoassay with specific mAbs directed against DNA and histones using the cell death detection ELISA (Roche) according to the manufacturer’s instructions. The specific DNA fragmentation (enrichment of mono- and oligonucleosomes released into the cytoplasm) was calculated using the following formula: absorbance of the sample − absorbance of the medium/absorbance of the control − absorbance of the medium. Neutrophils obtained immediately after preparation from blood samples (before incubation in medium or in cocultures with mock- or HCMV-infected RPE cells) were used as controls. In some experiments neutrophils were pretreated with 100 μM caspase-8-like Cbz-Ileu-Glu-Thr-Asp(Ome)-fluoro-methylketone (zIETD) caspase inhibitor (Enzyme Systems Products, Livermore, CA) 30 min before coculture.

Determination of statistical significance was conducted with Student’s t test. Data groups were considered significant when p < 0.05.

The purity of the RPE cell cultures was >99% as confirmed by immunostaining for pan-cytokeratin and cellular retinaldehyde binding protein (Fig. 1). Primary and low passage cultures used in this study contained morphologically different cell types, including large stationary, nondividing cells with a high concentration of melanolipofuscin granules and small hexagonal epithelial-shaped dividing cells with a significantly lower concentration of granules that diluted with repeated cell division and were visible up to three passages. Confluent layers of RPE cells were infected with two different HCMV laboratory strains (AD169 and Hi91). Infection was evaluated by the assessment of cell numbers expressing IEA or LA in cultures infected at a multiplicity of infection of 10 (Fig. 2). RPE cell cultures I, II, and III differed in sensitivity to HCMV infection, ranging from 15 to 51% and from 13 to 41% positive cells for IEA and LA, respectively. Moreover, all lines exhibited decreased sensitivity to HCMV after three subcultures (data not shown), and therefore, only RPE cells up to three passages were used in additional experiments. Overall, no significant difference between the permissiveness of RPE for AD169 and Hi91 was found.

FIGURE 1.

Photomicrographs of RPE-positive immunoperoxidase staining for pan-cytokeratin (A) and cellular retinaldehyde-binding protein (B). Cells did not stain positively for FVIIIvWr Ag, a specific endothelial cell marker used as a negative control (C). Note the typical dark pigment granules in the cytoplasm of RPE cells.

FIGURE 1.

Photomicrographs of RPE-positive immunoperoxidase staining for pan-cytokeratin (A) and cellular retinaldehyde-binding protein (B). Cells did not stain positively for FVIIIvWr Ag, a specific endothelial cell marker used as a negative control (C). Note the typical dark pigment granules in the cytoplasm of RPE cells.

Close modal
FIGURE 2.

Photomicrographs of HCMV AD169-infected (MOI 10) RPE-I cells expressing HCMV specific nuclear IEA (A) and cytoplasmic LA (B). Cultures were stained (red-brown color) for IEA and LA 24 and 72 h p.i., respectively. As a negative control, staining was also conducted with an irrelevant Ab recognizing HSV glycoprotein B 24 h (C) and 72 h (D) p.i. Note the typical dark pigment granules in the cytoplasm of RPE cells.

FIGURE 2.

Photomicrographs of HCMV AD169-infected (MOI 10) RPE-I cells expressing HCMV specific nuclear IEA (A) and cytoplasmic LA (B). Cultures were stained (red-brown color) for IEA and LA 24 and 72 h p.i., respectively. As a negative control, staining was also conducted with an irrelevant Ab recognizing HSV glycoprotein B 24 h (C) and 72 h (D) p.i. Note the typical dark pigment granules in the cytoplasm of RPE cells.

Close modal

Chemokine secretion was analyzed in supernatants of mock- and HCMV-infected RPE cultures I–III at different times p.i. In Fig. 3, data from RPE-I are representatively shown. Both IL-8 and MGSA/Gro α secretions were significantly increased 4 h p.i. (mock, 555 ± 56 pg/ml; AD169, 1210 ± 145 pg/ml; Hi91, 1455 ± 135 pg/ml) and (mock, 61 ± 6.2 pg/ml; AD169, 137 ± 12 pg/ml; Hi91, 122 ± 11 pg/ml), respectively. Enhanced levels were sustained for 24 and 72 h p.i. (Fig. 3, A and B).

FIGURE 3.

Time course of IL-8 (A) and MGSA/Gro α (B) secretion and of ICAM-1 expression (C) 4, 24, and 72 h after infection of RPE-I with HCMV strain AD169. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with mock-infected cells.

FIGURE 3.

Time course of IL-8 (A) and MGSA/Gro α (B) secretion and of ICAM-1 expression (C) 4, 24, and 72 h after infection of RPE-I with HCMV strain AD169. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with mock-infected cells.

Close modal

Flow cytometric analyses of RPE cells that were stained with fluorescence-conjugated mAb against ICAM-1 revealed that mAb-specific fluorescence intensity units (FIU; Fig. 3 C) were augmented as early as 4 h p.i. (mock, 168 ± 25 FIU; AD169, 249 ± 33 FIU; Hi91, 255 ± 35 FIU; p < 0.05 vs mock). Maximum values were found 72 h p.i. (mock, 141 ± 21 FIU; AD169, 452 ± 68 FIU; Hi91, 415 ± 59 FIU; p < 0.05 vs mock).

To test whether increased production of chemokines in HCMV-infected RPE cells is functional, the chemotactic potentials of the respective supernatants were evaluated. HUVEC were confluently grown on Transwell membranes. Supernatants collected 4, 24, and 72 h p.i. from mock- and AD169-infected RPE cultures were placed in the lower chamber. These supernatants did not impair the monolayer integrity throughout the chemotaxis experiments as confirmed microscopically. When the chemotactic peptide FMLP (positive control) was added to the lower chamber, neutrophil chemotaxis across endothelial cell monolayers was significantly enhanced up to 10-fold compared with that in culture medium alone or supernatants from mock-infected RPE cells. Supernatants obtained from infected (AD169) RPE cells enhanced chemotaxis about 2-fold compared with supernatants from mock-infected RPE cells (Fig. 4). For example, the percentages of the number of neutrophils transmigrating toward supernatants of mock-infected and AD169-infected RPE (collected 4 h p.i.) were 15.4 ± 1.3% (7.7 ± 0.65 × 104 cells) and 33.0 ± 3.0% (16.5 ± 1.5 × 104 cells), respectively. Increased chemotactic activity of neutrophils triggered by supernatants of virus-infected RPE could be significantly inhibited by neutralizing Abs directed against IL-8 (15.26 ± 1.8%; 7.63 ± 0.9 × 104 cells), but not by neutralizing Abs directed against MGSA/Gro α (25.79 ± 2.3%; 12.8 ± 1.15 × 104 cells). The migration of neutrophils toward the chemotactic peptide FMLP was not affected by either neutralizing Ab (not shown).

FIGURE 4.

Transwell neutrophil chemotaxis across endothelial monolayers. White bars, chemotactic capacity of mock-infected RPE-I supernatants (MOCK) and of AD169-infected RPE-I (AD169), respectively. Supernatants were harvested at different times p.i. and placed in the lower chambers of the Transwells. Data are from three experiments and are expressed as the mean ± SD. ∗, p < 0.05 compared with corresponding values from mock-infected RPE-I cell culture supernatants.

FIGURE 4.

Transwell neutrophil chemotaxis across endothelial monolayers. White bars, chemotactic capacity of mock-infected RPE-I supernatants (MOCK) and of AD169-infected RPE-I (AD169), respectively. Supernatants were harvested at different times p.i. and placed in the lower chambers of the Transwells. Data are from three experiments and are expressed as the mean ± SD. ∗, p < 0.05 compared with corresponding values from mock-infected RPE-I cell culture supernatants.

Close modal

To determine whether HCMV-induced ICAM-1 expression on the surface of RPE cells is functional, neutrophil adhesion experiments were conducted. The adhesion of neutrophils to RPE cells was quantified at different times after coculture (ranging from 10 to 60 min). When neutrophils were cocultured with RPE cells 4 or 24 h p.i. no difference between neutrophil adhesion to mock- or HCMV-infected RPE cells was seen (data not shown). However, the number of neutrophils adhered to a significantly lower degree when RPE cells were HCMV infected for 72 h before coculture (Fig. 5). Whereas the number of adherent neutrophils cocultured with mock-infected RPE cells increased from 22.0 ± 3.1% (11.0 ± 1.55 × 104 cells) to 38.0 ± 5.2% (19.0 ± 2.6 × 104 cells) between 10–60 min of coculture, the numbers of neutrophils adherent to AD169- and Hi91-infected RPE cells were significantly reduced by 10 min after the start of coculture. For example, the percentage of neutrophils adherent to AD169-infected RPE cells was only 13.5 ± 1.7% (6.75 ± 0.85 × 104 cells). These low levels were found throughout the coculture period up to 60 min.

FIGURE 5.

Number of neutrophil binding to mock- or HCMV (AD169, Hi91)-infected RPE-I cells after different incubation times given as a percentage of the total number of cocultured neutrophils. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with mock-infected cells.

FIGURE 5.

Number of neutrophil binding to mock- or HCMV (AD169, Hi91)-infected RPE-I cells after different incubation times given as a percentage of the total number of cocultured neutrophils. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with mock-infected cells.

Close modal

The percentage of those neutrophils that penetrated the RPE cell monolayers subsequent to adhesion was determined. Penetration of mock- vs HCMV-infected RPE monolayers was quantified 10, 30, and 60 min after the start of coculture. In cocultures with mock-infected RPE cells 5.2 ± 0.8% (5.72 ± 0.88 × 103 cells), 6.1 ± 1.2% (10.98 ± 2.16 × 103 cells), and 5.0 ± 0.9% (9.50 ± 1.71 × 103 cells) penetrated the monolayer after 10, 30, and 60 min, respectively. In contrast, in the same experimental set up, neutrophil penetration through monolayers of HCMV-infected RPE cells was totally abrogated at all time points measured.

To determine whether reduced neutrophil binding to HCMV-infected RPE may be due to the induction of neutrophil apoptosis, the expression of FasL in mock- and HCMV-infected RPE cells was investigated. As demonstrated by flow cytometry all three RPE cultures constitutively expressed FasL on the cell surface. HCMV-infected RPE cells did not exhibit modified FasL expression 4 h p.i., whereas at 24 and 72 h p.i. expression was augmented about 1.5- and 3-fold, respectively (Fig. 6). Up-regulated FasL expression was confirmed at the transcriptional level, as determined by RT-PCR (Fig. 7). FasL expression was not influenced by RPE cell incubation with ultrafiltrated supernatants from virus-infected cells or UV-inactivated virus (not shown), indicating that infectious virus is necessary for this effect.

FIGURE 6.

FasL expression in mock- or HCMV (AD169, Hi91)-infected RPE-I cells at different times p.i. as determined by flow cytometry. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with mock-infected cells.

FIGURE 6.

FasL expression in mock- or HCMV (AD169, Hi91)-infected RPE-I cells at different times p.i. as determined by flow cytometry. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with mock-infected cells.

Close modal
FIGURE 7.

Representative FasL mRNA expression in mock-infected (MOCK) and AD169-infected (AD169) cultures of RPE cells 4 h p.i. detected by means of RT-PCR. The negative control (NC) contained only RT mix without RNA. Amplification products of FasL and GAPDH were visualized with ethidium bromide on agarose gel.

FIGURE 7.

Representative FasL mRNA expression in mock-infected (MOCK) and AD169-infected (AD169) cultures of RPE cells 4 h p.i. detected by means of RT-PCR. The negative control (NC) contained only RT mix without RNA. Amplification products of FasL and GAPDH were visualized with ethidium bromide on agarose gel.

Close modal

Experiments were conducted to determine whether FasL induction is due to the IE gene expression of HCMV, which is known to exhibit trans-activating properties, or to viral DNA replication. Preincubation of RPE cells with ISIS 2922 (1 μM), an antisense oligonucleotide against HCMV-IE mRNA, but not with the unspecific antisense ISIS 26062 (1 μM) or the viral DNA replication inhibitor GCV (40 μM), prevented HCMV-induced FasL expression (Fig. 8,A). ISIS 2922 also prevented HCMV-induced ICAM-1 expression (Fig. 8,B) and reversed the inhibitory effect of HCMV-infected RPE cells on neutrophil binding (Fig. 8 C).

FIGURE 8.

Influence of GCV (viral DNA replication inhibitor), ISIS 2922 (IE mRNA-specific oligonucleotide), and ISIS 26062 (unspecific oligonucleotide) on HCMV (AD169)-mediated FasL expression on RPE-I (A), ICAM-1 expression on RPE-I (B), and neutrophil binding to RPE-I (C) as determined by flow cytometry (A and B) or microscopic evaluation 30 min after the start of coculture (C). Data are the mean ± SD from one representative experiment performed in triplicate.

FIGURE 8.

Influence of GCV (viral DNA replication inhibitor), ISIS 2922 (IE mRNA-specific oligonucleotide), and ISIS 26062 (unspecific oligonucleotide) on HCMV (AD169)-mediated FasL expression on RPE-I (A), ICAM-1 expression on RPE-I (B), and neutrophil binding to RPE-I (C) as determined by flow cytometry (A and B) or microscopic evaluation 30 min after the start of coculture (C). Data are the mean ± SD from one representative experiment performed in triplicate.

Close modal

To confirm our hypothesis that HCMV-induced FasL on RPE cells is responsible for the lack of neutrophil binding, blocking experiments with specific neutralizing mAb against FasL (clones NOK-1 or NOK-2) were performed. Alternatively, neutrophils were preincubated with mAb against Fas (clone B-D29). As shown in Fig. 9 both treatment strategies completely abolished HCMV-induced impairment of neutrophil binding to RPE cells as measured 72 h p.i.. In contrast, blocking of ICAM-1 function on RPE with the neutralizing mAb BBIG-I1 resulted in decreased neutrophil adhesion to mock- and HCMV-infected RPE-I cells to a similar degree (2.3- and 1.9-fold, respectively).

FIGURE 9.

Blocking of FasL-mediated reduction of neutrophil adhesion with neutralizing Abs against FasL (NOK-1 and NOK-2) on RPE-I or Fas (B-D29) on neutrophils and blocking of ICAM-1-mediated binding with clone BBIG-I1 on RPE-I as determined by microscopic evaluation 30 min after the start of coculture. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with corresponding values without neutralizing Abs (control).

FIGURE 9.

Blocking of FasL-mediated reduction of neutrophil adhesion with neutralizing Abs against FasL (NOK-1 and NOK-2) on RPE-I or Fas (B-D29) on neutrophils and blocking of ICAM-1-mediated binding with clone BBIG-I1 on RPE-I as determined by microscopic evaluation 30 min after the start of coculture. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with corresponding values without neutralizing Abs (control).

Close modal

To evaluate the impact of FasL-mediated neutrophil apoptosis on the observed reduction of neutrophil binding to RPE cells, the surface membrane binding of annexin V-FITC (early phase of apoptosis) and DNA fragmentation after different times of coculture were measured (Fig. 10). After neutrophil coculture with both mock- and AD169-infected cells (72 h p.i.) for 10–60 min, no evidence for RPE-induced apoptosis was found, i.e., similar amounts of annexin V-FITC binding cells (∼10–15%) or DNA fragmentation were detected compared with neutrophils incubated in medium without RPE cells. After 4 h, the percentage of annexin V-FITC binding to neutrophils that were cocultured with mock-infected or AD169-infected RPE cells was increased to 21 and 35%, respectively (Fig. 10,A). DNA fragmentation (Fig. 10,B) was increased by factors of 1.6 (mock) and 2.8 (AD169; p < 0.05). Preincubation of neutrophils with the caspase-8 inhibitor zIETD at a concentration of 100 μM for 30 min reduced annexin V-FITC binding, but did not prevent impaired neutrophil adhesion (Fig. 11). The percentage of neutrophils adhering to mock-infected RPE cells was 37.0 ± 4.2% (18.5 ± 2.1 × 104) without zIETD pretreatment and 39.2 ± 4.5% (19.5 ± 2.3 × 104) with zIETD pretreatment, respectively, whereas the percentages of neutrophils that were adherent to AD169-infected RPE cells were 12.3 ± 1.6% (6.0 ± 0.85 × 104 cells) and 11.3 ± 1.5% (5.5 ± 0.75 × 104 cells) without or with zIETD pretreatment, respectively (Fig. 11 B).

FIGURE 10.

Detection of apoptosis in neutrophils by means of annexin V-FITC staining (A) and DNA fragmentation (B) after different periods of incubation without RPE cells or with mock- or AD169-infected RPE cells. The results are expressed as the percentage of annexin-positive neutrophils (A) or as DNA fragmentation (B) relative to that of neutrophils obtained immediately after preparation from blood samples. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with previous incubation times.

FIGURE 10.

Detection of apoptosis in neutrophils by means of annexin V-FITC staining (A) and DNA fragmentation (B) after different periods of incubation without RPE cells or with mock- or AD169-infected RPE cells. The results are expressed as the percentage of annexin-positive neutrophils (A) or as DNA fragmentation (B) relative to that of neutrophils obtained immediately after preparation from blood samples. Data are the mean ± SD from one representative experiment performed in triplicate. ∗, p < 0.05 compared with previous incubation times.

Close modal
FIGURE 11.

Effects of leukocyte preincubation with zIETD, an inhibitor of caspase-8, on induction of apoptosis (annexin staining method; A) and neutrophil adhesion (B) to mock- or AD169-infected RPE cells after 30 min of coculture. ∗, p < 0.05.

FIGURE 11.

Effects of leukocyte preincubation with zIETD, an inhibitor of caspase-8, on induction of apoptosis (annexin staining method; A) and neutrophil adhesion (B) to mock- or AD169-infected RPE cells after 30 min of coculture. ∗, p < 0.05.

Close modal

The pathomechanisms involved in HCMV retinitis are not entirely defined. It is likely that in immunocompromised individuals the virus itself may exert its pathogenicity due to the failure of the immune system to kill virus-infected cells. On the other hand, some patients with AIDS and HCMV retinitis have foci of neutrophilic infiltrates in retinal tissue at autopsy, suggesting intact granulocyte function and chemotaxis (3, 14). In these patients HCMV retinitis may involve nonspecific cell-mediated immune mechanisms that are pathogenic but fail to specifically eliminate the virus. The retinal pigment epithelium is part of the blood-retina barrier and is important for retinal homeostasis. Alterations of this barrier are associated with vascular, degenerative, and inflammatory disease of the retina and choroid (29, 30, 31, 32). As a component of the blood-retina barrier, RPE cells are suggested to be critical to the regulation of ocular inflammation and thus important in sustaining the immune privilege of the eye. In this study cultured RPE cells were infected with HCMV to investigate the potential of infected RPE cells to influence proinflammatory responses.

Our in vitro results showed that infection with HCMV resulted in enhanced production and secretion of proinflammatory chemokines as well as enhanced surface membrane expression of ICAM-1. Whereas HCMV-induced secretion of chemokines was shown to be functional in terms of enhanced chemotactic activity, the binding and transepithelial migration of neutrophils to/through infected RPE cells were significantly reduced. The latter observation was in strong contrast to infected HUVEC and HFF, where HCMV-induced chemokine secretion and ICAM-1 expression augmented neutrophil adhesion/transmigration (33). No evidence for HCMV infection of neutrophils could be found in our experimental coculture, leading us to suggest that transfer of virus or viral proteins may not account for this phenomenon. We therefore assumed that infection of RPE cells may augment specific cellular features that down-regulate neutrophil/RPE interactions, a mechanism that probably prevents leukocyte-induced damage of the blood-retina barrier.

Evidence for the immune privilege of RPE has been reported by Greenwood et al. (34) using a model of autoimmune uveoretinitis. At the retinal pigment epithelium, there was little evidence of migration into the retina during the early stages of the disease, even though the choroid became packed with inflammatory cells. Other reports suggested that RPE cells constitute an immunologic functional barrier against potentially harmful T cells or other leukocytes (20, 35). Moreover, it has been proposed that RPE cells may induce TCR-independent apoptosis in activated human T cells through the Fas/FasL pathway (18).

Based on this knowledge, we evaluated whether Fas/FasL may play a cell-specific role in the observed reduction of neutrophil binding to HCMV-infected RPE cells. Indeed, we found FasL constitutively expressed on RPE cells. This constitutive expression was strongly augmented by HCMV infection on the transcriptional and protein levels, whereas in endothelial cells and fibroblasts no HCMV-induced expression could be detected (data not shown). To our knowledge, this is the first report showing that HCMV infection may induce FasL. However, this seems not to be a general feature of HCMV infection but, rather, a specific feature of HCMV-infected RPE cells. Furthermore, our data clearly indicate that HCMV-induced FasL expression on RPE cells mediates down-regulation of neutrophil binding with a maximum effect at 72 h p.i. At this time point both HCMV-induced FasL and ICAM-1 expressions are maximal. Therefore, it should be emphasized that FasL-mediated down-regulation of neutrophil binding is superior over the function of HCMV-induced ICAM-1 that mediates binding of neutrophils to RPE cells via interaction with integrins (36). As reported for other cell types, HCMV-induced ICAM-1 resulted in enhanced neutrophil adhesion (7, 11, 33).

Because Fas/FasL is a well-recognized pathway in the induction of neutrophil apoptosis (37, 38), we evaluated annexin V staining and DNA fragmentation in those neutrophils that lost their initial cell-to-cell contact to RPE cells. Decreased binding/penetration of neutrophils to/through HCMV-infected RPE cells occurred as early as 10 min after the start of coculture, whereas signs of apoptosis were not found before 4 h. On the first view, it seems to be a paradox that, on the one hand, FasL expression is responsible for neutrophil detachment and, on the other hand, neutrophil binding is required for Fas/FasL-mediated apoptosis. However, we were able to show that the initial short binding period of 10 min is sufficient to induce 1) neutrophil detachment (caspase-8 independent) and 2) apoptosis (caspase-8 dependent). Moreover, the finding that neutrophils do not exhibit signs of apoptosis before 4 h suggests that early steps in apoptosis are switched on during the initial binding phase and that mechanisms further downstream in the apoptotic cascade occur regardless of cell-to-cell contact to infected RPE cells.

These findings further suggest that Fas/FasL-dependent immune regulatory pathways (18) may not exclusively be attributed to the induction of apoptosis, but may also occur through the early regulating processes of neutrophil binding that precede apoptosis. On the other hand, it is possible that early downstream steps of the Fas/FasL pathway that do not follow the caspase-8 (a constituent of death-inducing signaling complex that activates effector caspases such as caspase-3) route are involved in the latter effect. For example, caspase-1, which may also be activated upon FasL/Fas interaction, might play a role in cytoskeletal derangement and modified interactions with leukocytes independent of apoptosis (39, 40). It is well established that cytoskeleton derangement may influence the distribution of cell surface adhesion molecules and cell-cell interactions (41).

Recently, we and others suggested that IE proteins play an important role in the regulation of cellular gene expression (11, 12, 24, 33, 42, 43). Here, we showed that HCMV-induced functional protein expression of FasL in RPE cells is probably due to HCMV IE gene products. HCMV IE is expressed in infected cells before viral DNA replication occurs and thus cannot be inhibited by the standard anti-HCMV drug GCV, an inhibitor of viral DNA replication. We showed that the antisense oligonucleotide ISIS 2922 that blocks IE mRNA prevented HCMV-induced FasL expression and function, whereas the irrelevant antisense oligonucleotide ISIS 26062 or GCV had no effect. The HCMV-induced stimulation of FasL expression may be discussed as a mechanism of virus escape from immune surveillance (44) and is probably a general feature of viruses restricted to specific host cells. For example, induction of FasL expression on HIV-1-infected lymphocytes, hepatitis B virus-infected hepatoma cell lines, EBV-infected B lymphocytes, and macrophages has been reported (45, 46, 47).

In conclusion, HCMV-induced FasL expression on RPE cells is a novel mechanism that precedes neutrophil apoptosis and down-regulates acute neutrophil binding activity. These mechanisms may be interpreted as a special feature of RPE cells to protect the eye from immune invasion, as it is known as an immune-privileged sites. On the other hand, FasL allows persistence of the virus specifically in RPE cells, which may entail chronic immune responses and disease, such as HCMV retinitis.

We appreciated very much the critical reviews of Dr. Stephanie Dimmeler and Prof. Leonhardt Rosenthal, and we are grateful to Gabi Bauer for the microphotographs, and to Elsie Oppermann, Karina Cordier, and Gesa Meincke for the technical assistance.

1

This work was supported in part by Frankfurter Stiftung für krebskranke Kinder and Hilfe für krebskranke Kinder Frankfurt e.V.

3

Abbreviations used in this paper: HCMV, human cytomegalovirus; FasL, Fas ligand; FIU, fluorescence intensity units; GCV, ganciclovir; HFF, human foreskin fibroblasts; IE, immediate early; IEA, IE Ag; LA, late Ag; RPE, retinal pigment epithelial cells; p.i., postinfection; MGSA, melanoma growth-stimulating activity; PI, propidium iodide.

1
Egbert, P. R., R. B. Pollard, J. G. Gallagher, T. C. Merigan.
1980
. Cytomegalovirus retinitis in immunosuppressed hosts. II. Ocular manifestations.
Ann. Intern. Med.
93
:
664
2
Jabs, D. A., J. G. Barlett.
1997
. AIDS and ophthalmology: a period of transition.
Am. J. Ophthalmol.
124
:
227
3
Pepose, J. S., C. Newman, M. C. Bach, T. C. Quinn, R. F. Ambinder, G. Holland, P. S. Hosstrom, H. M. Frey, M. Y. Foos.
1987
. Pathologic features of cytomegalovirus retinopathy after treatment with the antiviral agent ganciclovir.
Ophthalmology
94
:
414
4
Jabs, D. A., C. Enger, J. Haller, S. de Bustros.
1991
. Retinal detachments in patients with cytomegalovirus retinitis.
Arch. Ophthalmol.
109
:
794
5
Pepose, J. S., G. N. Holland, M. S. Nestor, A. J. Cochran, R. Y. Foos.
1985
. Acquired immune deficiency syndrome: pathogenic mechanisms of ocular disease.
Ophthalmology
92
:
472
6
Rao, N. A., J. Zhang, S. I. Ishimoto.
1998
. Role of retinal vascular endothelial cells in development of CMV retinitis.
Trans. Am. Ophthamol. Soc.
96
:
111
7
Scholz, M., A. Hamann, R. A. Blaheta, M. K. Auth, A. Encke, B. H. Markus.
1992
. Cytomegalovirus- and interferon-related effects on human endothelial cells: cytomegalovirus infection reduces upregulation of HLA class II antigen expression after treatment with interferon-γ.
Hum. Immunol.
35
:
230
8
Grundy, J. E., K. L. Downes.
1993
. Up-regulation of LFA-3 and ICAM-1 on the surface of fibroblasts infected with cytomegalovirus.
Immunology
78
:
405
9
Murayama, T., Y. Ohara, M. Obuchi, K. S. Khabar, H. Higashi, N. Mukaida, K. Matsushima.
1997
. Human cytomegalovirus induces interleukin-8 production by a human monocytic cell line, THP-1, through acting concurrently on AP-1- and NF-κB-binding sites of the interleukin-8 gene.
J. Virol.
71
:
5692
10
Craigen, J. L., K. L. Yong, N. J. Jordan, L. P. MacCormac, J. Westwick, A. N. Akbar, J. E. Grundy.
1997
. Human cytomegalovirus infection up-regulates interleukin-8 gene expression and stimulates neutrophil transendothelial migration.
Immunology
92
:
138
11
Grundy, J. E. 1998. Current antiviral therapy fails to prevent the pro-inflammatory effects of cytomegalovirus infection, whilst rendering infected cells relatively resistant to immune attack. In CMV-Related Immunopathology. M. Scholz, H. F. Rabenau, H. W. Doerr, and J. Cinatl, Jr., eds. Karger, Basel, pp. 67–89.
12
Cinatl, J., Jr, R. Kotchetkov, E. Weimer, R. A. Blaheta, M. Scholz, J.-U. Vogel, H. O. Gümbel, H. W. Doerr.
2000
. The antisense oligonucleotide ISIS 2922 prevents cytomegalovirus-induced up-regulation of IL-8 and ICAM-1 in cultured human fibroblasts.
J. Med. Virol.
60
:
313
13
Michelson, S., P. Dal Monte, D. Zipeto, B. Bodaghi, L. Laurent, E. Oberlin, F. Arenzana-Seisdedos, J. L. Virelizier, M. P. Landini.
1997
. Modulation of RANTES production by human cytomegalovirus infection of fibroblasts.
J. Virol.
71
:
6495
14
Palestine, A. G., M. M. Rodrigues, A. M. Macher, C. C. Chan, H. C. Lane, A. S. Fauci, H. Masur, D. Longo, M. C, M. Reichert, R. Steis.
1984
. Ophthalmic involvement in acquired immunodeficiency syndrome.
Ophthalmology
91
:
1092
15
Niederkorn, J. Y..
1990
. Immune privilege and immune regulation in the eye.
Adv. Immunol.
48
:
191
16
Streilein, J. W..
1993
. Immune privilege as the result of local tissue barriers and immunosuppressive microenvironments.
Curr. Opin. Immunol.
5
:
428
17
Streilein, J. W., B. R. Ksander, A. W. Taylor.
1997
. Immune deviation in relation to ocular immune privilege.
J. Immunol.
158
:
3557
18
Jorgensen, A., A. K. Wiencke, M. la Cour, C. G. Koestel, H. O. Madsen, S. Hamann, G. M. Lui, E. Scherfig, J. U. Prause, A. Svejgaard, et al
1998
. Human retinal pigment epithelial cell-induced apoptosis in activated T cells.
Invest. Ophthalmol. Vis. Sci.
39
:
1590
19
Griffith, T. S., S. M. T Brunner., D. R. Fletcher, D. R. Green, T. A. Ferguson.
1995
. Fas ligand-induced apoptosis as a mechanism of immune privilege.
Science
270
:
1189
20
Farrokh-Siar, L., K. A. Rezai, R. T. Semnani, S. C. Patel, J. T. Ernest, E. J. Peterson, G. A. Koretzky, G. A. van Seventer.
1999
. Human fetal retinal pigment epithelial cells induce apoptosis in the T-cell line Jurkat.
Invest. Ophthalmol. Vis. Sci.
40
:
1503
21
Cinatl, J., Jr, J. Cinatl, B. Weber, H. Rabenau, H. O. Gümbel, J. F. Chenot, M. Scholz, A. Encke, H. W. Doerr.
1995
. In vitro inhibition of human cytomegalovirus replication in human foreskin fibroblasts and endothelial cells by ascorbic acid 2-phosphate.
Antiviral Res.
27
:
405
22
Flood, M. T., P. Gouras, H. Kjeldbye.
1980
. Growth characteristics and ultrastructure of human pigment epithelium in vitro.
Invest. Ophthalmol. Vis. Sci.
19
:
1309
23
Bunt-Milam, A. H., J. C. Saari.
1983
. Immunocytochemical localization of two retinoid-binding proteins in vertebrate retina.
J. Cell Biol.
97
:
703
24
Cinatl, J., Jr, R. Kotchetkov, M. Scholz, J. Cinatl, J.-U. Vogel, P. Hernáiz-Driever, H. W. Doerr.
1999
. Human cytomegalovirus infection decreases expression of thrombospondin-1 independent of the tumor suppressor protein p53.
Am. J. Pathol.
155
:
285
25
Azad, R. F., V. B. Driver, K. Tanaka, R. M. Crooke, K. P. Anderson.
1993
. Antiviral activity of a phosphorothioate oligonucleotide complementary to RNA of the human cytomegalovirus major immediate-early region.
Antimicrob. Agents Chemother.
37
:
1945
26
Turley, J. M., T. Fu, F. W. Ruscetti, J. A. Mikovits, D. C. Bertolette, III, M. C. Birchenall-Roberts.
1997
. Vitamin E succinate induces Fas-mediated apoptosis in estrogen receptor-negative human breast cancer cells.
Cancer Res.
57
:
881
27
Dustin, M. L., T. A. Springer.
1989
. T-cell receptor cross-linking transiently stimulates adhesiveness through LFA-1.
Nature
341
:
619
28
Blaheta, R. A., N. P. Hailer, N. Brude, B. Wittig, E. Oppermann, K. Leckel, S. Harder, M. Scholz, S. Weber, A. Encke, et al
1998
. Novel mode of action of the calcium antagonist mibefradil (Ro 40–5967): potent immunosuppression by inhibition of T-cell infiltration through allogeneic endothelium.
Immunology
94
:
213
29
Cunha-Vaz, J...
1979
. The blood-ocular barriers.
Surv. Ophthalmol.
23
:
279
30
Vinores, S. A., E. Van Niel, J. L. Swerdloff, P. A. Campochiaro.
1993
. Electron microscopic immunocytochemical demonstration of blood-retinal barrier breakdown in human diabetics and its association with aldose reductase in retinal vascular endothelium and retinal pigment epithelium.
Histochem. J.
25
:
648
31
Vinores, S. A., M. Kuchle, N. L. Derevjanik, J. D. Henderer, J. Mahlow, W. R. Green, P. A. Campochiaro.
1995
. Blood-retinal barrier breakdown in retinitis pigmentosa: light and electron microscopic immunolocalization.
Histol. Histopathol.
10
:
913
32
Luna, J. D., C. C. Chan, N. L. Derevjanik, J. Mahlow, C. Chiu, B. Peng, T. Tobe, P. A. Campochiaro, S. A. Vinores.
1997
. Blood-retinal barrier (BRB) breakdown in experimental autoimmune uveoretinitis: comparison with vascular endothelial growth factor, tumor necrosis factor α, and interleukin-1β-mediated breakdown.
J. Neurosci. Res.
49
:
268
33
Scholz, M., J.-U. Vogel, R. A. Blaheta, H. W. Doerr, J. Cinatl, Jr.
1998
. Cytomegalovirus, oxidative stress and inflammation as interdependent pathomechanisms: need for novel therapeutic strategies. M. Scholz, Jr, and H. F. Rabenau, Jr, and H. W. Doerr, Jr, and J. Cinatl, Jr, eds.
CMV-Related Immunopathology
90
-105. Karger, Basel.
34
Greenwood, J., R. Howes, S. Lightman.
1994
. The blood-retinal barrier in experimental autoimmune uveoretinitis: leukocyte interactions and functional damage.
Lab. Invest.
70
:
39
35
Rezai, K. A., R. T. Semnani, L. Farrokh-Siar, K. J. Hamann, S. C. Patel, J. T. Ernest, G. A. van Seventer.
1999
. Human fetal retinal pigment epithelial cells induce apoptosis in allogenic T-cells in a Fas ligand and PGE2 independent pathway.
Curr. Eye Res.
18
:
430
36
Elner, S. G., V. M. Elner, M. A. Pavilack, R. F. Todd, L. Mayo-Bond, W. A. Franklin, R. M. Strieter, S. L. Kunkel, A. R. Huber.
1992
. Modulation and function of intercellular binding molecule-1 (CD54) on human retinal pigment epithelial cells.
Lab. Invest.
66
:
200
37
Liles, W. C., P. A. Kiener, J. A. Ledbetter, A. Aruffo, S. J. Klebanoff.
1996
. Differential expression of Fas (CD95) and Fas ligand on normal human phagocytes: implication for the regulation of apoptosis in neutrophils.
J. Exp. Med.
184
:
429
38
Hsieh, S. C., M. H. Huang, C. Y. Tsai, Y. Y. Tsai, S. T. Tsai, K. H. Sun, H. S. Yu, S. H. Han, C. L. Yu.
1997
. The expression of genes modulating programmed cell death in normal human polymorphonuclear neutrophils.
Biochem. Biophys. Res. Commun.
233
:
700
39
Kondo, T., K. Takeuchi, Y. Doi, S. Yonemura, S. Nagata, S. Tsukita.
1997
. ERM (ezrin/radixin/moesin)-based molecular mechanism of microvillar breakdown at an early stage of apoptosis.
J. Cell Biol.
139
:
449
40
Thornberry, N. A., Y. Lazebnik.
1998
. Caspases: enemies within.
Science
281
:
1312
41
Del Pozo, M. A., M. Nieto, J. M. Sarrador, D. Sancho, M. Vincente-Manzanares, C. Martinez, F. Sanchez-Madrid.
1998
. The two poles of the lymphocyte: specialized cell compartments for migration and recruitment.
Cell Adhes. Commun.
6
:
125
42
Speir, E., Z. X. Yu, V. J. Ferrans, E. S. Huang, S. E. Epstein.
1998
. Aspirin attenuates cytomegalovirus infectivity and gene expression mediated by cyclooxygenase-2 in coronary artery smooth muscle cells.
Circ. Res.
83
:
210
43
Burns, L. J., J. C. Pooley, D. J. Walsh, G. M. Vercellotti, M. L. Weber, A. Kovacs.
1999
. Intercellular adhesion molecule-1 expression in endothelial cells is activated by cytomegalovirus immediate early proteins.
Transplantation
67
:
137
44
Smyth, M. J., J. A. Trapani.
1998
. The relative role of lymphocyte granule exocytosis versus death receptor-mediated cytotoxicity in viral pathophysiology.
J. Virol.
72
:
1
45
Xu, X. N., B. Laffert, G. R. Screaton, M. Kraft, D. Wolf, D. Kolanus, J. Mongkolsapay, A. J. McMichael, A. S. Baur.
1999
. Induction of Fas ligand expression by HIV involves the interaction of Nef with the T cell receptor ζ chain.
J. Exp. Med.
189
:
1489
46
Shin, E. C., J. C. Sin, J. H. Park, S. J. Kim.
1999
. Expression of Fas ligand in human hepatoma cell lines: role of hepatitis-B virus X (HBX) in induction of Fas ligand.
Int. J. Cancer
82
:
587
47
Tanner, J. E., C. Alfieri.
1999
. Epstein-Barr virus induces Fas (CD95) in T cells and Fas ligand in B cells leading to T-cell apoptosis.
Blood
94
:
3439