Lipoxin A₄ inhibits immune cell binding to salivary epithelium and vascular endothelium

Abstract

Lipoxins are formed by leukocytes during cell-cell interactions with epithelial or endothelial cells. Native lipoxin A₄ (LXA₄) binds to the G protein-coupled lipoxin receptors formyl peptide receptor 2 (FPR2)/ALX and CysLT1. Furthermore, LXA₄ inhibits recruitment of neutrophils, by attenuating chemotaxis, adhesion, and transmigration across vascular endothelial cells. LXA₄ thus appears to serve as an endogenous “stop signal” for immune cell-mediated tissue injury (Serhan CN; Annu Rev Immunol 25: 101–137, 2007). The role of LXA₄ has not been addressed in salivary epithelium, and little is known about its effects on vascular endothelium. Here, we determined that interleukin-1β (IL-1β) and tumor necrosis factor-α (TNF-α) receptor activation in vascular endothelium and salivary epithelium upregulated the expression of adhesion molecules that facilitates the binding of immune cells. We hypothesize that the activation of the ALX/FPR2 and/or CysLT1 receptors by LXA₄ decreases this cytokine-mediated upregulation of cell adhesion molecules that enhance lymphocyte binding to both the vascular endothelium and salivary epithelium. In agreement with this hypothesis, we observed that nanomolar concentrations of LXA₄ blocked IL-1β- and TNF-α-mediated upregulation of E-selectin and intercellular cell adhesion molecule-1 (ICAM-1) on human umbilical vein endothelial cells (HUVECs). Binding of Jurkat cells to stimulated HUVECs was abrogated by LXA₄. Furthermore, LXA₄ preincubation with human submandibular gland cell line (HSG) also blocked TNF-α-mediated upregulation of vascular cell adhesion molecule-1 (VCAM-1) in these cells, and it reduced lymphocyte adhesion. These findings suggest that ALX/FPR2 and/or CysLT1 receptor activation in endothelial and epithelial cells blocks cytokine-induced adhesion molecule expression and consequent binding of lymphocytes, a critical event in the pathogenesis of Sjögren's syndrome (SS).

sjögren's syndrome (SS) is a chronic inflammatory autoimmune disorder that causes dysfunction in the salivary and lacrimal glands, leading to xerostomia (dry mouth) and keratoconjunctivitis sicca (dry eyes) (23, 82). SS is characterized by extensive lymphocytic infiltration into the exocrine glands that is associated with the decrease of saliva secretion (36). Glandular damage in SS is mediated by T lymphocytes, mainly primed CD4⁺ T helper lymphocytes with a CD4⁺/CD8⁺ ratio >2 (93). B lymphocytes make up ∼20% of the infiltrating cell population and monocytes/macrophages are also present (50). Infiltrating lymphocytes produce proinflammatory cytokines that have been shown to play a role in the pathogenesis of SS (10, 21, 35, 71, 73). During the evolution of the inflammatory response in SS, an array of adhesive interactions occurs between immune cells and target tissues (2, 22, 75, 89). Specifically, adhesion molecules, expressed on the surface of lymphocytes, and their cognate ligands, expressed on endothelial cells and exocrine glands, are thought to be the main molecules involved in lymphocytic infiltration into sites of inflammation in SS (3). Furthermore, these adhesion molecules trigger signaling cascades that regulate inflammation and immune responses (15, 87, 88). This plays a role in lymphocyte-endothelial or epithelial cell adhesion (3, 83).

The tethering of inflammatory cells to the vascular endothelium is the initial step in a cascade of events that eventually leads to cell migration into sites of injury or infection. This first step is mediated by an interaction between E-selectin and P-selectin expressed on activated endothelial cells and a variety of glycoproteins and potentially glycolipids expressed on the leukocyte surface (46, 57). Tethering is followed by leukocyte rolling, activation, and arrest. While the initial tethering and rolling interactions are largely selectin mediated, integrin binding to members of the immunoglobulin superfamily contribute to the firm arrest of leukocytes to the endothelium. In the salivary vasculature, infiltration is associated with the increased expression of adhesion molecules, including vascular cell adhesion molecule-1 (VCAM-1) and intercellular cell adhesion molecule-1 (ICAM-1) (2, 22, 75). ICAM-1 may be important for lymphocyte recruitment and glandular damage, and VCAM-1 may be important for the development of vasculitis in patients with SS (92). Furthermore, ICAM-1 and VCAM-1 are expressed in salivary gland epithelium of SS patients (2, 75, 89, 91), enabling direct interaction with infiltrating lymphocytes and regulating the inflammatory process (66).

Several studies demonstrate that human and animal cells convert arachidonic acid into lipoxins (LX), which are short-lived, highly potent antiinflammatory agents that control the duration and magnitude of inflammatory signals in many cell types (80). LX are formed via two sequential lipoxygenase-catalyzed oxygenations of arachidonic acid (7). LX formation mainly occurs by heterotypic transcellular synthesis between polymorphonuclear leukocytes (PMN) with platelets, endothelia, and epithelia during inflammation (79). In the presence of aspirin, the formation of an epimer of lipoxin A₄ (LXA₄), 15-epi-lipoxin A₄ (ATL), occurs via alteration of the enzymatic activity of cyclooxygenase-2 (18, 32). LXA₄ exerts its physiological action via activation of the G protein-coupled receptors (GPCRs) and LXA₄ receptors ALX/formyl peptide receptor 2 (FPR2) and CysLT1 (17, 29, 30).

In this study we used the human submandibular gland cell line (HSG) and human umbilical vein endothelial cells (HUVECs) to investigate the effects of LXA₄ on adhesion molecule expression and lymphocyte binding mediated by cytokines_. Our findings indicate that LXA₄ is able to block interleukin-1β (IL-1β)- and tumor necrosis factor-α (TNF-α)-induced upregulation of E-selectin and ICAM-1 as well as lymphocyte binding to vascular endothelium and TNF-α-mediated VCAM-1 upregulation and lymphocyte binding to salivary epithelium. These results suggest an important role for LXA₄ in regulating critical steps related to the pathogenesis of SS.

MATERIALS AND METHODS

HUVEC cell culture.

HUVECs were obtained from Lonza (Walkersville, MD). Cells were cultured using sterile endothelial growth medium (EGM-2, Clonetics, Walkersville, MD) in a 37°C humidified atmosphere containing 5% CO₂. HUVECs were preincubated for 30 min with or without LXA₄ from two different commercial sources (Cayman Chemicals, Ann Arbor, MI) or (EMD Chemicals Calbiochem, Gibbstown NJ), or with or without ATL (EMD Chemicals Calbiochem). Then, cells were incubated for 4–6 h in the absence or presence of IL-1β or TNF-α (both from R&D Systems, Minneapolis, MN), and live cells were used for enzyme-linked immunosorbent assay.

HSG cell culture.

The HSG cells are a human salivary intercalated duct cell line (HSG), which was established from an irradiated human salivary gland and exhibits morphological, biochemical, and functional characteristics of acinar cells (84). HSG cells were cultured in DMEM (Invitrogen, Carlsbad, CA) with 5% (vol/vol) FBS (Atlanta Biologicals, Lawrenceville, GA), 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were maintained at 37°C in a humidified atmosphere of 5% CO₂. HSG cells grown to 80% confluence were cultured in serum-free DMEM overnight before and during the addition of agonists. Serum-starved HSG cells were used to prevent nonspecific responses triggered by serum (which possess growth factors). Epidermal growth factor is known to stimulate VCAM-1 expression in HSG cells (6). HSG cells were preincubated for 30 min with or without LXA₄. Then, cells were incubated for 6 h in the absence or presence of TNF-α and lysed in 200 μl of 2× Laemmli buffer [120 mM Tris·HCl, pH 6.8, 10% (vol/vol) glycerol, 2% (wt/vol) SDS, 5% (vol/vol) β-mercaptoethanol, and 0.002% (wt/vol) bromophenol blue] for Western blot analysis.

Jurkat cell culture.

Jurkat CD4⁺ lymphocytic cells (American Type Culture Collection, Manassas, VA) were cultured in RPMI 1640 medium (Invitrogen) with 5% (vol/vol) FBS, 100 U/ml penicillin, and 100 μg/ml streptomycin. Cells were maintained at 37°C in a humidified atmosphere of 5% CO₂.

Peripheral blood CD4⁺ T lymphocytes (negatively selected).

CD4⁺ T lymphocytes were purchased from Astarte Biologics (Redmond, WA). Donors were negative for human immuodeficiency virus-1 and -2, hepatitis B, hepatitis C, and human T-cell leukemia virus I. Cells were purified (≥95%) by fluorescence-activated cell sorting and cryopreserved until used.

Enzyme-linked immunosorbent assay.

To determine antibody specificity, HUVEC cell monolayers were grown in 96-well plates at an initial density of 6.25 × 10⁴ cells per well. Confluent HUVEC cell monolayers were pretreated with LXA₄ (100 ng/ml, 30 min), followed by cytokines such as IL-1β (1 ng/ml, 4–6 h) or TNF-α (10 ng/ml, 4–6 h). Cells were washed with PBS and blocked with PBS containing 2% BSA (up to the top of the well) for 30 min. The plates were washed 5× with PBS and incubated with anti-mouse ICAM-1 antibody (1:400 dilution; Ancell, Bayport, MN), which recognizes the D1 domain of the CD54 molecule, and anti-human E-selectin antibody (1:400 dilution; Ancell), which recognizes the lectin domain of the CD62E molecule. Antibodies were diluted in PBS containing 0.1% BSA for 1 h at room temperature. The plates were washed 5× with PBS and incubated with horseradish peroxidase-conjugated secondary antibody for 1 h at room temperature. After washing 5× with PBS, peroxidase activity in monolayers was quantified by the addition of OPD (Sigma, St. Louis, MO) and read at 10 min and 15 min in a Bio-Tek Epoch spectrophotometer at 450 nm. Analysis was performed using the Gen 5 software, on three wells per condition, and experiments were repeated at least three times.

Western blot analysis.

Cell lysates were sonicated for 5 s with a Branson Sonifier 250 (microtip; output level 5; duty cycle 50%) and boiled for 3 min. The lysates were subjected to 7.5% (wt/vol) sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) on minigels, and transferred to nitrocellulose membranes. Membranes were blocked for 1 h with 5% (wt/vol) nonfat dry milk in Tris-buffered saline [0.137 M NaCl, 0.025 M Tris (hydroxymethyl)-aminomethane, pH 7.4] containing 0.1% (vol/vol) Tween 20 (TBST) and immunoblotted overnight with goat anti-rabbit ALX/FPR2 antibody (1:200 dilution; Alomone Labs, Jerusalem, Israel), which recognizes a sequence corresponding to amino acid residues 184–196 of human FPR2, and goat anti-rabbit VCAM-1 antibody (1:200 dilution; Protein Tech, Chicago, IL), which recognizes an epitope corresponding to the last 400 amino acids mapping the COOH terminus of VCAM-1 and goat anti-rabbit VCAM-1 (1:200 dilution; Abnova, Taipei City, Taiwan). Membranes were washed 3× with TBST during a 45-min period and incubated with peroxidase-linked goat anti-rabbit IgG antibody (1:2,000 dilution; Santa Cruz Biotechnology) at room temperature for 1 h. The incubations with primary antibodies were performed at 4°C; antibodies were diluted in TBST containing 3% (wt/vol) BSA and 0.02% (wt/vol) sodium azide. After incubation with the primary antibody, membranes were washed 3× for 15 min each with TBST and incubated with peroxidase-linked goat anti-rabbit IgG antibody (1:2,000 dilution; Santa Cruz Biotechnology), as appropriate, at room temperature for 1 h. The membranes were washed 3× for 15 min each with TBST and then treated with a chemiluminescence detection reagent containing 20 mM Tris buffer, pH 8.5, 250 mM luminol, and 90 mM coumaric acid (Sigma). Protein bands were visualized on X-ray film and their densities were quantified using a computer-driven scanner and Quantity One software (Bio-Rad, Hercules, CA). For signal normalization, membranes were treated with stripping buffer [0.1 M glycine, pH 2.9, and 0.02% (wt/vol) sodium azide] and reprobed with goat anti-rabbit (total) extracellular signal-regulated kinase (ERK-1/2) antibody (1:1,000 dilution; Santa Cruz Biotechnology). VCAM-1 was expressed as a ratio of normalized values of the band intensities of VCAM-1 to total ERK. All experiments were repeated at least three times.

Fluorescence microscopy analysis.

HSG cell monolayers were fixed in 4% paraformaldehyde for 10 min at room temperature, incubated with 0.1% Triton X-100 in PBS for 5 min, and washed 3× with PBS. Then, cells were incubated with 5% goat serum containing 10 μM digitonin for 2 h at room temperature and washed 3× with PBS. The cells were incubated overnight at 4°C with rabbit anti-CysLT1 receptor (Cayman Chemicals) or rabbit anti-ALX/FPR2 receptor (Alomone Labs) antibody at 1:200 dilution in 5% goat serum containing 10 μM digitonin. The next day, cells were washed 3× for 5 min with PBS (cells were warmed up to room temperature for 20 min before wash). Cells were incubated for 45 min with AlexaFluor 488-conjugated goat anti-rabbit (1:500 dilution in 5% goat serum containing 10 μM digitonin) and washed 3× with PBS. Cells were stained for 5 min with 1:10,000 dilution in PBS of Hoechst nuclear stain (Sigma). Images were obtained using a Carl Zeiss 510 confocal microscope and analyzed using the AxioVision software (release 4.8).

Jurkat and human CD4⁺ T lymphocyte rolling and binding to HUVECs in microfluidic flow cell.

HUVECs were grown on 60-mm plates and preincubated with LXA₄ (100 ng/ml) for 30 min at 37°C. Subsequently, cells were exposed to IL-1β (1 ng/ml) or TNF-α (10 ng/ml) for 4–6 h. The plates were positioned in a vacuum-sealed microfluidic flow-chamber and perfused with 2′,7′-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein (BCECF)-labeled lymphocytes at an estimated shear stress of 2 dyn/cm² for 5 min. Lymphocytes bound to the monolayer were visualized using a Zeiss Axio Observer Fluorescence Microscope. Interacting cells were calculated as the total of the rolling and adherent cells (63, 94). Rolling cells were calculated as the ones that move slowly, at velocity less than the free stream velocity, after attaching to the surface. Adherent cells were calculated as the ones that do not change position within 10 s of attaching to the surface. The entire period of perfusion was recorded using a Qimagin retiga 1300 cooled charge-coupled device camera and Streampix video acquisition software.

Binding of Jurkat cells and cryopreserved human CD4⁺ T lymphocytes to HSG cell monolayers.

Confluent serum-starved HSG cell cultures in 12-well plates were pretreated with LXA₄ (0.1–100 ng/ml) or ATL for 30 min, then monolayers were incubated both with and without TNF-α (100 ng/ml) for 6 h at 37°C. Lymphocytic cells (10⁶) were labeled with the green fluorescent dye PKH2 (Sigma) and were then added to 80% confluent cultures of HSG cells for 1 h at 37°C. Cells were washed in RPMI serum-free medium, and adherent lymphocytes were counted using fluorescence microscopy in a Zeiss Axio Observer Fluorescence Microscope. To investigate the involvement VCAM-1 in lymphocyte binding, HSG cells were treated with and without TNF-α (100 ng/ml) for 6 h at 37°C followed by the addition of 10 μg/ml mouse anti-human VCAM-1 (Ancell), or mouse anti-human ICAM-1 (P2A4) (Chemicon International, Tremula, CA) for 45 min before addition of lymphocytes. Cell numbers were determined in seven different fields of view in nine culture wells from three experiments.

Statistical analysis.

Data are means ± SE of results from three or more determinations. Data were analyzed by one-way ANOVA followed by pairwise post hoc Tukey's t-test where P < 0.05 represents significant differences between experimental groups.

RESULTS

HSG cells express LXA₄ receptor.

LXA₄ binds to the GPCRs ALX/FPR2, which is expressed on leukocytes, enterocytes, and HUVECs (4, 29, 32, 86), and CysLT1, which is expressed on HUVEC and mesangial cells (29, 30, 90). To evaluate whether the ALX/FPR2 receptor was expressed in HSG cell monolayers, Western blot analysis was performed as described in materials and methods. Results illustrate two protein bands with an apparent molecular mass of 44 kDa and 48 kDa in membranes probed with the ALX/FPR2 antibody indicating that this receptor is expressed in HSG and Jurkat cells (Fig. 1A). The ALX/FPR2 together with the CysLT1 receptors were detected in HSG cells by immunohistochemistry (Fig. 1, B and C). The ALX/FPR2 receptor was visualized both in the membrane and within the cytoplasmic compartment of HSG cells (Fig. 1B) while the CysLT1 was visualized only on the membrane of HSG cells (Fig. 1C). In the presence of secondary antibody alone, there was minimal or absent staining in HSG cell monolayers (data not shown). These results indicate that receptors for LXA₄ are expressed in the cell systems utilized in our studies.

Fig. 1.

ALX/formyl peptide receptor 2 (FPR2) is expressed in human submandibular gland (HSG) cells and Jurkat cells. A: protein extracts from Jurkat and HSG cells were subjected to Western blot analysis using rabbit anti-ALX/FPR2 receptor that recognizes a sequence corresponding to amino acid residues 184–196 of human FPR2. B and C: HSG cell monolayers were fixed, and expression of ALX/FPR2 and CysLT1 receptors was detected using immunofluorescence microscopy with rabbit anti-ALX/FPR2 (green) and rabbit anti-CysLT1 (green). Images were obtained and analyzed using a Carl Zeiss 510 confocal microscope. Left: rabbit anti-ALX/FPR2 stain (top, green) or rabbit anti-CysLT1 (bottom, green). Center: Hoechst nuclear stain (blue). Right: merged images (green/blue). Arrows indicate receptor localization.

LXA₄ inhibits IL-1β- and TNF-α- mediated upregulation of E-selectin and ICAM-1 in HUVECs.

Incubation of HUVECs with IL-1β and TNF-α induced upregulation of E-selectin and ICAM-1 after 4–6 h (Figs. 2 and 3), a time at which these molecules are usually upregulated by these proinflammatory cytokines (8, 76). Cytokine-mediated upregulation of adhesion molecules has been well characterized in HUVECs (44, 61, 77). E-selectin is rapidly induced after exposure of HUVECs to IL-1β and TNF-α, peaking at 4–6 h and returning to basal levels after 24 h (44). ICAM-1 is induced after exposure of HUVECs to IL-1β and TNF-α, peaking at 4–6 h and decreasing its expression after 72 h (44). Pretreatment of HUVECs with LXA₄ (100 ng/ml, 30 min) significantly inhibited the upregulation of E-selectin and ICAM-1 mediated by IL-1β (1 ng/ml, 4–6 h) and TNF-α (10 ng/ml, 4–6 h) (Figs. 2 and 3). These results suggest that LXA₄ is effective in inhibiting cytokine-mediated upregulation of adhesion molecule expression in endothelial cells.

Fig. 2.

Lipoxin A₄ (LXA₄) inhibits IL-1β-mediated E-selectin and ICAM-1 expression in human umbilical vein endothelial cells (HUVECs). Confluent HUVECs (grown in 96-well plates) were pretreated with LXA₄ (100 ng/ml, 30 min), followed by IL-1β (1 ng/ml, 4–6 h). E-selectin (CD62E) and ICAM-1 (CD54D2) expression was detected by ELISA and read at 450 nm absorbance. Data represent means ± SE of at least three independent experiments. *P < 0.05, significant difference from IL-1β-mediated E-selectin or ICAM-1 upregulation.

Fig. 3.

LXA₄ inhibits TNF-α-mediated E-selectin and ICAM-1 expression in HUVECs. Confluent HUVECs (grown in 96-well plates) were pretreated with LXA₄ (100 ng/ml, 30 min), followed by TNF-α (10 ng/ml, 4–6 h). E-selectin (CD62E) and ICAM-1 (CD54D2) expression was detected by ELISA and read at 450 nm absorbance. Data represent means ± SE of at least three independent experiments. *P < 0.05, significant difference from TNF-α-mediated E-selectin or ICAM-1 upregulation.

LXA₄ blocks lymphocyte rolling and binding to HUVECs.

Since we observed an effect of LXA₄ on adhesion molecule expression, we tested whether binding of immune cells mediated by these adhesion molecules was also affected. Therefore, we developed an in vitro model to study whether LXA₄ blocks lymphocyte rolling and binding to HUVECs under shear flow conditions. For these studies, HUVECs were incubated with LXA₄ (100 ng/ml, 30 min) and then exposed to IL-1β (1 ng/ml, 4–6 h) and TNF-α (10 ng/ml, 4–6 h). Subsequently, BCECF-labeled Jurkat cells or human CD4⁺ T lymphocytes were perfused onto the monolayers using a microfluidics-based shear assay. As shown in Fig. 4, A and C, pretreatment of HUVECs with LXA₄ abolished Jurkat (Fig. 4A) and CD4⁺ T lymphocyte (Fig. 4C) rolling and binding to HUVECs. These results were quantified and summarized in Fig. 4, B and D, for Jurkat cells and CD4⁺ T lymphocytes, respectively. Our data support the hypothesis that activation of lipoxin receptors by LXA₄ decreases IL-1β- and TNF-α-mediated upregulation of cell adhesion molecules and consequent binding of lymphocytes to vascular endothelium. A demonstrative view of these effects can be seen in Supplemental Videos S1–S4 (Supplemental Material for this article is available online at the Journal website).

Fig. 4.

LXA₄ inhibits IL-1β- and TNF-α-mediated immune cell rolling and attachment to HUVECs. HUVECs were grown on 100-mm plates and preincubated with LXA₄ (100 ng/ml) for 30 min at 37°C. Subsequently, cells were exposed to IL-1β (1 ng/ml) or TNF-α (10 ng/ml) for 4–6 h. The plates were positioned in a flow chamber, mounted, and visualized using a Zeiss Axio Observer Fluorescence Microscopy System. HUVECs were perfused with 2′,7′-bis-(2-carboxyethyl)-5-(and-6)-carboxyfluorescein (BCECF)-labeled Jurkat cells (A and B) or human CD4⁺ T lymphocytes (C and D) at a shear stress of 2 dyn/cm² (flow rate, 0.5 ml/min) for 5 min. The entire period of perfusion was digitally recorded. Interacting cells were calculated as the total of the rolling and adherent cells. Rolling cells were calculated as the ones that move slower than the free stream after attaching to the surface. Adherent cells were counted as the ones that do not change position within 10 s of attaching to the surface. Data summarized in B and D represent means ± SE of at least three independent experiments. *P < 0.05, significant difference with respect to corresponding stimulated HUVECs without LXA₄.

LXA₄ inhibits TNF-α-mediated VCAM-1 upregulation in HSG cells.

The diminished function of exocrine glands in SS is often associated with extensive lymphocytic infiltration of the tissue (45). Lymphocytic cells emigrate from the vascular endothelium into salivary glands (19). Our previous studies indicated that VCAM-1 plays an important role during lymphocyte binding to HSG cells (6). Therefore, we used the HSG model to study whether treatment with LXA₄ inhibited the cytokine-mediated VCAM-1 upregulation. As shown in Fig. 5, TNF-α (100 ng/ml, 6 h) caused a significant upregulation of VCAM-1 expression in HSG cells. However, when cells were preincubated with LXA₄ (100 ng/ml, 30 min), TNF-α failed to induce VCAM-1 upregulation (Fig. 5). These results are summarized in the bar graph and indicate that LXA₄ significantly inhibits the TNF-α-mediated effects on adhesion molecule expression in HSG cells.

Fig. 5.

LXA₄ inhibits TNF-α-mediated VCAM-1 expression in HSG cells. HSG cells were pretreated with LXA₄ (100 ng/ml, 30 min), followed by TNF-α (100 ng/ml, 6 h). Protein expression of VCAM-1 was determined by Western blot analysis; membrane was stripped and reprobed with an antibody against ERK-1/2. Data represent means ± SE of results from three different experiments. Results from a representative experiment are shown at top. *P < 0.05, significant difference from TNF-α-treated cells.

LXA₄ prevents lymphocyte binding to HSG cells.

TNF-α caused a ≈2.5-fold increase in the adherence of Jurkat lymphocytic cells (Fig. 6A) and human CD4⁺ T lymphocytes (Fig. 6C) to HSG cells as compared with untreated cells or LXA₄-treated cells. Strikingly, a 30-min pretreatment of the HSG cell monolayers with LXA₄ significantly blocked the effects of TNF-α on the immune cell binding (Fig. 6, B and D) for Jurkat and CD4⁺ T cells, respectively. These data can be visualized in Fig. 6, A and C, in which a higher amount of fluorescent-labeled Jurkat cells is observed in the TNF-α-treated group as compared with LXA₄-treated or untreated groups (fluorescent-labeled lymphocytes, right panel). In the left panel are shown light micrographs of the HSG cell monolayers at ×10 magnification. These data support the hypothesis that LXA₄ prevents the cytokine-mediated immune cell binding to salivary epithelium. The LXA₄ stable analog ATL effectively blocks as well as counterregulates cytokine production and the adhesion of leukocytes to several cell types at values of 0.1 and 10 ng/ml in vitro (20, 51). Therefore, we assessed whether LXA₄ and ATL prevented TNF-α-mediated Jurkat cell binding to HSG cell monolayers in a similar trend. As shown in Fig. 6, E and F, LXA₄ decreased TNF-α-mediated lymphocyte cell binding to HSG cells with an EC₅₀ of 0.96 ng/ml (Fig. 6E). Similarly, ATL blocked lymphocyte binding with an EC₅₀ of 0.75 ng/ml (Fig. 6F). These data indicate that both compounds are virtually equally potent inhibitors of cytokine-mediated lymphocyte binding to HSG cells.

Fig. 6.

LXA₄ inhibits TNF-α-mediated immune cell binding to HSG cells. Confluent serum-starved HSG cell cultures in 12-well plates were incubated with LXA₄ (100 ng/ml) for 30 min, and monolayers were then incubated both with and without TNF-α (100 ng/ml) for 6 h at 37°C. A–D: binding of PKH2-labeled Jurkat cells (A and B) or CD4⁺ T lymphocytes (C and D) to HSG cell monolayers was performed for 1 h at 37°C. PKH2-bound Jurkat cells or CD4⁺ T lymphocytes (A and C, right) and light micrographs of HSG cell monolayers at ×10 magnification (A and C, left) are shown. Data represent means ± SE of results for Jurkat cells (B) or CD4⁺ T lymphocytes (D) binding from three independent experiments. *P < 0.05, significant difference from TNF-α-treated cells. E and F: LXA₄ (E) and 15-epi-LXA₄ (ATL; F) dose-dependent inhibition of lymphocyte binding to HSG cells. Cells were preincubated with the indicated concentrations of LXA₄ for 30 min at 37°C. Binding of PKH2-labeled Jurkat cells to HSG cell monolayers was determined as described in materials and methods. Data represent means ± SE of results from three experiments.

VCAM-1 specifically mediates immune cell binding to HSG cell monolayers.

Our previous studies indicated that incubation of UTP-stimulated HSG cells with anti-VCAM-1 antibody, but not with anti-ICAM-1 antibody or anti-E-selectin antibody, significantly blocked Jurkat cell binding to HSG cell monolayers (6). To corroborate whether VCAM-1 was the molecule responsible for the cytokine-mediated immune cell binding, HSG cell monolayers treated with TNF-α (100 ng/ml, 4–6 h) were incubated with anti-VCAM-1 or anti-ICAM-1 blocking antibodies as described in materials and methods. The VCAM-1 blocking antibody almost completely abolished Jurkat cell binding to TNF-α-stimulated HSG cells (Fig. 7B). These results can be visualized in Fig. 7A, in which a low amount of Jurkat-labeled cells can be observed in monolayers treated with VCAM-1 blocking antibody as compared with cells incubated in the absence of the antibody or in cells incubated with anti ICAM-1 blocking antibody. In the left panel are shown light micrographs of the HSG cell monolayers at ×10 magnification. These results indicate that VCAM-1 is the adhesion molecule responsible for the majority of the Jurkat cell binding to HSG cells.

Fig. 7.

VCAM-1 blocking antibody completely block cytokine-mediated immune cell binding to HSG cells. A: confluent serum-starved HSG cell cultures in 12-well plates were incubated in the absence (Basal) or presence of TNF-α (100 ng/ml) for 6 h at 37°C. Binding of PKH2-labeled Jurkat cells to HSG cell monolayers was determined as described in materials and methods. For some HSG cell cultures, 10 μg/ml anti-VCAM-1 or anti-ICAM-1 was added to TNF-α-stimulated HSG cells for 45 min before the addition of Jurkat cells. B: data represent means ± SE of results from three experiments. *P < 0.05, significant difference from TNF-α-stimulated cells.

DISCUSSION

Aberrant expression of specific adhesion molecules in salivary tissues from SS patients has been described (2, 43, 65, 81, 89). Adhesion molecules are important contributors to salivary gland inflammation observed in SS because they allow the binding of immune cells that cause tissue destruction (2, 43, 89). The present study indicates that the lipid mediator LXA₄ is able to regulate inflammatory responses in vitro by blocking cytokine-induced upregulation of adhesion molecules and consequent binding of lymphocytes to vascular endothelium and salivary epithelium. These studies are highly significant in light of the prevalence of SS and the current lack of effective treatment.

LXA₄ is able to activate at least two classes of receptors, the ALX/FPR2 and the CysLT1 (which also binds to leukotriene C4 and D4) (67). Here we report that both of these receptors are expressed in HSG cells (Fig. 1, A–C). Additionally, we detected ALX/FPR2 receptor expression in Jurkat cells (Fig. 1A). Previous studies indicated that the ALX/FPR2 receptor is expressed in leukocytes (60), enterocytes (39), intestinal epithelial cells (39), and endothelial primary cells (4, 68). More recently, we found that the ALX/FPR2 is also expressed on the basolateral surface of mice submandibular glands (67a) and intestinal epithelial cells (53). This is interesting, since LXA₄ generation at the paracellular space can rapidly act on the receptor to downregulate local inflammation. The CysLT1 receptor has been shown to be expressed in endothelial native cells (49, 85) and in mesangial cells (90), however, it is absent in Jurkat cells (48). These results indicate that both the ALX/FPR2 and the CysLT1 receptors are expressed in the cell systems studied here.

Under normal conditions, the vascular endothelium serves to maintain blood fluidity and provides a barrier that separates blood cells and plasma factors from highly reactive elements in the underlying tissue (64, 74). Quiescent endothelial cells produce substances to maintain vascular homeostasis (25, 47, 59, 64). However, endothelial cells can shift from a quiescent to a proinflammatory status due to injury or stress in surrounding tissue, largely due to the increased release of cytokines (52, 62). Increased adherence of blood immune cells to cytokine-activated vascular endothelium surrounding the glands may facilitate proinflammatory lymphocyte migration into the salivary glands (2, 22, 75, 89).

Previous studies indicated that LXA₄ displays potent antiinflammatory and antiangiogenic actions on endothelial cells (4, 26). We found that LXA₄ is able to block the expression of E-selectin and ICAM-1 as well as immune cell binding mediated by IL-1β and TNF-α in HUVECs (Figs. 2–4). Our results are consistent with previous studies in PMN/HUVEC cocultures indicating that LXA₄ (10 nM) significantly reduced the LPS-mediated upregulation of LFA-1 in PMN and E-selectin in HUVECs (31). Furthermore, LXA₄ was able to prevent the TNF-α-mediated upregulation of VCAM-1 in HSG cells (Fig. 5). Previous studies indicated that LXA₄ analogs inhibited the lipopolysaccharide (LPS)- and TNF-α-mediated binding of neutrophils to human coronary artery endothelial cells (HCAECs) (27). However, this latter inhibition was mostly due to a downregulation of L-selectin and CD11b/CD18 in neutrophils (27). In the same study, LXA₄ (5 μmol/l) caused only 5 to 9% inhibition of the immune cell binding due to LPS- and TNF-α-mediated adhesion molecules in HCAECs (27). Collectively, these results indicate that LXA₄ effects are cell type specific and can be used to modulate adhesion molecule expression in different cell types.

Adhesion molecule upregulation in salivary glands with SS enables direct interaction with infiltrating lymphocytes (2, 75). Lymphocytic infiltration into the exocrine tissues is a key event in the pathogenesis of SS (1, 34). Binding of lymphocytes to salivary epithelium in SS is associated with an increased production of proinflammatory cytokines such as IL-1β, interleukin-6 (IL-6), TNF-α, and interferon-γ (10, 34, 42, 71). These proinflammatory cytokines cause alteration of epithelial integrity leading to salivary gland dysfunction (5). Previous studies indicate that activation of P2Y₂ nucleotide receptor, and TNF receptor in salivary epithelium upregulates VCAM-1 expression and stimulates lymphocyte adherence (6). Our current studies indicate that LXA₄ is able to prevent the TNF-α-mediated lymphocyte binding to salivary epithelium (Fig. 6, A–E) mediated by VCAM-1 (Fig. 7, A and B). These studies confirm the role of proinflammatory cytokines in immune cell adherence in salivary epithelium and provide the first evidence for direct antiinflammatory effects of LXA₄ in salivary epithelium. Previous studies in intestinal epithelial cell lines and colonic epithelium indicated that LXA₄ and its stable analogs downregulate chemokine secretion (38). LXA₄ also enhanced mucosal antimicrobial protection by stimulating the expression of the bactericidal/permeability-increasing protein (BPI) in oral and intestinal epithelial cells (14), and accelerated epithelial wound closure in murine cornea (40). Additionally, LXA₄ has been shown to be a potent inhibitor of basal to apical PMN transmigration across well-differentiated bronchial epithelial cells (9). Thus, LXA₄ regulates functions that promote inflammatory resolution in many systems in vivo and in vitro (79). The LXA₄ antiinflammatory effects have been observed in a variety of inflammatory diseases including asthma (56), dermal inflammation (41), and arthritis (28). It has been reported that the oral administration of an LXA₄ analog attenuated dextran sulfate sodium-induced colitis (37) and that the oral delivery of another LXA₄ analog (3-oxa-ATL analog ZK-192) exhibited antiinflammatory effects on a 2,4,6-trinitrobenzene sulfonic acid-induced colitis model (33). Additionally, intraperitoneal administration of LXA₄ attenuated the intestinal inflammatory responses induced by LPS (54). Given that LXA₄ also binds to the CysLT1 receptor (67), we determined whether LXA₄ counterregulated inflammatory events mediated by ligands of this receptor (e.g., leukotriene C4 and leukotriene D4). However, the CysLT1 ligands did not affect TNF-α, VCAM-1, or ICAM-1 expression or cytokine release in HSG cells (data not shown). These results suggest that CysLT1 receptor is not involved in the specific effects observed here; however, future comprehensive studies will be necessary to elucidate whether other adhesion molecules or cytokines are modulated by activation of the CysLT1 receptor.

Nanomolar concentrations of a LXA₄ analog reduced NF-κB-mediated transcriptional activation in a LXA₄ receptor-dependent manner and inhibited induced degradation of IκBα in intestinal epithelial cells (37). LXA₄ blocked IL-1β-mediated ICAM-1 expression in human astrocytoma cells via NF-κB (20). Additionally, LXA₄ reduced LPS-induced inflammation in macrophages and intestinal epithelial cells through inhibition of NF-κB activation (54). These studies indicate that antiinflammatory effects of LXA₄ may be dependent on suppression of NF-κB activation, and this may be a common mechanism between vascular endothelial cells and salivary epithelial cells. NF-κB activation is the rate-controlling step for inflammatory processes (54) and has been linked to the development of autoimmune diseases such as Type 1 diabetes (55), systemic lupus erythematosus (24), and rheumatoid arthritis (12). Interestingly, in minor salivary glands with SS, an increase in the nuclear translocation of factor NF-κB has been observed (58). Furthermore, IκBα^M/M mice have reduced life span and developed SS (72). These results suggest that NF-κB activation may be one of the main mechanisms behind exocrine inflammation found in SS and that the novel lipid mediator LXA₄ may contribute to regulation of adhesion molecule expression that contributes to immune cell infiltration in exocrine tissues.

In summary, our results demonstrate that resident LXA₄ circuits are key determinants for the degree of inflammatory cell binding in response to proinflammatory cytokines in both endothelium and epithelium. In view of the prominent expression of the lipoxin receptor in endothelial and epithelial cells and in lymphocytes, the endogenous role of LXA₄ circuits in the development of SS clearly warrants further studies.

GRANTS

This work was supported by National Institutes of Health (NIH)-National Institute of Dental and Craniofacial Research Grant R21-DE019721-01A1 and NIH-National Heart, Lung, and Blood Institute Grant R01-HL063014.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

S.C., O.O., N.M., J.N., S.N., and O.J.B. performed the experiments; S.C., O.O., N.M., S.N., and O.J.B. analyzed the data; S.C., J.N., S.N., and O.J.B. interpreted the results of the experiments; S.C., N.M., J.N., and O.J.B. prepared the figures; S.C. drafted the manuscript; S.C., O.O., N.M., J.N., S.N., and O.J.B. approved the final version of the manuscript; O.O., J.N., S.N., and O.J.B. edited and revised the manuscript; S.N. and O.J.B. conception and design of research.