Published ahead of print on February 5, 2003, doi:10.1164/rccm.200207-654OC
© 2003 American Thoracic Society
Biologically Active Intercellular Adhesion Molecule-1 Is Shed as Dimers by a Regulated Mechanism in the Inflamed Pleural SpaceInstitute of Respiratory Physiopathology, Italian National Research Council; Institute of General Medicine and Pneumology, University of Palermo, Palermo, Italy; and Department of Microbiology and Infectious Diseases and Department of Internal Medicine, University of Calgary, Calgary, Alberta, Canada Correspondence and requests for reprints should be addressed to Mark Gjomarkaj, M.D., Istituto di Fisiopatologia Respiratoria, Consiglio Nazionale delle Ricerche, Via Ugo La Malfa, 153-90146 Palermo, Italy. E-mail: marco{at}ifr.pa.cnr.it
Intercellular adhesion molecule-1 (ICAM-1) is an adhesion molecule that plays a crucial role in cell–cell interactions involved in the recruitment of cells and immune responses. Under some circumstances, ICAM-1 is found as a soluble protein that has the potential to influence the nature of immunoinflammatory responses. By examining cells and fluid from the pleural compartment of patients with cancer, tuberculosis, and congestive heart failure, the cellular source, conformation, control, and biological activity of soluble ICAM-1 (sICAM-1) were investigated. The results suggest that dimeric sICAM-1 was released locally in the pleural compartment of tuberculous and malignant effusions. sICAM-1 was shed from preexpressed surface ICAM-1 rather than produced de novo, and both CD45-positive leukocytes and cytokeratin-positive epithelial and mesothelial cells expressed ICAM-1, suggesting multiple cellular sources for sICAM-1. The expression of sICAM-1 was regulated because pleural macrophages caused release of sICAM-1 via a tumor necrosis factor-  -dependent mechanism. The functional significance of sICAM-1 was demonstrated by showing that pleural sICAM-1 interfered with conjugate formation between LAK cells and K562 cells, suggesting that pleural sICAM-1 plays an immunosuppressive role by inhibiting adhesion of cytotoxic lymphocytes and tumor cells. Thus, sICAM-1 is shed from the surface of cells in a regulated manner and has the potential to influence the immune response in the pleural space.
Key Words: intercellular adhesion molecule-1 • lung cancer • mononuclear phagocyte • pleural effusion • tuberculosis Despite the recognition of a soluble form of intercellular adhesion molecule-1 (sICAM-1), the cellular source, protein conformation, factors controlling its release, and physiologic importance of sICAM-1 in disease states remain unclear.
ICAM-1 is a member of the immunoglobulin supergene family that is widely distributed among hematopoietic and nonhematopoietic cells (1). Although both monomeric and dimeric forms of ICAM-1 are present on the cell membrane, the dimeric form is the predominant one (2, 3) and is able to bind leukocyte function associated-1 (LFA-1) molecules with greater affinity compared with monomers (3). Because of the ability of surface-expressed homodimers of ICAM-1 to act as a ligand for LFA-1 and MAC-1, it plays a crucial role in intercellular interactions involved in the development of inflammatory responses (2). Accordingly, ICAM-1 is actively expressed on the surface of leukocytes, and the expression of ICAM-1 is upregulated by inflammatory cytokines such as tumor necrosis factor (TNF)- Elevated concentrations of sICAM-1 have been detected in serum and various extracellular fluids in a variety of inflammatory and neoplastic diseases (6, 7) and it has been demonstrated to play an important role in the regulation of some immune response mechanisms. In this regard we have previously shown that sICAM-1 that is produced by a lung cancer cell line inhibits cell adhesion and cytotoxicity between LAK cells and a lung cancer cell line (8). The pleural space is frequently involved in inflammatory and neoplastic diseases that may compromise both pleural vascular and mesothelium permeability and lead to the accumulation of fluid enriched in cells and proteins into the pleural space, that is, the development of an exudative pleural effusion (9, 10). As a closed compartment that is involved in a variety of different human diseases, the pleural space has many advantages that facilitate the study of immune responses. Different diseases that result in pleural effusions present an ideal opportunity to study different models of the inflammatory mechanisms that are used within the pleural compartment. Although pleural effusions are common, the mechanisms involved in the pleural immune response are not well elucidated; in particular, little is known about the role that is played by adhesion molecules in promoting the intercellular interactions that may contribute or inhibit the development of pleural inflammation. It is not clear which cells produce sICAM-1 in the pleural compartment. Nor has it been clearly established whether sICAM-1 is produced de novo and immediately released or initially expressed on the surface of pleural cells and then shed from their surface. It is also not known whether this event is regulated and whether sICAM-1, which is produced in vivo, could play a role in the regulation of inflammatory and antitumor responses within the pleural space. The aims of this study were (1) to evaluate whether sICAM-1 is produced locally within the inflamed pleural space, (2) to identify the mechanism of release and cell source(s) and protein conformation of pleural sICAM-1, and (3) to investigate the possible role(s) of sICAM-1 within the inflamed pleural space. Moreover, because pleural macrophages (PleMs) have been demonstrated to play a role in the regulation of inflammatory and immune reactions within the pleural space (11, 12), we evaluated the role played by PleMs in the modulation of ICAM-1 expression and sICAM-1 release by pleural effusion cells.
Pleural Fluid and Blood Sample Collection Pleural fluid was collected by therapeutic thoracentesis from hospitalized patients with congestive heart failure (CHF) (n = 9 [6 males and 3 females]; age range, 56 to 84 years), tuberculosis (TB) (n = 8 [6 males and 2 females]; age range, 26 to 71 years), and cancer (n = 15 [10 males and 5 females]; age range, 39 to 81 years). All subjects gave informed written consent and the study was approved by the institutional review board for human studies. Standard clinical, laboratory, and radiological investigations established the diagnosis for each patient as previously described (12). In detail, CHF effusions were defined as transudates associated with an enlarged heart, distended neck veins, and cardiac gallop that improved with therapy for the congestive heart failure; tuberculous effusions were defined as exudates with positive culture for Mycobacterium tuberculosis or with positive smear for acid-fast bacilli of pleural fluids; malignant effusions were defined as exudates associated with a pathologic diagnosis of cancer from cytologic examination of pleural fluids and from pathologic examination of lung tissues. The effusions were classified as transudates or exudates by meeting at least one of the criteria described by Light and coworkers (13). The fluids were drawn into polypropylene tubes containing heparin (10–20 IU/ml). Red blood cells were differentiated from nucleated pleural cells by Turk's stain. The total cell counts and cell viability (more than 95%) were performed by trypan blue dye exclusion. Differential cell counts were performed by using the following monoclonal antibodies (mAbs): anti-CD68, specific for mononuclear phagocytes (Dako, Glostrup, Denmark); anti-cytokeratin (MNF116), specific for epithelial and mesothelial cells (Dako); anti-human neutrophil elastase (clone NP57; Dako), specific for neutrophils; anti-human eosinophil cationic protein (Pharmacia and Upjohn Diagnostics, Uppsala, Sweden), specific for eosinophils; anti-CD3, specific for T lymphocytes (Dako); anti-CD20, specific for B lymphocytes (Dako); anti-CD16 (clone DJ130c), specific for natural killer (NK) cells (Dako); and isotype-matched immunoglobulins (Dako) as negative controls. The identification of cancer cells was performed by the Department of Pathology of our hospital (Ospedale V. Cervello, Palermo, Italy) according to standard diagnostic procedures, using a panel of mAbs, including anti-human calretinin (Dako), as recommended by the current consensus (14, 15). The pleural fluids were centrifuged at 400 x g for 10 min. Cell-free fluids were immediately frozen at –70°C until they were analyzed for subsequent experiments. Autologous serum specimens were obtained from blood samples collected from patients on the same day as thoracentesis and from healthy control subjects (n = 11 [5 males and 6 females]; age range, 25 to 59 years) after informed written consent had been obtained.
Detection of sICAM-1 in Pleural Fluid and Serum
Immunoprecipitation and Western Blotting of sICAM-1 in Pleural Fluid
Depletion of sICAM from Pleural Fluids
Phenotypic Characterization of ICAM-1-positive Cells
Isolation of Macrophages from Pleural Effusions
Modulation of ICAM-1 Expression and sICAM-1 Release and Shedding
To evaluate the ability of PleMs to modulate ICAM-1 expression and sICAM-1 release and shedding, both unbiotinylated and biotinylated pleural cells were cultured in CM (24 hours, 37°C, 5% CO2) in the presence or absence of PleM culture supernatants. LPS from E. coli (Sigma) (1 µg/ml), recombinant TNF- Pleural cells were recovered from the bottom of the tubes and tested for the expression of surface ICAM-1 by cytofluorimetric analysis, as described above. Data from these experiments are expressed as mean fluorescence intensity.
Conjugate Formation Assay
Statistical Analysis
Expression of sICAM-1 in Pleural Effusions Table 1 shows the main characteristics of pleural effusions; as expected, effusions recovered from patients with TB (n = 8) and cancer (n = 15) had higher total cell numbers as well as sICAM-1 and albumin concentrations when compared with effusions from patients with CHF (n = 9).
To determine whether sICAM-1 was produced locally or accessed the compartment by extravasation from the blood into the pleural space, the concentrations of sICAM-1 and albumin were compared in pleural fluids and in sera. Albumin is of lower molecular weight and was selected as a marker of vascular/mesothelial permeability. Therefore, if the sICAM-1/albumin ratio were the same in pleural fluid and sera, it would suggest a vascular and a mesothelial leak, whereas a ratio that was higher in pleural fluid than in sera would suggest local production of sICAM-1. sICAM-1/albumin ratios were significantly higher in pleural fluids than in autologous serum in patients with TB (Figure 1A) and cancer (Figure 1B); in marked contrast, no significant difference was detected between sICAM-1/albumin ratios in pleural fluid and autologous serum in patients with CHF (Figure 1C). The serum sICAM-1 concentrations, evaluated in normal subjects, were 5.5 ± 1.5 ng/mg of albumin and were significantly (p < 0.05) lower than the serum sICAM-1 concentrations in the other three study groups. These data suggest that sICAM-1 is produced locally in tuberculosis and cancer, but that it is due to leak from the blood in patients with CHF.
Characterization of sICAM-1 in Pleural Effusions To evaluate whether sICAM-1 was present in pleural fluids in a monomeric or in a dimeric form, we performed Western blot analysis to characterize its molecular weight. Because the treatment with reducing agents (DTT) used in the Western blot analysis may affect the stability of dimeric sICAM-1 molecules, we tested the presence of sICAM-1 in pleural fluids under both nonreducing (in the absence of DTT) and reducing (in the presence of DTT) conditions. Under nonreducing conditions, native pleural sICAM-1 showed only a strong dimeric band at about 190 kD, and no other significant bands were detected at different molecular weights. Under reducing conditions, no band was detected at 190 kD whereas two bands were detected at about 60 and 95 kD, respectively, corresponding to different monomeric and/or fragmented dimeric sICAM-1 forms (Figure 2) . These results demonstrate that sICAM-1 is present, in inflammatory pleural effusions, in a dimeric form of about 190 kD and that the other forms detected under reducing conditions are due to technical artifacts.
Correlation between sICAM-1 Dimers and Membrane Surface ICAM-1 in Pleural Effusions sICAM-1 dimers could be released into the pleural fluid by two mechanisms. They could be expressed on the surface of cells and shed from cells into the fluid, or they could be synthesized and released de novo from cells that did not express ICAM-1 on their surface. To examine the possibility that ICAM-1 was shed from cells that express it on their cell surface, we first determined whether sICAM-1 levels correlated with the expression of surface ICAM-1. Cells from all three patient groups expressed ICAM-1 (Table 2) . Although there was no significant difference in the percentage of cells expressing ICAM-1 among the three study groups, the number of cells per volume of pleural fluid expressing ICAM-1 was significantly higher in patients with cancer and patients with TB than in patients with CHF (Table 2). Further, a positive correlation was observed between the number of cells expressing ICAM-1 on their surface and the sICAM-1/albumin ratio in pleural fluid (Figure 3) . This suggested that the increased levels of sICAM-1 dimers in pleural fluids of patients from these two study groups were due to cells expressing and releasing ICAM-1.
Cell Type Expressing ICAM-1 in Pleural Effusions To evaluate the cell type that expressed, and was therefore capable of releasing, sICAM-1 dimers, cells recovered from pleural effusions were labeled with anti-ICAM-1 and with anti-cytokeratin mAb specific for mesothelial and epithelial cells or anti-CD45 for leukocytes. In patients with TB the ICAM-1 molecule was present almost exclusively on the surface of CD45-positive cells (Figure 4A) , suggesting that lymphocytes, the overwhelmingly predominant cell population in TB effusions (Table 1), were responsible for the presence of sICAM-1 dimers within the pleural space in patients with TB. In contrast, in patients with cancer, cells that were CD45 positive expressed a low level of surface ICAM-1 (Figure 4B), whereas cells that were cytokeratin positive expressed high levels of surface ICAM-1 (Figure 4B), suggesting that a variety of cells of both bone marrow and mesenchymal origin could contribute to the release of sICAM-1 dimers within the pleural space in patients with cancer.
Mechanism of sICAM-1 Dimer Production: Role of Pleural Macrophages We next determined whether the sICAM-1 dimers present in the fluids were shed from the surface of ICAM-1-positive cells and whether this mechanism was regulated. For this purpose, pleural macrophages were used as a source of soluble factors that might alter surface ICAM-1 expression and sICAM-1 dimer production by pleural effusion cells.
Only negligible concentrations of sICAM-1 were detected in conditioned medium from PleMs (data not shown). However, incubation of pleural cells with PleM-conditioned medium dramatically increased the release of sICAM-1 by pleural effusion cells (Figure 5A)
. Anti-TNF-
To determine whether the sICAM-1 detected in pleural culture supernatants was shed from the surface of pleural cells, the cell surface proteins were biotinylated and culture supernatants were analyzed for biotinylated ICAM-1. We reasoned that if sICAM-1 was released from the cell surface, it would be biotinylated, whereas if it were released from within the cell or by de novo synthesis, it would not be biotinylated. Immunoblot analysis, performed under nonreducing conditions, for biotinylated proteins that had been immunoprecipitated with anti-ICAM-1 demonstrated the presence of a biotinylated band at about 190 kD, corresponding to the native sICAM-1 dimers, only after incubation with PleM-conditioned medium (Figure 5B). In addition, conditioned medium from LPS-stimulated PleMs caused a statistically significant increase in surface ICAM-1 expression on pleural effusion cells when compared with cells cultured in unconditioned medium (Figure 6) . Moreover, anti-TNF- mAb significantly reduced the percentage of cells expressing ICAM-1, suggesting that the presence of TNF- in PleM culture supernatants was responsible, at least in part, for the expression of ICAM-1 by pleural effusion cells. However, incubation with either recombinant TNF- or LPS alone was also able to modestly, but significantly (p < 0.05), upregulate ICAM-1 expression by pleural cells when compared with baseline. This upregulation was significantly lower (p < 0.05) than that generated by PleM culture supernatants (Figure 6). Thus, pleural macrophages were able to induce the shedding of dimeric sICAM-1 as a 190-kD protein from the surface of pleural cells and to upregulate the expression of ICAM-1 on the surface of pleural cells. Moreover, these results suggest that the control of ICAM-1 shedding is more complex than the control of ICAM-1 surface expression and is not merely mediated by TNF and LPS alone.
Role of Pleural sICAM-1 in Adhesion of Cytotoxic Lymphocytes and Tumor Cells To determine whether the sICAM-1 might be active and play a role in the immune response, experiments were performed to determine whether sICAM-1 inhibited conjugate formation between LAK effector cells, which were generated from healthy adults, and K562 target cells. K562 cells were used because these cells express high levels of surface ICAM-1 molecules, which are required for cellular cytotoxicity (23). We compared the ability of untreated pleural effusion with the ability of sICAM-1-depleted pleural effusion to affect the formation of conjugates. When untreated pleural fluid was added to the LAK effector cells and K562 target cells there was a slight augmentation in conjugate formation, possibly because of immunostimulation by the pleural fluid. On depletion of sICAM-1 there was a significant increase in conjugate formation when compared with the addition of untreated pleural fluids (Figure 7A) , indicating that sICAM-1 present in pleural fluids plays a role in the inhibition of cell–cell adhesion. As expected, the addition of recombinant sICAM-1, anti–ICAM-1 or anti–LFA-1 mAbs significantly inhibited the formation of conjugates when compared with baseline (Figure 7B).
The current studies have utilized cells and fluid from the pleural space to demonstrate that (1) sICAM-1 is produced locally by cells within an inflammatory compartment (the pleural space); (2) sICAM-1 is present in pleural fluid as a dimer and it is shed from the surface of cells that express surface ICAM-1; (3) shedding of sICAM-1 dimers occurs in a regulated fashion via the products of stimulated macrophages; and (4) sICAM-1 is capable of influencing immune responses. We first demonstrated that pleural effusions recovered from patients with TB and cancer have significantly higher concentrations of sICAM-1 than effusions recovered from patients with CHF. These findings confirm and extend the previous observations by Hoffmann and coworkers (24), who demonstrated a similar increase in exudative pleural effusions of heterogeneous etiologies, compared with transudative effusions.
The pleural compartment is separated from the blood circulation by two different limiting membranes, represented by the vascular endothelium and the pleural mesothelium, respectively, both contributing to limit the passage of cells and of soluble proteins. To clarify whether increased sICAM-1 concentration in pleural effusions was due to local production of this molecule or simply due to increased extravasation of serum sICAM-1 into the pleural space, we normalized sICAM-1 to albumin concentrations both in pleural effusions and in sera. Albumin is considered a good marker of capillary permeability in the pleura, because there is no evidence that albumin is produced or stored in pleural mesothelium. Further, the pleural microvascular endothelium is semipermeable (9); albumin originates only from serum and reaches different organ compartments via passive diffusion (25); the concentrations of total protein and albumin in pleural fluid are lower than in serum (26); and albumin and sICAM-1 have similar molecular weights. Therefore, sICAM-1 concentrations were normalized to albumin to gain insights into local sICAM-1 production. The ratio of sICAM-1 to albumin was elevated in TB and cancer, suggesting that sICAM-1 is produced locally in the pleural space during inflammatory processes, whereas no local production of sICAM-1 could be demonstrated in the absence of active inflammation, that is, in transudative effusions from patients with CHF. Interestingly, the serum sICAM-1 concentrations in patients with CHF were higher than in normal subjects. In this regard, it has been demonstrated that the concentrations of TNF- To characterize the molecular weight of sICAM-1 present within the pleural fluids we performed immunoblot analysis under both nonreducing and reducing conditions because it was demonstrated that under reducing conditions, the ICAM-1 dimeric molecule dissociates, yielding monomers or fragments that might not natively be present (28). Under nonreducing conditions the immunoblot identified a band of about 190 kD, with no other significant bands identified at lower molecular weights. Under reducing conditions two bands were identified at about 60 and 95 kD, corresponding to different monomeric and/or fragmented dimeric sICAM-1 forms. In this regard it has been demonstrated that the dimeric extracellular forms of sICAM-1 may be represented by U-shaped and by ringlike dimers. Moreover, two U-shaped dimers may compose a W-shaped tetramer. All these forms of polymeric sICAM-1 have molecular weights ranging from 112 and 200 kD (28). Accordingly, a reducing agent may break a dimeric native pleural sICAM-1 into monomers (MW 60 kD) or into a partially fragmented dimer (MW 95 kD). These findings demonstrated that sICAM-1 was present in pleural fluid in a native dimeric form and that the presence of smaller sICAM-1 molecules is due to technical artifacts related to the use of reducing agents (DTT). Soluble dimers of ICAM-1 may have important implications for immune responses. Membrane ICAM-1 is expressed predominantly as a dimer rather than as monomers (2, 3). Dimers have increased affinity for their ligand as demonstrated by an engineered form of soluble ICAM-1 dimer that has increased affinity for rhinovirus (one of the ligands for ICAM-1) compared with monomers of ICAM-1 (29), and reduced LFA-1-dependent transendothelial chemotaxis of peripheral blood mononuclear cells (30). ICAM-1 has two dimerization sites. The first is Val-51 in strand E of domain 1 (31). This appears to be critical for dimerization because this position is occupied by valine or isoleucine in all ICAM-1 molecules, whereas in ICAM-2, which does not dimerize, the amino acid at this position is hydrophobic (3, 31). The second dimerization site is within the transmembrane domain (2, 3). Although recombinant ICAM-1 is usually present in solution as monomers, we show that in a biological fluid it is a dimer. There are a number of possible mechanisms for this dimerization. The first is that Val-51 may be sufficient for dimerization, which is contrary to prior opinion (31). The second is that ICAM-1 dimers in pleural fluid may contain a portion of the transmembrane domain that contributes to dimerization. Finally, a ligand for ICAM-1 may contribute to the dimerization. The events that lead to this process will require further study. Because the dimeric form is the most biologically active form of sICAM-1, it prompted us to study the mechanisms of release and activity of this molecule within the pleural space. It has been demonstrated that cells that are capable of expressing membrane ICAM-1 may upregulate the expression of this molecule after stimulation, and may also increase the subsequent release of sICAM-1 (3). To test the hypothesis that sICAM-1 might be due to shedding of ICAM-1 from the cell surface, we first sought a correlation between the expression of ICAM-1 on cell membranes and sICAM-1 levels. We found that the sICAM-1/albumin ratio in pleural fluid correlated with the number of pleural effusion cells expressing surface ICAM-1. Further, by labeling the surface molecules of the pleural cells with biotin before the incubation with PleM culture supernatants, we were able to show that an approximately 190-kD band was immunoprecipitated with an anti-ICAM-1 antibody. This suggests that the sICAM-1 is cleaved from the surface of pleural cells and shed as an intact surface extracellular dimeric molecule. To verify whether structural or immunocompetent pleural cells expressed surface ICAM-1, we labeled cells with an anti-ICAM-1 mAb together with an anti-leukocyte (CD45) or an anti-cytokeratin mAb. In TB effusions, surface ICAM-1 was expressed, almost exclusively, on CD45-positive cells. Because it has been demonstrated that activated T lymphocytes express ICAM-1 (6) and T lymphocytes represent the majority of cells present in tuberculous pleurisy (32), it suggests that a compartmentalized immune response occurs at the site of disease and that T lymphocytes may actively contribute to the release of sICAM-1. Moreover, it has been demonstrated that in pulmonary sarcoidosis, another granulomatous disorder characterized by the activation of T lymphocytes, sICAM-1 concentrations correlate with the percentage of lymphocytes in bronchoalveolar lavage (33).
With regard to cancer effusions, we demonstrated that surface ICAM-1 is present on CD45-positive as well as cytokeratin-positive cells. Lymphocytes and macrophages represent the main cell population expressing surface ICAM-1 among CD45-positive leukocytes (6). With regard to cytokeratin-positive cells, we characterized our cytokeratin-positive cell population with an anti-cytokeratin mAb specific for cytokeratin MNF16 present on the surface of both mesothelial and cancer cells. In cancer effusions, all the cytokeratin-positive cells were also positive for the ICAM-1 molecule. Because every cancer effusion was specifically characterized in terms of the percentage of cancer cells by using a panel of mAbs including calretinin (see METHODS) and because both cancer (8) and mesothelial cells (34) express ICAM-1 on their surface, it is conceivable that both mesothelial and cancer cells constitute the ICAM-1-positive/cytokeratin-positive cell population present in cancer effusions. We have previously demonstrated that lung cancer cell lines express membrane ICAM-1 and release sICAM-1 after incubation with cytokines, including TNF-
The data presented here suggest that PleMs play a crucial role in regulating ICAM-1 signaling. In the presence of TNF- Although ICAM-1 is considered to play an important role in the immunologic host response against tumor cells and in the metastatic process (1), the role of sICAM-1 in cell-mediated host defense, within the pleural space, was not previously defined. In this regard, it has been suggested that the release of sICAM-1 might be one of the mechanisms by which cancer cells escape physical interactions with LFA-1-positive cytotoxic cells (5, 8). Consistent with this hypothesis, it has been demonstrated that, in patients with lung cancer, serum levels of sICAM-1 are associated with advanced tumor stages and metastatic potential (7, 37). Our finding that the in vitro adhesion between cancer and LAK cells was reduced by the addition of pleural fluid containing sICAM-1 compared with fluid that had been depleted of sICAM-1 suggests that, within the pleural space, sICAM-1 downregulates the interaction between cytotoxic T cells and tumor cells. To our knowledge, this is the first report demonstrating a functional role of pleural sICAM-1 as suppressive factor in host–cancer interactions. Further, it has been demonstrated that dimerization of sICAM-1 is required for optimal binding to its receptor (28). Accordingly, our data demonstrate that dimeric sICAM-1, the form of sICAM-1 present in pleural fluids, is capable of inhibiting intercellular adhesion. In conclusion, our study suggests that in the pleural space, during inflammation, the compartmentalized production of dimeric sICAM-1 is related to the expression of membrane ICAM-1 by heterogeneous inflammatory, structural, and neoplastic cells. sICAM-1, within the pleural space, may play a suppressive role in the ICAM-1-dependent interaction between cytotoxic lymphocytes and cancer cells. PleMs play a crucial role in the regulation of this phenomenon by releasing soluble factors capable of upregulating ICAM-1 expression and sICAM-1 release by other pleural effusion cells.
Supported by the Italian National Research Council and by the Alberta Lung Association. C.H.M. is a Senior Scholar of the Alberta Heritage Foundation for Medical Research. This article has an online supplement, which is accessible from this issue's table of contents online at www.atsjournals.org Received in original form July 4, 2002; accepted in final form January 15, 2003
This article has been cited by other articles:
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
TOP
ABSTRACT
METHODS
(
= 0.86; p < 0.0002.
