Detection of soluble interleukin-2 receptor and soluble intercellular adhesion molecule-1 in the effusion of otitis media with effusion

We measured sIL-2R, TNF-α and sICAM-1 in the sera and middle ear effusions (MEEs) of patients with otitis media with effusion (OME). Although there was no signmcant difference between the sIL-2R levels of the serous and mucoid MEEs, they were significantly higher than serum sIL-2R levels of OME patients and healthy controls. TNF-α levels of the mucoid MEEs were significantly higher than those of the serous type. However, TNF-α was rarely detected in the sera of OME patients or healthy controls. We observed significant differences between the serous and mucoid MEEs with respect to their sICAM-1 levels, which were also higher than serum slCAM-1 levels of OME patients and healthy controls. Our findings suggested that IL-2, TNF-α and ICAM-1 could be significantly involved in the pathogenesis of OME through the cytokine network.


Introduction
submucosa of the middle ears. 18 '19 Cytokines form a network together with inflammatory cells Interleukin-2 (IL-2) is one of the most well and adhesion molecules, and play important characterized cytokines. It stimulates the pro-roles in the inflammatory disorders. It is necesliferation of activated T-lymphocytes, 1'2 and binds sal T to investigate the effects of cytokines and to the IL-2 receptor (IL-2R) which is composed adhesion molecules on the pathogenesis of OME. of three distinct subunits, namely a, [3 and T. However, there have been few reports about slLchains. [3][4][5] In patients with rheumatoid arthritis, 6 2R and slCAM-1 in cases with OME. Therefore, adult T cell leukaemia, 7 Kawasaki disease 8 and we studied slL-2R, TNF-a and slCAM-1 in the bronchial asthma, 9 the serum levels of soluble ILmiddle ear effusion and the sera of patients with 2R (slL-2R), which is the released extracellular OME.
domain of the a-chain of IL-2R, are elevated.
Moreover, activated T-lymphocytes are reported not only to elaborate interferon-gamma (INF-T) Materials and Methods and to cause macrophages to produce turnout necrosis factor-alpha (TNF-a), 11 but also to reg-Subjects: We collected middle ear effusions ulate the expression of intercellular adhesion (MEEs) using Juhn Tyro-Taps (Xomed-Treace, molecule-1 (ICAM-1) on the endothelium indir-Jacksonville, FL, USA) from 31 patients with ectly through such media as TNF-a and INF-T, OME (17 males and 14 females, ranging in age resulting in the recruitment of inflammatory from 3 to 79 years with a mean of 22.9 years) cells. 2 ICAM-1 belongs to the immunoglobulin when they had undergone myringotomy. The supergene family 13 and acts a ligand for leuko-MEEs were classified into two groups, the cyte function-associated antigen-1 (LFA-1) 14 and serous and the mucoid types. While 16 patients integrin corresponding to the [32 group (Mac-(nine males and seven females with a mean age 1).15 It plays important roles in intercellular inter-of 22.9 years, ranging from 4 to 76 years) action when leukocytes adhere and transmigrate belonged to the serous type, 15 others (eight to a focus. Thus, soluble ICAM-1 (slCAM-1), a males and seven females with a mean age of soluble form of ICAM-1, is detected in the focus 23.0 years, ranging from 3 to 79 years) belonged because of the shedding of ICAM-1 as well as the to the mucoid type. Serum samples were also expression of lCAM-1 on the cell surface induced obtained from all 31 patients with OME, and by TNF-a and INF-T. 16'7 from ten healthy volunteers who served as con-In otitis media with effusion (OME), pre-trols, MEEs and sera were collected with the dominant neutrophils with lymphocytes and macinformed consent of the patients and/or their rophages are seen in the effusion and the families. PBS. These samples were centrifuged at 1500 x g for 10 min, and the supernatants were collected and stored at-80C until they were assayed. The sera were also stored in the same way.
The assays of slL-2R, TNF<z and slCAM-1 were performed in duplicate with cell-free interleukin-2 receptor bead assay kit (T cell Diagnostic Co., MA, USA), human TNF-z immunoassay kit (Research and Diagnostics Systems, MN, USA) and human ICAM-1 ELISA kit (Serotec, Oxford, England), respectively. The concentrations of slL-2R, TNF-z and slCAM-1 in the test samples were determined from the calibration curve.
For the measurement of slL-2R, TNF-z and slCAM-1 in the MEE, the effect of the MEE itself on these assays was preliminarily tested but no significant effect was found. Briefly, three kinds of stock solutions obtained from surplus MEEs were prepared for the test. An equal volume of each of standard slL-2R, TNF<z and slCAM-1 was added to each stock solution and the actual increase in slL-2R, TNF<z and slCAM-1 concentrations was compared with the calculated value.
Statistical analysis: The statistical significance of the difference between the recorded values was determined at p < 0.01 by Student's t-test.

Results
The levels of sIL-2R in the MEEs and sera are illustrated in Fig. 1. The mean level of sIL-2R in the mucoid MEEs was 15000U/ml, while the mean level in the serous MEEs was 9500 U/ml. There was no significant difference between the slL-2R levels of the two types of MEEs. However, the slL-2R levels in both types were significantly higher than those of the sera of OME patients, which was 326U/ml on the average. Moreover, there was no significant difference between the sera of OME patients and those of the healthy controls (310 U/ml).
The levels of TNF<z in the MEEs and sera are illustrated in Fig. 2. The mean level of TNF-z in the mucoid MEEs was 300pg/ml, whereas the  The levels of slCAM-1 in the MEEs and sera are illustrated in Fig. 3. The mean level of sICAM-1 in the mucoid MEEs was 1440 ng/ml, whereas the mean level in the serous MEEs was 430 ng/ ml. Thus, a significant difference was observed between slCAM-1 levels of the two types of MEEs. Furthermore, the levels of slCAM-1 in the two types of MEEs were significantly different from those of the sera of the OME patients because the mean serum level was 180ng/ml. However, there was no significant difference between the sera of the OME patients and those of the healthy controls (180 ng/ml).

Discussion
The slL-2R levels in the serous and mucoid types of MEEs were high, their mean values being 9500 U/ml and 15 000 U/ml, respectively. These values were extremely high as compared with the sIL-2R levels of bronchoalveolar lavage fluid (BALF) of asthma (42U/ml) and miliary tuberculosis (1320U/ml), 9 nasal secretion of nasal allergy (1510 U/ml) 2 and synovial effusion of rheumatoid arthritis (1170 U/ml). 6 Even in the severe cases of acute asthma, 22 extensive pulmonary tuberculosis 2 and acute-phase Kawasaki disease, s it was reported that the mean values of serum slL-2R levels were about 570 U/ml, 2700 U/ ml and 3700 U/ml, respectively. However, slL-2R level of acute adult T cell leukaemia which was beyond 10000 U/ml in serum, 7 was as high as that of MEE. In our study, the mean level of slL-2R in the sera of OME patients was 326 U/ml, which was the same as that of the healthy controis (310U/ml). These data suggest that the expression of IL-2R on activated T lymphocytes and the binding of IL-2 to its receptor might be concerned with local pathogenesis of OME in the middle ear. Yellon et a/. 24 also reported that IL-2 was detected in the MEEs of 54% of OME patients. However, there is a possibility that IL-2 might be biologically a little unstable or technically difficult to detect as compared with slL-2R, because slL-2R was detectable in all our cases.
There are a few reports which claim that TNFcz was detected in the MEEs of OME, [24][25][26] although its values were expressed with the unit of pg/mg of total protein. Particularly, the mucoid type effusion is too viscous to dissolve in PBS. In our methods, we used dithiothreitol and liquefied the mucoid pellet to dissolve it in PBS. 2 We expressed TNF<z levels in the effusion with the unit of pg/ml and could compare TNF<z values with those of other diseases. In the BAIs 27 and sputa 28 of patients with active asthma, mean TNF-z levels were 578 pg/ml and 1783 pg/ml, respectively. In children with leukaemia, the mean level of TNF-cz in the sera was 63.6pg/ml while it was 21.5 pg/ml in the solid tumours. 29 We found that the mean TNF-z level in the mucoid type effusions was similar to that in BALFs of asthma patients.
Chihara et al. reported that slCAM-1 levels of sputa and sera of asthma patients were 8.6 ng/ml and 350ng/ml, respectively, and that the slCAM-1 level in the normal human sera was 200 ng/ml. It means that the mean serum level of slCAM-1 in asthma was 1.75 times higher than that in the healthy control and that its sputum level could be disregarded as it was too low as compared with the serum level of control subjects. In our study, slCAM-1 levels in the serous and mucoid MEEs were 5.3 and 7.8 times higher than those in the control sera. slCAM-1 levels in various other diseases have also been reported. In patients with rheumatoid arthritis, serum and synovial fluid levels of slCAM-1 were 1.2 and 1.5 times higher than the serum levels of the normal controls, respectively. Serum levels of sICAM-1 in patients with nasal allergy 32 and malignant dis- 16 eases with metastasis were 1.4 and 5.7 times higher than those of the healthy controls, respectively. These data indicate that slCAM-1 levels detected in the MEEs of OME were very high while the serum levels were the same as those of the healthy controls. Himi et al. also reported that slCAM-1 levels of MEEs were significantly higher than those of the healthy control sera. However, they could not make any distinction between the various types of the MEEs because sIL-2R and sICAM-I in OME their sICAM-1 levels were expressed in ng/mg of total protein, which was different from our unit ng/ml. However, it is not surprising that sICAM-1, a soluble component of ICAM-1, was highly detectable in the MEEs. It is supposed that ICAM-1 might play an important role in inflammatory disorders of the middle ear. Our investigation suggested that ICAM-1 might cause a difference in the production and/or accumulation of MEE because there was a significant difference between the slCAM-1 levels of the serous and mucoid MEEs. This supposition was highly acceptable also for the fact that there was significan difference between the TNF-0t levels of the two types of MEEs and that TNF<z also activates endothelium, causing the expression of adhesion molecules such as ICAM-1.12 The physiological significance of the shedding of sIL-2R and sICAM-1 remains unknown.
However, we supposed that they might be playing some role in inflammatory disorders. There is a possibility that they could serve as useful markers of inflammation, and may also have inhibitory effects on the inflammatory cascade of reactions, because slL-2R inhibits the functional response elicited by IL-2, 6 and slCAM-1 prevents cell adherence to native ICAM-1. 6-7 IL-2 induces the proliferation of activated lymphocytes and ICAM-1 also facilitates adhesion and transmigration of neutrophils. 4 These inflammatory cells invade the middle ear of OME patients. In cases of serous types of OME, it was reported that a larger number of neutrophils and lymphocytes were observed in the MEEs and that more severe inflammatory signs were shown in the submucosa, as compared with mucoid types. 18 '19 However, particularity in our study, slCAM-1 levels of mucoid MEEs were significantly higher than those of serous MEEs. These morphological findings of cellular infiltration, which might be induced by ICAM-1, disagreed with slCAM-1 levels detected in the MEEs. Therefore, it is necessary to investigate the more detailed roles of slL-2R and slCAM-1 in the pathogenesis of OME.
It has been reported that various chemical mediators such as leukotrienes, prostaglandins 5 and platelet activating factor were detectable in the MEEs of OME. These chemical mediators are responsible at least partly for the pathogenesis of OME because they could induce hypersecretion and inhibit mucociliary transport systems in the middle ear, causing an accumulation of the effusion. v' However, these mediators could also be secreted by inflammatory cells that infiltrated into the focus. Furthermore, these inflammatory cells are regulated by the cytokine network that consists of cytokines and related factors. Our investi-gation indicated the possibility that IL-2, TNF<z and ICAM-1 could affect the pathogenesis of OME through the inflammatory cells and the chemical mediators derived from these cells.