INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The function of E-selectin and P-selectin in leukocyte rolling, thrombosis, and inflammation, as well as their possible use for therapeutic intervention in reperfusion injury, inflammation, allergy, autoimmunity, and cancer has been described (1). The leukocyte rolling is mediated by the interaction of E- and P-selectin expressed on endothelium with their carbohydrate-bearing physiologic ligand, sialyl-Lewis X (SLe^x) (10).

Hypersensitivity pneumonitis (HP) is an immunologically mediated disease that results from repeated exposure to various environmental organic materials (20). Farmer's lung disease, a common example of HP, is caused by exposure to Saccharopolyspora rectivirgula (SR), and is characterized by granuloma formation and interstitial infiltrates consisting primarily of lymphocytes and plasma cells, with occasional neutrophils and eosinophils (20). Previous studies have shown that T-helper-1 (Th1) lymphocytes play a critical role in the pathogenesis of HP by producing cytokines such as interferon- (IFN)- and tumor necrosis factor (TNF)- (24). Th1 cells may immigrate into the site of inflammation in HP through selective recruitment mediated by E-selectin and P-selectin expressed by endothelial cells (24, 25). Recent reports of in vivo experiments have suggested that inhibition of the interaction between selectins and their ligands substantially reduces the inflammatory response (26). Therefore, we hypothesized that blocking of the binding between E- and P-selectin and SLe^x might suppress the infiltration of Th1 lymphocytes into lung exposed to antigen. To examine this hypothesis, we evaluated the effect of SLe^x ganglioside analogues on the pathologic changes in an experimental murine model of HP.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Antigen and SLe^x Ganglioside Analogues

SR was obtained from the Osaka Fermentation Institute, Osaka, Japan. SR was cultured in Bennett's agar growth medium at 52° C for 3 d in a shaking incubator. The organism was harvested by centrifugation and washed three times with sterile distilled water, and the cell walls were disrupted with ultrasonication (UR-20; Tomy Seiko, Ltd., Tokyo, Japan) for 25 min. The antigen was kept lyophilized until use.

The structure of SLe^x ganglioside analogues (27) is shown as follows:

Animals

Specific pathogen-free male, 6-wk-old C3H/He mice were purchased from Japan Animal Laboratory Inc., Tokyo. They were housed in laminar-flow hoods with high efficiency particular air-filtered air and were given water ad libitum. Mice were housed six to a cage before experimental manipulation and in individual cages thereafter.

Induction of Experimental Hypersensitivity Pneumonitis

Mice were divided into four groups. In Groups 1, 2, and 3 the mice were given 90 µl (2 mg/ml) of SR antigen in saline, intranasally while under light anesthesia with diethyl ether. For controls, physiologic saline was administered instead of SR (Table 1). The materials were applied at the tip of the nose and aspirated involuntarily. This was done for three consecutive days per week for 3 wk. This method was based on previous work done by others (22, 24, 28). We also confirmed in preliminary studies that this was the optimal dose schedule for inducing an inflammatory response in murine lung. In addition, 200 µg particulate SR in normal saline, emulsified in an equal volume of complete Freund's adjuvant (Difco, Detroit, MI), was injected into the base of the tail and two subcutaneous sites (total volume = 0.1 ml) on the first day of the transnasal administration. Seven days later, the mice received 200 µg of SR mixed with an equal volume of incomplete Freund's adjuvant subcutaneously. During the third week, SLe^x ganglioside analogues were administrated continuously into the skin with a microosmotic pump (Alza Corporation, Palo Alto, CA) at 0.5 µl/h for 7 d to the mice in Group 2 until they were killed. The mice in Group 3 were given 90 µl (2 mg/ml) SLe^x ganglioside analogues transnasally for five consecutive days. The mice in Groups 2 and 3 were simultaneously given SR transnasally during the third week, as described earlier. Four days after the final administration of SR the mice were killed by exsanguination while under anesthesia with diethyl ether.

Bronchoalveolar Lavage

For quantitative cell differential counting in the alveolar epithelial lining fluid, bronchoalveolar lavage (BAL) was performed on the basis of previous work done by others (22, 28). Briefly, after the mice were killed, a 21-gauge catheter was inserted into the trachea. Bronchoalveolar lavage fluid (BALF) samples were obtained by three instillations of 1,200 µl each of physiologic saline. The cells in the BALF were suspended in RPMI-1640 medium and spun down in a cytocentrifuge (Auto Smear CF-120; Sakura, Tokyo, Japan). Cytospin preparations were fixed with 95% ethanol and stained using Diff-Quik staining (Baxter, Inc., McGaw Park, IL). With a microscope, differential counts were made of at least 1,000 cells, using standard morphologic criteria to identify the cells as neutrophils, lymphocytes, eosinophils, or macrophages.

Lung Index

Lungs were removed, trimmed of extraneous tissue, rinsed, and weighed. Lung indices were calculated as follows (22):

Histologic Evaluation

Lungs were perfused and fixed in periodate-lysine-paraformaldehyde for 2 h at 4° C and embedded in paraffin or OCT compound (Miles Laboratories, Naperville, IL) for frozen sectioning. The tissue was cut into sections 3 µm thick, stained with hematoxylin and eosin (H&E), and evaluated light microscopically. A histologic score for each lung was determined according to the following criteria: 0 = no lung abnormality; 1 = presence of inflammation and granulomas involving < 10% of the lung parenchyma; 2 = lesions involving 10 to 30% of the lung; 3 = lesions involving 30 to 50% of the lung; and 4 = lesions involving > 50% of the lung (29, 30).

Immunohistochemical Staining

Immunohistochemical staining was performed according to a standard protocol. An immunohistological staining kit using the Histomouse-SP Kit (Zymed Laboratories Inc., CA) was adopted for frozen sections. The biotinylated streptavidin system in a Histofine Kit (Nichirei, Tokyo, Japan) was adopted for staining of paraffin-embedded sections. In brief, the paraffin sections were deparaffinized and treated with 0.3% hydrogen peroxide in methanol for 15 min to block endogenous peroxidase activity. Frozen tissue sections were treated with Peroxo-Block (Zymed) for 45 s. The sources and working dilutions of the antibodies used are listed in Table 2. The sections were then incubated overnight at 4° C with primary antibodies. Aminoethyl carbazole (AEC) or 3'3-diaminobenzidine (DAB) was used as the chromogenic substrate.

SLe^x-Positive T Lymphocytes in Peripheral Blood

Peripheral blood was obtained from the right femoral artery of mice under anesthesia. The number of leukocytes was counted with a hemocytometer. Peripheral blood was smeared on glass slides and stained with Giemsa stain for the analysis of cell types such as polymorphonuclear cells, monocytes, and lymphocytes. SLe^x-positive T lymphocytes in peripheral blood were identified with antimouse CD3 and antihuman SLe^x antibodies and their populations were analyzed with a flow cytometry in a fluorescence-activated cell sorter (Becton Dickinson, Inc., Mountain View, CA) with an argon-ion laser.

Statistical Analysis

Statistical analysis was done with the Mann-Whitney U test. Values of p < 0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Histopathologic Changes in Experimental HP and the Effect of SLe^x Ganglioside Analogues

Although no inflammatory responses were observed in the lungs of control mice (Figure 2a), in the mice of Group 1 exposed to antigen, the histopathologic examination showed a strong inflammatory response with moderate to marked interstitial and especially perivascular and peribronchiolar infiltration with lymphocytes and granuloma formation (Figure 1b). In the mice in Groups 2 and 3, which were given SLe^x ganglioside analogues after antigen exposure, lymphocyte infiltration into the lung parenchyma and granuloma formation were less marked than in Group 1 (Figures 1c and 1d). Although the histologic scores were higher than those of the control group, they were also significantly lower than those of Group 1 (histologic scores: control: 0 ± 0.00 [mean ± SD] [n = 9]; Group 1: 3.89 ± 0.11 [n = 9]; Group 2: 2.60 ± 0.49 [n = 6], Group 3: 2.00 ± 0.63 [n = 6]) (Figure 2). In accord with these results, the lung indices of Groups 2 and 3 were significantly lower than that of Group 1 (lung index: control: 1 ± 0.00; Group 1: 1.55 ± 0.38; Group 2: 1.26 ± 0.21; Group 3: 1.27 ± 0.20) (Figure 3).

View larger version (14K): [in this window] [in a new window]   Figure 2.   Quantification of inflammation and granuloma formation in different groups of mice. Data for histologic scores are given as mean ± SEM. Controls consisted of mice exposed to saline alone (n = 9), with a histologic score of 0 ± 0.00. Group 1 consisted of mice exposed to antigen (n = 9), with a histologic score of 3.89 ± 0.11. Group 2 consisted of mice exposed to antigen and treated with SLex ganglioside analogues by osmotic pump (n = 6), with a histologic score of 2.60 ± 0.49. Group 3 consisted of mice exposed to antigen and treated with SLex ganglioside analogues transnasally (n = 6), with a the histologic score of 2.00 ± 0.63. There was a statistically significant difference between the antigen-exposure group and SLex ganglioside analogue-treated groups (p < 0.05).

View larger version (154K): [in this window] [in a new window]   View larger version (174K): [in this window] [in a new window]   View larger version (169K): [in this window] [in a new window]   View larger version (170K): [in this window] [in a new window]   Figure 1.   Representative H&E-stained histology sections (× 50). (a) Mice exposed to saline alone. (b) Mice exposed to antigen. (c) Mice exposed to antigen and treated subcutaneously with SLex ganglioside analogues delivered by microosmotic pump. (d ) Mice exposed to antigen and treated intranasally with SLex ganglioside analogues.

View larger version (13K): [in this window] [in a new window]   Figure 3.   Changes in lung weight (lung index) in groups of mice exposed to antigen and treated with SLex ganglioside analogues. Data are presented as mean ± SEM. The lung index of Group 1, exposed to antigen alone was 1.56 ± 0.38. The lung index of Group 2, exposed to antigen and treated with SLex ganglioside analogues given by microosmotic pump, was 1.26 ± 0.21. The lung index of Group 3, exposed to antigen and treated with SLex ganglioside analogues given transnasally, was 1.27 ± 0.20. There was a statistically significant difference between the antigen-exposure group and SLex ganglioside analogue-treated groups (p < 0.05).

Analysis of Cell Differential Counts in BALF

In the mice in Group 1, the total cell count in BALF increased (83.1 ± 25.6 × 10⁵; n = 9), over normal control values (7.8 ± 2.2 × 10⁵; n = 9). The predominant cells were lymphocytes (63.25 ± 4.07%), macrophages (20.95 ± 1.02%), and neutrophils (7.6 ± 2.6%), and their percentages differed from those of the control group (lymphocytes: 8.06 ± 0.76%; macrophages: 90.3 ± 0.95%; and neutrophils: 1.01 ± 4.1%). In Groups 2 and 3, the ratio of neutrophils to total numbers of cells was close to that in Group 1. The ratios of lymphocytes to total numbers of cells in Group 2 (31.18 ± 7.00%) and Group 3 (29.16 ± 10.71%) were higher than that of the control group, but were significantly lower than that of Group 1 (Table 3).

Accumulated Mononuclear Cells Bearing SLe^x in Granulomas

Immunohistochemical staining was done with antihuman SLe^x antibody. In mice of Group 1, the infiltrating mononuclear cells included a large number of lymphocytes, and a substantial number of other cells had the morphology of monocytes or macrophages. A significant percentage of mononuclear cells were positive for SLe^x in mice of Group 1 (Figure 3). Some neutrophils in the vessels were also positive for SLe^x (data not shown). In mice given with SLe^x ganglioside analogues by microosmotic pump (Group 2) and transnasally (Group 3), the number of accumulated mononuclear cells was diminished. Few mononuclear cells positive for SLe^x were observed (Figures 4b and 4c). The ratios of mononuclear cells expressing SLe^x to whole cells in granulomas of Group 2 (9.5 ± 9.0%; n = 17) and Group 3 (8.9 ± 8.0%; n = 15) were significantly lower than that of Group 1 (21.7 ± 10.2%; n = 27) (Figure 5).

View larger version (125K): [in this window] [in a new window]   View larger version (128K): [in this window] [in a new window]   View larger version (127K): [in this window] [in a new window]   Figure 4.   Staining with anti-SLex antibody, using DAB as a chromogen and methyl green as counterstain (× 500). (a) Mice exposed to antigen. A portion of the infiltrating lymphocytes were positive (some are shown by arrows). (b) Mice given SLex ganglioside analogues by microosmotic pump. Few infiltrating mononuclear cells were positive. (c) Mice given SLex ganglioside analogues intranasally; only a few infiltrating lymphocytes were positive (shown by arrows).

View larger version (13K): [in this window] [in a new window]   Figure 5.   Comparison of ratio of SLex-positive cells to total numbers of cells in granulomas in Groups 1, 2, and 3. The percentage of SLex-positive cells in Group 1 was 21.7 ± 10.2% (n = 27 granulomas investigated in histologic sections), that in Group 2 was 9.5 ± 9.0% (n = 17), and that in Group 3 was 8.9 ± 8.0% (n = 15).

SLe^x-Positive T Lymphocytes in Peripheral Blood

There was no significant change in the numbers of total leukocytes or lymphocytes in the peripheral blood of control mice or mice in Group 1. Furthermore, transnasal treatment with SLe^x ganglioside analogues (Group 3) did not significantly change these numbers, as shown in Table 4. Nor did the population of SLe^x-positive T lymphocytes in peripheral blood change significantly among these three groups (Table 4).

Expression of E- and P-Selectin

In control mice, the endothelial lining of vessels failed to express E-selectin or P-selectin (Figures 6a and 6b). In contrast, vascular endothelium expressed E- and P-selectin strongly in all mice in Group 1 (Figures 6c and 6d). Strong expression of E- and P-selectin was also observed in Groups 2 and 3 (data not shown).

View larger version (128K): [in this window] [in a new window]   View larger version (118K): [in this window] [in a new window]   View larger version (130K): [in this window] [in a new window]   View larger version (135K): [in this window] [in a new window]   Figure 6.   Staining with anti-E- and P-selectin antibodies using AEC as chromogen and hematoxylin as counterstain (× 500). (a) Mice exposed to saline alone. Vascular endothelium failed to express E-selectin. (b) Mice exposed to saline alone. Vascular endothelium failed to express P-selectin. (c) Mice exposed to antigen. Vascular endothelium expressed E-selectin (arrow). (d ) Mice exposed to antigen. Vascular endothelium expressed P-selectin (arrow).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

This study evaluated the effects of the inhibition of E- and P-selectin function with SLe^x ganglioside analogues on the inflammatory response in the lung in experimental HP. We established an experimental model of HP in mice, which showed marked interstitial and especially perivascular and peribronchiolar infiltrates consisting primarily of lymphocytes, in addition to granuloma formation, after antigen exposure. These pathologic changes were significantly suppressed in mice treated with SLe^x ganglioside analogues, which reduced the infiltration of lymphocytes.

It is well known that E-selectin and P-selectin share the same modular structure, characterized by N-terminal lectin, which is postulated to recognize the carbohydrate ligand SLe^x (10). It is very important pathophysiologically that the binding of E-selectin and P-selectin to SLe^x mediates cell-cell interactions in inflammatory diseases (14, 25, 31). In a previous study, SLe^x ganglioside analogues were shown to contain, between the sugar and the lipophilic moieties, a spacer of various lengths that is necessary for expressing the potent activity of these substances as selectin inhibitors (27). It is known that other analogues of SLe^x have significant protective effects against lung injury (32). In the present study, significant differences were found in Group 1, which was exposed to antigen, and Groups 2 and 3, which were treated with SLe^x ganglioside analogues. It is suggested that SLe^x ganglioside analogues exert an antiinflammatory effect by reducing the numbers of lymphocytes in BALF, and dramatically reduce the lung index and diminish the tissue accumulation of lymphocytes and granuloma formation.

In Groups 2 and 3 in our study, the tissue accumulation of lymphocytes and granuloma formation were diminished, but the inflammatory response was not eliminated completely. There are at least three possible explanations for this. First, the pulmonary inflammation may have occurred before the administration of SLe^x analogues. Second, since lung leukocyte recruitment relies on adhesion pathways that are partly redundant (33, 34), mononuclear cell migration may not have depended only on the binding of selectins to SLe^x in our study. Third, the concentration of SLe^x analogues may not have been sufficient to produce the maximum effect on the migration of leukocytes.

SLe^x analogues significantly reduced lymphocyte recruitment into the lungs of mice after exposure to SR antigen in our study. In P- and E-selectin-deficient mice, the number of peritoneal neutrophil influxed after thioglycollate injection was suppressed for 8 h compared with that in wild-type mice (35). These results suggested a critical role of binding of SLe^x to P- and E-selectin in the recruitment of leukocytes to inflammatory sites. However, because the analysis was done only at a single time point in the present study, it is possible that the effect of SLe^x analogues is limited to delaying the leukocyte recruitment, and that the ultimate magnitude of lung inflammation is not influenced by SLe^x analogues, because of selectin-independent pathways of leukocyte recruitment.

In the present study, we showed by immunostaining the strong expression of P- and E- selectin in endothelial cells of mice exposed to antigen, suggesting a role for these selectins in the recruitment of mononuclear cells. The strong expression of E- and P-selectin was also observed in Groups 2 and 3 in our study (data not shown). These results suggested that SLe^x ganglioside analogues did not influence the expression of E- and P-selectin. Both anti-P- and anti-E-selectin antibodies might bind epitopes not affected by the binding of these selectins to SLe^x analogues, because the immunostaining for E- and P-selectin with the antibodies used in our study was not influenced by the in vitro addition of SLe^x analogues. In addition, administration of SLe^x analogues in vivo did not seem to facilitate the shedding of E- or P-selectin.

Th1 cells preferentially secrete IFN- and TNF-, which are thought to activate macrophages and to be responsible for cell-mediated immune reactions (24, 25). Schuyler and colleagues reported that Th1 CD4⁺ cell lines could adoptively transfer experimental HP, which suggested that Th1 cells and their cytokines might be important in the HP of experimental murine models (24). Gudmundsson and associates showed that IFN- is essential for the pathogenesis of HP by comparing the responses to inhaled antigen of mice lacking the IFN- gene with that of their normal littermates (22). Austrup and coworkers reported that Th1 cells but not Th2 cells could efficiently immigrate into inflamed sites, such as sensitized skin or arthritic joints, and that their recruitment was mediated by E- and P-selectin (25). In our study, the administration of SLe^x ganglioside analogues suppressed the inflammatory response in our HP model, which included a reduction in the number of infiltrating lymphocytes, suggesting that SLe^x ganglioside analogues may inhibit the recruitment of Th1 cells, which play a critical role in the pathogenesis of HP, although we did not evaluate Th1 cells in the lung in the present study.

Munro and colleagues reported that from 5% to 15% of circulating T lymphocytes expressed SLe^x, and that a higher proportion of T cells in inflamed skin were SLe^x-positive (31). In the present study, there was unexpectedly no significant difference in the number of lymphocytes and in the ratio of SLe^x-positive T cells in peripheral blood in the control group and the groups of mice given SR alone and SR followed by SLe^x analyogues given intranasally, as shown in Table 4. Previous studies have shown that leukocytes accumulate in the circulation transiently as a result of inhibition of binding between selectins and SLe^x (35). With regard to this point, there was a possibility that the changes in the numbers of lymphocytes and the ratios of SLe^x-positive T cells in peripheral blood might have been subtle on the days when these measures were analyzed, and could not be detected, since antigen exposure was conducted chronically for 3 wk and SLe^x analogues were administered five times transnasally for 5 d. These results also suggested that at least the decrease in SLe^x-positive lymphocytes in granulomas in the mice treated with SLe^x analogues was not directly caused by a decrease in the number of SLe^x-positive lymphocytes in the peripheral blood. In addition, to eliminate the possibility that the decrease in SLe^x-positive lymphocytes in granulomas of the mice treated with SLe^x analogues was due to interference of administered SLe^x analogues with the binding between SLe^x and anti- SLe^x antibody, we performed immunostaining for SLe^x on lung sections of Group 1 mice to which SLe^x analogues were added in vitro. The result revealed that in vitro treatment of lung sections with SLe^x analogues did not inhibit the immunostaining of mononuclear cells for SLe^x with anti-SLe^x antibody.

In BALF, the absolute number of lymphocytes in mice exposed to SR was reduced by the administration of SLe^x analogues. On the other hand, the absolute numbers of macrophages tended to increase in Groups 2 and 3, which were treated with SLe^x analogues, as compared with Group 1, which was not so treated. From what we know about the effects of SLe^x analogues on the number of alveolar macrophages, there is no information to explain these findings.

No significant differences were observed between the two different routes of administration of SLe^x analogues, consisting of transnasal administration and subcutaneous administration with a microosmotic pump. It is possible that SLe^x analogues are absorbed both through the lining of epithelial cells of the airway and through the blood-circulatory pathway.

In summary, we suggest that the cellular responses in our experimental lung model of HP may have been principally mediated by Th1 lymphocytes recruited from the blood through the interaction between E- and P-selectin and SLe^x preferentially, and that the lymphocyte-mediated inflammatory responses were significantly suppressed by the competitive inhibitory action of SLe^x ganglioside analogues.