New Aspects of Degranulation and Fates of Airway Mucosal Eosinophils

JONAS S. ERJEFÄLT and CARL G. A. PERSSON

Departments of Physiological Sciences and Clinical Pharmacology, Lund University Hospital, Lund, Sweden

	INTRODUCTION

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Eosinophils are bone marrow-derived granulocytes with a richness of surface receptors for immunologic ligands (1). Eosinophils characteristically harbor specific granules which, besides some 20 known chemokines, cytokines, and growth factors, contain large amounts of the tissue-toxic cationic proteins major basic protein, eosinophil cationic protein (ECP), eosinophil peroxidase, and eosinophil-derived neurotoxin (5, 6). Given these properties, eosinophils may have many roles in tissue, ranging from immunoregulation (7) to tissue remodelling and repair (8) and the killing of foreign organisms (1). Eosinophils abounding and degranulating in airway mucosal and pulmonary tissues are also believed to have pathogenic roles in asthma, nasal polyposis, allergic rhinitis, and eosinophilic pneumonia (11, 12). Increased levels of granule proteins such as ECP in body fluids (13), and specific ultrastructural features of tissue eosinophils (14, 15) (Table 1), reflect eosinophil degranulation activity.

View larger version (117K): [in this window] [in a new window]   Figure 2.   Transmission electron micrograph of nasal biopsy specimen obtained during active allergic rhinitis (22), demonstrating airway tissue eosinophils with signs of eosinophil cytolysis (right) and PMD (left). PMD involves protein release (electron lucency) from the specific granules residing within the intact cell (left). In contrast, eosinophil cytolysis via rupture of the cell membrane (arrow), causes the release of electron-dense, specific granules into the extracellular matrix (right).

View larger version (39K): [in this window] [in a new window]   Figure 3.   Eosinophils may be stimulated in vitro to undergo cytolysis (a) or apoptosis (b). The fluorescence micrograph (a) shows eosinophils adhering to sIgA-coated Sepharose beads (whole-mount preparation). Cell nuclei are stained red (thick arrow) by the DNA-binding stain propidium iodide (PI), and eosinophil specific granules are stained green. Cytolytic eosinophils are characterized by chromatolysis (loss of PI staining) and scattered free granules (thin arrows). The transmission electron micrograph displays an apoptotic eosinophil observed in vitro after 24 h culture in the absence of eosinophil growth factors (b).

View larger version (122K): [in this window] [in a new window]   Figure 1.   Once recruited, airway tissue eosinophils have different fates in vivo. Red and blue arrows represent pathways for granule protein release and cell traffic, respectively. Without losing viability, tissue eosinophils undergo PMD. During active disease processes, many tissue eosinophils are "ultimately activated" to undergo eosinophil cytolysis. Classical exocytosis of eosinophils has yet to be demonstrated in airway tissues in vivo. Viable tissue eosinophils migrate into the airway lumen; some may even move to regional lymph nodes (RLN). The luminal eosinophils are further cleared through cytolysis, mucociliary clearance, and apoptosis. However, data on the occurrence of apoptotic eosinophils in blood-perfused airway mucosal tissue are scarce.

At the light-microscopic level, immunohistochemical staining with the monoclonal antibody EG2 (23) has been widely used to identify degranulating tissue eosinophils. Similarly, terminal deoxynucleotidyl transferase-uridine nucleotide end-labeling (TUNEL) technique has recently been used to examine the occurrence of apoptotic eosinophils in the airways (24, 25). However, these commonly used markers of eosinophil activation and death, respectively, may not always be specific. For example, EG2 staining may merely indicate eosinophils in general (26), and the capacity of TUNEL techniques to clearly distinguish between apoptosis and necrosis (cytolysis) is currently debated (27). Indeed, the TUNEL technique must be combined with analysis of cell morphology in the assessment of apoptosis. In the absence of acceptable molecular markers, and to complement the determination of eosinophil proteins, morphologic changes will have to be assessed to identify and quantify modes of degranulation and fates of airway mucosal eosinophils (15) (Table 1).

Many eosinophils that dwell in the airway mucosa may enter the airway lumen (Figure 1). Some eosinophils may move to the regional lymph nodes (30). Others remain in the airway mucosa until they die. As demonstrated in in vitro test systems, death of eosinophils may occur silently through apoptosis (33, 34). On the other hand, observations of diseased airway mucosae indicate that these cells may instead die through the mechanism of eosinophil cytolysis, which simultanously acts as an ultimate mode of degranulation (16). Airway tissue eosinophils may thus have several different fates in vivo (Figure 1), each with its own potential significance.

The present article deals with clearance pathways for airway mucosal eosinophils, such as "luminal entry," eosinophil apoptosis, and eosinophil cytolysis. We further discuss eosinophil cytolysis as a major mode of degranulation of eosinophils in airway diseases, in addition to piecemeal degranulation (PMD; Table 1, Figure 2). As in other areas of airway research (35, 36), in vitro research standards relating to eosinophil events may partly diverge from what actually occurs in blood-perfused airway tissue. Hence, the present emphasis is on in vivo observations.

	LUMINAL ENTRY AND LUMINAL DEATH OF AIRWAY EOSINOPHILS

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Twelve decades of studies of eosinophils appearing in airway secretions/exudations (Table 2) have successfully focused on the utility of luminal cells for diagnosing diseases such as asthma and allergic rhinitis. The luminal cells are also harvested for further examination; as do test systems involving cultured eosinophils, studies of luminal eosinophils will disclose the appearance and biology of eosinophils removed from tissue. In vitro studies may demonstrate interesting differences between eosinophils sampled from different sites. For example, Sedgwick and colleagues (38), examining cells in vitro after allergen challenge, demonstrated that luminal and blood eosinophils display different oxygen radical responses upon stimulation by formylmethionylleucylphenylalanine (fMLP). The significant possibility that luminal cells as well as cultured eosinophils, irrespective of sampling site, may differ from eosinophils present in diseased airway tissue in vivo is discussed later.

To some extent, luminal entry of granulocytes, particularly neutrophils, may occur in association with sites of patchy epithelial shedding and repair, which constitute "hotbeds" of intense inflammatory, exudative, and remodeling activities in vivo (10, 32, 39). However, viable eosinophils could migrate between any given set of connecting epithelial cells (40, 41). In vivo data show that adhesive, plasma-derived proteins are laid down all around epithelial cells (32, 42). In addition, in vitro data suggest that the airway epithelium facilitates the trans-epithelial (43) migration of leukocytes in a physiologic basolateral-to-apical direction (44, 45). Further promoting the paraepithelial migration of leukocytes into the lumen may be local chemokines such as eotaxin, which may create a chemoattractant gradient across the epithelium (46). The capacity of the luminal entry mechanism is strongly suggested by observations in guinea-pig tracheal explants in which luminal application of the chemotaxin fMLP produced migration of more than 150,000 eosinophils/cm² into the lumen within 120 min (47).

Working with sensitized and allergen-challenged rats, Schneider and colleagues (48) examined late-phase kinetics of peribronchial eosinophils and of luminal eosinophils that were retrievable by bronchoalveolar lavage (BAL). Whereas the number of eosinophils in lavage fluid peaked at Day 3, there was an earlier peak (Day 2) in lung tissue eosinophils. The data reported by Schneider and colleagues suggest that a large proportion of the eosinophils that left the peribronchial tissue between Days 2 and 3 entered the airway lumen. Superficial airway mucosal eosinophils, even more readily than peribronchial eosinophils, may end up in the lumen. This possibility is supported by observations in guinea pigs. Allergen challenge of sensitized guinea pig trachea in vivo, a condition in which many resident eosinophils hover in the epithelium, increased the number of luminal eosinophils by 10-fold within 10 min. Simultaneously, mucosal eosinophilia in the tracheal area of interest was significantly reduced (32).

As was particularly evident in the era before steroid treatment, episodes of asthma may be followed by pronounced sputum eosinophilia (35, 49). In a recent study Aalbers and associates (50) found that bronchial tissue eosinophils peaked a few hours after allergen challenge. At this early time point, eosinophils in bronchial lavage fluid were still increasing toward a peak at 24 h, when bronchial tissue eosinophils had already declined significantly. These data support the notion that luminal entry is an efficient elimination pathway for the mucosal leukocytes. However, movement of eosinophils across the airway epithelial lining has so far almost exclusively been acknowledged as a proinflammatory event (43, 50)

It is noteworthy that the plasticity of cell junctions in the epithelial lining apparently allows the swift passage of bulk plasma (including fibronectin, fibrinogen, ₂-macroglobulin). Prompt luminal entry of nonsieved plasma macromolecules may thus occur in guinea pig and human airways without compromising the integrity of the mucosa as an absorption barrier (51). Moreover, granulocytes may move to the mucosal surface without harming the epithelial lining cells (52). Once they are in the lumen, the fate of the vast majority of eosinophils may be sealed (Figure 1). They may be either coughed up or blown out. Or, when remaining for some time among the waste material in the airway lumen, the cells would be subjected to degenerative processes. Luminal eosinophils in asthma and rhinitis thus undergo apoptosis (53) as well as cytolysis (16). Current data support the possibility that eosinophils develop susceptibility to apoptotic mechanisms after their entry into the airway lumen (see the subsequent discussion). By its capacity to eliminate eosinophils from diseased bronchial tissue, the luminal entry mechanism would have an important role in the resolution of airway inflammation. Clearance of viable granulocytes from the airway mucosa, without enhancing inflammation, may thus occur either through an epithelial crossing maneuver or, theoretically, through apoptosis.

	EOSINOPHIL APOPTOSIS: IN VITRO OBSERVATIONS

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Death of cells via the apoptotic pathway is a widely recognized mechanism critically involved in cellular homeostasis (56). As a corollary, it has also been proposed that apoptosis is central to the regulation of inflammatory cell accumulation in human airway diseases (57). Eosinophil apoptosis in particular has become a topical area of research. On the basis of observations in cell culture test systems, eosinophil apoptosis is now considered central to the resolution of airway inflammation, and proapoptotic effects are being adduced as desirable properties of future antiasthma drugs (57). Indeed, the possibility that drugs already available, such as steroids and xanthines, may operate by increasing eosinophil apoptosis has received attention (60, 61). In culture test systems, most eosinophils die through apoptosis within a few days, providing ample opportunity for reductive biologic research into this process (34). Eosinophils in vitro also provide clear illustrations of the ultrastructural features of apoptosis (Figure 3b).

The biologic control of apoptosis is an explosive area of molecular medicine. As suggested by detailed schemes (62) for the execution of death (proapoptotic) and survival (antiapoptotic) signals, the subcellular pathways of apoptosis involve many levels of regulation. Common to most cell types is induction of apoptosis through the Fas-Fas ligand system (63) and enhancement of apoptosis through Bax, with involvement of apoptosis effectors such as p53 (62). Inhibition of apoptosis is mediated by Bcl-2, although in eosinophils Bcl-xL is expressed rather than Bcl-2 (62). Eosinophils present in airway tissues in vivo may also express these regulatory molecules (24, 64), demonstrating a potential capacity of these cells to undergo apoptosis (16, 58, 65).

Several antiapoptotic factors for eosinophil survival have been identified. Among them interleukin (IL)-5 may be of particular interest. Its signaling pathway has been explored (66), and its anti-apoptotic effect on eosinophils in vitro (33) may be prevented by steroids and xanthines (67, 68). IL-5 commonly occurs in the airway mucosa in asthma and allergic rhinitis (69). Furthermore, preliminary data suggest that IL-5 antagonists may inhibit allergen challenge-induced eosinophilia in the blood and sputum of allergic asthmatic individuals (without inhibiting the induced hyperresponsiveness and late asthmatic reaction [70]). However, it is at present unclear whether blocking of antiapoptotic actions of IL-5 contributes to the antieosinophilic effect. Besides being an acknowledged survival factor, IL-5 is clearly implicated in eosinophilopoiesis and in eosinophil trafficking to the airways (69, 71). Hence, antagonism of these latter mechanisms alone may explain the pharmacologic inhibition of eosinophilia in vivo.

Numerous excellent reviews have recently been published on the mechanisms of eosinophil apoptosis (58, 59, 72, 73). The advanced in vitro work in this field has established eosinophil apoptosis as a strong research paradigm. Naturally, this cutting-edge, reductive research also underpins the current notion that regulation of eosinophil apoptosis is a major mechanism in eosinophilic airway diseases.

	APOPTOTIC EOSINOPHILS IN HUMAN AIRWAY LUMEN AND IN ANIMAL MODELS

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Woolley and coworkers (55) demonstrated that the relative proportion of apoptotic eosinophils in sputum samples was increased in asthmatic individuals during the period of their recovery from an acute episode of asthma requiring steroid treatment. (Without doubt, apoptotic eosinophils can readily be detected in nasal discharges as well as sputum from asthmatic individuals [54, 55].) Interestingly, there appears to be unanimous agreement (61, 64, 74, 75) that the data reported by Woolley and coworkers (55) constitute "compelling evidence" for the importance in vivo of eosinophil apoptosis in asthma, especially in resolution of the inflammation in the disease. However, this may be a misconception.

The apoptotic eosinophils detected by Woolley and coworkers (55) would have been viable as long as they remained in the mucosal tissue compartments; dead cells would probably not migrate across the epithelium and into the airway lumen. Inferentially, the clear occurrence of apoptotic eosinophils in the airway lumen cannot indicate the occurrence of apoptosis in airway tissue. However, the luminal data on eosinophil apoptosis would support the proposal, discussed earlier, that luminal entry may be a significant mechanism for elimination of viable airway tissue granulocytes.

Work by Tsuyuki and colleagues (53) and Kawabori and associates (76) provides standard animal model references on eosinophil apoptosis in the resolution of mucosal inflammation. Tsuyuki and colleagues (53) employed a mouse model of allergic asthma. However, in their study, eosinophil apoptosis was in fact determined ex vivo, in cultured eosinophils obtained from the allergic mice. A quite recent study by Kodama and coworkers (25), involving immunohistochemical and ultrastructural examinations of the airways of allergic mice, could thus tentatively claim to be "the first report to show direct evidence for apoptosis of airway tissue eosinophils in vivo." The "apoptotic eosinophils" in this report (25) were emphasized by an accompanying editorial (75). However, Kodama and coworkers (25) actually depict (in the selected transmission electron micrograph shown in Figure 6 of their article) the ultrastructural features of a nonapoptotic (sic) eosinophil. Moreover, these authors (25) explicitly state that their light microscopic data cannot specifically reveal apoptotic eosinophils in the mouse airways.

The study by Kawabori and associates (76), an early study in this field, examined steroid-induced effects on intestinal eosinophilia in rats. These authors quantified deranged eosinophils (some of these deranged eosinophils were clearly apoptotic and some were cytolytic). Kawabori and associates (76) did not use TUNEL techniques. However, their interesting illustrations indicated that apoptotic eosinophils may occur in the gut (rat) mucosa, and clearly showed that if apoptotic eosinophils were present, they would be readily detectable by transmission electron microscopic analysis. Additionally, phagocytosis by macrophages appeared to be an efficient mechanism for removing eosinophils from the intestine of steroid-treated rats. Notably, this was so whether or not the eosinophils were apoptotic and whether or not the rat intestine was inflamed. Kawabori and associates were commendably cautious in their conclusions, and their preliminary proposal that eosinophil apoptosis could explain the steroid-induced attenuation of gut eosinophilia in both healthy and inflamed rat intestine may need to be confirmed. In particular, extrapolation of their findings to extraintestinal tissues may not yet be justified.

Indeed, rat gut and lung seem to differ from one another; thus recent work involving Sephadex bead-induced, high permeability, eosinophilic lung edema in vivo in rats demonstrated prompt resolution of established edemogenic inflammation after steroid treatment was begun (77). However, apoptotic eosinophils could not be detected in lung tissue samples obtained at various time points. Eosinophils were examined by double staining for TUNEL-positive nuclei and EG2 immunoreactivity. Interestingly, luminal eosinophils exhibiting apoptosis were observed, supporting the observations in human airways of the selective occurrence of apoptotic eosinophils in the airway lumen (54).

	THE SEARCH FOR APOPTOTIC EOSINOPHILS IN AIRWAY TISSUES IN VIVO IN ASTHMA AND RHINITIS

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

In the rich literature on airway eosinophilia published around 1900, we have discovered but one intriguing description of a cell morphology that would agree with the occurrence of apoptotic eosinophils (37) (Table 2). The question is whether the particular eosinophil features described, in postmortem airway tissues from an asthmatic child, can be confirmed by more clear-cut in vivo observations. Simon and associates reported "delayed apoptosis as a mechanism causing tissue eosinophilia" (78). Indeed, these authors report that apoptotic eosinophils do not appear in nasal polyp tissues until after several days of culture ex vivo; if survival-factor antibodies are present, their appearance can be detected as early as after 3 d in culture (78). In vivo investigations involving transmission electron microscopic analysis have also failed to detect even a single apoptotic eosinophil in freshly excised polyps (15, 54). A further in vivo study with polyps, focusing on tissues obtained from steroid-treated individuals, involved TUNEL and EG2 staining as well as transmission electron microscopic analysis (79). Although in vitro data suggest that steroids should theoretically increase the occurrence of apoptotic eosinophils, none was detected. Again, however, apoptotic eosinophils were observed on the luminal aspect of the airway epithelium. A reanalysis of an extensive study done with transmission electron microscopy (TEM) (14), involving subjects with asthma of different degrees of severity, most of whom were receiving significant steroid therapy, revealed no eosinophils with the ultrastructural features of apoptosis (P. K. Jeffery and colleagues, unpublished observations). Similarly, in cases of eosinophilic allergic rhinitis with and without steroid therapy, many airway tissue eosinophils were scrutinized with TEM and were found to exhibit features of eosinophil cytolysis and PMD, but not of apoptosis (22). In our studies (15, 22, 54) involving quantitative assessment of ultrastructural features of airway tissue eosinophils as well as immunohistochemistry, local macrophages only exceptionally contained eosinophilic material. Hence, our work also failed to support the idea that apoptotic eosinophils "are not visible at any one time because of rapid clearance in tissues" (74). Furthermore, we know of no obvious reason why any mechanism should eliminate apoptotic eosinophils much more rapidly than the apoptotic neutrophils, mast cells, and lymphocytes that are readily detected in the human airway tissues in vivo (15, 22, 54). These attempts at finding apoptotic eosinophils may provide a basis for discussing two current reports on eosinophil apoptosis in asthma.

Comparing healthy subjects with two groups of asthmatic subjects (steroid-treated and untreated), Druilhe and associates (24) reported a disease- and steroid treatment-independent occurrence of apoptosis-regulating molecules (Bcl-2. FAS, and FAS ligand) with airway tissue eosinophils. Interestingly, Druihe and associates also reported that TUNEL-positive eosinophils were present in all three groups of their test subjects, again with no significant difference between the groups (24). Unfortunately, Druilhe and associates published no illustration of apoptotic eosinophils (double staining for EG2 and the TUNEL technique were used). They also provided no information about any apoptotic shape change at the light- or electron-microscopic level (24).

In an extensive immunohistochemical study, Vignola and coworkers (64) included a search for apoptotic tissue eosinophils in the airways of healthy, asthmatic, and chronic bronchitic subjects. These authors also reported finding apoptotic eosinophils (TUNEL + EG2 staining), but these eosinophils are not illustrated in asthmatics. Importantly, however, Vignola and coworkers did depict three double-stained "apoptotic eosinophils" (both EG2 and TUNEL positive) in human airway tissue in chronic bronchitis (64) (shown in Figure 4b of their article). Two features of this illustration are of note. First, nearly all tissue cells in the figure appear to be TUNEL positive; this implies that a very high number of dying cells was present or, more likely, that in this case, nonapoptotic cell nuclei were also stained by the TUNEL-technique. Second, the three apoptotic eosinophils shown in their selected illustration may not exhibit the required shape changes ("marked loss of cytoplasm and nuclear condensation" [64]) for apoptosis. Potential weaknesses in the identification of apoptotic eosinophils were not discussed (64), nor were such critical matters brought up in the accompanying editorial that highlighted this work (80).

View larger version (33K): [in this window] [in a new window]   Figure 4.   Upon binding of immunologic ligands, primed eosinophils may undergo cytolysis involving chromatolysis and rupture of the cell membrane. Thus, major cell constituents, notably the protein-laden specific granules (Cfegs), are liberated into the surrounding extracellular matrix, where they may produce high local concentrations of granule proteins, including ECP, major basic protein, eosinophil-derived neurotoxin, and eosinophil peroxidase.

Yet apoptotic eosinophils may well have been observed in other human tissues in vivo. It has, for example, been reported that cell apoptosis, including the occurrence of apoptotic eosinophils, was increased in allergic skin after local allergen challenge (81). Perhaps eosinophils in the skin, where other noninflammatory paths of elimination may be scarce, are more likely than mucosal eosinophils to undergo apoptosis (clearly, skin eosinophils may also undergo extensive cytolysis [82]).

In summary, currently available data may not quite amount to compelling evidence for the importance or even the occurrence of eosinophil apoptosis in airway tissues in vivo. The important message at this stage is probably that further, careful clinical studies be done to explore in vivo conditions under which apoptotic eosinophils may appear in the airway mucosa and, most importantly, to explore whether apoptosis, spontaneous or drug-induced, can be a major event in the elimination of these significant granulocytes from the blood-perfused airway mucosae. A significant part of the further work in this field concerns continued and improved validation of immunohistochemical and other methods used in the detection of apoptotic eosinophils in respiratory tract tissues in vivo. For example, the important TUNEL technique can evidently be employed with different degrees of specificity, ranging from positive staining of most cells in a specimen to failure to stain even those cells that are apoptotic (as shown by TEM; unpublished observations). Hence, in this relatively young research field, an acceptable in vivo research standard may remain to be established.

If apoptosis of eosinophils is a rare event in the airway mucosa in health and disease, alternative routes ought to be available for eliminating these potent cells without inflammation. As discussed earlier luminal entry is proposed to be one such innocuous means for clearing airway tissue eosinophils.

	EOSINOPHIL CYTOLYSIS

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

One significant but still poorly recognized proinflammatory fate of tissue eosinophils concerns their proclivity to undergo cytolysis (Figures 2-4) (16, 83). As reviewed elsewhere (16, 83), cytolytic eosinophils have frequently been observed and imaged in tissues subjected to eosinophilic inflammation, and also in association with parasitic infections. However, this phenomenon was not extensively discussed in the individual reports in which these data were presented, nor has eosinophil cytolysis received any attention at all in many leading reviews of eosinophil functional mechanisms. The little attention hitherto given to eosinophil cytolysis, as well as to its product of clusters of free extracellular granules (Cfegs) (Table 1, Figures 3 and 4), may reflect commendable consideration of methodologic flaws. Thus, artifactual causes for eosinophil cytolysis, such as in the taking of biopsy specimens, may not have been excluded in studies in which these phenomena have been previously depicted (16). A more explicit but equally valid reason for disinterest in eosinophil cytolysis may be the idea that this process is exclusively a mechanism secondary to intensive degranulation. Weiler and colleagues (18), discussing disease conditions that exhibit lysis of eosinophils at the site of lesions, thus noted: "eosinophil lysis was characterized by extensive degranulation, dissolution of the nuclear membrane, and disruption of the cellular membrane." On the basis of TEM of nasal mucosal eosinophils, Watanabe and associates (84) also conjectured that intensive intracellular release of toxic granule proteins caused lysis of eosinophils.

It is clear that eosinophil cytolysis would be of less interest as a degranulation mechanism if Cfegs consisted from their inception of more or less empty granules with negligible capacity to release eosinophil proteins. However, as subsequently discussed free eosinophil granules may not be an artifact. Moreover, cytolysis may be a primary event happening to eosinophils that exhibit few signs of prior degranulation.

	EOSINOPHIL CYTOLYSIS AS A PATHOGENIC MECHANISM

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Extracellular eosinophil granules have now been observed in whole-mount airway preparations well away from cut or handled tissue surfaces (10, 15, 32). Analysis with TEM further shows that cytolytic eosinophils and Cfegs are scattered among intact cells in inflamed airway mucosa (10, 15, 32). Hence, eosinophil cytolysis and the occurrence of Cfegs are not part of accidentally induced cell damage or tissue necrosis. As shown by ultrastructural analyses of human eosinophil-rich nasal tissues, cytolytic eosinophils may have a high number of unaltered granules. Cytolysis thus occurs in vivo in eosinophils that are in fact significantly less degranulated than are nonlysed eosinophils of the same tissue (Figure 2) (15, 22). These data depict eosinophil cytolysis as a primary mechanism by which granules and, subsequently, granule proteins are released in diseased airway tissues.

Cytolysis of mucosal tissue eosinophils was evoked within 1 h after allergen challenge of the airways of sensitized guinea-pigs. Furthermore, in this model, the number of Cfegs correlated well with the degree of epithelial fragility and shedding (32). Cytolytic eosinophils also occur in airway tissues of allergic rats, dogs, and primates (16) (unpublished observations), but this information may not be reported, presumably because of the general lack of recognition of eosinophil cytolysis (16).

Unfortunately, the immunologically important mouse models of human allergic airways may not exhibit asthmalike eosinophil activity (85). Accordingly, although significant airway and pulmonary eosinophilia develops in allergic mice, neither eosinophil cytolysis nor other signs of degranulation of tissue eosinophils, such as PMD (see the following discussion), have been clearly demonstrated in the current mouse models (31, 86, 87). By this same token, there has also been limited documentation of eosinophil-mediated tissue toxicity, such as epithelial damage-repair processes (32, 39), in these widely used models (31, 85, 88).

By contrast, in human allergic upper airways, repeated allergen exposures cause not only eosinophilia but also eosinophil cytolysis and PMD. In the nasal mucosa of subjects with mild 24-h symptoms of allergic rhinitis, the occurrence of eosinophil cytolysis and Cfegs indicated that one-third of the entire tissue population of eosinophils were affected by the cytolytic mechanism (22). The naturally occurring disease (seasonal allergic rhinitis) produced eosinophilia, but there was an even greater relative increase in the number of mucosal Cfegs (89). Tissue Cfegs seem also to abound and to reflect disease severity in the bronchial mucosa in asthma (14, 16, 83, 90). Further studies are warranted to test this hypothesis.

Being both a mechanism of degranulation and a pathway of cell demise, eosinophil cytolysis could readily create airway conditions with high levels of granule proteins but few eosinophils. Hence, one may speculate that a cytolytic disappearance of eosinophils may contribute to high ECP levels not only in asthma and allergic rhinitis, but also in diseases that are currently not recognized as typically eosinophilic diseases. For example, both chronic obstructive pulmonary disease (93) and rhinovirus infections have been associated with high local levels of ECP (94, 95). Exploratory studies of the occurrence of eosinophil cytolysis in these airway conditions seem warranted.

Photographs of airways in vivo (16, 83) (see the previous discussion) suggest that eosinophil cytolysis and Cfegs commonly occur in allergic rhinitis, asthma, nasal polyposis, and eosinophilic pneumonia, as well as in other eosinophilic diseases including atopic dermatitis, episodic angiodema, and parasitic infections (16, 83). Pictorial evidence for the occurrence and importance of eosinophil cytolysis in eosinophilic pneumonia has, in fact, accumulated during the past 50 yr (16, 96), and for even longer in asthma (16, 35) (Table 2). Furthermore, quantitative TEM studies in challenged and diseased airways now indicate that eosinophil cytolysis can be evoked rapidly, and that the cytolytic event is a primary activation mechanism distinct from other forms of degranulation (Figures 1, and 3; Table 1). The tissue distribution of Cfegs, the release of proteins from Cfegs in diseased tissues (22, 90), the rarity of eosinophil granules in phagosomes of macrophages (15), the association between Cfegs and epithelial injury (10, 32), and associations between Cfegs and disease severity (22, 89, 90) support the notion that eosinophil cytolysis is a potent proinflammatory mechanism, and that the released eosinophil granules are important effector organelles involved in the pathogenesis of significant respiratory tract diseases.

	REGULATION OF EOSINOPHIL CYTOLYSIS

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Airway conditions under which eosinophil cytolysis occurs may generally exhibit a highly complex mucosal molecular milieu, to which both plasma-derived proteins and cell-derived mediators contribute. The plasma exudation process may intermittently dominate the scene by dynamically flooding the extracellular matrix with pluripotent proteins as well as their bioactive degradation molecules. Importantly, plasma exudation occurs not only in severe conditions but also in mild, noninjurious and nonedematous airway inflammation (51). Hence, one may speculate that plasma-derived molecules might contribute to the cytolytic mechanism. Supporting this possibility is that blood plasma in conjunction with Sephadex beads is a powerful stimulus for rapid (within 1 h) cytolysis of eosinophils in primary culture (19). Similar in vitro experiments involving protein-coated Sephadex beads further suggest that eosinophil cytolysis is a major mechanism for prompt secretory IgA (sIgA)-induced eosinophil degranulation (Figure 3a) (18). Preliminary in vitro data suggest that anti-CD18 antibodies inhibit sIgA-induced eosinophil cytolysis (and release of granule proteins) without affecting PMD (19). The in vitro observations thus support the importance of firm adherence of eosinophils through ₂-integrins for effecting the eosinophil cytolysis mechanism.

Recent studies involving other cell types than eosinophils have shown that cytolysis may be executed along highly organized pathways (97). Thus, Fas receptor and tumor necrosis factor receptor type 1 may be coupled to either apoptotic or "necrotic" pathways in mouse fibrosarcoma cells (100). Further, in human T cells (Jurkat cells) exposed to known apoptogenic stimuli, the intracellular energy level may determine which type of cell death (apoptosis or cytolysis) is executed (97). The cytolysis of eosinophils may involve distinct molecular mechanisms that have yet to be unravelled. The idea of programmed eosinophil lysis is not entirely novel. Fourteen years ago, Fukuda and colleagues (101), who examined calcium ionophore-induced effects in vitro, mentioned the possibility that "eosinophils are programmed to lyse upon stimulation under certain conditions," along with the statement that "eosinophil lysis may occur as a consequence of intracellular release of cationic granule components." So far, putative agonists and antagonists of programmed cell lysis may not have been examined in experimental systems involving eosinophils. There is also no evidence that current airway-active drugs exert any specific actions against eosinophil cytolysis (16). Indeed, the entire field of biologic, physiologic, and pharmacologic control of eosinophil cytolysis remains unexplored.

	PMD AND THE PMD INDEX

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

By definition, PMD occurs in those intact eosinophils that contain more or less empty, specific granules (Table 1; Figures 1 and 2) (21). Airway mucosal eosinophils exhibiting lucent granules indicative of PMD have been demonstrated in asthma (14, 49, 91, 102, 103), allergic rhinitis (22, 104), and nasal polyposis (15, 105). As indicated by recent observations in allergen-exposed nasal airways, virtually all of the viable eosinophils in mucosal tissue may show signs of PMD under active disease conditions (22). Eosinophil granule alterations reflecting PMD have also been described in guinea pig models of asthma (32, 106). By contrast, the commonly used mouse models of asthma, which exhibit extensive pulmonary eosinophilia, appear to lack significant PMD (31, 86, 87). As already noted, other signs of eosinophil degranulation also remain to be demonstrated in allergic mice (85, 88). A careful search for eosinophil cytolysis and PMD in murine airway tissues is warranted, along with development of novel mouse models of asthma and rhinitis (85).

Different patterns of granular lucency can define the features of PMD in some further, systematized detail. At least four distinct subtypes of specific granules may be discerned (15), three of which are based on different sites of electron lucency in the granule (matrix, core, or both matrix and core) and one of which is based on the intact granule. Eosinophil granule subtypes may exhibit characteristic patterns in individual diseased airway mucosa (15), but this has not been extensively examined.

A useful PMD index (PMDi) for assessing the extent of PMD in tissues can be obtained by determining the percentage of altered granules (granules exhibiting ultrastructural signs of protein release) in a given eosinophil (15). The use of this method in studies of human diseased nasal mucosa showed that eosinophil cytolysis occurred in eosinophils exhibiting a relatively low degree of prior PMD (15, 22), supporting the notion that eosinophil cytolysis and PMD are distinct mechanisms for granule protein release in the airways.

Quantitative TEM data, obtained in studies of eosinophilic nasal polyps, were used to compare the PMDi with the common procedure of equating any reduced numbers of granules with degranulation (15). Although a statistically significant, negative correlation was found between the number of granules and the PMDi in individual cells, the scatter of data points was marked and the slope of the correlation line was shallow (15). Importantly, it is not known whether eosinophils with reduced numbers of granules migrate to the airways, from other sites, such as when generation and traffic of eosinophils to the airways are stimulated by inflammatory processes (71, 107). Hence, the number of specific granules alone may not be a sufficient indicator of eosinophil degranulation.

Large differences in mean PMDi have been observed under different eosinophilic conditions, and even between individual patients (15, 22). This finding, and the independence of the PMDi from the most common measure of eosinophilic inflammation (i.e., tissue numbers of eosinophils), further underscore the potential utility of the PMDi. Further clinical studies assessing PMDi are now warranted to explore the relationship between PMD and the severity of airway diseases. The pharmacology of inducement and inhibition of PMD also remains unexplored.

	FUNCTIONAL ROLES OF DIFFERENT MODES OF DEGRANULATION

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Extensive in vitro research has generated molecular information about signal transduction pathways potentially controlling granule protein release (108). However, few studies have explored the mode (PMD, eosinophil cytolysis, or exocytosis) of protein release after stimulation in vitro. PMD and eosinophil cytolysis may occur commonly and have variably been described, but not quantified, after in vitro stimulation of eosinophils with the calcium ionophore A23187 (112) and with complement fragment 5a (C5a), platelet activating factor (113), interleukin-5 (114), and respiratory syncytial viruses (115). Indeed, the occasional quantitative ultrastructural in vitro study indicates that significant PMD is induced during the purification of eosinophils (114, 116). If PMD is already extensive under baseline conditions, current in vitro studies will not be able to unravel the mechanisms that initiate PMD in eosinophils. It is also not known how the various experimental approaches now used to seek molecular information about eosinophil regulation will affect either PMD or eosinophil cytolysis.

It may frequently be presumed that degranulation of eosinophils in vitro occurs via a classical exocytosis mechanism (Table 1). However, beyond a few in vitro studies describing either shape and membrane changes compatible with eosinophil exocytosis (117, 118) or ultrastructural characteristics of exocytosis (112), little specific information about this mechanism of eosinophil degranulation is currently available. An interesting report suggests that eosinophil cytolysis is much more prominent in vitro than is the exocytosis mechanism (18). If eosinophil exocytosis is a major mechanism in isolated eosinophils, this may be yet another example of differences between in vitro and in vivo; currently, a significant participation of classical eosinophil exocytosis remains to be demonstrated in diseased airway mucosae.

In vivo observations suggest that eosinophil cytolysis and PMD may be independent major modes of eosinophil granule protein release in respiratory tract diseases (15, 22, 91, 96). It has been conjectured that PMD provides the opportunity for slow and selective release of proteins (119), whereas eosinophil cytolysis would produce an ultimate and rapid release of granule products (16). Such distinct modes of degranulation would agree with the proposed disparate roles of the eosinophil, ranging from fulminant cytotoxicity in the killing of parasites (120) and causing tissue disturbances in allergic airway diseases (16, 22, 121) to immunoregulation (7, 10, 122). (As we have briefly reviewed elsewhere [16] eosinophil-parasite interactions provide numerous illustrations of Cfegs and eosinophil cytolysis. Such ultrastructural evidence in favor of a role of eosinophil cytolysis in defense against parasites continues to be reported [120, 126], but eosinophil cytolysis has so far received little attention in this field [16].)

	SUMMARY

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

Although the modes of granule protein release and disappearance of eosinophils from the airway mucosa may determine the role of these granulocytes, they have not been extensively examined in vivo. For example, it would be of interest to learn more about potential connections between mucosal eosinophils and those eosinophils that may abound in tracheobronchial lymph nodes in allergy and asthma (32, 37). Several questions also remain about such established in vivo fates of eosinophils as their entry into the airway lumen. Traversing the epithelium may be important in "first line" airway defense, but may also be a clearance route whereby eosinophils could leave the mucosal tissue without causing inflammation. Perhaps mucosal inflammation can in part be resolved by efficiently inducing the luminal entry of airway tissue granulocytes.

Eosinophils in the airway lumen, like eosinophils in culture, are readily available for investigation. It is attractive, but not always possible, to translate appearances and biologic features of cells removed from tissue to diseased tissues in vivo. Thus, for example, although a classical form of eosinophil degranulation, exocytosis (Table 1), clearly occurs in vitro, it may remain to be demonstrated in airway tissues in vivo. Moreover, it cannot be excluded that apoptosis as a fate may be more valid for eosinophils of culture and airway luminal phenotype than for eosinophils in diseased airway tissues, whether or not the latter are treated with steroids. On the other hand, important phenomena may exist, such as eosinophil cytolysis, that do not fall within current avenues of in vitro research, but which may nevertheless may play significant roles in vivo. This latter possibility underscores the continued utility of disease-like in vivo test systems involving classical clinical approaches (physiology, clinical pharmacodynamics, and histopathology) in original biomedical research (35, 36, 127). As a further point, PMD may exemplify an acknowledged but little investigated mechanism that is common both in vitro and in vivo.

Growing evidence suggests that eosinophil cytolysis, producing Cfegs as protein-laden free granules in diseased airway/ pulmonary tissues, constitutes a major mode of activation of tissue eosinophils in vivo. This mode of degranulation is distinct from classical exocytosis and the common PMD (Figures 1 and 4). The extracellular eosinophil granules, appearing as Cfegs, are not doomed to phagocytosis, but appear to release their potent granule proteins close to target cells in diseased airway mucosa. The biologic regulation of eosinophil cytolysis (including the mode of protein release from free granules) emerges as a promising area of investigation to further the understanding of leukocyte activation and fate, and potentially to find novel treatments for eosinophilic conditions.

	Footnotes

Supported by The Medical Faculty, Lund, Sweden, The Swedish Medical Research Council, Astra Draco, The Heart and Lung Foundation, and The Vårdal foundation, Sweden.

Correspondence and requests for reprints should be addressed to Carl Persson, Department of Clinical Pharmacology, Lund University Hospital, S-221 85, Lund, Sweden. E-mail: carl.persson{at}klinfarm.lu.se

(Received in original form June 17, 1999 and in revised form October 29, 1999).

	References

TOP INTRODUCTION LUMINAL ENTRY AND LUMINAL... EOSINOPHIL APOPTOSIS: IN VITRO... APOPTOTIC EOSINOPHILS IN HUMAN... THE SEARCH FOR APOPTOTIC... EOSINOPHIL CYTOLYSIS EOSINOPHIL CYTOLYSIS AS A... REGULATION OF EOSINOPHIL... PMD AND THE PMD... FUNCTIONAL ROLES OF DIFFERENT... SUMMARY REFERENCES

1. Rothenberg, M. E.. 1998. Eosinophilia. N. Engl. J. Med. 338: 1592-1600 [Free Full Text].

2. Strek, M. E., and A. R. Leff. 1997. Eosinophils. In P. J. Barnes, M. M. Grunstein, A. R. Leff, and A. J. Woolcock, editors. Asthma, Vol. 1. Lippincott Raven, New York. 399-417.

3. Weller, P. F.. 1997. Human eosinophils. J. Allergy Clin. Immunol. 100: 283-287 [Medline].

4. Giembycz, M. A., and M. A. Lindsay. 1999. Pharmacology of the eosinophil. Pharmacol. Rev. 51: 213-340 [Abstract/Free Full Text].

5. Kroegel, C., J. C. Virchow Jr., W. Luttmann, C. Walker, and J. A. Warner. 1994. Pulmonary immune cells in health and disease: the eosinophil leucocyte (Part I). Eur. Respir. J. 7: 519-543 [Abstract].

6. Martin, L. B., H. Kita, K. M. Leiferman, and G. J. Gleich. 1996. Eosinophils in allergy: role in disease, degranulation, and cytokines. Int. Arch. Allergy Immunol. 109: 207-215 [Medline].

7. Weller, P. F., K. Lim, H. C. Wan, A. M. Dvorak, D. T. Wong, W. W. Cruikshank, H. Kornfeld, and D. M. Center. 1996. Role of the eosinophil in allergic reactions. Eur. Respir. J. Suppl. 22: 109s-115s [Medline].

8. Erjefält, J. S., I. Erjefält, F. Sundler, and C. G. A. Persson. 1994. Microcirculation-derived factors in airway epithelial repair in vivo. Microvasc. Res. 48: 161-178 [Medline].

9. Wong, D. T. W., R. B. Donoff, J. Yang, B. Song, K. Mattosian, N. Nagura, A. Elovic, J. McBride, G. T. Gallegher, R. Todd, T. Chiang, L. S. Chou, M. Y. Chou, S. J. Galli, and P. F. Weller. 1993. Sequential expression of TGF-a and TGF-B1 by eosinophils during cutaneous wound healing in the hamster. Am. J. Pathol. 143: 130-142 [Abstract].

10. Erjefält, J. S., F. Sundler, and C. G. A. Persson. 1996. Eosinophils, neutrophils, and venular gaps in the airway mucosa at epithelial removal-restitution. Am. J. Respir. Crit. Care Med. 153: 1666-1674 [Abstract].

11. Gleich, G. J., S. Motojima, E. Frigas, G. M. Kephart, T. Fujisawa, and L. P. Kravis. 1987. The eosinophilic leukocyte and the pathology of fatal bronchial asthma: evidence for pathologic heterogeneity. J. Allergy Clin. Immunol. 80: 412-415 [Medline].

12. Venge, P., and L. Håkansson. 1991. Eosinophils. In M. A. Kaliner, P. J. Barnes, and C. G. A. Persson, editors. Asthma: Its Pathology and Treatment. Marcel Dekker, New York. 477-502.

13. Fahy, J. V., J. Liu, H. Wong, and H. A. Boushey. 1993. Cellular and biochemical analysis of induced sputum from asthmatic and from healthy subjects. Am. Rev. Respir. Dis. 147: 1126-1131 [Medline].

14. Jeffery, P. K., R. W. Godfrey, E. Adelroth, F. Nelson, A. Rogers, and S. A. Johansson. 1992. Effects of treatment on airway inflammation and thickening of basement membrane reticular collagen in asthma: a quantitative light and electron microscopic study. Am. Rev. Respir. Dis. 145: 890-899 [Medline].

15. Erjefält, J. S., M. Andersson, L. Greiff, M. Korsgren, M. Gizycki, P. K. Jeffery, and C. G. A. Persson. 1998. Cytolysis and piecemeal degranulation as distinct modes of activation of airway mucosal eosinophils. J. Allergy Clin. Immunol. 102: 286-294 [Medline].

16. Persson, C. G. A., and J. S. Erjefält. 1997. Eosinophil lysis and free granules: an in vivo paradigm for cell activation and drug development. Trends Pharmacol. Sci. 18: 117-123 [Medline].

17. Majno, G., and I. Joris. 1995. Apoptosis, oncosis, and necrosis: an overview of cell death. Am. J. Pathol. 146: 3-15 [Abstract].

18. Weiler, C. R., H. Kita, M. Hukee, and G. J. Gleich. 1996. Eosinophil viability during immunoglobulin-induced degranulation. J. Leukoc. Biol. 60: 493-501 [Abstract].

19. Erjefält, J. S., G. M. K. Malm, and C. G. A. Persson. 1999. Eosinopil cytolysis (ECL) as a major mechanism for sIgA-induced granule protein release (abstract). Am. J. Respir. Crit. Care Med. 159: A99 .

20. Dvorak, A. M.. 1999. Images in clinical medicine: an apoptotic eosinophil. N. Engl. J Med. 340: 437 [Free Full Text].

21. Dvorak, A. M., T. Furitso, L. Letourneau, T. Ishizaka, and L. Ackerman. 1997. Mature eosinophils stimulated to develop in human cord blood mononuclear cell cultures with human recombinant interleukin-5: I. Piecemeal degranulation of specific granules and distribution of Chacot-Leiden crystal protein. Am. J. Pathol. 138: 69-82 [Abstract].

22. Erjefält, J. S., L. Greiff, M. Andersson, E. Matsson, H. Petersen, M. Linden, T. Ansari, P. K. Jeffery, and C. G. A. Persson. 1999. Allergen-induced eosinophil cytolysis is a primary mechanism for granule protein release in human upper airways. Am. J. Respir. Crit. Care Med. 160: 304-312 [Abstract/Free Full Text].

23. Tai, P. C., C. J. Spry, C. Peterson, P. Venge, and I. Olsson. 1984. Monoclonal antibodies distinguish between storage and secreted forms of eosinophil cationic protein. Nature 309: 182-184 [Medline].

24. Druilhe, A., B Wallaert, A. Tsicopoulos, J. R. Lapa e Silva, R. Tillie, I. Leblond, A. B. Tonnel, and M. Pretolani. 1998. Apoptosis, proliferation, and expression of Bcl-2,Fas, and Fas ligand in bronchial biopsies from asthmatics. Am. J. Respir. Cell Mol. Biol. 19: 747-757 [Abstract/Free Full Text].

25. Kodama, T., T. Matsuyama, S. Miyata, H. Nishimura, Y. Niashioka, O. Kitada, and M. Sugita. 1998. Kinetics of apoptotis in the lung of mice with allergic airway inflammation. Clin. Exp. Allergy 28: 1435-1443 [Medline].

26. Jahnsen, F. L., T. S. Halstensen, and P. Brandtzaeg. 1994. Erroneous immunohistochemical application of monoclonal antibody EG2 to detect cellular activation. Lancet 344: 1514-1515 [Medline].

27. Frankfurt, O. S., J. A. Robb, E. V. Sugarbaker, and L. Villa. 1996. Monoclonal antibody to single-stranded DNA is a specific and sensitive cellular marker of apoptosis. Exp. Cell Res. 226: 387-397 [Medline].

28. Huppertz, B., F. Hans-Georg, and P. Kaufmann. 1999. The apoptosis cascademorphological and immunohistochemical methods for its visualisation. Anat. Rec. 200: 1-18 .

29. Labat, M. F., C. Guillermet, P. Lorimier, C. Robert, S. Lantuejoul, E. Brambilla, and A. Negoescu. 1998. TUNEL apoptotic cell detection in tissue sections: critical evaluation and improvement. J. Histochem. Cytochem. 46: 327-334 [Abstract/Free Full Text].

30. Kato, M., G. M. Kephart, N. J. Talley, J. M. Wagner, M. G. Sarr, M. Bonno, T. W. McGovern, and G. J. Gleich. 1998. Eosinophil infiltration and degranulation in normal human tissue. Anat. Rec. 252: 418-425 [Medline].

31. Korsgren, M., J. S. Erjefält, O. Korsgren, F. Sundler, and C. G. A. Persson. 1997. Allergic eosinophil-rich inflammation develops in lungs and airways of B cell-deficient mice. J. Exp. Med. 185: 885-892 [Abstract/Free Full Text].

32. Erjefält, J. S., M. Korsgren, M. C. Nilsson, F. Sundler, and C. G. A. Persson. 1997. Association between inflammation and epithelial damage-restitution processes in allergic airways in vivo. Clin. Exp. Allergy 27: 1345-1357 .

33. Yamaguchi, Y., T. Suda, S. Ohta, K. Tominaga, Y. Miura, and T. Kasahara. 1991. Analysis of the survival of mature human eosinophils: interleukin-5 prevents apoptosis in mature human eosinophils. Blood 78: 2542-2547 [Abstract/Free Full Text].

34. Stern, M., L. Meagher, J. Savill, and C. Haslett. 1992. Apoptosis in human eosinophils: programmed cell death in the eosinophil leads to phagocytosis by macrophages and is modulated by IL-5. J. Immunol. 148: 3543-3549 [Abstract].

35. Persson, C. G. A.. 1997. Centennial notions of asthma as an eosinophilic, desquamative, exudative, and steroid-sensitive disease. Lancet 350: 1021-1024 [Medline].

36. Persson, C. G. A., J. S. Erjefält, M. Andersson, I. Erjefält, L. Greiff, M. Korsgren, M. Linden, F. Sundler, and C. Svensson. 1997. Epithelium, microcirculation, and eosinophilsnew aspects of the allergic airway in vivo. Allergy 52: 241-255 [Medline].

37. Huber, F. K., and K. K. Koessler. 1922. The pathology of bronchial asthma. Arch. Intern. Med. 30: 689-760 [Abstract/Free Full Text].

38. Sedgwick, J. B., W. J. Calhoun, R. F. Vrtis, M. E. Bates, P. K. McAllister, and W. W. Busse. 1992. Comparison of airway and blood eosinophil function after in vivo antigen challenge. J. Immunol. 149: 3710-3718 [Abstract].

39. Persson, C. G. A., and J. S. Erjefält. 1996. Airway epithelial restitution following shedding and denudation. In R. G. Crystal, J. B. West, P. J. Barnes, E. R. Weibel, and N. S. Cherniak, editors. The Lung, Scientific Foundation. Lipincott-Raven, Philadelphia. 2611-2627.

40. Hansel, T. T., and C. Walker. 1992. The migration of eosinophils into the sputum of asthmatics: the role of adhesion molecules. Clin. Exp. Allergy 22: 345-356 [Medline].

41. Miller, L. A., and E. C. Butcher. 1998. Human airway epithelial monolayers promote selective transmigration of memory T cells: a transepithelial model of lymphocyte migration into the airways. Am. J. Respir. Cell Mol. Biol. 19: 892-900 [Abstract/Free Full Text].

42. Erjefält, J. S., I. Erjefält, F. Sundler, and C. G. A. Persson. 1995. Epithelial pathways for luminal entry of bulk plasma. Clin. Exp. Allergy 25: 187-195 [Medline].

43. Liu, L., A. E. Zuurbier, F. P. Mul, A. J. Verhoeven, R. Lutter, E. F. Knol, and D. Roos. 1998. Triple role of platelet-activating factor in eosinophil migration across monolayers of lung epithelial cells: eosinophil chemoattractant and priming agent and epithelial cell activator. J. Immunol. 161: 3064-3070 [Abstract/Free Full Text].

44. Liu, L., F. P. Mul, R. Lutter, D. Roos, and E. F. Knol. 1996. Transmigration of human neutrophils across airway epithelial cell monolayers is preferentially in the physiologic basolateral-to-apical direction. Am. J. Respir. Cell Mol. Biol. 15: 771-780 [Abstract].

45. Taguchi, M., D. Sampath, T. Koga, M. Castro, D. C. Look, S. Nakajima, and M. J. Holtzman. 1998. Patterns for RANTES secretion and intercellular adhesion molecule 1 expression mediate transepithelial T cell traffic based on analyses in vitro and in vivo. J. Exp. Med. 187: 1927-1940 [Abstract/Free Full Text].

46. Humbles, A. A., D. M. Conroy, S. Marleau, S. M. Rankin, R. T. Palframan, A. E. Proudfoot, T. N. Wells, D. Li, P. K. Jeffery, D. A. Griffiths, Johnson, T. J. Williams, and P. J. Jose. 1997. Kinetics of eotaxin generation and its relationship to eosinophil accumulation in allergic airways disease: analysis in a guinea pig model in vivo. J. Exp. Med. 186: 601-612 [Abstract/Free Full Text].

47. Munoz, N. M., I. Douglas, D. Mayer, A. Herrnreiter, X. Zhu, and A. R. Leff. 1997. Eosinophil chemotaxis inhibited by 5-lipoxygenase blockade and leukotriene receptor antagonism. Am. J. Respir. Crit. Care Med. 155: 1398-1403 [Abstract].

48. Schneider, T., D. van Velzen, R. Moqbel, and A. C. Issekutz. 1997. Kinetics and quantitation of eosinophil and neutrophil recruitment to allergic lung inflammation in a Brown Norway rat model. Am. J. Respir. Cell Mol. Biol. 17: 702-712 [Abstract/Free Full Text].

49. Metzger, W. J., D. Zavala, H. B. Richerson, P. Moseley, P. Iwamota, M. Monick, K. Sjoerdsma, and G.W. Hunninghake. 1987. Local allergen challenge and bronchoalveolar lavage of allergic asthmatic lungs: description of the model and local airway inflammation. Am. Rev. Respir. Dis. 135: 433-440 [Medline].

50. Aalbers, R., J. G. de Monchy, H. F. Kauffman, M. Smith, Y. Hoekstra, B. Vrugt, and W. Timens. 1993. Dynamics of eosinophil infiltration in the bronchial mucosa before and after the late asthmatic reaction. Eur. Respir. J. 6: 840-847 [Abstract].

51. Persson, C. G. A., J. S. Erjefält, L. Greiff, M. Andersson, I. Erjefält, R. W. Godfrey, M. Korsgren, M. Linden, F. Sundler, and C. Svensson. 1998. Plasma-derived proteins in airway defence, disease and repair of epithelial injury. Eur. Respir. J. 11: 958-970 [Abstract].

52. Milks, L. C., M. J. Brontoli, and E. B. Cramer. 1983. Epithelial permeability and the transepithelial migration of human neutrophils. J. Cell Biol. 96: 1241-1247 [Abstract/Free Full Text].

53. Tsuyuki, S., C. Bertrand, F. Erard, A. Trifilieff, J. Tsuyuki, M. Wesp, G. P. Anderson, and A. J. Coyle. 1995. Activation of the Fas receptor on lung eosinophils leads to apoptosis and the resolution of eosinophilic inflammation of the airways. J. Clin. Invest. 96: 2924-2931 .

54. Erjefält, J. S., L. Greiff, E. Matsson, M. Anderson, M. Linden, A. Rogers, P. K. Jeffery, and C. G. A. Persson. 1998. Lack of apoptotic airway tissue eosinophils in vivo, even in ste