|
|
||||||||
Published online before print March 12, 2004
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||



,1
* Division of Molecular Biosciences, John Curtin School of Medical Research, Australian National University, Canberra;
School of Biomedical Sciences, Faculty of Health, University of Newcastle, Australia;
Department of Biochemistry and Molecular Biology, Mayo Clinic, Scottsdale, Arizona; and
Department of Molecular Biosciences, University of Adelaide, Australia
1 Correspondence: Division of Molecular Biosciences, The John Curtin School of Medical Research, The Australian National University, Canberra, ACT, 0200, Australia. E-mail: Paul.Foster{at}anu.edu.au and School of Biomedical Sciences, Faculty of Health, University of Newcastle, Newcastle, NSW, Australia 2300. E-mail: Paul.Foster{at}newcastle.edu.au
| ABSTRACT |
|---|
|
|
|---|
Key Words: asthma cellular activation eosinophil
| INTRODUCTION |
|---|
|
|
|---|
Cationic, cytotoxic proteins are stored in the secondary granules of eosinophils, and their release, termed degranulation, may proceed via three mechanisms: classical exocytosis, whereby secondary granules fuse directly with the plasma membrane of the cell to release their entire contents into the extracellular environment [6 ]; piecemeal degranulation (PMD), whereby small vesicles bud from the secondary granules and subsequently transport a subset of the granule proteins to the cell surface, resulting in the progressive loss of secondary granule constituents [7 , 8 ]; and cytolysis, a highly organized process of cell death, where loss of the plasma membranes integrity leads to the release of cellular contents [9 ]. These mechanisms of eosinophil degranulation have been well characterized by transmission electron microscopy (TEM), and all three processes have been observed in tissues from patients suffering from a range of inflammatory disorders [9 10 11 ]. Notably, in airway tissues from atopic patients, PMD was predominantly observed, although some eosinophils released granule contents by cytolysis [9 , 11 ].
Recently, mouse models of CD4+ T helper 2 lymphocyte immunity have emerged as central tools to study disease mechanisms that underpin allergic disorders [12 ]. Mouse models of allergic airways disease display several characteristic features of human asthma and have been used extensively to define eosinophil functions in pathogenesis [13 , 14 ]. Indeed, in mouse models of experimental asthma, the recruitment of eosinophils to the lung has been linked directly to the development of pathogenic features that are hallmarks of the disorder in humans [15 16 17 ]. However, attenuation (but not abolition) of eosinophil trafficking to the lung does not always correlate with inhibition of disease processes [18 19 20 ], and eosinophil degranulation in the submucosa of the allergic lung of mice [21 , 22 ] (a key feature of human disease [4 , 9 , 11 ]) has not been observed. The failure to observe eosinophil degranulation has questioned the relevance of mouse models of allergic disease to understanding the contribution of eosinophils to allergic disorders [21 ].
Although eosinophil degranulation is not evident in subepithelial regions of the lungs of allergic mice, MBP and EPO have been detected in cell-free extracts taken from the bronchoalveolar lavage fluid (BALF) [17 , 23 , 24 ]. It is interesting that eosinophils residing in the airway lumen of allergic mice have been shown to actively participate in immune processes [25 , 26 ]. Collectively, these investigations suggest that eosinophils may selectively receive activation signals in the lumenal compartment. Notably, only limited studies have investigated the activation status of eosinophils in the lungs of allergic mice, and these have focused on tissue-dwelling cells and not of those residing in the airway lumen [16 , 21 , 22 ].
In this investigation, we extend our characterization of the role of the eosinophil in the pathogenesis of allergic airways disease in mice by analyzing the ultrastructure and activation status of this cell as it migrates from the bone marrow to the airway lumen during inflammation. We detected cell-free EPO in respiratory secretions (BALF) and showed that eosinophils in the airway lumen are highly activated and release granule contents by PMD (the primary mechanism of degranulation in human asthmatic tissue). Furthermore, we link eosinophil activation (evidenced by PMD and the presence of cell-free EPO) in the airway lumen with enhanced airways reactivity to methacholine. We speculate that the differences in the degree and site of eosinophil activation/degranulation between mouse models of asthma and human-diseased tissue may arise, in part, because of higher thresholds for the activation of the mouse leukocyte in response to secretagogues and also because of the acute (mouse) versus chronic (human) nature of the inflammatory processes. However, degranulation of eosinophils in respiratory cavities, which is a feature of some allergic respiratory disorders [4 , 5 , 27 ], can be induced in the inflamed lungs of mice, providing a model to understand this process.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Induction of allergic airways disease
Mice were sensitized by an intraperitoneal injection of ovalbumin (OVA; Grade V, Sigma Chemical Co., St. Louis, MO) on days 0 and 12 at a dose of 50 µg in 1 mg Alhydrogel (CSL Ltd., Parkville, Australia) in isotonic saline (SAL) [15
]. Nonsensitized mice received 1 mg Alhydrogel in isotonic solutions (SAL). On days 24, 26, 28, and 30, all groups of mice were exposed to an aerosol formed from a 1% w/v solution of OVA in SAL for three periods of 30 min and a 30-min interval between each exposure. The aerosol was generated by a RapidFlow nebulizer bowl (Allersearch, Melbourne, Australia) [15
]. This regime produced maximal, inflammatory cell numbers in the bone marrow, blood, and pulmonary tissue (ref. [15
]; data not shown). Cellular responses were characterized 16 h after the last aerosol exposure as described previously [15
].
Collection and processing of samples for TEM
Lung tissue and cells isolated from bone marrow, blood, and BALF were processed for TEM. Bone marrow, blood, and BALF cells and lung tissue were isolated (six mice per group) by standard techniques [15
, 24
, 29
, 30
]. Briefly, individual cell populations were pooled, and the red blood cells were lysed using an ammonium chloride buffer (0.16 M NH4Cl, 0.17 M Tris, pH 7.2). Subsequently, the cell populations were washed in phosphate-buffered saline (PBS) containing 1 mM EDTA before being fixed in 2% glutaraldehyde in 0.1 M sodium cocadylate, pH 7.4, at 4°C overnight. Isolated lung tissue (3x3x3 mm) was placed immediately in fixative (as above) and incubated at 4°C overnight. All samples were washed in 0.1 M sodium cocadylate buffer. The samples were postfixed in 1% osmium tetroxide for 90 min, stained en bloc in 2% uranyl acetate, dehydrated in graded alcohol solutions, and embedded in Spurs resin (ProSciTech, Australia). Sections (1 µm) were stained with toluidine blue to identify intact regions of pulmonary tissue. Ultrathin sections (8085 nm) were cut on a Reichert-Jung Ultracut E ultramicrotome. The samples were stained with lead citrate before analysis using an Hitachi transmission electron microscope (H-7000, Hitachi, Japan).
Quantification of eosinophil degranulation by TEM
Initially, the number of eosinophils that was resting, degranulating, or apoptotic was defined in various compartments. Degranulating eosinophils were then delineated further from one another based on the mechanism of granule protein release (exocytosis, PMD, and cytolysis). Subsequently, the secondary granules of each eosinophil were counted and classified into four categories (type I, intact; type II, loss of crystal core; type III, loss of matrix; and type IV, loss of crystal core and matrix). The definitions for all groups are based on the method devised by Erjefalt et al. [9
]. The extent of degranulation was quantified by means of a degranulation index, calculated using the following formula: degranulation index = 100 x (number of activated granules, types IIIV/total granules) [9
].
Measurement of AHR
Airways reactivity to methacholine was measured in conscious, unrestrained mice using a barometric plethysmograph (Buxco, Troy, NY) as described previously [30
].
Eosinophil purification
Eosinophils from the peritoneal cavity of IL-5 Tg BALB/c mice were purified by fluorescence-activated cell sorter (FACS; FACStar Plus, Becton Dickinson, San Jose, CA) according to their characteristic forward-scatter versus side-scatter plot and light polarization properties as described previously [29
]. The purity of the enriched population was
95%, as determined by differential staining with Giemsa-May-Grünwald. Greater than 98% of the purified eosinophils was deemed viable by trypan blue exclusion. The contaminating cell population was macrophages.
Adhesion assay
Eosinophil adhesion was monitored in triplicate samples by measuring eosinophil peroxidase content of adherent cells. The cells were cultured in 96-well flat-bottom tissue-culture plates, which were coated with 2.5% bovine serum albumin (BSA) in PBS for 2 h at 37°C. After purification, eosinophils were washed and resuspended in RPMI 1640 at a concentration of 57.5 x 105 cells/ml. A 100-µl aliquot of the eosinophil suspension was added to each well. Stimulation of the eosinophils was initiated by the addition of an equal volume of phorbol 12-myristate 13-acetate (PMA) dissolved in RPMI 1640. The eosinophils were cultured for various periods of time at 37°C in a humidified atmosphere containing 5% CO2. After the incubation, each well was gently washed three times with 200 µl warm PBS to remove nonadherent cells. EPO was extracted by lysing the eosinophils with 200 µl 0.22% cetyltrimethylammonium bromide (CTAB), dissolved in 10 mM HEPES buffer, pH 8.0. Eosinophil adhesion was reported as a percentage of the EPO extracted from adherent cells according to the formula: % adhesion = (EPO in adherent fraction/total EPO added to well) x 100. The denominator was determined by measuring EPO content in an equivalent aliquot of cells added at the start of the experiment. The enzyme was extracted using 0.22% CTAB dissolved in 10 mM HEPES buffer, pH 8.0. The samples were frozen at 70°C until EPO content was measured using a colorimetric assay.
In vitro degranulation assay
Purified eosinophils from the peritoneal cavity of IL-5 Tg mice were cultured in 24-well flat-bottom tissue-culture plates, which were coated with a 2.5% BSA solution. Approximately 5 x 105 cells were added to each well in a volume of 250 µl RPMI 1640 to initiate degranulation, an equal volume of culture media, which contained PMA was added. The leukocytes were cultured for 04 h at 37°C in a humidified atmosphere containing 5% CO2. After incubation, the plates were centrifuged at 300 g for 5 min. The supernatant was collected and stored at 70°C until assay for MBP.
To assess the morphology of cultured eosinophils by TEM, the cells were cultured on a glass coverslip, which was coated in succession with poly-L-lysine (10 mg/ml) and then BSA (2.5% w/v in PBS). The cells were stimulated for 3 h in the presence or absence of 107 M PMA. After the incubation, the coverslips were gently rinsed with warm PBS three times, and the cells were fixed in 2% glutaraldehyde dissolved in 0.1 M sodium cacodylate buffer, pH 7.4, at 4°C overnight. The subsequent steps were performed as described above.
Detection of EPO and MBP
EPO content in cell-free BALF and eosinophil lysates was measured using a colorometric assay [31
, 32
]. Briefly, a 75-µl aliquot of sample was transferred to a 96-well microtiter plate, and the reaction was initiated by the addition of 75 µl substrate solution (12 mM o-phenylene diamine, 0.005% H2O2, in 10 mM HEPES, pH 8.0). CTAB is not only a detergent used to disrupt membranes but also serves as a source of bromine ions required by EPO. The final concentration of CTAB in the reaction was 0.11% w/v and was present in the lysates or added to the substrate solution when analyzing BALF. The enzymatic reaction was stopped after a 30-min incubation at room temperature by the addition of 50 µl 4 N H2SO4. Absorbance at 490 nm was measured using a Thermomax microplate reader, and data were analyzed using SoftMax (version 2.01, Molecular Devices, Menlo Park, CA).
Eosinophils were purified from the spleens of IL-5 Tg mice as described previously [29 ], and intracellular content was obtained by freeze-thawing in PBS. MBP was standardized to 100 U/ml, which is equivalent to 1 x 106-purified eosinophils/µl supernatant. The presence of MBP in supernatants from cultured eosinophils was assessed using an immunodot-blot assay, developed with Western blue as described previously [24 ]. The dot-blot was digitalized by densitometry, and the image was analyzed using National Institutes of Health (NIH; Bethesda, MD) Image 1.62 software.
Statistical analysis
Data are presented as the mean ± SEM except for the data describing a frequency of occurrence in a sample population, which were presented as a percentage. Statistical significance of differences between experimental groups was assessed with Students unpaired t-test or
2 analysis. Differences in means were considered significant if P< 0.05.
| RESULTS |
|---|
|
|
|---|
|
|
|
Only a small fraction of eosinophils in the airway lumen of allergic mice was apoptotic (Table 1) . However, we often observed that macrophages in the airway lumen had engulfed eosinophil constituents including intact secondary granules (Fig. 2D) .
Eosinophils in the airway lumen have the highest activation status
Concurrently with morphological description of eosinophil degranulation, we attempted to quantify their activation status by grading the degree of degranulation per cell in a defined compartment relative to eosinophils isolated from naive (nonallergic) mice (Fig. 3
). Resting eosinophils contained only type I granules, and activated eosinophils possessed intact granules and those displaying at least a partial loss of granule content (types IIIV). When the data were transformed into the degranulation index (on a scale of 0100, where 0 indicates a resting eosinophil and 100, a fully activated cell with no type I granules), distinct differences in the degree of activation were observed between eosinophils residing in the various compartments (Fig. 3A) . Eosinophils isolated from naive mice had a degranulation index equal to bone marrow-derived eosinophils from allergic mice. Furthermore, in the case of WT mice, the degranulation index of circulating eosinophils increased significantly after allergen provocation (9.9±1.7 vs. 16.2±2.6, P<0.05). Although eosinophils in lung tissue of allergic mice were activated (in contrast to bone marrow-derived cells), the highest activation status was observed in the airways lumen. This is particularly evident after classifying the eosinophils as low, moderately, or highly activated (Fig. 3B)
and when the proportion of type IV granules is compared between eosinophil populations in the various compartments (Fig. 3C)
. Movement from the tissue to the airway lumen significantly promoted the loss of the crystal core and matrix from individual granules. These data directly correlate with the morphological assessment of eosinophil activation (Fig. 2
and degranulation index). Several differences were observed between WT and IL-5 Tg eosinophils in the various compartments of allergic mice. Notably, although the proportion of activated eosinophils in IL-5 Tg mice was only slightly increased within any one compartment (Table 1)
, those cells that became activated reached higher levels of activation (Fig. 3A
and 3B)
. This difference was particularly evident in comparison with BALF populations.
|
|
|
|
| DISCUSSION |
|---|
|
|
|---|
Migration of eosinophils into the airway lumen during allergic inflammation resulted in marked activation and the induction of PMD. The ultrastructural features and in particular, the fraction of granules with a partial loss of matrix and core (type IV granules) and the degranulation index strongly indicated a significant loss of granule proteins into the airway lumen in response to antigen inhalation. Indeed, concomitant with ultrastructural changes, significant levels of noncell-associated EPO were detected in the BALF of allergic mice. Eosinophil recruitment into the lungs and activation correlate with the induction of AHR, supporting earlier studies with rats and primates that directly link eosinophils and their constituents with the induction of disease processes in the lung [2 , 34 ]. A recent report has also established a temporal association between the release and persistence of EPO in the airway lumen with the induction and resolution of AHR [17 ]. However, it should be noted that AHR is not attenuated in the allergic lung of mice deficient in MBP or EPO, indicating that eosinophils may also regulate airways reactivity independently of these proteins or that other pathways may operate in parallel with granule-induced processes in the allergic lung [33 , 35 ].
The release of granule proteins in BALF was not a result of sample processing or cellular degenerative processes, as the ultrastructural morphology of BALF eosinophils showed that degranulation occurred via an active process that was reproducibly confined to a distinct percentage of cells. Immunohistochemical studies have also described the deposition of extracellular MBP on the apical surface of epithelial cells during allergic inflammation of the lungs of mice exposed to filarial antigens, indicating that eosinophil degranulation occurs in vivo and not ex vivo in the lavage fluid [36 ]. Furthermore, recruitment of eosinophils to the airways of naive mice by the overexpression of IL-5 and eotaxin does not subsequently result in eosinophil degranulation in isolated BALF [24 ]. We also routinely isolate large numbers of eosinophils from body compartments (peritoneal cavity and blood of IL-5 Tg mice) and have not observed degranulation ex vivo [29 ].
Although eosinophils were activated in our model, similarities and differences were observed between mouse and human eosinophils participating in allergic disease. In our mouse model, as in clinical studies, free granule proteins were observed in BALF (which correlated with the recruitment of eosinophils to this compartment) [5 ], allergen challenge increased the activation level of circulating eosinophils [37 ], PMD was the primary mechanism for granule decomposition [9 , 11 ], and apoptotic eosinophils were rare [9 , 11 ]. By contrast to human studies [9 , 11 , 38 ], eosinophils resident in the tissues of the allergic lung of mice displayed only a low level of activation, cytolytic eosinophils were not observed, and lipid bodies were rare. These latter observations are in agreement with previous studies [21 , 22 ] and suggest that the mechanisms regulating degranulation of mouse and human eosinophils and/or the microenvironment within allergic tissues differ between these species.
Molecules that prime eosinophils can also induce eosinophil degranulation but to a considerably lower degree [39 , 40 ]. According to the criteria used in this study, many of the eosinophils in the blood and tissues of allergic mice are primed and become further activated to degranulate in the airway lumen. Indeed, eosinophils chronically stimulated with IL-5 in IL-5 Tg mice possessed higher basal levels of activation and attained greater levels of activation after immune stimulation in the airway lumen. The compartmentalization of degranulation primarily to the lumen of the lung may reflect distinct spatial and temporal aspects of acute allergic responses, which are in marked contrast to chronic inflammation observed in human asthma. However, the mechanism underpinning the degranulation of eosinophils in the airway lumen may be highly relevant to asthma and other respiratory disorders (e.g., respiratory syncytial virus-induced bronchiolitis), where this process is a feature of disease [4 , 5 , 41 , 42 ].
To characterize the similarities and differences in the regulatory mechanisms of cellular activation between mouse and human eosinophils, we investigated the effect of PMA on adhesion and degranulation processes. Our data indicate that several features of activation have been conserved between mouse and human eosinophils. First, mouse eosinophils adhered and degranulated in a time- and dose-dependent manner in response to PMA. Although the process of adhesion was slower with mouse eosinophils, like human cells, adhesion preceded degranulation. Thus, adhesion molecules may play a central role in the regulation of mouse eosinophil degranulation as is observed with human cells [43
, 44
]. It is interesting that human and mouse eosinophils have similar adhesion molecule profiles, and both species express lymphocyte function-associated antigen-1 (
Lß2), Mac-1 (
Mß2), and very-late antigen-4 (
4ß1). These molecules have been implicated in the regulation of various cellular functions including cellmatrix and cellcell contacts, migration, and activation. Second, in response to PMA, mouse eosinophils released MBP by PMD, a process widely observed in human eosinophil populations in vitro (in response to various stimuli including PMA) and in vivo [7
8
9
, 11
, 21
]. PMD provides a mechanism for the preferential release of granule proteins [45
]. In response to PMA, mouse eosinophils appear to release, preferentially, proteins associated with the core, notably MBP. After activation, the largest proportion of secondary granules progressed from type I to type II granules.
Although, the kinetics of adhesion and degranulation and the mechanism for the release of granule contents (PMD) were similar to human cells, higher concentrations of PMA were required to induce these responses [21 , 44 , 46 ]. The optimal concentration of PMA required to stimulate adhesion and degranulation would appear to be significantly different between species. Mouse eosinophil activation occurred at concentrations as low as 1 nM but only reached a maximum at 100 nM. In contrast, human eosinophils only require 1.5 nM of PMA to secrete maximal amounts of granule proteins [44 , 46 ]. The threshold of responsiveness highlights an important difference between human and mouse eosinophils.
It is interesting that eosinophils recruited to the lung by the local expression of exogenous IL-5 and eotaxin in naive mice could only induce AHR after the concomitant delivery of OVA to the airways, and this was coincident with the induction of MBP release from eosinophils (a marker of eosinophil activation). Antigens also activate eosinophils recovered from atopic patients [24 , 45 , 47 ]. In our mouse model, antigen is primarily localized to the airway lumen during the acute inflammatory reaction [25 ], and it is tempting to speculate that this phenomena results in eosinophil degranulation occurring primarily in this compartment. The dose and dispersal of antigen within the lung may be important determinants of where and if further eosinophil degranulation occurs in mouse models of allergic lung disease. Indeed, in studies where eosinophil degranulation was not observed in the allergic lung of mice, only modest levels of antigen were delivered to the lung, at markedly reduced exposure times in comparison with our investigation [21 , 22 ].
In summary, we have provided evidence for eosinophil degranulation in our model of allergic airways disease and linked eosinophil activation in the airway lumen to the induction of AHR. EPO was detected in the BALF of sensitized mice, and electron microscopy of BALF eosinophils depicted morphological changes characteristic of PMD. Although the eosinophils in the airway submucosa of sensitized mice were not releasing large amounts of granule contents, these cells were activated by comparison to control and bone marrow-derived eosinophils. However, eosinophils were highly activated in the airway lumen. Previous investigations have shown that deposition of eosinophil products in the airways lumen can affect lung function [2 ]. Our in vitro studies indicate that like human cells, mouse eosinophil activation follows adhesion, but these cells are more resistant to activation by PMA. The differences (degree and site of activation) observed in vivo may be, in part, a result of the higher thresholds for activation of mouse eosinophils, dispersal of antigen in the lung, and the acute nature of the disease process in mice compared with humans. Understanding the mechanisms regulating mouse eosinophil degranulation in the airway lumen may provide important insights into how this process occurs in human allergic respiratory disorders.
| ACKNOWLEDGEMENTS |
|---|
Received August 19, 2003; revised January 21, 2004; accepted January 22, 2004.
| REFERENCES |
|---|
|
|
|---|
and independent of eosinophil infiltration Am. J. Respir. Cell Mol. Biol. 16,325-334[Abstract]
4 Integrin-dependent eotaxin induction of bronchial hyperresponsiveness and eosinophil migration in interleukin-5 transgenic mice Am. J. Respir. Cell Mol. Biol. 20,992-1000
RII induce eosinophil degranulation J. Clin. Invest. 95,2813-2821This article has been cited by other articles:
![]() |
F.-X. Ble, C. Cannet, S. Zurbruegg, H. Karmouty-Quintana, R. Bergmann, N. Frossard, A. Trifilieff, and N. Beckmann Allergen-induced Lung Inflammation in Actively Sensitized Mice Assessed with MR Imaging Radiology, September 1, 2008; 248(3): 834 - 843. [Abstract] [Full Text] [PDF] |
||||
![]() |
K. Nakagome, M. Dohi, K. Okunishi, R. Tanaka, T. Kouro, M. R. Kano, K. Miyazono, J.-i. Miyazaki, K. Takatsu, and K. Yamamoto IL-5-Induced Hypereosinophilia Suppresses the Antigen-Induced Immune Response via a TGF-beta-Dependent Mechanism J. Immunol., July 1, 2007; 179(1): 284 - 294. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. I. Ochkur, E. A. Jacobsen, C. A. Protheroe, T. L. Biechele, R. S. Pero, M. P. McGarry, H. Wang, K. R. O'Neill, D. C. Colbert, T. V. Colby, et al. Coexpression of IL-5 and Eotaxin-2 in Mice Creates an Eosinophil-Dependent Model of Respiratory Inflammation with Characteristics of Severe Asthma J. Immunol., June 15, 2007; 178(12): 7879 - 7889. [Abstract] [Full Text] [PDF] |
||||
![]() |
D. J. Tumes, J. Cormie, M. G. Calvert, K. Stewart, C. Nassenstein, A. Braun, P. S. Foster, and L. A. Dent Strain-dependent resistance to allergen-induced lung pathophysiology in mice correlates with rate of apoptosis of lung-derived eosinophils J. Leukoc. Biol., June 1, 2007; 81(6): 1362 - 1373. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Lang, C. Murgia, M. Leong, L.-W. Tan, G. Perozzi, D. Knight, R. Ruffin, and P. Zalewski Anti-inflammatory effects of zinc and alterations in zinc transporter mRNA in mouse models of allergic inflammation Am J Physiol Lung Cell Mol Physiol, February 1, 2007; 292(2): L577 - L584. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. M. Swartz, K. D. Dyer, A. W. Cheever, T. Ramalingam, L. Pesnicak, J. B. Domachowske, J. J. Lee, N. A. Lee, P. S. Foster, T. A. Wynn, et al. Schistosoma mansoni infection in eosinophil lineage-ablated mice Blood, October 1, 2006; 108(7): 2420 - 2427. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. Homma, J. H. T. Bates, and C. G. Irvin Airway hyperresponsiveness induced by cationic proteins in vivo: site of action Am J Physiol Lung Cell Mol Physiol, September 1, 2005; 289(3): L413 - L418. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||