TLR-induced inflammation in cystic fibrosis and non-cystic fibrosis airway epithelial cells

Cystic fibrosis (CF) is a genetic disease characterized by severe neutrophil-dominated airway inflammation. An important cause of inflammation in CF is Pseudomonas aeruginosa infection. We have evaluated the importance of a number of P. aeruginosa components, namely lipopeptides, LPS, and unmethylated CpG DNA, as proinflammatory stimuli in CF by characterizing the expression and functional activity of their cognate receptors, TLR2/6 or TLR2/1, TLR4, and TLR9, respectively, in a human tracheal epithelial line, CFTE29o(-), which is homozygous for the DeltaF508 CF transmembrane conductance regulator mutation. We also characterized TLR expression and function in a non-CF airway epithelial cell line 16HBE14o(-). Using RT-PCR, we demonstrated TLR mRNA expression. TLR cell surface expression was assessed by fluorescence microscopy. Lipopeptides, LPS, and unmethylated CpG DNA induced IL-8 and IL-6 protein production in a time- and dose-dependent manner. The CF and non-CF cell lines were largely similar in their TLR expression and relative TLR responses. ICAM-1 expression was also up-regulated in CFTE29o(-) cells following stimulation with each agonist. CF bronchoalveolar lavage fluid, which contains LPS, bacterial DNA, and neutrophil elastase (a neutrophil-derived protease that can activate TLR4), up-regulated an NF-kappaB-linked reporter gene and increased IL-8 protein production in CFTE29o(-) cells. This effect was abrogated by expression of dominant-negative versions of MyD88 or Mal, key signal transducers for TLRs, thereby implicating them as potential anti-inflammatory agents for CF.     

Mucous solids and liquid secretion by airways: studies with normal pig, cystic fibrosis human, and non-cystic fibrosis human bronchi

To better understand how airways produce thick airway mucus, nonvolatile solids were measured in liquid secreted by bronchi from normal pig, cystic fibrosis (CF) human, and non-CF human lungs. Bronchi were exposed to various secretagogues and anion secretion inhibitors to induce a range of liquid volume secretion rates. In all three groups, the relationship of solids concentration (percent nonvolatile solids) to liquid volume secretion rate was curvilinear, with higher solids concentration associated with lower rates of liquid volume secretion. In contrast, the secretion rates of solids mass and water mass as functions of liquid volume secretion rates exhibited positive linear correlations. The y-intercepts of the solids mass-liquid volume secretion relationships for all three groups were positive, thus accounting for the higher solids concentrations in airway liquid at low rates of secretion. Predictive models derived from the solids mass and water mass linear equations fit the experimental percent solids data for the three groups. The ratio of solids mass secretion to liquid volume secretion was 5.2 and 2.4 times higher for CF bronchi than for pig and non-CF bronchi, respectively. These results indicate that normal pig, non-CF human, and CF human bronchi produce a high-percent-solids mucus (>8%) at low rates of liquid volume secretion (≤1.0 μl·cm(-2)·h(-1)). However, CF bronchi produce mucus with twice the percent solids (~8%) of pig or non-CF human bronchi at liquid volume secretion rates ≥4.0 μl·cm(-2)·h(-1).     

T cells of atopic asthmatics preferentially infiltrate into human bronchial xenografts in SCID mice

T cells play an important role in the pathogenesis of bronchial asthma. However, it is not completely known how circulating lymphocytes infiltrate into the airways of asthmatic patients. Because SCID mice are unable to reject xenogenic transplants, many xenotransplant models using various human tissues have been developed. Therefore, to examine the interaction between bronchi and T lymphocytes of asthma, it may be possible to use the human bronchial xenograft and PBMC xenograft in SCID mice. We transplanted human bronchi into the subcutaneum of SCID mice and i.p. injected PBMCs that were obtained from patients with atopic asthma, atopic dermatitis and rheumatoid arthritis, and normal subjects (asthmatic, dermatitis, rheumatic, and normal huPBMC-SCID mice). There was no difference in the percentage of CD3-, CD4-, CD8-, CD25-, CD45RO-, CD103-, and cutaneous lymphocyte Ag-positive cells in PBMCs among the patients with asthma, dermatitis, rheumatoid arthritis, and normal subjects, and CD3-positive cells in peripheral blood of asthmatic, dermatitis, rheumatic, and normal huPBMC-SCID mice. The number of CD3-, CD4-, and CD8-positive cells in the xenografts of asthmatic huPBMC-SCID mice was higher than those of dermatitis, rheumatic, and normal huPBMC-SCID mice. IL-4 mRNA and IL-5 mRNA were significantly higher in the xenografts of asthmatic huPBMC-SCID mice than those in the xenografts of normal huPBMC-SCID mice, but there were no significant differences in the expressions of IL-2 mRNA or IFN-gamma mRNA between them. These findings suggest that T cells, especially Th2-type T cells, of asthmatics preferentially infiltrate into the human bronchi.     

Accessible xenografts of human synovium in the subcutaneous tissues of the ears of SCID mice

This work was undertaken to examine whether human synovium could be engrafted into subcutaneous pouches in the ears of severe combined immunodeficient (SCID) mice. Synovium was transplanted into surgically constructed ear pouches. The grafts were examined by histological and immunohistochemical methods after varying periods after engraftment, or after percutaneous injection of TNF-alpha. Normal, osteo-arthritic and rheumatoid synovium was engrafted successfully in subcutaneous ear pouches. The general morphology and cellular compositions of xenografts were retained including human endothelial cells. In rheumatoid xenografts, macrophages, fibroblasts and lymphocytes persisted for at least 4 weeks. Vascular expression of intercellular adhesion molecule-1 (ICAM-1) was maintained but expression of vascular adhesion molecule-1 (VCAM-1), E-selectin and MHC class II diminished with time. Percutaneous injection of TNF-alpha induced up-regulation of VCAM-1. Human synovium can be engrafted into subcutaneous ear pouches in SCID mice. The xenografts are accessible and respond to injection of a pro-inflammatory cytokine.     

Recruitment of mononuclear leucocytes to osteoarthritic human synovial xenografts in the ears of SCID mice

A system has been established to assess the recruitment of 99mTc-hexamethylpropylene amine oxamine (99mTc-HMPAO)-labelled PBMC and [125I]iododeoxyuridine-labelled Con A stimulated lymphoblasts to allogeneic human synovial xenografts in the ears of SCID mice. Successful engraftment of osteoarthritic synovium was achieved in approximately 90% of cases and a connection between the human microvasculature of the xenograft and the circulation of the mouse was shown. Cells were delivered to the xenograft by a system of regional vascular perfusion, thus avoiding the major murine vascular beds. The accumulation of 99mTc-HMPAO-labelled PBMC in mouse ears was monitored in real time. Direct injection of xenograft-bearing ears with recombinant human TNF-alpha, 7 h prior to perfusion, increased the accumulation of both PBMC and lymphoblasts in cytokine-injected ears compared to contralateral control-injected ears. Autoradiography revealed the presence of [125I]iododeoxyuridine-labelled lymphoblasts associated with human microvasculature within the xenograft. However, the increased accumulation of lymphoblasts in cytokine-injected ears occurred in the tissues surrounding the xenograft, where lymphoblasts were associated more often with murine than human vessels. Although the system described offers advantages over similar models, the propensity for mouse endothelium to interact with human leucocytes is likely to be a generic disadvantage for models of human leucocyte recruitment to xenografts in immunodeficient mice.     

CTLA-4 blockade augments human T lymphocyte-mediated suppression of lung tumor xenografts in SCID mice

Previous studies by others using transplantable murine tumor models have demonstrated that the administration of antibodies that block CTLA-4 interaction with B7 can provoke the elimination of established tumors, and that the tumor suppression is mediated by T-cells and/or cells expressing NK1.1. Studies from our lab have established in a human/severe combined immunodeficient (SCID) mouse chimeric model that autologous peripheral blood leukocytes (PBL) can suppress the growth of tumor xenografts in a PBL dose-dependent fashion, and that this suppression is dependent upon the patients T and NK cells. Using this human/mouse chimeric model, we sought to determine whether an antibody blockade of CTLA-4 would enhance the anti-tumor response of a patients PBL. It was first important to determine whether the tumor suppression observed in the SCID model was dependent upon CD28/B7 co-stimulation. Blockade of B7 with a human CTLA-4-Ig fusion protein completely abrogated the lymphocyte-mediated tumor suppression, confirming in this model that tumor suppression is dependent upon a CD28/B7 co-stimulation. Using two different CTLA-4 specific monoclonal antibodies, we observed that CTLA-4 blockade significantly enhanced the human lymphocyte-mediated tumor suppression in mice co-engrafted with PBL and tumor cells. This enhancement was observed in both an allogeneic setting (in which the PBL were allogeneic with respect to the tumor) and an autologous setting (in which the PBL and tumor were from the same patient). These results sustain the notion that human anti-tumor immune response can be augmented (in vivo) by blocking the interaction between CTLA-4 and B7.     

Proteomic analysis of nasal cells from cystic fibrosis patients and non-cystic fibrosis control individuals: search for novel biomarkers of cystic fibrosis lung disease

Potential biological markers for cystic fibrosis (CF) lung disease were identified by comparative proteomics profiling of nasal cells from deletion of phenylalanine residue 508 (F508del)-homozygous CF patients and non-CF controls. From the non-CF 2-DE gels, 65 spots were identified by MS, and a reference 2-DE map was thus established. The majority of those correspond to ubiquitously expressed proteins. Consistent with the epithelial origin of this tissue, some of the identified proteins are epithelial markers (e.g. cytokeratins, palate lung and nasal epithelium clone protein (PLUNC), and squamous cell carcinoma antigen 1). Comparison of this protein profile with the one similarly obtained for CF nasal cells revealed a set of differentially expressed proteins. These included proteins related to chronic inflammation and some others involved in oxidative stress injury. Alterations were also observed in the levels of cytoskeleton proteins, being probably implicated with cytoskeleton organization changes described to occur in CF-airways. Lower levels were found for some mitochondrial proteins suggesting an altered mitochondrial metabolism in CF. Differential expression was also found for two more enzymes that have not been previously associated to CF. Further studies will clarify the involvement of such proteins in CF pathophysiology and whether they are targets for CF therapy.     

Differentiated and functional human airway epithelium regeneration in tracheal xenografts

To investigate the regeneration process of a well-differentiated and functional human airway epithelium, we adapted an in vivo xenograft model in which adult human nasal epithelial cells adhere and progressively repopulate denuded rat tracheae grafted in nude mice. The proliferating activity, the degree of differentiation, and the barrier integrity of the repopulated epithelium were studied during the regeneration process at optical and ultrastructural levels with immunocytochemistry and a permeability tracer. Three days after implantation in nude mice, tracheal xenografts were partially repopulated with a flattened nonciliated and poorly differentiated leaky epithelium. By the end of the first week after the graft, cell proliferation produced on the entire surface of the rat trachea an epithelium that was stratified into multiple layers and tightly sealed. During successive weeks, cell proliferation dramatically decreased. Moreover, the epithelium became progressively columnar, secretory, ciliated, and transiently leaky. At 4-5 wk, a fully differentiated pseudostratified functional epithelial barrier impermeable to a low-molecular-weight tracer was reconstituted. The regeneration of a well-differentiated and functional human airway epithelium in rat tracheae grafted in nude mice includes several steps that mimic the regeneration dynamics of airway epithelium after injury.     

Bioelectric properties and ion transport of airways excised from adult and fetal sheep

Segments of fetal and maternal trachea, maternal bronchi from near-term sheep, and trachea and bronchi from nonpregnant adult sheep were excised and mounted as sheets in Ussing chambers. The conductance (G) for each group of tissues was similar (approximately 4 mS/cm-2); the short circuit current (Isc) ranged from 45-90 microA/cm-2. Under short-circuit or open-circuit conditions trachea and bronchi from pregnant and nonpregnant adult animals absorbed Na+, whereas fetal trachea secreted Cl-. Short-circuited maternal bronchi secreted K+, whereas maternal and fetal trachea did not. Isoproterenol induced an increase in Isc, G, and Cl- secretion of fetal trachea. Maternal trachea and bronchi were not affected. Amiloride reduced Na+ absorption and Isc of maternal trachea and bronchi, but had little effect on fetal trachea. The permeability of fetal trachea to 14C-mannitol was 17 X 10(-7) cm/s and was not affected by isoproterenol. The permeation of dextran (10 K) and horseradish peroxidase across fetal trachea and of all three probes across maternal airways did not reach steady state, but the relative rates were compatible with an equivalent pore radius greater than 4 nm. We conclude that ion transport in fetal large airways contributes to the Cl- and liquid secretion by the entire fetal pulmonary epithelium, whereas resting ion transport of large airways from adult sheep, like that of mature airways of many species, is dominated by Na+ absorption. All of these airway epithelia are characterized by large paracellular aqueous paths.     

Bioelectric properties and ion transport across excised canine fetal and neonatal airways

Knowledge of liquid secretion by fetal lung stems from studies of sheep. We extended these studies to dogs and examined the persistence of the fetal pattern of airway epithelial permeability and ion transport in the neonatal animal. Plasma and lung liquid from fetal dogs were analyzed for Na+, K+, Cl-, and HCO3-. Only the Cl- concentration of fetal lung liquid (129 meq/l) was significantly different from that of fetal plasma (111 meq/l). Segments of trachea from fetal and neonatal (less than 1, 7-10, and 21-46 days after birth) dogs were excised and mounted in flux chambers. The transepithelial potential difference (PD) of all tissues was oriented lumen negative (9.8-14.8 mV). Under short-circuit conditions, unidirectional Na+ flows were symmetrical. Cl- was secreted, and the secretion was equivalent to short-circuit current (Isc). Cl- secretion persisted under open-circuit conditions. Lobar bronchi from 21- to 46-day neonates absorbed Na+ (1.9 mueq.cm-2.h-1), but unidirectional flows of Cl- were symmetrical. Amiloride (10(-4) M) reduced Isc of neonatal bronchi by 47% but did not affect fetal bronchi. Isoproterenol increased Isc of both fetal (33%) and neonatal (40%) bronchi. These responses suggest that fetal bronchi do not absorb Na+ but can be stimulated to secrete Cl-. We conclude that Cl- secretion by epithelium of large airways may contribute to fetal lung liquid production, but it is unlikely that the tracheal epithelium is involved in fluid absorption at birth. Whereas fetal bronchi appear to secrete Cl-, neonatal bronchi absorb Na+.(ABSTRACT TRUNCATED AT 250 WORDS)     

Antioxidant properties of cystic fibrosis sputum

Oxidative stress is a likely contributor to the pathogenesis of cystic fibrosis (CF) lung disease. However, hydrogen peroxide (H(2)O(2)), a physiological oxidant, is not elevated in CF exhalates. H(2)O(2) may be neutralized by antioxidants in CF airway secretions. The H(2)O(2)-detoxifying capacity of CF airway secretions, obtained via sputum induction, was studied in an in vitro H(2)O(2) cytotoxicity model. 16HBE14o- cells were exposed to H(2)O(2) in culture medium containing either 0 or 10% fetal bovine serum (FBS) or 10% CF sputum supernatant (extracted without use of dithiothreitol). The efficiency of H(2)O(2) neutralization was estimated by measuring intracellular oxidant levels (dihydrorhodamine 123) after 2 h and cell viability (propidium iodide) after 24 h of H(2)O(2) exposure. Furthermore, the presence of reduced thiols (DTNB assay) and reduced glutathione (recycling assay) in CF sputum samples was evaluated. CF sputum extracts completely prevented intracellular oxidant accumulation seen in cells incubated with H(2)O(2) in both control media (i.e., 0 or 10% FBS). Furthermore, CF sputum abolished cell death in 16HBE14o- cells exposed to up to 1 mM H(2)O(2). In contrast, there was 100% cytotoxicity in cells exposed to 600 microM H(2)O(2) in both control media. The H(2)O(2)-detoxifying potential of CF sputum was sustained after catalase and heme peroxidases were inactivated by sodium azide, which does not affect glutathione peroxidase. In addition, reduced protein thiols were found in abundance in CF sputum. In conclusion, CF sputum is capable to neutralize H(2)O(2) and abundant reduced thiols and/or glutathione peroxidase are fully sufficient to detoxify H(2)O(2).     

Surface-lining layer of airways in cystic fibrosis mice

Lung disease is the major cause of death in individuals suffering from cystic fibrosis (CF), with abnormal lung-lining fluids occurring as early as early infancy. However, the precise etiology of CF lung disease is still poorly understood. We investigated the structural components of the airway surface-lining layer in targeted Cftrtm1HGU/Cftrtm1HGU mutant mice and non-CF controls. Five lungs per animal group were fixed by intravascular triple perfusion. The ultrastructure of the surface-lining layer of large and small intrapulmonary conducting airways was systematically investigated according to a standard protocol in transmission and scanning electron micrographs. In both animal groups, the surface-lining layer consisted of an aqueous phase and an osmiophilic film of variable thickness at the air-fluid interface. The aqueous phase usually did extend <1 microm beyond the uppermost tips of the epithelial cells in both animal groups. The aqueous phase of the small airways was slightly more electron dense in Cftrtm1HGU/Cftrtm1HGU than in non-CF mice. Neither the ultrastructure of the surfactant film at the air-fluid interface nor the forms assumed by the osmiophilic structures associated with surfactant turnover in the aqueous layer differed significantly in Cftrtm1HGU/Cftrtm1HGU and non-CF mice. Hence, there were no signs of any ultrastructural abnormalities in the surface-lining layer of young adult Cftrtm1HGU/Cftrtm1HGU mice before infection with CF-related pathogens.     

Airway surface fluid volume and Cl content in cystic fibrosis and normal bronchial xenografts

We describe theuse of an in vivo human bronchial xenograft model of cystic fibrosis(CF) and non-CF airways to investigate pathophysiological alterationsin airway surface fluid (ASF) volume (V_s) and Cl content.V_s was calculated based on thedilution of an impermeable marker,[³H]inulin, duringharvesting of ASF from xenografts with an isosmotic Cl-free solution.These calculations demonstrated thatV_s in CF xenographs (28 ± 3.0 µl/cm²;n = 17) was significantly less thanthat of non-CF xenografts (35 ± 2.4 µl/cm²;n = 30). The Cl concentration of ASF([Cl]_s) wasdetermined using a solid-state AgCl electrode and adjusted for dilutionduring harvesting using the impermeable[³H]inulin marker.Cumulative results demonstrate small but significant elevations(P < 0.045) in[Cl]_s in CF (125 ± 4 mM; n = 27) compared with non-CF(114 ± 4 mM; n = 48) xenografts.To investigate potential mechanisms by which CF airways may facilitatea higher level of fluid absorption yet retain slightly elevated levelsof Cl, we sought to evaluate the capacity of CF and non-CF airways toabsorb both ²²Na and³⁶Cl. Two consistent findings wereevident from these studies. First, in both CF and non-CF xenografts,²²Na and³⁶Cl were always absorbed in anequal molar ratio. Second, CF xenografts hyperabsorbed (~1.5-foldhigher) both ²²Na and³⁶Cl compared with non-CFxenografts. These results substantiate previously documented findingsof elevated Na absorption in CF airways and also suggest that theslightly elevated[Cl]_s found in thisstudy of CF xenograft epithelia does not occur through a mechanism ofdecreased apical permeability to Cl.     

Bioelectric properties and ion flow across excised human bronchi

Bioelectric properties and ion transport of excised human segmental/subsegmental bronchi were measured in specimens from 40 patients. Transepithelial electric potential difference (PD), short-circuit current (Isc), and conductance (G), averaged 5.8 mV (lumen negative), 51 microA X cm-2, and 9 mS X cm-2, respectively. Na+ was absorbed from lumen to interstitium under open- and short-circuit conditions. Cl- flows were symmetrical under short-circuit conditions. Isc was abolished by 10(-4) M ouabain. Amiloride inhibited Isc (the concentration necessary to achieve 50% of the maximal effect = 7 X 10(-7) M) and abolished net Na+ transport. PD and Isc were not reduced to zero by amiloride because a net Cl- secretion was induced that reflected a reduction in Cl- flow in the absorptive direction (Jm----sCl-). Acetylcholine (10(-4) M) induced an electrically silent, matched flow of Na+ (1.7 mueq X cm-1 X h-1) and Cl- (1.9 mueq X cm-12 X h-1) toward the lumen. This response was blocked by atropine. Phenylephrine (10(-5) M) did not affect bioelectric properties or unidirectional ion flows, whereas isoproterenol (10(-5) M) induced a small increase in Isc (10%) without changing net ion flows significantly. We conclude that 1) Na+ absorption is the major active ion transport across excised human bronchi, 2) Na+ absorption is both amiloride and ouabain sensitive, 3) Cl- secretion can be induced by inhibition of the entry of luminal Na+ into the epithelia, and 4) cholinergic more than adrenergic agents modulate basal ion flow, probably by affecting gland output.     

Secretory activity of hamster tracheal explants and isolated tracheal epithelial cells and the effects of cystic fibrosis serum

Serum from cystic fibrosis patients has been shown by scanning electron microscopy to cause release of large quantities of mucus from the cultured tracheal rings of 3–4-month-old male Golden Syrian hamsters. In order to study this phenomenon on single cells, an epithelial (HTE) cell culture has been established from the hamster tracheal rings using the cell rescue method of Goldman and Baseman (1980a, In Vitro, 16:313). The cells were demonstrated to be epithelial by histochemical staining and immunofluorescent detection of laminin. Proteins secreted by HTE cells were partially characterized and shown to consist, at least in part, of acidic glycoproteins. The proteins were precipitated by addition of buffered alcian blue (AB) to the cell-free medium under conditions in which all of a polyanionic protein [³H]-labeled mucin, was precipitated without carrier. [¹⁴C] galactosamine-labeled AB precipitate was β-eliminated and, after neutralization and centrifugation, the material in the supernatant was sized by chromatography on a calibrated Bio-Gel P₂ column. The label eluted with a molecular weight close to a disaccharide. HTE cells pulse--labeled for 1.0 hr with [³H] leucine or [¹⁴C] galactosamine secreted increasing amounts of labeled glycoprotein during the chase. Sodium dodecylsulfate-polyacrylamide gel electrophoresis and fluorography of labeled AB precipitates revealed three major bands, two with molecular weights greater than 100 kd. Secretion was stimulated by retinoate (50% increase), but not by retinol. Exposure of HTE cells to whole sera from cystic fibrosis patients resulted in heightened secretion rates as compared to results obtained with normal sera. Heterozygote sera produced secretion rates intermediate between the two extremes.