|
|
||||||||
1Department of Pediatrics, University of Virginia Health System, Charlottesville, Virginia; and 2Department of Molecular Biology and Biochemistry, Universidad de Málaga, Málaga, Spain
Submitted 22 September 2005 ; accepted in final form 16 November 2005
| ABSTRACT |
|---|
|
|
|---|
and tumor necrosis factor-
, but not direct rhinovirus infection, decreased pH, an effect partly associated with decreased ammonium concentrations. This effect was 1) prevented by nitric oxide synthase inhibition; 2) independent of cyclic GMP; 3) associated with an increase in endogenous airway epithelial cell S-nitrosothiol concentration; 4) mimicked by the exogenous S-nitrosothiol, S-nitroso-N-acetyl cysteine; and 5) independent of glutaminase expression and activity. We then confirmed that decreased epithelial pH inhibits human rhinovirus replication in airway epithelial cells. These data suggest that a nitric oxide synthase-dependent host response to viral infection mediated by S-nitrosothiols, rather than direct infection itself, plays a role in decreased airway surface pH during human rhinovirus infection. This host immune response may serve to protect the lower airways from direct infection in the normal host. In patients with asthma, however, this fall in pH could be associated with the increased mucus production, augmented inflammatory cell degranulation, bronchoconstriction, and cough characteristic of an asthma exacerbation. asthma; S-nitrosothiol; airway epithelium; glutaminase
During an experimental HRV upper respiratory infection, HRV mRNA can be detected in the intrathoracic airways (16), but there are also modifications of lower airway reactivity and of bronchial inflammation that may be independent of direct lower airway infection (12, 13, 19). For example, cytokines produced by T helper (Th) 1 lymphocytes in response to viral infection can affect the lower airways (24), and it has been proposed that upper airway inflammation can also signal lower airway responses through neuronal signaling pathways (26). Thus it is possible that HRV infection could decrease airway pH, and so contribute to an acute asthma exacerbation, through a direct infection of bronchial epithelial cells and/or through the indirect signaling effects of inflammatory and neuronal mediators.
One common mechanism by which both inflammation and neuronal signaling could affect the airways following upper airway HRV infection is through upregulation or activation of nitric oxide synthases (NOS) (44). NOS activation forms both NO and S-nitrosothiols (SNOs) (17, 21, 48). S-nitrosothiol bioactivities can be mediated by homolytic cleavage to form NO, with downstream activation of guanylate cyclase to form cyclic guanosine monophosphate (cGMP). On the other hand, SNO effects can also be mediated by transnitrosation reactions involving S-nitrosylation of specific target cysteine residues (30, 41, 44). In eukaryotic organisms, NH3 is synthesized by glutaminase, S-nitrosoglutathione reductase (GSNOR) and other enzymes expressed by airway epithelial cells (24, 25, 39) that may have a role in maintaining pH homeostasis in the airways (26). Because glutaminase has an active site cysteine (5), and GSNOR is involved in airway nitrogen oxide metabolism (39), one mechanism by which NOS activation could lower pH could be through impaired NH3 production.
Here, we show that 1) HRV infection does not directly inhibit airway epithelial cell pH buffering in vitro; 2) interferon (IFN)-
and tumor necrosis factor (TNF)-
inhibit airway epithelial cell pH buffering; 3) the pH effect of IFN-
and TNF-
is mimicked by that of S-nitroso-N-acetyl cysteine (SNOAC), is inhibited by the NOS inhibitor nitro-L-arginine methyl ester (L-NAME), and is cGMP independent; 4) IFN-
and TNF-
exposure results in SNO production in airway epithelial cells; 5) glutaminase activity in the airway epithelium is primarily characterization of the renal isoform (5, 6); 6) however, the effects of IFN-
, TNF-
, and SNOs are independent of glutaminase; and 7) decreasing airway epithelial pH inhibits HRV replication. These data suggest that airway host defense mechanisms that increase NOS activity could both retard HRV replication and exacerbate asthma, in part, through an SNO-associated decrease in airway pH.
| METHODS |
|---|
|
|
|---|
In vitro viral infection. Both BEAS-2B and A549 cells express ICAM-1, to which HRV binds to cause infection (37, 45). Cells were infected with HRV-16 as previously described (28) using a multiplicity of infection (MOI) of 2 for the BEAS-2B cells and MOI of 5 for the A549 cells. An equal volume of medium alone was added to control wells. After 1-h incubation at 34°C in a CO2-free incubator, the media were replaced with HCO3 buffer-depleted media as described (28). Samples were collected at specified time points after the infection. HRV infection was confirmed and quantitated in MRC-5 cells by quantitative viral culture as previously described (28).
Biochemistry and pharmacology.
Selected wells were treated with TNF-
and IFN-
(10 ng/ml each) with the S-nitrosylating agent (SNOAC) (1, 10, or 100 µM) and/or with L-NAME (1 mM) in HCO3-depleted media. During the experiments, the wells were incubated at 37°C in a CO2-free incubator. Ammonium concentration was determined spectrophotometrically (Shimadzu, Kyoto, Japan) by reaction with
-ketoglutarate, catalyzed by glutamate dehydrogenase, in the presence of NADPH (Diagnostic Chemicals, Charlottetown, Prince Edward Island, Canada) (34). pH was measured with an Orion model 98-02 microelectrode attached to a model 410A pH meter (Thermo Orion, Waltham, MA). Protein concentrations were determined by the method of Lowry.
Airway lining fluid levels of H+ and other electrolytes in vivo are difficult to 1) recapitulate in vitro and 2) measure without altering the result (7, 24, 25). This issue is particularly challenging with regard to H+ because of the abundant buffering capacity of cells and cell culture medium. To measure changes in airway epithelial buffering, we have chosen to measure the change in pH of a fixed volume of buffer-free medium on the surface of cells; this system is, by necessity, somewhat arbitrary but has previously been shown to reflect relative changes in airway epithelial buffering by NH3 (24).
To determine which glutaminase isoform was active in airway epithelial cells, BEAS-2B cells were trypsinized and eluted with PBS plus fetal calf serum (5%) for assay as previously described (38). Cells were pelleted and washed twice with TES buffer (25 mM Tris·HCl, 0.2 mM EDTA, 0.33 M sucrose, pH 8). Cells were lysed with two freeze-thaw cycles, and samples of 25 µl were added to 35 µl of 1) a mixture of 99 mM potassium phosphate, 171 mM glutamine, 171 mM glutamate; 2) a mixture of 99 mM potassium phosphate, 171 mM glutamine; 3) deionized water; or 4) in these conditions in the presence of 10 µM SNOAC. The samples were incubated for either 30 or 60 min at 37°C. The reaction was terminated by adding 10 µl of 10% trichloracetic acid, kept on ice for 15 min, and then centrifuged. Ammonium concentration in the supernatant was determined as described above.
Immunoblotting. Immunoblotting was performed as described previously (30). In brief, whole cell extracts were prepared in 1% Nonidet P-40 (Sigma) lysis buffer containing 50 mM Tris·HCl buffer at pH 8.0, 1% Nonidet P-40, 150 mM NaCl, 2 µM leupeptin, 1 µM aprotinin, 1 µM pepstin, 1 µM phenylmethylsulfonyl fluoride (PMSF), and 2 µM Na2VO4 (Boehringer, Mannheim, Indianapolis, IN and Roche Diagnostics, Mannheim, Germany). Insoluble material was recovered and sheared through a 25-gauge needle. Protein was quantified by the BCA protein assay kit. Protein (100 µg), separated by 10% SDS/polyacrylamide gel in 1x electrode buffer, was transferred on to pure nitrocellulose membrane (Bio-Rad Laboratories, Hercules, CA) in Tobin Transfer buffer. Blots were blocked in Tris-buffered saline/Tween 20 containing 5% (wt/vol) nonfat dried milk and probed overnight (4°C) with a 1:200 dilution of rabbit polyclonal anti-human glutaminase antibody (6), followed by horseradish peroxidase-conjugated goat anti-rabbit antibody (1:3,000; Pierce). Glutaminase was visualized by enhanced chemiluminescence (ECL; Amersham Biosciences, Piscataway, NJ) using Hyperfilm (Amersham Bioscience). Membranes were stripped and reprobed with antibody to actin (1:100; Sigma Immuno Chemicals, St. Louis, MO). The gel films were scanned on a Molecular Dynamics Densitometer using ImageQuant.
Cellular S-nitrosothiol assay.
We collected cells treated with TNF-
and IFN-
and/or L-NAME by scraping in 1 ml of protease inhibitor buffer [2 µM leupeptin, 1 µM aprotinin, 1 µM pepstatin, 1 µM PMSF, and 2 µM Na2VO4 plus 1 mM EDTA and N-ethyl maleinide in PBS, pH 7.4]. The cells underwent centrifugation at 13,000 rpm for 10 min at 4°C. The cell pellet was frozen (80°C) and then was resuspended immediately before assay in 50 µl of protease inhibitor buffer plus 1% Triton X-100 and assayed for SNO content. The SNO assay involved reductive chemiluminescence in 100 µM cysteine and saturated CuCl as previously described (10).
Statistical analysis. Data are expressed as means ± SE, if normally distributed, and as median and interquartile range if not. Differences between two groups of cells were evaluated by the t-test for independent samples if normally distributed or with Mann-Whitney test if not. Differences among treatment groups were evaluated by ANOVA or ANOVA on ranks with appropriate pairwise comparisons (Holm-Sidak or Dunn's test). Differences were considered significant at P values < 0.05.
| RESULTS |
|---|
|
|
|---|
Th1 cytokine-stimulated decrease in airway epithelial pH is NOS dependent.
BEAS-2B cells in modified LHC9 medium at initial pH of 7 were treated with IFN-
and TNF-
(10 ng/ml each). Consistent with our previous report (24), cells (five wells each) treated with the cytokines had, relative to control cells, a greater fall in pH (0.30 ± 0.006 in treated vs. 0.28 ± 0.006 in control, P = 0.008 at 48 h) and a lower medium [NH4+] (627 ± 18 in treated vs. 707 ± 19 in control, P = 0.015). In previous studies, we have also observed induction of NOS2 in BEAS-2B cells by IFN-
and TNF-
, particularly in early cell passages. We hypothesized that IFN-
and TNF-
might affect the pH response through an inducible NOS (iNOS)-dependent mechanism. Indeed, in freshly purchased cells, there was a more robust decrease in pH with IFN-
and TNF-
, and this effect was inhibited by the NOS inhibitor L-NAME (Fig. 1, P < 0.05). Of note, the effect of L-NAME alone on pH (0.09 ± 0.11 at 5 h) did not differ from control. Moreover, the effect was not mimicked by a cell-permeable cGMP analog, 8-bromoguanosine 3',5'-cyclic monophosphate (8-BrcGMP), that is known to mimic guanylate cyclase activation in airway epithelial cells in vitro (47) (Fig. 1, P = NS). IFN-
and TNF-
treatment also resulted in an increased epithelial cell concentration of endogenous SNO (Fig. 2A, P > 0.05), and the cell-permeable SNO, SNOAC, caused a decrease in pH (Fig. 2B). There was a similar effect on [NH4+], but, importantly, it was not always in parallel with the effect on pH (Fig. 2C), suggesting that SNOAC and Th1 cytokines may affect pH both by NH3-dependent and NH3-independent mechanisms. Indeed, multiple buffering modalities (26) may explain the relative heterogeneity of results within and between cell types, a heterogeneity that could have introduced
-error with regard to identifying differences in pH and NH3. Cells were >90% viable by Trypan blue after 5 h under all conditions.
|
|
and TNF-
(10 ng/ml each) (Fig. 3B) substantially affected expression of glutaminase (n = 3 each, P = NS by densitometry), suggesting that neither the effect of SNOAC nor the Th1 cytokine response is mediated through an effect on glutaminase expression.
|
5 before they enter the alveolar space (46). We have proposed that this low pH is buffered in the more proximal airway by the activity of glutaminase and other enzymes (26). Therefore, we measured whether attenuated buffering on the airway epithelial surface could affect HRV replication in airway epithelial cells. One-hour exposure of BEAS-2Bs to HRV16 at pH 5 attenuated viral replication as assayed at 48 and 72 h (P < 0.05 and P < 0.01, respectively; Fig. 4). Cells were >90% viable at all time points.
|
| DISCUSSION |
|---|
|
|
|---|
During experimental nasal HRV infection, HRV mRNA is detectable in the lower airways (16). However, the role of direct HRV infection in causing asthma exacerbations is debated. With regard to the potential role of pH in HRV-associated asthma exacerbations, we did not find a direct effect of the virus on airway epithelial cell medium pH or on NH4+ levels, despite a significant infection in vitro. Thus HRV infection of bronchial epithelial cells does not appear directly to affect epithelial pH regulatory activity in these cell types.
On the other hand, consistent with our previous results (24), we found that IFN-
and TNF-
reduced airway epithelial cell culture medium pH and [NH4+]. These cytokines can be produced following Th1 lymphocyte activation in response to HRV and other infections (13, 19). We and others have previously shown that IFN-
and TNF-
increase iNOS expression and activity in BEAS-2B cells, A549 cells, and other lung epithelial cell lines (3, 29). Indeed, this NOS induction appears to have important immune effector functions (33). Here, we have shown for the first time that 1) NOS inhibition prevents the IFN-
and TNF-
-induced fall in airway epithelial pH, suggesting that NOS activity can mediate, in part, the fall in epithelial pH and 2) the fall in epithelial pH caused by Th1-stimulated airway epithelial NOS activation is cGMP independent, is associated with increased cellular SNO production, and is mimicked by the exogenous SNO, SNOAC. NOS-derived SNOs could inhibit pH regulatory enzymes through a mechanism involving cysteine S-nitrosylation (14, 17, 30, 44).
Though upregulation of iNOS is considered characteristic both of uncontrolled asthma (9) and of response to viral infection, it is also possible in vivo that there is a neural reflex response to upper airway HRV infection leading to activation of neuronal NOS (nNOS) in nonadrenergic, noncholinergic neurons and that nNOS signals glutaminase inhibition through SNO production. Indeed, SNO production associated with both nNOS and iNOS activity is localized (30, 32).
We hypothesized that one pH-regulatory target of SNOs might be glutaminase (5). Airway epithelial cells express glutaminase (24), and our current data are the first to suggest that this enzyme activity is mainly that of the kidney (KGA) isoform (Fig. 3). In the kidney, this enzyme catabolizes glutamine, producing glutamate and NH4+. This activity contributes to the pH buffering in the kidney by two mechanisms. One involves dissociation of NH4+ into NH3 and H+; NH3 diffuses into the lumen and binds protons, buffering the pH, whereas H+ is eventually eliminated through the basolateral Na+/H+ exchange (39). The second involves downstream breakdown of product glutamate to form HCO3 (and other buffers) (18). However, neither the expression nor the activity of glutaminase was inhibited by SNOAC, suggesting that the NH3 production in the intact cells that was inhibited by SNOAC was a different enzyme, perhaps GSNOR (39). Work is ongoing to characterize this biochemistry. Importantly, our data also suggest that there are mechanisms, independent of NH3, by which Th1 cytokines can regulate airway pH.
The contents of alveolar type II cell lamellar bodies are secreted into the airways at low pH (46). Furthermore, the distal airways and gas exchange units are heavily populated with acidifying macrophages (20), and the alveolar PCO2 is
40 mmH2O. The airway lining fluid in distal human airways may therefore be acidic; indeed, this acidity may have an antimicrobial effect in the distal airway as it does in the stomach. During mucociliary transport of airway lining fluid to the proximal airway, airway pH may then be buffered by glutaminase, in addition to other mechanisms such as HCO3 transport, H+/Na+ exchange, and epithelial carbonic anhydrase activity (4, 7, 8). Thus, since both asthma attacks and HRV infections are associated with a reduction in EBC pH (to
5), inhibition of pH buffering could be both an antimicrobial defense in the more proximal airway and a mechanism by which HRV infections trigger acute asthma exacerbations. Indeed, nebulization of mild base has been reported to improve lung function during asthma exacerbations (1, 2).
Consistent with previous results (23), we show here that decreased pH inhibits HRV infection in airway epithelial cells in vitro. This suggests the possibility that cytokine/NOS-associated airway epithelial pH buffering could help prevent spread of HRV in the lower airway; HRV rarely causes true pneumonia in immunocompetent hosts. Of note, NOS activation itself, presumably independently of pH, can inhibit viral replication, both through S-nitrosylation-based inhibition of viral cysteine proteases (31, 43) and through impaired host-defense gene regulation (49). Furthermore, a fall in pH results in 1) augmented protein S-nitrosylation (43) and 2) conversion of nitrite to NO (25), both of which could augment the antiviral effect of NOS products.
In conclusion, we have found that airway epithelial cell pH can be affected, in part, by Th1 cytokines; downstream, this effect appears to be a cGMP-independent NOS effect that may involve SNO signaling. Therefore, we suggest that the HRV infection of the upper airways may affect the activity of pH-regulatory mechanisms indirectly through the immune response, causing a decrease in airway epithelial pH. In nonasthmatic subjects, this decrease in pH can provide an innate host defense, inhibiting viral replication in the lower airways. In asthmatic patients, however, this mechanism may also have a role in triggering asthma symptoms and could be involved the pathophysiology of HRV infection-induced asthma exacerbations.
| GRANTS |
|---|
|
|
|---|
| FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
| REFERENCES |
|---|
|
|
|---|
z de Castro, and Marquez J. Expression of recombinant human L-glutaminase in Escherichia coli: polyclonal antibodies production and immunological analysis of mouse tissues. Biochim Biophys Acta 1648: 1723, 2003.[Medline]
-glutamyl transpeptidase. Biochem J 323: 477481, 1997.This article has been cited by other articles:
![]() |
K. W. Raines, M. G. Bonini, and S. L. Campbell Nitric oxide cell signaling: S-nitrosation of Ras superfamily GTPases Cardiovasc Res, July 15, 2007; 75(2): 229 - 239. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Carraro Noninvasive markers of airway inflammation and redox biochemistry applied to asthma Eur. Respir. Rev., December 1, 2006; 15(101): 202 - 203. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |