INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Keywords: 11-hydroxysteroid dehydrogenase type II enzyme; asthma; inhaled corticosteroids; endobronchial biopsies; immunohistochemistry

11-Hydroxysteroid dehydrogenases (11-HSD) are tissue-specific enzymes responsible for the interconversion of bioactive endogenous and/or synthetic glucocorticoids to and from their receptor-inactive metabolites, such as cortisol and cortisone in humans, and corticosterone and 11-dehydrocorticosterone in rodents. There exist two isoforms of 11-HSD: 11-hydroxysteroid dehydrogenase type I (11-HSD1) is a low-affinity nicotinamide adenine nucleotide diphosphate (NADP)-dependent enzyme that can function bidirectionally but acts mainly as an oxidoreductase to produce active glucocorticoid, cortisol, or corticosterone; 11-hydroxysteroid dehydrogenase type II (11-HSD2) has a high affinity for glucocorticoids (K_m ~ 47 nM), is NAD⁺ dependent, and acts on natural glucocorticoids as a dehydrogenase to produce inactive metabolites (1). Paradoxically, increasing evidence suggests that 11-HSD2 acts as an oxidoreductase on some metabolites of synthetic and fluorinated glucocorticoids, therefore potentially enhancing their biologic potency (2-5).

In the distal tubules of the kidney, 11-HSD2 acts as a protector of the nonselective mineralcorticoid receptors by preventing their occupation by glucocorticoids, and it modulates access of glucocorticoids to glucocorticoid receptors. Mutations in this enzyme result in the syndrome of apparent mineralcorticoid excess, with sodium retention, severe hypertension, growth retardation, edema, renal hyperplasia and hypertrophy (6, 7). 11-HSD2 has also been found in other tissues controlling local concentrations of glucocorticoids, such as in the placenta, where it protects the fetus from the high circulating levels of maternal glucocorticoids (7, 8), but also in the skin (9-11), brain, gastrointestinal tract (11), thymus, spleen, and lung (12-15).

Inhaled synthetic glucocorticoids (ICS) remain the primary effective antiinflammatory treatment agents for asthma, but little is known about the metabolism of ICS in the lung. Schleimer (13) did not observe metabolism of cortisol in resected samples taken from three patients with carcinoma of the lung, nor in minces of airway samples or in pulmonary blood vessels, but they did find evidence of this process in the pulmonary parenchyma in these patients. 11-HSD2 immunoreactivity localized to epithelial cells has been detected in formalin-fixed, paraffin-embedded airway tissue (14, 15). In vitro studies have shown 11-HSD activity in human small cell lung cancer (DMS-79) cell lines (16). Other studies have demonstrated 11-HSD activity in the human lung fibroblast LU-19 cell line (17) and also in a human tracheal epithelial cell line (13).

If 11-HSD, and especially the type II enzyme, exist within the airways, they could influence the effectiveness of ICS by converting metabolites back into the active form, whether such metabolism occurs locally, or if the metabolites return to the airways, via the circulation. There has been very little research into such potentially important airway pharmacodynamic issues.

Furthermore, there is evidence of substantial interperson variability in the tissue expression of 11-HSD enzyme activity (18), which could potentially relate to the wide spectrum of patients' needs for ICS for asthma control. This potential is emphasized by inhibitors of 11-HSD2, such as licorice and carbenoxolone (with the active compounds in both cases being glycyrrhetinic acid or its derivatives), which have been shown to have antiinflammatory properties (9, 19).

We conducted a study with the aims of determining whether 11-HSD2 could be localized to airway tissue as represented by endobronchial biopsy specimens and whether differential expression of this enzyme may contribute to asthma severity and to the relative ICS needs of different patients with asthma.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subjects

Twenty-two asthmatic subjects (meeting American Thoracic Society guidelines [20] for a diagnosis of asthma) were recruited and were divided into two groups: (1) a group not treated with ICS (n = 7); and (2) a group treated with ICS (200 to 1,500 µg/d beclomethasone dipropionate [BDP]; n = 15) (Table 1). There were also nine nonasthmatic volunteer subjects. All subjects were nonsmokers, and the study was approved by the Alfred Hospital Ethics Review Committee. Written informed consent was obtained before commencement of the study. At screening, atopy status, spirometry and airway responsiveness to inhaled methacholine in terms of the provocative dose of methacholine producing a 20% reduction in FEV₁ (PD₂₀M) (21) were measured.

Fiberoptic Bronchoscopy and Endobronchial Biopsy Procedures

Subject preparation for fiberoptic bronchoscopy and endobronchial biopsy procedures, and specimen processing before immunoperoxidase staining, have been described previously (22).

Immunohistochemical Staining with Anti-11-HSD2 Antibody

Seven-micron-thick sections of bronchial tissue were cut on a Cryocut 1800 cryostat (Reichert-Jung, Heidelberg, Germany), mounted in duplicate and approximately 21 µm apart, on poly-L-lysine-coated slides, and fixed with paraformaldehyde-lysine periodate (PLP) for each patient. Immunopurified, polyclonal rabbit primary antibody directed to the last 16 residues in the carboxy-terminus of human 11-HSD2 (HUH23) (7) was applied. Staining was done with a standard three-layered, indirect immunoperoxidase technique with biotinylated goat antirabbit immunoglobulin antibody followed by a peroxidase-conjugated streptavidin antibody (DAKO, Glostrup, Denmark). Metal-enhanced diaminobenzidine (DAB) (Pierce, Rockford, IL) was used as the chromogen substrate, and specimens were counterstained with hemotoxylin. Each staining run included a negative rabbit immunoglobulin (DAKO) control slide and nasal polyp positive control slide. Slides were coded randomly.

Slides were assessed by a single blinded observer (B.E.O.) and were analyzed with a method previously described (23), using a computerized image analysis system and Image-Pro Plus 3.0 Windows software (Media Cybernetics, Silver Spring, MD).

Five or more consecutive, nonoverlapping high-power fields (×40 objective) of one section were assessed per patient. HUH23 immunoreactivity was localized to vessel endothelium. The total number of vessels stained in consecutive sections versus those stained specifically with HUH23 was determined by using a collagen type IV antibody. These absolute and relative vessel data have already been published (22), with the current analysis being a further development of this study.

Some immunoreactivity was observed in the epithelial cells lining the airways, and a semiquantitative method of analysis was used, involving a visual scale for both the extent and intensity of staining (24).

Statistical Analysis

Data are presented as mean ± SD, except for PD₂₀M values, which were log₁₀ transformed and are presented as geometric means and ranges. Raw data were fitted to a general linear model (25), with account for potential confounders (total area measured, age, sex, FEV₁ percent predicted before bronchodilator treatment, PD₂₀M, and atopic status). Spearman's rank correlation test (two-tailed) was used in determining the relationship between physiologic indices and expression of 11-HSD2. Values of p < 0.05 were considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Staining with the anti-11-HSD2 (HUH23) antibody was localized to the endothelium of vascular structures in the lamina propria and the epithelial cells of the endobronchial biopsy specimens that were analyzed (Figure 1) both in asthma patients and in nonasthmatic control subjects. The four vascular indices obtained for the localization of HUH23 staining were: (1) number of 11-HSD2-positive vessels/mm² of lamina propria measured; (2) percentage of vessels positive for 11-HSD2 (i.e., number of 11-HSD2-positive vessels/mm² of lamina propria measured, divided by the total number of vessels stained for collagen type IV/mm² of lamina propria measured, multiplied by 100); (3) percentage of lamina propria area stained positively for 11-HSD2; and (4) percentage of total vessel area stained positively for 11-HSD2 (i.e., 11-HSD2-positive area, divided by the total positive vessel area, multiplied by 100). The extent and intensity of positive epithelial staining with 11-HSD2 were based on a visual scale of 0 to 4 in increments of 0.5 throughout the biopsy section under analysis. Thus, for extent of staining observed, 0 = none and 4 > 75% of epithelial cells stained; whereas for intensity, 0 = no staining and 4 = very strongly stained. No statistically significant differences were observed in any of the staining indices measured for asthma patients receiving and not receiving ICS nor for asthma patients versus nonasthmatic controls (Tables 2 and 3).

View larger version (83K): [in this window] [in a new window]   Figure 1.   Representative photomicrographs. (a) Polyclonal anti-11-HSD2 (HUH23) antibody staining in endobronchial biopsy specimen, showing vessel (V) and epithelial (E) staining. (b) Negative control of same endobronchial biopsy field.

Epithelial basement membrane staining with HUH23 was also observed in all patients and controls, and no obvious differences were observed between the study groups.

There was no significant relationship between clinical physiologic features (PD₂₀M and FEV₁% predicted) and 11-HSD2 staining indices when the two asthma groups were combined (n = 22) (data not shown). Nor was there any relationship between steroid dose and indices of vascular 11-HSD2 staining. However, there was a relatively weak but statistically significant inverse relationship between steroid dose and extent of epithelial 11-HSD2 staining (r = 0.44; p = 0.04).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The most important outcome of this cross-sectional study was the finding that 11-HSD2 is present in the human airways. Furthermore, we demonstrated an inverse relationship between epithelial staining for 11-HSD2 and the ICS dose required for asthma control. This observation is consistent with the previously published finding that, paradoxically, 11-HSD2 can act as an oxidoreductase and reactivate synthetic glucocorticoids such as dexamethasone (2-5). There was no statistical difference in airway staining in asthma patients treated and those not treated with ICS nor between asthma patients and volunteer controls. This suggests that although 11-HSD2 may have a role in normal airway physiology, it is unlikely to play a significant part in asthma pathophysiology in the range of asthma severity of our study patients. The relative expression of 11-HSD2 in asthma patients receiving different doses of ICS would suggest that the local breakdown of ICS to inactive metabolites does not contribute to the greater need for ICS in some patients as compared with others. Further studies are warranted in patients who are particularly resistant to ICS therapy, and evaluating the activity of 11-HSD in the airways will also be important in determining the effects of ICSs.

In this study, 11-HSD2 immunoreactivity was seen in the endothelial cells of vascular structures in the lamina propria and in the epithelium of endobronchial biopsy specimens. Smith and colleagues (10) found staining of vascular smooth-muscle cells around the dermal arterioles of skin, arterioles of myocardium, and saphenous vein and renal interlobular arteries when using the same HUH23 antibody in formalin-fixed and paraffin-embedded tissue samples and with a three-layered immunoperoxidase technique similar to ours (7). In their study, 11-HSD2 was apparent in the endothelial cells lining vascular structures, but was not directly discussed. Brem and coworkers (26) found the endothelial cells of vessels in rat aorta to contain 11-HSD2 messenger RNA. Endothelium could control entry of endogenously produced steroid from the blood into the airways under physiologic conditions. It may also potentially modulate the efficacy of systemic glucocorticoids given therapeutically.

However, Schleimer (13) found no metabolism of cortisol in minced pulmonary blood vessels or in airways (2 to 3 mm in diameter) from the lungs of three cancer patients. Hirasawa and associates (14) and Suzuki and colleagues (15), using the same antibody as ourselves but staining human formalin-fixed and paraffin-embedded autopsy airway specimens, found 11-HSD2 expressed only in the epithelium and not in the subepithelial vascular plexus. This could have been due to a postmortem artifact and/or to different processing of the tissue samples.

Our finding of 11-HSD2 in the epithelial lining of the airways supports the previous findings of Hirasawa and Suzuki and their coworkers. In addition, 11-HSD2 has been found in other epithelia, such as those of the skin, gastrointestinal tract, and glomerular distal tubules of the kidney (7, 27). The presence of 11-HSD2 in the airway epithelium may again provide a barrier to the penetration of ICS into deeper structures, as well as aiding clearance of excess ICS from the airway.

The patients in our study had stable, mild-to-moderate but symptomatic asthma. Our initial hypothesis was that those receiving ICS treatment, and especially those taking higher doses of ICS, would have greater expression of 11-HSD2 than those not receiving ICS therapy and/or controls (i.e., that increased enzyme in the airways would be a cause of their relative steroid needs). However, this did not seem to be the case. Immunohistologically detected expression of an enzyme does not necessarily equate with its activity. Indeed, there are a number of factors that may influence the activity of 11-HSD2 and thus the bioavailability of ICS to their receptors. Such factors include inflammatory cell activation and local presence of cytokines. One group has found that cell cultures of leukocytes in the presence of interleukin (IL)-5, IL-6, IL-4, and interferon-, individually or in combination, increased 11-HSD activity (28). Thus, asthmatic inflammation itself may conceivably increase ICS breakdown in the airway. Another group (29) has postulated that 11-HSD may regulate the antiinflammatory action of endogenous cortisol but not that of synthetic glucocorticoids (such as beclomethasone dipropionate or fluticasone proprionate) in cultured airway epithelial cells. These are interesting data, but the study that served as the basis for this last suggestion had its shortfalls. It investigated the collective activity of both the 11-HSD1 and 11-HSD2 enzymes, without differentiating between the two, and the epithelial cells used in the study were obtained at autopsy up to 24 h after death, when enzyme activity may well be altered and/or deteriorating. Moreover, the study did not determine the effect of the enzymes in intact tissue, and 11-HSD has been shown to be influenced by a number of local factors, including growth factors such as epidermal growth factor, which have been shown to decrease 11-HSD2 activity (30), and by local NAD⁺ concentrations. Conversely, certain steroid hormones have been shown to increase the activity of this enzyme (30), an effect that could potentially play a role in premenstrual asthma.

Given that our data suggest an inverse relationship between the extent of epithelial 11-HSD2 staining and the ICS dose needed for effective asthma control, we should consider the possibility that this enzyme operates in the airways as an oxidoreductase of ICS, thus making these steroids more readily available. Best and coworkers (3) have shown that dexamethasone is variably metabolized in the human kidney, with about 40% of this steroid usually being metabolized. Furthermore, approximately 70% of 11-dehydrodexamethasone is reconverted to dexamethasone by 11-HSD2 in vitro. In vivo, the 11-dehydrodexamethasone metabolite can be detected in relatively low concentrations in plasma after patients have received intravenous injections of dexamethasone. However, rapid turnover of dexamethasone and its 11-dehydrodexamethasone metabolite in different tissues and variations in this metabolic balance may affect tissue concentrations of the two steroids without necessarily affecting their circulating concentrations. It is possible that plasma levels of metabolites as determined in clinical trials do not accurately reflect tissue kinetics. In addition, there may be considerable oxidation of glucocorticoids by other enzymes that is masked by 11-HSD2 reconversion. Further, there is evidence that the equilibrium for the action of 11-HSD2 shifts toward reduction in the metabolism of 9-fluorinated steroids (4, 5), which might play an important role in local regulation of these synthetic glucocorticoids.

Our study raises the possibility that such local pharmacodynamic factors could be important in influencing how much ICS a person with asthma needs in order to control their active disease. We would suggest that future studies should focus on the oxidoreductase, 11-HSD1, in addition to 11-HSD2 and the balance between the two isozymes. There is evidence that 11-HSD1 mRNA is more abundant than that for 11-HSD2 in rat aortic endothelial cells (26), but there are no data on this balance in pulmonary tissues. The two isoforms may function together in determining the bioavailability of ICS in the lung, either in the same dynamic direction when acting on ICS or in opposition to one another, and it may be the balance of effects that is crucial.

More studies are needed to investigate the potentially important 11-HSD enzyme system and the complex cascade of events that may modulate ICS bioavailability to determine their ultimate therapeutic effectiveness in managing asthma and indeed in managing other respiratory diseases as well. These enzymes need to be tested in longitudinal studies of ICS effectiveness in corticosteroid-naïve subjects and also in relatively steroid-resistant asthmatic subjects in whom our hypothesis would suggest the balance of metabolism could be especially adverse.