Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Inhaled and systemic glucocorticoids are highly effective anti-inflammatory agents that are the mainstay of therapy for numerous inflammatory conditions, such as atopic dermatitis (1), inflammatory bowel disease (2, 3), collagen vascular disease (4, 5), and asthma (6). These drugs exert their influence by binding to the glucocorticoid receptor in target cells, resulting in programmed cell death and/or altered secretion of inflammatory mediators (7, 8). Glycyrrhetinic acid, a compound derived from licorice extracts, is similar in structure to glucocorticoids and is also known to possess anti-inflammatory properties (9). Licorice extracts, for example, have been used for centuries in the treatment of asthma and were recommended by such historically prominent physicians as Jacobus Sylvius in the sixteenth century and William Withering in the eighteenth century (10). Glycyrrhetinic acid has also been suggested to be helpful in other diseases typically treated by glucocorticoids, including eczema (11) and Addison's disease (12). The mechanism by which glycyrrhetinic acid exerts its anti-inflammatory activity, however, has not been well understood. Despite its structural similarity to glucocorticoids, glycyrrhetinic acid binds poorly to the glucocorticoid receptor (13) and is therefore unlikely to operate via the same pathway.

Recent evidence suggests that the pharmacologic effects of glycyrrhetinic acid are due to its influence on 11-hydroxysteroid dehydrogenase (11HSD), an enzyme that reversibly converts hydrocortisone, the principal active glucocorticoid in humans, to its inactive product cortisone (14). This enzyme, the expression of which varies greatly throughout the body, is known to exist in isoforms that have both high and low affinities for hydrocortisone (15, 16). Its presence is thought to regulate the activity of hydrocortisone in an organ-specific fashion. In the kidney, for instance, hydrocortisone (which in plasma is found at concentrations several orders of magnitude higher than aldosterone) has an affinity for the mineralocorticoid receptor equal to that of aldosterone. 11HSD is now known to prevent hydrocortisone from binding the renal mineralocorticoid receptor by transforming it locally to cortisone (17, 18). Glycyrrhetinic acid is a powerful inhibitor of 11HSD; ingestion of excess quantities results in the unimpeded binding of the mineralocorticoid receptor by hydrocortisone, manifested by marked salt retention and hypertension (19).

The lung is another organ in which significant 11HSD activity has been detected (20). Previous studies from our laboratory have shown that bronchial epithelial cells convert radiolabeled hydrocortisone to cortisone and that this activity is suppressed by glycyrrhetinic acid (9). Because bronchial epithelium is now believed to be both an active participant in airway inflammation and an important target of glucocorticoids, local manipulation of 11HSD activity may be potentially relevant to the treatment of allergic pulmonary diseases, such as asthma (21). We therefore sought to examine whether inhibition of 11HSD by glycyrrhetinic acid would modulate responses to hydrocortisone in primary bronchial epithelial cell (PBEC) cultures. Secretion of granulocyte macrophage colony-stimulating factor (GM-CSF) was chosen as a marker for PBEC activity. In previous studies involving bronchial epithelial cells, GM-CSF secretion was markedly enhanced by the inflammatory cytokine interleukin (IL)-1 and effectively suppressed by glucocorticoids (22). The present studies show that 11HSD inhibition with glycyrrhetinic acid significantly increases the potency of hydrocortisone as an inhibitor of epithelial GM-CSF release in PBEC cultures. These findings could not be repeated in BEAS-2B, a cell line that lacks 11HSD activity, or in the presence of synthetic steroids not susceptible to 11HSD degradation, demonstrating that the anti-inflammatory properties of glycyrrhetinic acid are specific to 11HSD inhibition. Together, these findings suggest that 11HSD may regulate the suppressive effects of endogenous hydrocortisone on lung inflammation.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Reagents: Cytokines and Glucocorticoids

The cytokine IL-1 and the paired antibodies against GM-CSF used for enzyme-linked immunosorbent assay (ELISA) experiments were obtained from R&D Chemical Company (Minneapolis, MN). Hydrocortisone and glycyrrhetinic acid were obtained from Sigma Chemical Company (St. Louis, MO). Budesonide was a generous gift from Dr. Ralph Brattsand of Astra Draco (Lund, Sweden); beclomethasone dipropionate (BDP) and fluticasone propionate (FP) were obtained from Glaxo Wellcome (Research Triangle Park, NC); and mometasone furoate (MF) and triamcinolone acetonide (TAA) were obtained from Schering-Plough (Kenilworth, NJ). All steroids, as well as glycyrrhetinic acid, were stored in the vehicle dimethyl sulfoxide (DMSO) at 0.1 M and at 20°C until use. [³H]hydrocortisone ([1,2-(³H)(N)]-2.0 Tb_q/mmol; 55 Ci/mmol) was obtained from New England Nuclear (Boston, MA).

Tritiated cortisone is not commercially available and was therefore biosynthetically produced by the oxidation of [³H]hydrocortisone with placental tissue. Twenty grams of finely minced human placental tissue were washed and filtered three times with PAGCM buffer ([1,4-piperazinebis (ethane sulfonic)]-buffered saline containing 0.003% human serum albumin, 0.1% D-glucose, 1 mM CaCl₂, and 1 mM MgCl₂) through 150 µm mesh Nitex (Tetko, Inc., Briarcliff Manor, NY). The tissue was then washed once with RPMI medium, filtered through Nitex, and incubated with 50 µCi of [³H]hydrocortisone in 3 ml of RPMI medium. The mixture was allowed to incubate at 37°C under an atmosphere of 5% CO₂/5% air for 24 h with periodic mixing. The reaction was stopped and the glucocorticoids were extracted with ethyl acetate. Tubes were capped, vortexed, and centrifuged to separate the phases. The upper organic layer, which contained [³H]hydrocortisone and [³H]cortisone, was separated from the lower aqueous layer, which contained cellular debris, and dried in a Savant vacuum concentrator (Savant, Farmingdale, NY). After resuspension in methanol, [³H]cortisone was purified from the extract by thin-layer chromatography on silica gel G plates using a developing system consisting of 90% chloroform and 10% methanol. The tritiated cortisone product was scraped from the plate, extracted from the silica with ethyl acetate, and rechromatographed to assess radiochemical purity, which exceeded 98%.

Cell Purification and Culture

PBECs. Mainstem and lobar human bronchi (Anatomical Gift Foundation, Athens, GA) were obtained at autopsy within 24 h of death and incubated for 48 h at 4°C in F12 medium supplemented with penicillin (200 U/ml), streptomycin (200 µg/ml), and pronase (1 mg/ml) (Calbiochem, La Jolla, CA). Tissue was then placed in serum-free F12 media supplemented with 10% fetal calf serum and dissected longitudinally. Epithelial cells were dislodged from the bronchial surface by jets of medium delivered from a 10-cc pipette. Cells were centrifuged, washed, centrifuged again, and resuspended in F12/Dulbecco's modified Eagle's medium (DMEM) media (medium containing equal parts F12 and DMEM, supplemented with 5% fetal bovine serum as well as penicillin [100 U/ml] and streptomycin [200 µg/ml]). Cell cultures were incubated at a concentration of 50,000 cells/ml on collagen-coated 1.5-cm plates (Collagen Biomaterials, Palo Alto, CA) at 37°C under an atmosphere of 5% CO₂/95% air until reaching confluence approximately 7 d later. Identity of PBECs was confirmed by cytokeratin staining as described elsewhere (23). Purity exceeded 98% in all cases and no contaminating cells could be identified. All cell cultures were also assayed for viability and spontaneous 11HSD activity. No difference in viability was found between cells treated with glycyrrhetinic acid and cells treated with DMSO alone (data not shown).

BEAS-2B, H441, and A549. The SV-40 virus-transformed cell line BEAS-2B was the generous gift of Dr. Curtis Harris (Bethesda, MD). The malignant-transformed cell lines H441 and A549 were obtained commercially from the American Type Culture Collection (Rockville, MD). All cell cultures were incubated in F12/DMEM at 37°C under an atmosphere of 5% CO₂/95% air until confluent.

Biochemical Assays

Cell cultures and measurement of GM-CSF release. Cells were cultured as previously described in F12/DMEM on 24-well plates until approximately 70% confluent, at which time they were treated with hydrocortisone or budesonide at the indicated concentrations for 24 h in the presence or absence of 10⁶ M glycyrrhetinic acid. We have found that preincubation with glucocorticoid before the addition of cytokine optimizes its suppressive effects (24). Moreover, in a previous study, we found that 10⁶ M glycyrrhetinic acid maximally inhibits 11HSD activity in PBEC cultures (9). Control cultures contained the diluent DMSO at the same concentration as in corresponding experimental cultures. After washing all cell cultures three times with Hanks' balanced salt solution (Biofluids, Inc., Rockville, MD), the medium was replaced with appropriate glucocorticoid and glycyrrhetinic acid-containing media, and cells were stimulated with 5 ng/ml IL-1 for 24 h. This concentration of IL-1 has previously been shown to maximally stimulate PBEC GM-CSF release (22). After cellular confluence was visually confirmed, the supernatants were removed and centrifuged at 1,200 rpm for 6 min to remove cellular debris. GM-CSF levels were quantitated by ELISA. Inhibition of IL-1-induced GM-CSF was calculated by the formula: [1  (GM-CSF produced in the presence of glucocorticoid ± glycyrrhetinic acid/GM-CSF produced by controls)] × 100.

Assay of 11HSD activity. 11HSD activity was quantified as the percentage conversion of [³H]hydrocortisone to [³H]cortisone. Cells were grown on 24-well plates until 70 to 80% confluent, at which time they were incubated for 24 h with 1,500 cpm of [³H]hydrocortisone under an atmosphere of 5% CO₂/95% air. After confluence was confirmed, cells and supernatant together were removed from the plates by incubating with 0.05% trypsin (GIBCO, Inc., Grand Island, NY) for 40 min and transferring to glass tubes. Glucocorticoids were extracted by the addition of 2.5 ml ethyl acetate. Tubes were capped, vortexed, and centrifuged to separate the upper organic layer, which contained [³H]hydrocortisone and [³H]cortisone, from the aqueous phase and cellular debris. Samples were dried in a Savant vacuum concentrator (Savant) and resuspended in 100 µl of methanol. [³H]hydrocortisone and [³H]cortisone were separated by thin-layer chromatography on silica gel plates (Silica Gel G; Analtech, Newark, DE) using a 90% chloroform/10% methanol solvent mixture as described elsewhere (8). The plates were scraped in 5-mm fractions and the samples transferred to scintillation vials. Scintillation cocktail was added (Budget-Solve; RPI, Mount Prospect, IL) and radioactivity was determined in a Beckman LS1701 beta counter. 11HSD activity was expressed as the percentage conversion of [³H]hydrocortisone to [³H]cortisone after subtracting background. In the absence of tissue, no conversion to [³H]cortisone was observed (data not shown).

Statistical Analysis

Data are expressed as means ± standard error of the mean (SEM). Comparisons between the study groups were analyzed using two-way analysis of variance (ANOVA) for repeated measures. Comparisons among cells exposed to synthetic glucocorticoids other than budesonide were analyzed using a paired t test. Statistical significance was assumed at P < 0.05.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Cell Cultures

To test whether 11HSD influences the effects of hydrocortisone on cytokine release, we first screened several immortalized cell lines in the hope of identifying one that expressed 11HSD enzymatic activity. Average 11HSD activity, expressed as the percent conversion of [³H]hydrocortisone to [³H]cortisone, was 21.3 ± 4.7% for PBECs (range 11.8 to 29.7%, n = 7) and less than 5% in BEAS-2B, H441, and A549. PBECs were therefore selected for all subsequent studies.

Influence of 11HSD on PBEC Response to Hydrocortisone

If 11HSD reduces the intracellular concentrations of hydrocortisone available to the glucocorticoid receptor, then inhibition of this enzyme with glycyrrhetinic acid should increase PBEC sensitivity to hydrocortisone. Responsiveness to hydrocortisone was determined by the inhibition of IL-1-induced GM-CSF release. The concentration of GM-CSF in basal PBEC cultures was 1,075 ± 301 pg/ml and rose to 2,041 ± 319 pg/ml after stimulation with IL-1. Incubation of PBECs with hydrocortisone led to a dose-dependent decrease in GM-CSF expression (concentration that produces 50% inhibition of maximum effect [IC_1/2max] 5.0 × 10⁸ M) which was significantly potentiated by the presence of glycyrrhetinic acid (IC_1/2max 1.5 × 10⁹ M) (Figure 1). Maximum inhibition of GM-CSF release was 55.5 ± 13.7% and 48.7 ± 12.2% in the presence and absence, respectively, of glycyrrhetinic acid. As expected, maximum inhibition was not significantly affected by glycyrrhetinic acid because the maximal effect is determined by receptor saturation.

View larger version (23K): [in this window] [in a new window]   Figure 1.   Glycyrrhetinic acid potentiates the inhibition of GM-CSF release by hydrocortisone in PBECs. PBECs were preincubated with hydrocortisone at the indicated concentrations ± glycyrrhetinic acid (106 M). Cells were stimulated with 5 ng/ml IL-1 and GM-CSF levels were quantitated by ELISA. Data shown are means ± SEM percent inhibition of GM-CSF release in the presence and absence of glycyrrhetinic acid from seven experiments. Incubation with hydrocortisone resulted in a dose-dependent suppression of GM-CSF release that was significantly potentiated by coincubation with glycyrrhetinic acid (two-way ANOVA, P < 0.05). Glycyrrhetinic acid suppressed GM-CSF release in unstimulated PBECs by 10.3 ± 8.8% in the absence of hydrocortisone, but this was not statistically significant (t test).

To further test the necessity of 11HSD activity to develop an effect with glycyrrhetinic acid, identical experiments were performed using BEAS-2B, a transformed cell line that does not express 11HSD activity. Basal levels of GM-CSF in BEAS-2B were 39 ± 5 pg/ml, which rose to 264 ± 17 pg/ml after stimulation with IL-1. Although GM-CSF release was again inhibited by hydrocortisone in a concentration-dependent fashion (IC_1/2max 5.2 × 10⁸ M), glycyrrhetinic acid failed to influence the concentration-response curve (Figure 2). In addition, the synthetic glucocorticoid budesonide was tested in IL-1-stimulated PBECs. Budesonide is probably not metabolized by 11HSD in humans because no keto-derivative has been identified in patients treated with this glucocorticoid (25). Baseline and IL-1-stimulated levels were 585 ± 143 and 1,243 ± 253 pg/ml, respectively, in the absence of glycyrrhetinic acid; and 645 ± 144 and 1,270 ± 265 pg/ml in the presence of glycyrrhetinic acid. As with hydrocortisone, budesonide caused a concentration-dependent inhibition of GM-CSF release that was not significantly affected by glycyrrhetinic acid (IC_1/2max 6.0 × 10¹⁰ M) (Figure 3). In sum, these results suggest that glycyrrhetinic acid is effective only in cells expressing a functional 11HSD enzyme and when an 11HSD-susceptible steroid is used (i.e., hydrocortisone).

View larger version (20K): [in this window] [in a new window]   Figure 2.   Lack of effect of glycyrrhetinic acid on the inhibitory effect of hydrocortisone on GM-CSF release in BEAS-2B. BEAS-2B cultures were preincubated with hydrocortisone at the indicated concentrations ± glycyrrhetinic acid (106 M). Cells were stimulated with 5 ng/ml IL-1, and GM-CSF levels were quantitated by ELISA. Data shown are means ± SEM percent inhibition of GM-CSF release from three experiments. Incubation with hydrocortisone resulted in a dose-dependent suppression of GM-CSF release that was not significantly altered by glycyrrhetinic acid (two-way ANOVA).

View larger version (23K): [in this window] [in a new window]   Figure 3.   Lack of effect of glycyrrhetinic acid on the inhibition of GM-CSF release by budesonide. PBECs were preincubated with budesonide at the indicated concentrations ± glycyrrhetinic acid (106 M). Cell cultures were stimulated with 5 ng/ml IL-1, and GM-CSF levels were quantitated by ELISA. Data presented are means ± SEM percent inhibition of GM-CSF release from five experiments. Incubation with budesonide resulted in a dose- dependent suppression of GM-CSF release that was not significantly altered by glycyrrhetinic acid (two-way ANOVA).

The susceptibility of other inhaled steroids to 11HSD is not well established. Therefore, we tested the effect of glycyrrhetinic acid on the activity of BDP, FP, MF, and TAA (Figure 4). PBECs were exposed to 10⁹ M of each steroid except hydrocortisone, where 10⁸ M was used. These concentrations were chosen because they have been shown to reduce basophil histamine release by approximately 50% (10⁹ M for all but hydrocortisone, which was 10⁶ M) (31). Glycyrrhetinic acid did not significantly alter the ability of any synthetic steroid tested to alter GM-CSF release. These results provide circumstantial evidence that the steroids tested are also insensitive to degradation by 11HSD.

View larger version (33K): [in this window] [in a new window]   Figure 4.   Inhibition of GM-CSF release by BDP, FP, MF, and TAA was not altered by glycyrrhetinic acid. PBECs were preincubated with the indicated glucocorticoid ± glycyrrhetinic acid (106 M). The concentration of steroid used was 109 M for all glucocorticoids except hydrocortisone (HC), for which 108 M was used. Cell cultures were stimulated with 5 ng/ml IL-1, and GM-CSF levels were quantitated by ELISA. Data shown are means ± SEM percent inhibition of GM-CSF release from three experiments. Glycyrrhetinic acid significantly potentiated the ability of hydrocortisone to suppress GM-CSF release but had no significant effect on any of the other glucocorticoids tested. *Significant difference from cells stimulated in the absence of glycyrrhetinic acid (paired t test).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

Studies in the kidney and skin have shown that 11HSD can profoundly diminish glucocorticoid activities (11, 18). The present studies suggest that this enzyme may also blunt the response of airway epithelial cells to hydrocortisone. We found that the potent 11HSD inhibitor glycyrrhetinic acid enhanced the ability of hydrocortisone to suppress GM-CSF release in IL-1-stimulated PBEC cultures. In the presence of glycyrrhetinic acid, the potency of hydrocortisone increased approximately 33-fold. The effect of glycyrrhetinic acid was most pronounced in the presence of 10⁸ M hydrocortisone, a concentration roughly equivalent to levels of free hydrocortisone in circulating plasma (26). To ensure that these observations are indeed related to the inhibition of 11HSD and not other enzymes or processes that may alter hydrocortisone activity, these experiments were repeated in BEAS-2B, a cell line that lacks 11HSD activity, and with the glucocorticoids budesonide, BDP, FP, MF, and TAA, synthetic steroids that are not subject to 11HSD degradation. Glycyrrhetinic acid had no effect under these other conditions, suggesting that the observations made in PBECs treated with hydrocortisone were specifically related to 11HSD.

Several previous studies have similarly demonstrated potentiation of anti-inflammatory activity by 11HSD inhibition. Whorwood and colleagues (27) demonstrated that the combination of the rat glucocorticoid corticosterone and licorice derivatives (which contain glycyrrhizin, a compound related to glycyrrhetinic acid) inhibited expression of the glucocorticoid target gene prolactin more than did either agent alone. This effect was blocked by the glucocorticoid antagonist RU486 (27). Similar studies by Escher and associates (28) using a renal model showed that corticosterone inhibits expression of the inflammatory enzyme group II phospholipase A2 in IL-1- and tumor necrosis factor--stimulated rat mesangial cells; this effect was potentiated by glycyrrhetinic acid (28).

It is worth noting that most measurements reported here were performed in media supplemented with fetal calf serum, which contains low levels of hydrocortisone. We have found that such supplementation with serum is necessary to maintain cell confluence and optimal viability of PBEC cultures. In separate experiments using gas chromatography/mass spectrometry, we measured levels of free hydrocortisone in F12/DMEM containing 5% serum to be approximately 1.2 × 10⁹ M (range 6 × 10¹⁰ M to 1.9 × 10⁹ M) (data not shown). These low levels of glucocorticoids are unlikely to affect our observations significantly. Nonetheless, in separate experiments we found glycyrrhetinic acid to potentiate the effect of hydrocortisone even in the absence of serum (data not shown).

Several studies now suggest that endogenous glucocorticoids, such as hydrocortisone, may play a greater role in inflammation than previously realized. Recent studies from our laboratories found higher levels of circulating hydrocortisone in the plasma of asthmatic patients who did not manifest a late asthmatic response to inhaled allergen than in the plasma of those who did (29). Further, Sasaki and coworkers (30) found that the inhibition of adrenal glucocorticoid production with metyrapone dramatically potentiates the late-phase airway response in dogs. 11HSD inhibitors may therefore be of some therapeutic value in inflammatory diseases currently treated with synthetic glucocorticoids by augmenting the properties of endogenous hydrocortisone, and may consequently allow reductions in their dosing.

The present findings indicate the potential value of studies to determine whether 11HSD activity varies among individuals with pulmonary disease as compared with individuals without pulmonary disease, and whether local inhibition of this enzyme in the airways can exert an anti-inflammatory effect. Information gained from such studies will be important in clarifying the role of 11HSD in regulating pulmonary inflammation and determining whether pharmacologic manipulation of its activity can be exploited therapeutically.