Glucocorticoid (GC) metabolism by the 11β-hydroxysteroid dehydrogenase (HSD) system is an important prereceptor regulator of GC action. The HSD enzymes catalyze the interconversion of the endogenous, biologically active GC cortisol and its inactive 11-dehydro metabolite cortisone. The role of the HSD enzymes in the metabolism of synthetic GCs, such as dexamethasone (Dex), is more complex. The human lung is a classic GC-sensitive organ; however, the roles of the HSD enzymes (HSD1 and HSD2) in the human lung are poorly understood. In the present study, we examined the expression of the HSD enzymes in human adult and fetal lung tissues and the human lung epithelial cell line NCI-H441. We observed that human adult and fetal lung tissues, as well as H441 cells, express HSD2 protein and that it is upregulated by Dex (10−7 M). By contrast, HSD1 protein was undetectable. We also show that the Dex-mediated regulation of surfactant protein A is attenuated by inhibition of HSD2 activity. Furthermore, we demonstrate that unlike the inactive, 11-dehydro metabolite of cortisol (i.e., cortisone), the 11-dehydro metabolite of Dex, 11-dehydro-Dex, competes for binding to the GC receptor (GR) in human lung epithelial cells and retains GR agonist activity. Together, these data suggest that differences exist in the biological activities of the metabolites of cortisol and Dex.
- H441 cells
pulmonary surfactant is a complex mixture of phospholipids and proteins produced and secreted by type II alveolar epithelial cells. Surfactant is secreted into the alveolar space, where it forms a monolayer that reduces surface tension at the air-liquid interface, thereby preventing alveolar collapse at end-expiration (12). Infants born prematurely (less than ∼36 wk gestation) often lack sufficient pulmonary surfactant and may develop respiratory distress syndrome (RDS), a leading cause of infant morbidity and mortality in the US (11). The production of the protein and phospholipid components of pulmonary surfactant is regulated by natural and synthetic glucocorticoids (GCs), such as cortisol and dexamethasone (Dex) (9, 18, 29, 31). Surfactant protein A (SP-A) is the most abundant of the surfactant-associated proteins and plays a role in the innate host defense of the lung by enhancing pathogen clearance and modulating inflammation (13). GC treatment increases the transcription of the SP-A gene but decreases the stability of the nascent mRNA transcripts, leading to a net decrease in SP-A mRNA and protein levels (8, 34). Additionally, SP-A is one of only two genes to be identified as an RDS susceptibility gene (22).
In humans, GCs are present in two forms, depending on the chemical group occupying the 11β position of the steroid. The endogenous, biologically active GC, cortisol, has a hydroxyl group at this position, whereas the biologically inactive form, cortisone, has a ketone group at this position. High affinity binding of cortisol to the GC receptor (GR) is thought to depend on the presence of a hydroxyl group at the 11β position (23). Steroids containing a ketone group at the 11β position (i.e., cortisone and prednisone) tend to have extremely low agonist activity (7, 17).
An important method of modulating GC action involves the prereceptor metabolism of GCs in peripheral tissues (for review, see Ref. 27). This metabolism is catalyzed by the 11β-hydroxysteroid dehydrogenase (HSD) system, which consists of two distinct enzymes, HSD type 1 and type 2 (HSD1 and HSD2). HSD1 functions in vivo as an 11β-reductase, converting the inactive 11β-keto GC, cortisone, to the active 11β-hydroxy GC, cortisol. HSD1 expression is especially prominent in the liver and adipose tissue, where it acts to amplify GC action by converting some of the inactive cortisone pool to cortisol (27). By contrast, HSD2 functions in vivo as an 11β-oxidase, converting active cortisol to inactive cortisone. HSD2 is most highly expressed in aldosterone-sensitive, sodium-transporting tissues such as the kidney, colon, and salivary glands. In these tissues, HSD2 catalyzes the inactivation of cortisol to cortisone, thus preventing illicit occupation of the mineralocorticoid receptor (MR) by cortisol. This action is physiologically significant because the MR has equal binding affinities for both aldosterone and cortisol (27). The metabolism of the commonly used synthetic GC, Dex, differs from that of the endogenous GC, cortisol. Fluorine and methyl group substitutions in Dex limit its oxidoreduction by HSD1 and oxidation by HSD2 (15). Interestingly, HSD2 actively catalyzes the oxidoreduction of 11-dehydro-Dex (DH-Dex) to active Dex but does not oxidoreduce cortisone to cortisol (15). Thus significant differences exist in the metabolism of cortisol vs. Dex by HSD2.
Synthetic GCs, such as Dex and betamethasone, are able to cross the placental barrier largely unmetabolized. As a result, women at risk for premature delivery are often treated with synthetic GCs in an effort to accelerate fetal lung maturation and reduce the incidence of RDS (1). However, numerous studies have reported significant postnatal side effects in infants receiving repeated doses of antenatal GCs. These side effects include decreased somatic growth, hypertension, and altered neurological development later in life (41). A better understanding of the activity and pharmacokinetics of synthetic GCs used in clinical settings may offer insights that could lead to more efficacious treatments for RDS and other pulmonary disorders.
The first goal of the present study was to examine the expression and regulation of the HSD enzymes in human lung epithelial cells. The second goal was to examine the potential role of these enzymes in the Dex-mediated regulation of SP-A. Natural and synthetic GCs, such as cortisol and Dex, downregulate the expression of SP-A mRNA in a dose-dependent manner (19, 34). To address the goals of our study, we have utilized the human lung epithelial cell line NCI-H441, as well as several previously described chemical inhibitors of the HSD enzymes (14, 30). The H441 cell line is a lung epithelial adenocarcinoma cell line that has been used extensively as a model system to study the GC-mediated regulation of the surfactant protein genes. Studies from our laboratory and others’ (4, 29) have demonstrated that GCs affect the expression of SP-A in H441 cells in the same manner as in cultured human lung explants and in isolated human adult type II cells. One recent study (36) compared directly the regulation of SP-A gene expression by Dex in H441 cells and in isolated human lung type II cells and found that the responses were identical in the two model systems. Finally, we compared the biological activites of cortisol and Dex with their 11-keto metabolites cortisone and DH-Dex.
H441 cells (American Type Culture Collection, Manassas, VA) were cultured in RPMI 1640 medium supplemented with 10% fetal bovine serum (FBS) and 1% antibiotic-antimycotic solution (Invitrogen, Carlsbad, CA). The cells were grown in 100-mm culture dishes at 37°C in a humidified atmosphere of 95% air-5% CO2. Media were changed biweekly and the cells were passed weekly, typically ≤15 passages. For most experiments, ∼1 × 106 cells were plated into each well of a six-well culture dish. The cells were serum starved for 24 h before treatment to avoid the influence of endogenous GCs found in FBS. The cells were then incubated with serum-free medium containing either vehicle [0.0001% ethanol or dimethyl sulfoxide (DMSO) vol/vol], an HSD inhibitor [10−6 M carbenoxolone (Carb), 10−6 M chenodeoxycholic acid (CDCA), or 5 × 10−5 M tetramethylthiuram disulfide (Thiram)], corticosteroid alone, or Dex plus an HSD inhibitor. Steroids were purchased from Sigma (St. Louis, MO), and 1 mM stock solutions were prepared in ethanol. HSD inhibitors were also purchased from Sigma and resuspended in either ultrapure water (Carb), ethanol (CDCA), or DMSO (Thiram) as 1 mM stock solutions.
One milliliter of TRIzol reagent (Invitrogen) was added to the wells, and the lysed cells were scraped into sterile microfuge tubes. Total RNA was then isolated according to the manufacturer’s directions. The RNA was resuspended in ultrapure water and quantified by determining the absorbance at 260 nm. Purity was assessed by determining the ratio of absorbance at 260 nm to that at 280 nm. The A260 to A280-nm ratios were typically between 1.7 and 1.9. RNA integrity was monitored by examining the 28S and 18S rRNA bands after electrophoresis in agarose gels stained with ethidium bromide.
One microgram of total RNA was reverse transcribed using oligo(dT) 18-mers as primers in a reaction volume of 30 μl that contained 2-deoxynucleotide 5′-triphosphate mix (2.5 mM each), 5 units of mouse murine leukemia virus RT, and the supplied reaction buffer. The RT reaction was carried out at 37°C for 1 h. Five microliters of the resulting cDNAs or 5 μl of water (no template control) were used as a template in PCR reactions, using HSD2-specific primers that span an exon (forward: 5′-CGGCTGTGACTCTGGTTTTGGCAA-3′; reverse: 5′-ATAGGCCCCCAAGCACGGATAT-3′). Amplification was performed for 35 cycles (95°C denaturation for 30 s, 54°C annealing for 90 s, 72°C extension for 120 s) followed by a final extension cycle of 10 min at 72°C. PCR reaction products were separated on a 2% agarose gel that contained ethidium bromide, exposed to UV light, and photographed. Purified PCR products were sequenced by the DNA Core Facility of the University of Iowa and compared with sequence data published in GenBank.
H441 cells were incubated for 1 h at 37°C with [1,2,4-3H]Dex (50 Ci/mmol, Amersham Life Sciences, Piscataway, NJ) at a final concentration of 10 nM. Cells were treated with no additions (control), a 500-fold excess of unlabeled GC (Dex or cortisol), increasing concentrations of Dex or DH-Dex (1 nM to 10 μM), 1 μM Carb, or 10 μM CDCA. After the incubation, the cells were washed three times in Hank’s buffered salt solution without Mg2+ or Ca2+ (Invitrogen) at room temperature and were then resuspended in 5 ml of scintillation cocktail (Budget-Solve; RPI, Mount Prospect, IL) before counting in a scintillation counter. Specific steroid binding was calculated by subtracting the amount of labeled GC bound in the presence of a 500× molar excess of unlabeled steroid (nonspecific) from all other treatments. Specific steroid binding was normalized to 1.0, and the specific binding detected in the presence of the various steroids or chemicals (Dex, DH-Dex, Carb, or CDCA) was compared with this normalized value. Assays were performed in replicates of six in a minimum of three independent experiments.
HSD2 enzyme activity assays.
Control and treated cells subcultured in six-well plates were washed with PBS and then incubated in serum-free medium containing [3H]Dex or [1,2,4-3H]cortisol (41 Ci/mmol; American Radiolabeled Chemicals, St. Louis, MO) at a final concentration of 1 nM for 24 h at 37°C. The enzyme assay was terminated by adding two volumes of ethyl acetate to the wells. The medium-ethyl acetate mixtures were then collected and the steroids extracted into the ethyl acetate phase by vortexing for 1 min followed by centrifugation for 10 min at 600 g. The ethyl acetate phase was dried under N2 gas and resuspended in 100 μl of ethanol, and 50 μl were spotted onto silica-coated glass plates (silica gel G; Analtech, Newark, DE) followed by resolution in an equilibrated thin-layer chromatography chamber using chloroform-ethanol (92:8) as the solvent. Parallel lanes were loaded with 3H-labeled steroids as standards. 3H-labeled steroids were detected and quantified using a TLC plate reader (Bio Scan AR-2000, Bio Scan, Washington, DC) that was capable of detecting tritium. HSD2 oxidase activity was assessed by calculating the conversions of [3H]Dex or [3H]cortisol to their respective [3H]11-dehydro metabolites (DH-Dex or cortisone). Parallel assays were performed in wells containing no cells to control for the potential spontaneous interconversion between 11-hydroxy and 11-dehydro forms of the steroids.
Protein isolation and Western blot analysis.
Total homogenate protein was isolated by homogenizing H441 cells, human adult lung tissue, or human fetal lung tissue in lysis buffer (1 nM phenylmethylsulfonyl fluoride in phosphate-buffered saline). The homogenate was centrifuged for 10 min at 600 g to pellet cellular debris. Microsomal proteins were obtained by homogenizing whole lung tissue or H441 cells in 2 ml of homogenization buffer (20 mM Tris·HCl, pH 7.4, supplemented with 250 mM sucrose and 1 mM EDTA). The resulting homogenate was centrifuged for 10 min at 600 g, 10 min at 10,000 g, and then 1 h at 170,000 g to sediment microsomes. The microsomal pellet was resuspended in lysis buffer. Protein concentrations were quantified using Bio-Rad protein assay reagent (Bio-Rad, Hercules, CA). Fifty to 100 μg of protein were separated via one-dimensional sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to nitrocellulose membrane. Blotting was performed using human HSD1- and HSD2-specific primary antibodies (HUH13 and HUH23, for HSD1 and HSD2, respectively) (28, 40) at a concentration of 2 μg/ml. The blots were washed and incubated in anti-rabbit horseradish peroxidase-conjugated secondary antibody (MP Biomedicals, Irvine, CA) at a 1:10,000 dilution. Immunoreactive bands were detected by use of enhanced chemiluminescence (Amersham Biosciences, Piscataway, NJ) followed by autoradiography. Exposed films were scanned and the immunoreactive bands semiquantified using Quantity One image analysis software (Bio-Rad). Protein levels were normalized to β-actin levels on blots stripped and reprobed with a β-actin monoclonal antibody at a dilution of 1:7,500 (Chemicon, Temecula, CA). Loading was monitored by staining the blots with Ponceau S, a nonspecific protein dye. Homogenate proteins of human liver (BioChain, Hayward, CA) and Caco2 cells (human colon epithelial cell line, a gift of Dr. F. Jeffrey Field, University of Iowa) were used as positive controls for the expression of HSD1 and HSD2, respectively.
Northern blot analysis.
Twenty micrograms of total RNA that was isolated from H441 cells were suspended in formamide-based RNA loading buffer and electrophoresed through a 1.2% MOPS-formaldehyde-containing agarose gel followed by transfer to a nylon membrane. Photographs of the ethidium-stained gels before and after transfer were examined to assess transfer efficiency. Northern blotting was performed as previously described using a human SP-A cDNA labeled with [32P]dCTP (Amersham Life Sciences) (20). The membranes were exposed to X-ray film for 8–24 h at −80°C with an intensifying screen. The resulting autoradiograms were scanned and the bands quantified by densitometry. 18S rRNA was also quantified by densitometry of the ethidium-stained gel before transfer to allow for correction of any loading artifacts.
Expression of HSD1 and HSD2 in human lung tissues.
Immunoblot analyses of total homogenate protein were used to detect HSD1 and HSD2 proteins in the various lung tissues and cells. As shown in Fig. 1A, HSD1 was detected via immunoblot analysis as an immunoreactive band of ∼34 kDa in human liver total protein homogenates (positive control). HSD1 protein was undetectable in total protein homogenates (100 μg) as well as in microsomal protein homogenates (100 μg) of both human fetal lung and H441 cells (Fig. 1A). The human lung tissues and H441 cells expressed HSD2 protein as a monomer of ∼40 kDa (Fig. 1B). Immunoreactive bands for HSD2 were also observed at ∼80 and ∼120 kDa (Fig. 1B). These bands are likely dimers and trimers of the HSD2 protein because HSD2 has been previously reported to oligomerize (21). HSD2 protein was also detected in protein homogenates of Caco2 cells, a human colon epithelial cell line that was used as a positive control for HSD2 expression. RT-PCR analyses confirmed that HSD2 mRNA was present in both the H441 cell line and in human adult and fetal lung tissues (Fig. 2). The identity of the PCR product was verified by sequencing followed by comparison to sequence data deposited in GenBank. Additionally, we observed that a 24-h treatment of H441 cells with Dex (10−7 M) increased HSD2 monomeric protein (the reported active form of the enzyme) (21) levels by ∼2-fold (Fig. 3).
Effects of HSD inhibition on GC-mediated regulation of SP-A.
In an effort to determine whether GC metabolism by the HSD system plays a role in the Dex-mediated regulation of a GC-sensitive gene, we tested the effects of several known HSD inhibitors on the Dex-mediated downregulation of pulmonary SP-A mRNA levels. In our initial experiments we used Carb, which inhibits both HSD1 and HSD2 activities (5). Carb has been shown previously (38) to inhibit HSD2 activity in a dose-dependent manner. We chose to use a single dose of Carb (10−6 M) because that dose has been used effectively to inhibit HSD2 activity in other studies (30). As shown in Fig. 4, treatment of Dex-stimulated H441 cells with Carb led to a statistically significant attenuation of the Dex-mediated downregulation of SP-A mRNA levels. Similar results were obtained when using the endogenous GC, cortisol (data not shown). Carb treatment alone had no effect on SP-A mRNA expression (Fig. 4). Although HSD2 protein was the only HSD protein detected in H441 cells (HSD1 protein was undetectable in H441 cells), we sought to confirm that the effects of Carb were in fact due to inhibition of HSD2 activity through the use of another, more specific, inhibitor. CDCA is a naturally occurring bile salt that has been previously reported (14) to specifically inhibit HSD1 activity. As expected, treatment of Dex-stimulated H441 cells with CDCA (10−6 M) did not alter the Dex-mediated inhibition of SP-A gene expression (Fig. 4). Treatment of H441 cells with CDCA alone did not affect SP-A mRNA expression (Fig. 4). To our knowledge, Thiram is the lone reported HSD2-specific inhibitor (6). However, we observed that Thiram, in addition to inhibiting HSD2 activity, inactivates the ligand binding site of the GR and was therefore unsuitable for studies examining the role of HSD2 activity on downstream GR signaling (manuscript in preparation).
To confirm that the results shown in Fig. 4 were due to the inhibition of HSD2 enzyme activity, we measured the effects of the HSD inhibitors on HSD2 enzyme activity in H441 cells. Because HSD2 is traditionally thought of as an oxidase that converts active 11-hydroxy steroids to their 11-dehydro metabolites, we measured HSD2 activity by assessing the conversion of [3H]Dex to [3H]DH-Dex. H441 cells were incubated in the presence of 10−9 M tritiated Dex with and without Carb (10−6 M) or CDCA (10−6 M). The concentrations of inhibitors used in these experiments have been previously reported (14, 30) to effectively reduce HSD activity. As is shown in Fig. 5A, HSD2 oxidase activity was significantly reduced by Carb (65 ± 12% reduction vs. control), whereas CDCA (a specific inhibitor of HSD1) did not affect HSD2 activity. Carb inhibited HSD2 activity in H441 cells to a similar degree, as previously reported (30). Similar results were obtained when tritiated cortisol was used, the endogenous GC in humans (data not shown). Additionally, we performed steroid-binding assays to rule out the possibility that the observed effect of Carb was the result of interactions with the GR. As is shown in Fig. 5B, neither inhibitor significantly altered the specific binding of labeled Dex to the GR in H441 cells. Taken together, our preliminary findings were suggestive that HSD2 activity might play a limited role in maintaining levels of active Dex in human lung epithelial cells.
GR agonist activity is retained by 11-dehydro metabolite of Dex.
Our observation that inhibition of HSD2 activity by Carb only slightly attenuated the response of the SP-A gene to GCs was initially unexpected. Because HSD2 has been shown to efficiently convert DH-Dex to active Dex, we anticipated that the inhibition of HSD2 activity would attenuate the effects of Dex to a greater degree (16). To address this issue, we treated H441 cells with the 11β-hydroxy (cortisol and Dex) and the 11β-keto (cortisone and DH-Dex) forms of these GCs. If HSD2 does in fact function to oxidoreduce Dex, but not cortisol, we would anticipate that DH-Dex, but not cortisone, would be converted to its active form and inhibit SP-A gene expression. As is shown in Fig. 6, both Dex and cortisol effectively reduce SP-A gene expression. Furthermore, DH-Dex also potently reduces SP-A gene expression, whereas cortisone does not. These data are suggestive that either HSD2 acts as an 11β-reductase in Dex metabolism, converting DH-Dex to active Dex in H441 cells, or that DH-Dex is a GR agonist.
Experiments were also performed in the presence or absence of the HSD inhibitor Carb to determine whether the inhibition of HSD2 activity would block the ability of DH-Dex to reduce SP-A gene expression. Interestingly, the ability of DH-Dex to affect SP-A expression was unaltered by Carb treatment (Fig. 7). This observation suggested one of two possibilities. First, the reduction of DH-Dex may be catalyzed by a previously uncharacterized member of the HSD family that is not inhibited by Carb. Alternatively, DH-Dex may be capable of binding to the GR and affecting SP-A gene expression directly, i.e., without oxidoreduction by HSD2. To assess the potential ability of DH-Dex to compete for ligand binding to the GR in H441 cells, we performed competitive steroid-binding assays by incubating intact H441 cells with labeled Dex in the presence or absence of increasing concentrations of unlabeled Dex or unlabeled DH-Dex. As is shown in Fig. 8, DH-Dex competed for the specific binding of ligand to the GR as effectively as Dex. By use of the equation of Cheng and Prusoff (10) and an estimated dissociation constant (Kd) of Dex for the human GR of 1.6 × 10−8 M (24), a Kd of ∼2.57 × 10−8 M was determined for DH-Dex. GCs such as cortisol and Dex have been shown to downregulate SP-A mRNA levels in a dose-dependent manner (19, 34). To test the hypothesis that Dex and DH-Dex are equally potent GR agonists, as is suggested by their relative binding affinities, we compared the ability of these two steroids to downregulate the expression of SP-A mRNA. Interestingly, as is shown in Fig. 9, SP-A mRNA expression was downregulated to a greater degree in the presence of Dex compared with DH-Dex at all concentrations tested. Half-maximal downregulation of SP-A levels was observed at a Dex concentration of ∼2 nM compared with ∼9 nM for DH-Dex. Taken together, these data suggest that the 11-dehydro form of Dex binds to the GR in human lung epithelial cells with the same relative affinity as Dex, although it is a less potent GR agonist than Dex.
In the present study, we have shown that human adult and fetal lung tissues, as well as H441 lung epithelial cells, express HSD2 mRNA and protein at detectable levels. By contrast, HSD1 protein was undetectable in H441 cells and human fetal lung tissue. In the intact human fetal lung, the oxidation of cortisol to cortisone (i.e., HSD2 activity) predominates, with little or no oxidoreduction of cortisone to cortisol (i.e., HSD1 activity) occurring (33). Significant conversion of cortisone to cortisol has been reported (3) only in dissociated human fetal lung cells maintained in culture for an extended period of time. During such culture conditions, the proportion of human fetal lung cells changes so that, by the time the dissociated cells reach confluence, fibroblast-like cells constitute the vast majority of cells remaining in culture (2). Thus low levels of HSD1 protein expressed in the human fetal lung may be restricted to fibroblasts and may, therefore, be below the level of detection in fresh, intact fetal lung tissue. We have also shown that HSD2 protein levels are upregulated by the synthetic GC Dex. This regulation suggests the presence of an autofeedback mechanism by which lung epithelial cells might alter the expression of GC-metabolizing enzymes in the face of elevated GC levels.
Previous studies (32) have demonstrated the expression of HSD1 in murine lung tissues. Furthermore, a role for HSD1 in the regulation of murine SP-A expression and fetal lung development has been established (25). It is important to note that the regulation of the surfactant system differs in rodents vs. humans. Notably, GCs stimulate SP-A expression in rodents and inhibit SP-A expression in humans (26, 34). Thus there are important differences in both the regulation of SP expression and GC metabolism in lung epithelial cells of rodents vs. humans, and this raises concerns about drawing direct parallels between lung development in animal model systems and in humans.
The reported differences in the metabolism of natural and synthetic GCs by HSD2 and our observation that nonspecific HSD inhibition by Carb slightly attenuated the response of the SP-A gene to Dex initially suggested that HSD2 may primarily act as an oxidoreductase to convert DH-Dex (which was presumed to be inactive) to biologically active Dex. In support of this concept, we observed that in the HSD2-expressing H441 cell line, cortisol, Dex, and DH-Dex all significantly reduced SP-A mRNA levels. Cortisone, which is not oxidoreduced by HSD2, had no effect on SP-A mRNA expression. Our observation that the inhibition of HSD2 activity by Carb slightly attenuated the effects of Dex on SP-A expression are in contrast to a previous study (39) showing that the inhibition of HSD activity in airway epithelial cells potentiated, rather than attenuated, the downstream effects of Dex. The molecular mechanisms by which GCs regulate the expression of SP-A are complex and, as of yet, not entirely understood. GCs increase the transcription of SP-A but also dramatically decrease the stability of the nascent SP-A mRNA transcripts, and this results in a net decrease in SP-A mRNA and protein (8). In our studies, we examined the role of HSD2 activity on the Dex-mediated regulation of SP-A over a 24-h period. The Dex-mediated production or activation of protein factors involved in the transcriptional and posttranscriptional regulation of SP-A may be particularly stable in the first 24 h, and this might have masked the effects of HSD2 inhibition. As a result, HSD2 activity may be more important in the long-term, rather than acute, effects of Dex on SP-A expression. However, this seems unlikely, given our observation that both Dex and DH-Dex are biologically active. Another possible explanation is that HSD2 activity plays a more prominent role in modulating the effects of natural GCs, as opposed to its effects on synthetic GCs. Indeed, it is known that endogenous GCs, such as cortisol, are better substrates for HSD2-mediated metabolism than synthetic GCs such as Dex (16). Alternatively, inhibition of HSD2 activity may limit the oxidoreduction of DH-Dex to Dex, which is a more potent GC (Fig. 9).
We observed that inhibition of HSD2 activity with Carb had no effect on the ability of DH-Dex to decrease SP-A mRNA levels, suggesting that DH-Dex agonist activity was independent of HSD2 function. In support of this, we have shown that DH-Dex competes for GR binding as well as Dex in intact H441 cells. Indeed, near the completion of our studies it was reported elsewhere (37) that DH-Dex binds to the GR in cell-free binding assays in vitro and is capable of stimulating the transcription of a heterologous GC-sensitive promoter in transfected cells. Although DH-Dex competes for GR binding as well as Dex, we observed that DH-Dex was markedly less effective as a GR agonist than Dex was in intact human lung epithelial cells. The apparent discrepancy between the affinity of DH-Dex for the GR and its ability to affect the expression of a GC-sensitive gene, compared with Dex, may be due to a number of factors. A rate-limiting step for the nuclear translocation of an activated GR-ligand complex is the induction of a conformational change in the GR protein leading to the displacement of cytosolic GR chaperones, including heat shock protein 90 (35). DH-Dex may be less efficient at inducing this required conformational change in the GR and may result in decreased levels of activated GR being translocated into the nucleus to affect gene expression. Alternatively, once translocated into the nucleus, a GR-DH-Dex complex may be less efficient than a GR-Dex complex at triggering the events that lead to a decrease in the half-life of the SP-A mRNA.
Our finding that a synthetic GC such as Dex is biologically active in both its 11β-hydroxy- and 11β-dehydro-forms in human lung epithelial cells is significant because synthetic GCs are routinely used to treat many pulmonary disorders as well as accelerate fetal lung maturation. Our results suggest that the biological activity of Dex may be less dependent on HSD activity than the activity of natural GCs, such as cortisol, or even other synthetic GCs. The retention of biological activity by the 11β-dehydro form of Dex appears to be unique to Dex and may not apply to other synthetic GCs. Indeed, although the synthetic GC prednisolone is biologically active, its 11β-dehydro form, prednisone, is biologically inactive (17). Further studies will be required to determine whether other synthetic GCs, such as betamethasone, are biologically active in their 11β-dehydro forms. Taken together, our data suggest that the HSD system may affect the response of human lung tissue to synthetic GCs differently than the response to natural GCs, such as cortisol.
This work was supported by National Heart, Lung, and Blood Institute Grant no. RO1-HL-50050 to J. M. Snyder. M. R. Garbrecht is supported by an American Heart Association Predoctoral Fellowship.
We thank Dr. Jonathan Klein, Department of Pediatrics, and Dr. Michael Welsh, Department of Internal Medicine, for supplying human fetal and adult lung tissues, respectively. We also thank Dr. F. Jeffrey Field, Department of Internal Medicine, for supplying Caco2 cells. Additionally, we thank Dr. Gary Snyder, Department of Biochemistry, for assistance in performing HSD2 enzyme activity assays. A portion of this work was presented in abstract form at the 2005 Experimental Biology Meeting in San Diego, CA (Garbrecht MR, Schmidt TJ, and Snyder JM. Dexamethasone metabolism and the regulation of SP-A in lung epithelial cells. FASEB J 19: A630, 2005).
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.
- Copyright © 2006 by American Physiological Society