Main

Bronchopulmonary dysplasia (BPD), a leading cause of morbidity in preterm infants (1), is characterized by arrested alveolarization. Part of this process, secondary septation, involves subdivision of terminal air sacs into alveoli. Multiple transcription factors regulate alveolarization, including hormone receptors that activate or repress gene expression (2,3). Glucocorticoids (GCs) and retinoids (RAs) act through the GC and RA receptors (GR, also called Nr3c1, RAR/RXR, respectively). GR signaling inhibits secondary septation by establishing a single-layered capillary network (4). GCs have been used to treat or prevent BPD (5). However, the beneficial effects of GC are now considered to be outweighed by increased risk of impaired lung and brain growth (6).

RA has been implicated in alveolar development and regeneration (7). Postnatal (PN) RA treatment counteracts GC-induced inhibition of alveolarization (8). By contrast, RA potentiates GC stimulation of expression of mesenchymal mitogens of alveolar type II cells. Vitamin D (VitD) is an important regulator of lung growth in utero (9).

In a search for downstream targets of GC that regulate lung maturation, we cloned late gestation lung 1 (Lgl1; also called Crespld2), a mesenchyme-specific gene (10,11) maximally expressed in late gestation and early PN life (10,12). Inhibition of Lgl1 impaired airway branching (13). In Lgl1 null mice, absence of Lgl1 is lethal before lung formation (14). Lgl1± heterozygotes display a complex respiratory phenotype including delayed histologic maturation, features of inflammation in the newborn period, and altered lung mechanics at maturity. In oxygen toxicity models of BPD, reduced levels of Lgl1 were restored during recovery (12). Therefore, we hypothesized that Lgl1 modulates alveolarization, subject to regulation by GC, RA, and VitD.

We report on the coordinate transcriptional regulation of Lgl1 by GC, RA, and VitD. Multiple genes important in human lung development are enriched for binding sites for GC, RA, and VitD that are in close proximity. Complex integration of the effects of GC, RA, and VitD on gene expression in the PN lung is likely to contribute to the timely advance of alveolarization without attendant inflammation.

MATERIALS AND METHODS

Fetal rat lung primary cell culture.

All procedures involving animals were conducted according to criteria of the Animal Care Committees of the Canadian Council and the Montreal Children's Hospital Research Institute, McGill University Health Centre. Timed pregnant Sprague-Dawley rats (day 0, mating; term, day 22) and postnatal (PN, day 1, and birth) were from Charles-River. Primary fibroblasts originating adjacent to the epithelium were isolated as described (10).

Cell treatment protocols.

Lung primary cells reached 75–80% confluence in minimal essential medium with Earle's balanced salts (MEM-EBSS) containing 10% fetal bovine serum (FBS), penicillin (100 units/mL), and streptomycin (100 μg/mL). The culture conditions were changed for 24 h to MEM-EBSS media containing 5% charcoal-stripped FBS, penicillin (100 units/mL), and streptomycin (100 μg/mL) before being treated with hydrocortisone (cortisol, 10 pM–100 nM), all-trans retinoic acid (ATRA, 1 μM–100 μM), 9-cis-retinoic acid (9-cis-RA, 10 μM), 1α, 25-dihydroxyvitamin D3 (VitD, 0.1–1000 μM) for 12–72 h in serum-free media. Control plates contained equal volumes of ethanol or DMSO in serum-free media (n ≥4).

RNA isolation.

Cells were lysed using 1 mL Trizol. Total isolated RNA was resuspended in RNASecure and quantified using an ND-1000 spectrophotometer.

Quantitative real-time PCR.

Real-time/quantitative real-time PCR (qRT-PCR) was performed in triplicate as described (12).

Transfection and dual reporter gene assays.

A 901bp sequence of the 5′ flanking region of rat Lgl1 [analyzed using MatInspector® promoter inspector software (15)] was inserted between the XhoI and HindIII sites upstream of the luciferase gene in the pGL3 basic plasmid to generate pLgl1-luc. Embryonic day (E) 20 and PN7 rat lung fibroblasts were seeded (2.5 × 105 cells/well) in 24-well plates. The next day, cells were cotransfected in serum and antibiotic-free DMEM using lipofectamine 2000 reagent, with 0.8 μg pLgl1-luc or with the empty pGL3Basic plasmid (control), and 4 ng of pRSV40-luc (Renilla). After 6 h, media was changed to antibiotic-free DMEM containing 5% charcoal-stripped FBS, and cells were grown for 24 h. Transfected cells were treated with 1 μM Dex, 100 nM cortisol, 10 μM ATRA, 10 μM 9-cis-RA, 100 μM VitD in DMEM containing 5% charcoal-stripped FBS (n = 3). After 24–72 h, cells were washed with PBS and lysed in 100 μL passive lysis buffer. Firefly- and renilla-luciferase activities were measured using the Dual Luciferase Reporter assay system kit and lysates measured using a Monolight 3010 luminometer. Relative promoter activities are reported as ratio of firefly- to renilla-luciferase activity to normalize for transfection efficiency. Assays were performed in triplicate.

Generation of deletion constructs.

Deletion constructs were generated using a reverse primer on the pGL3 basic plasmid and two forward primers GRE+ (-572/-552) and GRE− (-512/-492). Resulting PCR fragments were cloned in pGL3 basic plasmid using XhoI and HindIII sites. Plasmids obtained (GRE+ and GRE−) were transfected in NIH3T3 cells (in triplicate) and luciferase activity assayed as above.

Analysis of proximity of VitD receptor (VDR)/RXR or VDR/RAR consensus binding sites to GR.

Analysis was carried out for 66 genes (Table 1) with a previously reported role in lung development (16). One thousand randomly chosen Refseq genes were used as controls. The regulatory framework specified overlap of bindings sites of VDR and either RXR or RAR and that the GR binding site lies within 50bp of the overlapping sites. Nineteen motifs were surveyed in http://www.biobaseinternational.com/pages/index.php?id=transfac and http://thebrain.bwh.harvard.edu/uniprobe/about.php. Clover software (17) was used to search for consensus binding sites of VDR, RXR, RAR, and GR. The Regulatory Framework Enrichment (RFE) Score was computed as the proportion of selected genes that contained the regulatory framework. The Motif score threshold determines how well a motif-sequence match suggests a true binding site. RFE of 6 is the default.

Table 1a Genes involved in human lung development
Full size table

In vivo GC-treatment of rats.

Sprague-Dawley rat pups were injected intraperitoneally with 100 μL of either saline or 3 mg/kg budesonide (Bud, Astra Zeneca, Lund, Sweden) on PN4, PN5, and PN6. After 72 h, (at PN7) the lungs were collected and flash frozen. RNA was isolated with Trizol, and Lgl1 mRNA was measured by qRT-PCR. n = 6 for each group.

Statistical analysis.

All data are presented as mean ± SEM. Statistical significance was determined by two-way analysis of variance. Pair-wise group comparisons were then assessed using Student-Neuman-Keuls test. Significance was defined as p < 0.05.

RESULTS

GC stimulation of Lgl1 in rat lung fibroblasts is temporally regulated.

During maximal septation in rat, circulating levels of plasma corticosteroid are low. Given that lung Lgl1 levels are high during this period (12), we explored the temporal pattern of GC regulation of Lgl1 mRNA. Primary cultures of rat lung fibroblasts from E18 to PN14 were exposed to cortisol (100 nM) for 24 h and Lgl1 mRNA levels were determined by qRT-PCR. Cortisol stimulated Lgl1 mRNA levels at all time points (Fig. 1 A) with maximal stimulation of 100-fold at PN4.

Figure 1
figure 1

A, Temporal regulation of Lgl1 by cortisol in rat lung fibroblasts. Rat lung fibroblasts isolated from E18 to PN14 were treated with 100 nM cortisol for 24 h as described in Materials and Methods section. Lgl1 mRNA levels were quantified by qRT-PCR (n = 6). *p ≤ 0.01 Control (▪) vs GC-treated (□). B, Temporal regulation of Lgl1 by ATRA in rat lung fibroblasts. Rat lung fibroblasts isolated from E18 to PN14 were treated with 10 μM ATRA for 24 and 48 h as described in Materials and Methods section. Lgl1 mRNA levels were quantified by qRT-PCR (n = 4). †p ≤ 0.05 and *p ≤ 0.01 control (▪) vs ATRA-treated 24 h (□) and () ATRA-treated 48 h. C, Effects of coordinate treatment with cortisol and ATRA on Lgl1 mRNA in PN4 rat lung fibroblasts. Rat lung fibroblasts isolated at PN4 were treated with 100 nM cortisol, 10 μM ATRA or both for 24 to 72 h as described in Materials and Methods section. Lgl1 mRNA levels were quantified by qRT-PCR (n = 4). †p ≤ 0.05 and *p ≤ 0.01 control () vs treated 100 nM cortisol (▾), 10 μM ATRA (□), and 100 nM cortisol + 10 μM ATRA (♦). D, Suppression of Lgl1 by VitD in rat lung fibroblasts at PN4, 7, and 14. Rat lung fibroblasts isolated from E20 to PN14 were cultivated for 24 h in media containing charcoal-stripped fetal bovine serum (FBS) and treated with 10 μM VitD for 24 h. Lgl1 mRNA levels were quantified by qRT-PCR (n = 6). *p ≤ 0.01 control (▪) vs VitD treated (□).

Full size image

Retinoids regulate Lgl1 expression in PN lung.

Metabolically active forms of RA in lung fibroblasts increase after birth (7) with a parallel increase in RA receptors from PN8-PN12 (18). To determine whether RAs regulate Lgl1 expression during this period by RARs (activated by ATRA and 9-cis-RA) or RXRs (activated by 9-cis-RA), we initially analyzed the effects of ATRA and 9-cis-RA on Lgl1 mRNA levels in PN7 and PN10 rat lung fibroblasts. At both time points, 9-cis-RA or ATRA had a biphasic effect on Lgl1 mRNA: weakly stimulatory after 24 h but strongly inhibitory after 48 h.

Primary rat lung fibroblasts from E18-PN14 were exposed to 10 μM ATRA for 24 or 48 h. At 24 h, ATRA stimulated Lgl1 mRNA twofold on PN7 and PN10 (Fig. 1B). By contrast, after 48 h exposure, ATRA suppressed Lgl1 levels from E18 to PN10. Maximal inhibition by ATRA after 48 h at PN10 was 15-fold.

The effects on Lgl1 mRNA expression of GC and ATRA are antagonistic.

Given GC stimulation and ATRA suppression of Lgl1 mRNA levels, we asked whether ATRA inhibits cortisol-induced upregulation of Lgl1 mRNA. Primary rat lung fibroblasts isolated on PN4, when GC effects were maximal, were exposed to GC, ATRA, or both for 12–72 h (Fig. 1C). GC stimulated and ATRA suppressed Lgl1 mRNA expression. Although ATRA alone had no effect on Lgl1 mRNA levels after 12–24 h exposures, ATRA suppressed cortisol-induced Lgl1 levels as early as 12 h. ATRA suppression of cortisol-stimulated Lgl1 mRNA levels was stronger after 48 h but did not completely abrogate GC stimulation. Maximal suppression of GC stimulation was from 5.5- to 2-fold.

VitD suppression of Lgl1 in rat lung fibroblasts.

VitD mediates maturation of late gestation alveolar type II cells (9). To explore the effects of VitD on Lgl1 mRNA expression, rat lung fibroblasts isolated from E20-PN14 were exposed to VitD (10 μM) for 24 h, and Lgl1 mRNA levels were determined by qRT-PCR (Fig. 1D). VitD inhibited Lgl1 mRNA levels from PN4 to PN14 (Fig. 1D). Maximal inhibition at PN7 was twofold.

Combined treatment with RA and VitD does not suppress GC induction of Lgl1 mRNA.

In the PN lung, synergistic action between VitD and RA stimulates growth of immature fibroblasts (19). As both of these agents are present in the lung during alveolarization, we determined the effects on Lgl1 mRNA of combined treatment with VitD and 9-cis-RA in the presence and absence of GC. Rat lung fibroblasts isolated at PN4 were exposed to VitD, 9-cis-RA, or VitD plus 9-cis-RA for 24 h. Although RA alone did not impact Lgl1 mRNA expression at 24 h, RA significantly suppressed GC induction of Lgl1 mRNA at this time (Fig. 1C and data not shown). Combined treatment with VitD plus 9-cis-RA led to a modest inhibition of Lgl1 mRNA at PN4 and PN7, which was not additive (Fig. 2 A). VitD, similar to RA, suppressed GC induction of Lgl1 mRNA by ∼50% (Fig. 2A). Most interestingly, combined treatment with VitD and 9-cis-RA had no effect on GC induction of Lgl1 mRNA at all time points (Fig. 2A).

Figure 2
figure 2

A, Effects of coordinate treatments with VitD, cortisol, and 9-cis-RA on Lgl1 expression in rat lung fibroblasts. Rat lung fibroblasts isolated at E20, PN4, and PN7 were treated with 100 nM GC, 10 μM VitD, or combinations of 100 nM GC, 10 μM VitD, and 10 μM 9-cis-RA for 24 h as described in Materials and Methods section. Lgl1 mRNA levels were quantified by qRT-PCR (n = 6). *p ≤ 0.001, control vs treated cells. † p ≤ 0.02, combined treatments vs treatment with GC alone combination. B, Analysis of the 901bp 5′-flanking region of Lgl1. Representation of the Lgl1 5′-flanking promoter region showing binding sites of interest. C, Summary of luciferase assay results. Graph depicting the fold change in Lgl1 promoter activity at 48 h in treated compared with control (untreated) cells at E20 andPN7. *p ≤ 0.003, combined treatments vs treatment with GC alone.

Full size image

GC, ATRA, 9-cis-RA, and VitD regulate Lgl1 promoter activity.

We next explored whether GC, RA, and VitD regulate Lgl1 expression by effects on gene transcription. Consensus recognition sequences for putative binding sites for GR, RXR, RAR, and VDR are illustrated in Figure 2B. Five VDR/RXR heterodimer binding sites surround the single consensus binding site for GR with the most proximal located 38bp downstream. Because a weak stimulatory effect of RA on Lgl1 expression occurred at PN7 and PN10, we carried out these experiments at E20 and PN7.

Fibroblasts isolated at E20 or PN7 were cotransfected with the luciferase expression plasmids pLgl1-Luc and pRSV40-luc (Renilla). Transfected cells were exposed to GC, ATRA, 9-cis-RA, or VitD individually or in combination. Luciferase reporter activities were normalized to Renilla in two sets of experiments. In the first set, we confirmed the effects of GC and RA (ATRA and 9-cis-RA) individually or in combination. In the second set, we evaluated the effects of combined treatments with VitD and 9-cis-RA in the presence or absence of GC.

The Lgl1-reporter plasmid demonstrated significant basal activity (Fig. 3 A). GC, ATRA, and 9-cis-RA effects on Lgl1-promoter-driven luciferase reporter activities were entirely consistent with the observed effects of these agents on Lgl1 mRNA expression levels. In E20 cells, cortisol activated the Lgl1 promoter at all time points. No significant effect on promoter activity was observed after 24 h exposure to ATRA or 9-cis-RA. Exposure to cortisol plus ATRA or 9-cis-RA suppressed GC-induced promoter activity, which was time dependent.

Figure 3
figure 3

A, The effects of GC, ATRA, and 9-cis-RA on Lgl1 promoter activity in E20 rat lung fibroblasts. Rat lung fibroblasts isolated at E20 were cotransfected with pLgl1-luc and pRSV40-luc (Renilla) were treated with 100 nM cortisol, 10 μM ATRA, 10 μM 9-cis-RA, or combinations of GC plus RA or GC for 24–72 h as described in Materials and Methods section. Firefly- and renilla-luciferase activities were measured. Relative promoter activities are reported as a ratio of firefly- to renilla-luciferase activity to normalize for transfection efficiency (n = 3). †p ≤ 0.05 and *p ≤ 0.01 control vs treated. §p. ≤0.02, combined treatments vs treatment with GC alone. B, The effects of GC, ATRA, and 9-cis-RA on Lgl1 promoter activity in PN7 rat lung fibroblasts. Rat lung fibroblasts isolated at PN7 were cotransfected, treated, and assessed as in Fig. 3A (n = 3). †p ≤ 0.05 and *p ≤ 0.01, control vs treated. §p ≤ 6 × 10−4, combined treatments vs treatment with GC alone. Control (•), 100 nM cortisol (), 10 μM 9-cis-RA (▿), 10 μM ATRA (▾), 100 nM cortisol + 10 μM 9-cis-RA (□), and 100 nM cortisol + 10 μM ATRA (▪).

Full size image

At PN7 (Fig. 3B), GC effects on luciferase promoter activity were more pronounced. At 24 h, exposure to ATRA or 9-cis-RA for 24 h was associated with weak activation of Lgl1 expression; combined treatment with GC plus RA-augmented luciferase transcriptional activity above that seen with GC alone, consistent with our findings that 24 h exposure to ATRA at PN7 or PN10, weakly stimulated Lgl1 mRNA expression.

We next assessed the effects of VitD with or without RA on GC induction of Lgl1 promoter activity (Fig. 4 A and B). VitD effects on Lgl1 promoter driven luciferase reporter activity were consistent with those on Lgl1 mRNA expression levels. At E20, VitD alone had no significant effect and VitD plus 9-cis-RA weakly suppressed promoter activity at 72 h only (Fig. 4A, insert). By contrast, VitD suppressed GC induction by ∼50% at all time points. Similar to the case of Lgl1 mRNA, treatment with both VitD and 9-cis-RA failed to suppress GC induction of promoter activity. At PN7, VitD alone modestly inhibited Lgl1 promoter-driven luciferase activity (Fig. 4B, insert). VitD plus 9-cis-RA failed to suppress GC induction of the Lgl1 promoter. The results of Lgl1 promoter-driven luciferase activity are summarized in Figure 2C.

Figure 4
figure 4

A, The effects of GC, 9-cis-RA, and VitD on Lgl1 promoter activity in E20 rat lung fibroblasts. Rat lung fibroblasts isolated at E20 were cotransfected as in Figure 3A and treated with 100 nM cortisol, 10 μM 9-cis-RA, 10 μM VitD, or combinations thereof for 24 to 72 h and assessed as in Figure 3A (n = 3). †p ≤ 0.05 and *p ≤ 0.01, control vs treated. §p ≤ 0.009, combined treatments vs treatment with GC alone. B, The effects of GC, 9-cis-RA, and VitD on Lgl1 promoter activity in PN7 rat lung fibroblasts. Rat lung fibroblasts isolated at PN7 were cotransfected, treated, and assessed as in Fig. 3A (n = 3). †p ≤ 0.05 and *p ≤ 0.01 control vs treated. §p ≤ 6 × 10−4, combined treatments vs treatment with GC alone. Control (•), 100 nM cortisol (), 10 μM VitD (□), 100 nM cortisol + 10 μM VitD (▾), 10 μM 9-cis-RA + 10 μM VitD (▿), and 100 nM cortisol + 10 μM 9-cis-RA + 10 μM VitD (▪).

Full size image

Deletion of the promoter binding site for GR suppresses GC regulation of Lgl1.

As an independent confirmation of direct transcriptional regulation of Lgl1 by GC, we assessed the promoter activity of two constructs in which the binding site for GR was either eliminated (GRE−) or retained (GRE+) (Fig. 5 A). Lgl1 promoter-driven luciferase activity was measured after transfection of constructs into NIH 3T3 cells in the absence or presence of GC (Fig. 5A). In these cells, the Lgl1-reporter plasmid shows a threefold increase in luciferase activity over renilla in the presence of GC (Fig. 5B). When transfected with the GRE− vector, luciferase activity was suppressed 1.5-fold. Residual Lgl1 activity above that of pLgl1-Luc may reflect concomitant loss of inhibitory transcription factor effects such as the three upstream VDR/RXR sites. This interpretation is also supported by the finding that GRE+ shows a greater stimulation than pLgl1-Luc.

Figure 5
figure 5

Deletion of GRE in the promoter of Lgl1 abolishes the response to GC in NIH 3T3 cells. A, Schematic representation of luciferase reporter constructs. B, Effect of GC on Lgl1 promoter activity in NIH3T3 cells. NIH3T3 cells were cotransfected with pRSV40-luc (Renilla) and one of the constructs depicted in A as described in Materials and Methods section. Cells were treated with serum-free media with (□) or without (▪) 100 nM GC, for an additional 48 h. Firefly-luciferase activity, normalized to Renilla, was measured.

Full size image

Enrichment of GC responsive elements adjacent to overlapping VitD and RA binding sites in the promoters of lung development genes.

The individual effects of RA and VitD on GC stimulation of Lgl1 are abrogated when both were present, suggested that steric hindrance of transcription factor binding may occur when overlapping VDR/RXR sites closely abut GR binding sites. Because GC, RA, and VitD regulate multiple aspects of respiratory development, we interrogated the 5′ proximal promoters of 66 genes with known roles in respiratory development (Table 1) for GREs within a fixed distance of overlapping VDR/RXR (or VDR/RAR) compared with a set of 1000 random Refseq genes. The regulatory framework included genes for which the binding site of VDR and RXR (or RAR) are overlapping, and within 50 bp of a GRE (Fig. 6 A). The RFE Score describes the proportion of the 66 selected genes that satisfy the regulatory framework (Fig. 6B). The “motif score threshold” is the likelihood that a motif represents a true binding site. Using a threshold of 6.0, 81% (Table 1, Fig. 6) of the Development Gene Set met the enrichment criteria compared with only 48% in the Random Set. When the threshold is raised to 7, 20% of the Development Set met the enrichment criteria compared with 6% in the Random Set.

Figure 6
figure 6

The promoters of genes important in lung development are enriched for VDR/RXR or VDR/RAR binding sites in close proximity to binding sites for GR. A, Depiction of the regulatory framework defining the gene enrichment set—VDR/RXR (or VDR/RAR) within 50bp of GR binding site. B, The percent of genes in the Gene Enrichment Set using motif matching score thresholds of 6, 6.5, and 7. Selected (▪) and Random (□).

Full size image
Table 1b Genes involved in human lung development (Continued)
Full size table

In vivo GC stimulation of Lgl1 expression in PN rat.

To determine the stimulatory effects of GC on Lgl1 expression in developing lung in vivo, rat pups were injected daily with 3 mg/kg budesonide from PN4-PN6. Lgl1 mRNA expression in whole lung RNA of GC-treated pups at PN7 was 4.5-fold (Fig. 7).

Figure 7
figure 7

GC stimulation of Lgl1 mRNA in postnatal rat lung in vivo. Rat lungs isolated at PN7 after daily treatment with 3 mg/kg budesonide from PN4-PN6 (n = 6). Lgl1 mRNA levels were quantified by qRT-PCR. p ≤ 0.01, control (▪) and treated (□).

Full size image

DISCUSSION

The alveolar stage of lung development represents a particularly vulnerable period in lung maturation. At birth, failure to successfully adapt to air leads to respiratory distress and may be associated with long-term sequelae (BPD). A growing body of evidence points to roles of GC, RA, and VitD in regulating both lung maturation and inflammation in the newborn period. Our studies focused on the regulation of a single gene, Lgl1, while at the same time providing novel information about the complex interactions of GC, RA, and VitD in the regulation of PN lung development.

Information regarding the direct molecular targets of GC and RA in lung development is limited. We were the first to report on the coincident regulation of midkine by GC and RA, providing a potential mechanism for the integration of GC and RA effects on fetal lung development (20). We now report on the coordinate transcriptional regulation of Lgl1 by GC, RA, and VitD in the PN lung. We show that the interaction of GC and RA on Lgl1 expression can be additive or antagonistic and that this interaction is dependent on both duration of exposure and developmental stage. Each of VitD and RA is inhibitory when present alone, but together they have no effect on the induction of Lg1l by GC, a finding with pharmacologic implications.

GC is the most potent known modulator of Lgl1, stimulating expression over 100-fold at PN4. The sensitivity of Lgl1 to GC may account in part for the high levels of Lgl1 maintained in rat lung (12) from PN7 to PN14 when circulating levels of GC are low (21). The biphasic effects of 9-cis-RA and ATRA between PN7 and PN10 are also of interest. During this period of increasing levels of active forms of RA, these agents can have opposing effects, stimulating Lgl1 expression after a short exposure (24 h) and suppressing Lgl1 levels after longer exposure.

PN GC treatment inhibits alveolarization (22). In animal models, some of the detrimental effects of GC on lung development may be ameliorated by simultaneous RA treatment (23,24). However, RA does not normalize Dex-induced changes in lung function (25). The potential role of VitD as a modulator of Lgl1 in PN lung is also of interest. To our knowledge, this is the first report of VitD modulating mesenchymal gene expression in developing lung. We speculate that Lgl1 may be a mesenchymal mediator of VitD effects on epithelial maturation.

GC, RA, and VitD regulation of pulmonary development likely operates through both direct and indirect mechanisms. We believe that this is the first example of coordinate temporal regulation of target gene promoter activity by GC, RA, and VitD during lung development. Interactions between RA and VitD increase combinatorial possibilities for gene regulation of Lgl1 by VDR, RXR, and RAR (26). In the presence of increased concentrations of these hormones, steric hindrance by VDR/RXR heterodimers may compromise the effects of either agent alone.

The enrichment of consensus binding sites for GR, VDR, RXR, and RAR in the promoter region of genes essential to normal human lung development suggests that these agents are likely to impact multiple critical developmental processes in the lung. The proximity of GR and VDR/RXR or VDR/RAR sites offers a context for complex fine tuning of such regulatory mechanisms.

Deficiency of Lgl1 in the newborn period is associated with features of neonatal lung injury in rat (12). Lgl1 is an essential gene, required before lung formation (14). Haplo-insufficiency for Lgl1 in knockout mice is associated with a complex respiratory phenotype including delayed histologic maturation, features of inflammation in the newborn period, and altered lung mechanics at maturity. Hormones that regulate lung development also function as mediators of inflammation. Imbalance of these factors in the newborn period is likely to alter respiratory health. The effects of GC and RA in this period are well established. It is also now becoming clear that VitD is an important modulator of immune/inflammatory response (27,28). VDR expression in the lung microenvironment is required for maximal induction of lung inflammation (27).

Specific interventions that augment pulmonary maturation in a temporally appropriate manner have the potential to ameliorate hyperoxic lung injury. Abnormal lung development is also a susceptibility factor for respiratory diseases that present later in life, including chronic obstructive pulmonary disease (COPD) and asthma (29). As such, our findings may also have implications for long-term respiratory health.