Glucocorticoid receptor isoform–specific regulation of development, circadian rhythm, and inflammation in mice

Abstract

Glucocorticoids are primary stress hormones, and their synthetic derivatives are widely used clinically. The therapeutic efficacy of these steroids is limited by side effects and glucocorticoid resistance. Multiple glucocorticoid receptor (GR) isoforms are produced from a single gene by alternative translation initiation; however, the role individual isoforms play in tissue-specific responses to glucocorticoids is unknown. We have generated knockin mice that exclusively express the most active receptor isoform, GR-C3. GR-C3 knockin mice die at birth due to respiratory distress. Microarray analysis of fibroblasts from wild-type and GR-C3 mice indicated that most genes regulated by GR-C3 were unique to this isoform. Antenatal glucocorticoid administration rescued GR-C3 knockin mice from neonatal death. Dual-energy X-ray absorptiometry revealed no major alterations in body composition for rescued knockin mice. Rescued female, but not male, GR-C3 mice exhibited increased wheel running activity in the light portion of the day. LPS administration induced premature mortality in rescued GR-C3 knockin mice, and gene expression studies revealed a deficiency in the ability of GR-C3 to repress a large cohort of immune and inflammatory response genes. These findings demonstrate that specific GR translational isoforms can influence development, circadian rhythm, and inflammation through the regulation of distinct gene networks.—Oakley, R. H., Ramamoorthy, S., Foley, J. F., Busada, J. T., Lu, N. Z., Cidlowski, J. A. Glucocorticoid receptor isoform–specific regulation of development, circadian rhythm, and inflammation in mice.

Glucocorticoids are steroid hormones necessary for life that regulate many different physiologic processes, including growth, reproduction, intermediary metabolism, immune and inflammatory responses, cognition, behavior, and cardiovascular function (1). Because of their powerful anti-inflammatory and immunosuppressive actions, synthetic glucocorticoids comprise a large class of drugs that are indispensable in treating inflammation, autoimmune disorders, cancer, and organ transplant rejection. The therapeutic benefit of glucocorticoids, however, is limited by severe side effects and by the development of glucocorticoid resistance. The actions of glucocorticoids are mediated by the glucocorticoid receptor (GR), a member of the nuclear receptor superfamily of ligand-dependent transcription factors (2). Understanding the molecular factors that control cell- and tissue-specific responses to glucocorticoids is critical for the development of novel glucocorticoids and treatment regimens with improved therapeutic profiles.

We recently discovered that the single NR3C1 gene gives rise to multiple translational GR isoforms (3). Exon 2 of the human GR gene contains 8 alternative translation initiation sites that give rise to 8 GR subtypes with progressively shorter N-terminal domains (NTDs) (GR-A, GR-B, GR-C1, GR-C2, GR-C3, GR-D1, GR-D2, and GR-D3). Ribosomal leaky scanning and shunting mechanisms generate the N-terminal isoforms, and all 8 start sites are highly conserved across species. The central DNA binding domain and C-terminal ligand binding domain are identical in all the isoforms, and the isoforms display a similar affinity for glucocorticoids and equivalent capacity to interact with DNA after ligand activation. Microarray analysis of osteosarcoma cells stably overexpressing individual GR isoforms revealed that each receptor subtype displays a distinct gene regulatory profile (3, 4). The GR-C3 isoform was found to have the greatest transcriptional activity on GRE-driven promoters, whereas the GR-D3 isoform had the weakest activity. Functionally, expression of the more active GR-C3 correlated with increased sensitivity to glucocorticoid-induced apoptosis, whereas cells expressing the less active GR-D3 were the most resistant (4).

The GR translational isoforms also exhibit unique expression patterns. In many cell types, the GR-A and GR-B isoforms are most abundant. However, GR-C and GR-D are the predominant isoforms in trabecular meshwork cells from the human eye (5), and GR-D is the most abundant isoform in immature dendritic cells (6). In rodents, the levels of the GR-C isoforms are significantly higher in the pancreas and colon, whereas levels of GR-D are highest in spleen and lungs (3). Interestingly, in the human brain, the composition of GR translational isoforms varies during development and during the ageing process (7). Moreover, patients with schizophrenia and bipolar disorder were found to have a selective increase in GR-D isoform expression in certain brain regions (8, 9). However, to what extent the unique gene-regulatory profiles and tissue-specific expression patterns of the GR translational isoforms contribute to the diverse actions of glucocorticoids in vivo remains unknown.

To investigate whether a single GR isoform is sufficient for survival and is responsible for regulating distinct biological processes in vivo, we generated knockin mice that exclusively express the GR-C3 isoform. The GR-C3 knockin mice exhibited developmental defects in the lung and died soon after birth due to respiratory distress. Microarrays performed on wild-type (WT) and GR-C3 mouse embryonic fibroblasts (MEFs) revealed major differences in the transcriptome regulated by glucocorticoids. The GR-C3 knockin mice were rescued from neonatal lethality by antenatal glucocorticoid administration. Rescued GR-C3 knockin mice appeared normal but exhibited alterations in circadian rhythm and sensitivity to LPS-induced inflammation. These data show for the first time in vivo that the GR isoform composition within cells can govern the glucocorticoid-dependent response and that specific GR translational isoforms are crucial for proper development and viability of mice.

MATERIALS AND METHODS

Generation of GR-C3 knockin mice

Three main steps were involved in the generation of the GR-C3 knockin mice. First, embryonic stem (ES) cells in which exon 2 of the Nr3c1 gene was replaced by a Neo cassette flanked by an upstream lox71 site and downstream lox2272 site were generated by standard gene targeting procedures. ES cell clones with homologous recombination were confirmed by Southern blotting. Second, an exchange vector was generated containing exon 2 of the Nr3c1 gene in which the ATG start codons for the translational isoforms GR-A, GR-B, GR-C1, GR-C2, GR-D1, GR-D2, and GR-D3 were mutated to ATC (isoleucine). The GR-C3 start codon was left intact. These point mutations were introduced into exon 2 by site-directed mutagenesis using the QuikChange Kit (Stratagene, San Diego, CA, USA) and confirmed by DNA sequencing. A lox66 site followed by a hygromycin cassette flanked by Frt sites was positioned upstream of the mutated exon 2, and a lox2272 site was positioned downstream of the mutated exon 2 in the exchange vector. The exchange vector and Cre-recombinase were transfected into the targeted ES cells by electroporation, and positive ES cells were selected using G418 (sensitive) and hygromycin (resistant). ES cells undergoing successful cassette exchange were confirmed by PCR. Third, the GR-C3–positive ES cell clone was injected into C57Bl/6 blastocysts to create chimeric mice. Chimeric male mice were bred with C57Bl/6 female mice for germline transmission, and the hygromycin cassette was deleted by crossing mice with Flp-deleter [B6-Tg(ACTB-Flpe) <2Arte >N10] mice. The resulting Nr3c1^WT/C3 mice were maintained on a C57Bl/6 background and intercrossed to produce WT (Nr3c1^WT/WT), heterozygous GR-C3 knockin (Nr3c1^WT/C3), and homozygous GR-C3 knockin (Nr3c1^C3/C3) mice. The mutated translational start codons and the intact GR-C3 codon in exon 2 were confirmed by DNA sequencing of amplified genomic DNA isolated from GR-C3 knockin mice. Mice were genotyped by PCR using the following primers: forward, 5′-CTGTTAGAGCATTTCAGTGTGTAGGGACC-3′; reverse, 5′-GGTACTTGCTAACTGAATCCTGAAAATTCTAATC-3′. All experiments were approved and performed according to the guidelines of the Animal Care and Use Committee at the National Institute of Environmental Health Sciences.

Generation of WT and GR-C3 MEFs

After removing the heart and liver from whole embryonic day (E)12.5 embryos, the remaining tissues were rinsed in PBS and dissociated with sterile tweezers. The minced embryos were trypsinized overnight at 4°C. The suspension from each embryo was plated and cultured with 10% fetal bovine serum supplemented DMEM. Cells were subcultured 2 times to achieve a homogenous population and genotyped by PCR using the previously mentioned primers.

Histology analysis

E18.5 embryos were fixed in 4% paraformaldehyde and processed for paraffin embedding. Sections were stained with hematoxylin and eosin for histopathology analysis, and immunohistochemistry was performed on adrenal glands using an antityrosine hydroxylase antibody (AB152; MilliporeSigma, Burlington, MA, USA).

Immunoblot analysis

Whole cell lysates from WT and GR-C3 MEFs were prepared, and equivalent amounts of protein were electrophoresed and transferred onto nitrocellulose membrane. Blots were incubated with anti-GR (3660; Cell Signaling Technology, Danvers, MA, USA), anti-GR(S211) (4161; Cell Signaling Technology), anti-GR(S226) (97285; Cell Signaling Technology), and/or anti-actin (MAB1501; EMD Millipore) primary antibodies followed by goat anti-rabbit Alexa Fluor 680–conjugated secondary antibody (A21109; Thermo Fisher Scientific, Waltham, MA, USA) and goat anti-mouse IRDye800-conjugated secondary antibody (926-32210; Li-Cor Biosciences, Lincoln, NE, USA). Blots were visualized and quantitated using the Li-Cor Odyssey Imaging System.

Real-time PCR

Total RNA was extracted from cells using the RNeasy Mini Kit (Qiagen, Germantown, MD, USA) and from tissues using Trizol reagent (Thermo Fisher Scientific). Individual mRNA abundance was determined using a TaqMan 1-Step RT-PCR procedure on the 7900HT Sequence Detection System (Thermo Fisher Scientific). All primer/probe sets were from Thermo Fisher Scientific. The relative amount of mRNA for each gene normalized to cyclophilin B was calculated using the ΔΔC_t analysis method.

Microarray analysis

Total RNA was prepared from WT and GR-C3 MEFs treated with vehicle or 100 nM dexamethasone (Dex) for 6 h (n = 3/group). Gene expression analysis was conducted using Whole Mouse Genome oligo arrays (014868; Agilent Technologies, Santa Clara, CA, USA) following the Agilent 1-color microarray-based gene expression analysis protocol. Log₂ transformed data were imported into Partek Genomics Suite for statistical analyses (ANOVA with unadjusted value of P < 0.01). Differentially regulated genes were analyzed using Ingenuity pathway analysis (IPA). Gene enrichment significance values (P < 0.05) for biological functions were determined using Fisher’s exact test. The microarray data have been deposited in the Gene Expression Omnibus (National Center for Biotechnology Information, Bethesda, MD, USA; https://www.ncbi.nlm.nih.gov/geo/) with accession number GSE103675.

Immunocytochemistry

WT and GR-C3 MEFs were treated with vehicle or 100 nM Dex for 1 h and then fixed in paraformaldehyde. Cells were incubated with anti-GR primary antibody (3660S; Cell Signaling Technology) and goat anti-rabbit Alexa Fluor 488 secondary antibody (A11034; Thermo Fisher Scientific). Images were collected using a Zeiss 780 Laser-Scanning Confocal Microscope.

LPS-induced endotoxemia

Male WT and rescued GR-C3 knockin mice (2–3 mo old) were injected intraperitoneally with LPS (Escherichia coli 0111:B4; MilliporeSigma) at a dosage of 10 μg/g body weight and monitored for 48 h. Mice exhibiting moribundity were euthanized and considered to be mortalities.

Nanostring analysis

Total RNA was isolated from spleens removed from male WT and rescued GR-C3 knockin mice injected intraperitoneally with PBS or LPS (10 μg/g body weight) for 3 and 24 h (n = 4 mice/group). The Nanostring nCounter Analysis System (Nanostring Technologies, Seattle, WA, USA) was performed for each RNA sample using the Mouse Immunology Code Set. Data were normalized using the manufacturer’s positive and negative control probes and all housekeeping genes. To identify significant differences in RNA expression, an ANOVA was performed with post hoc Benjamini-Hochberg false discovery rate–corrected significance values (P < 0.05). Data have been deposited in the NCBI Gene Expression Omnibus with accession number GSE103721.

Antenatal Dex treatment

Pregnant heterozygous GR-C3 knockin mice were injected intramuscularly with Dex (2 mg/kg) or saline consecutively on E15.5 and E16.5.

MRI

MRI was performed on E18.5 WT and GR-C3 knockin embryos by Aspect Imaging (Shoham, Israel).

Corticosterone measurements

E18.5 WT and GR-C3 knockin embryos were decapitated, and serum was collected. Corticosterone was measured using an Enzyme Immunoassay Kit (Arbor Assays, Ann Arbor, MI, USA) according to the manufacturer’s instructions.

Dual-energy X-ray absorptiometry

Body composition of anesthetized 3-mo-old WT and GR-C3 knockin mice was determined using the UltraFocus High-Resolution Digital Radiography System (Faxitron Bioptics, Tucson, AZ, USA).

Running wheel analysis

Male and female WT and rescued GR-C3 knockin mice were individually housed with a running wheel apparatus. After an acclimation period, mice were allowed to run freely on the wheel. Rotations were continuously collected for 2 wk and analyzed for dark cycle and light cycle activity using Vital View Software (Starr Life Sciences, Oakmont, PA, USA).

Statistical analysis

A Student’s t test or 1-way ANOVA with Tukey’s post hoc analysis was used to determine the statistical significance between groups. For Kaplan-Meier survival curves, the log-rank test was used.

RESULTS

Generation of GR-C3 knockin mice

To evaluate the impact of single GR translational isoforms on survival and glucocorticoid-dependent responses in vivo, we used a knockin approach to generate mice exclusively expressing only one of the GR isoforms. The GR-C3 translational isoform, which lacks the first 97 aa of the NTD, was chosen for these studies because it exhibits the strongest activity in gene expression and functional assays performed in vitro in comparison to the other translational isoforms (3, 4). Development of the GR-C3 knockin mice involved 3 primary steps (Fig. 1A). In the first step, ES cells were generated by homologous recombination in which exon 2 of the Nr3c1 gene was replaced by a selection marker (Neo^r). In the second step, the targeted ES cells were transfected with Cre recombinase and an exchange vector containing exon 2 of the Nr3c1 gene in which all ATG start codons were mutated to ATC except the start codon for GR-C3, which was left intact (Fig. 1A, B). Cre catalyzed a LoxP site-dependent recombination, resulting in the excision of the Neo^r selection marker and the directional insertion of the mutated Nr3c1 exon 2 and adjacent Hyg^r selection marker. In the third step, GR-C3–positive ES cells were injected into blastocysts derived from C57Bl/6 mice, and chimeras were mated to C57Bl/6 female mice for germline transmission. The Hyg^r cassette was deleted using Flp deleter mice (Fig. 1A). The resulting heterozygous Nr3c1^WT/C3 mice were maintained on a C57Bl/6 background and intercrossed to produce WT (Nr3c1^WT/WT), heterozygous GR-C3 knockin (Nr3c1^WT/C3), and homozygous GR-C3 knockin (Nr3c1^C3/C3) mice (Fig. 1C).

Figure 1.

Generation of the GR-C3 knockin mice. A) Three major steps were involved in the development of GR-C3 knockin mice. In step 1, ES cells were generated by homologous recombination in which exon 2 of the Nr3c1 gene was replaced with a Neo^r selection cassette flanked by lox71 and lox2272 sites. In step 2, the targeted ES cells were transfected with Cre and an exchange vector containing a mutated version of exon 2 of the Nr3c1 gene that permits expression only of the GR-C3 isoform. The exchange vector also contained a Hyg^r selection marker flanked by Frt sites. Flanking the Hyg^r and mutant exon 2 cassettes were lox66 and lox2272 sites that permit Cre-dependent directional insertion of DNA in conjunction with the lox71 and lox2272 sites on the targeted allele. Cre-mediated recombination at the lox71 and lox66 sites generates a double-mutant lox site (DMlox). In step 3, the GR-C3–positive ES cells were injected into blastocysts derived from C57Bl/6 mice. Chimeric male mice were bred with C57Bl/6 female mice for germline transmission, and the resulting GR-C3 knockin progeny were bred with Flp deleter mice to remove the Hyg^r cassette. B) Schematic showing mutated Nr3c1 exon 2. The ATG start codons for GR-A, GR-B, GR-C1, GR-C2, GR-D1, GR-D2, and GR-D3 were mutated to ATC. The ATG start codon for GR-C3 was left intact. The initiator methionine (M) amino acid number is depicted for each of the GR translational isoforms. C) PCR genotyping assay on neonates reveals WT, heterozygous GR-C3 knockin, and homozygous GR-C3 knockin mice.

Early neonatal death of GR-C3 knockin mice

Homozygous GR-C3 knockin mice were generated by intercrossing heterozygous GR-C3 knockin mice. Genotyping of offspring at 3 wk of age revealed a reduced number of homozygous GR-C3 mice in comparison to the expected Mendelian frequency (Table 1). In contrast, no deficits were observed in the number of heterozygous GR-C3 mice at 3 wk of age. To further investigate the timing of the lethality in the homozygous GR-C3 knockin mice, we examined 4 litters from heterozygous intercrosses at birth. Homozygous GR-C3 mice were born alive at the expected Mendelian frequency, but most of them died within 18 h (Table 1). The newborn homozygous GR-C3 mice did not suckle, as revealed by the absence of visible milk in the stomach. Although the WT and homozygous GR-C3 pups had a comparable birth weight, the GR-C3 knockin mice deteriorated rapidly and appeared to be in respiratory distress just before death. Gross examination of E18.5 embryos showed no obvious difference in morphology between WT and homozygous GR-C3 knockin mice (Supplemental Fig. 1). However, histologic analysis of E18.5 GR-C3 embryos revealed delayed lung and liver development (Fig. 2A, B). The homozygous GR-C3 lungs did not fill the thoracic cavity and exhibited diminished saccular airspaces with a concomitant increase in saccular wall thickness, all characteristic of immature lung development. In the liver, the homozygous GR-C3 embryos had an increased number of myeloid hematopoietic cells surrounding the portal veins and central veins, which is characteristic of an earlier gestational age (Fig. 2B) (10). Histologic analysis also revealed alterations in the adrenal glands of the GR-C3 embryos (Fig. 2B, C). GR-C3 adrenal glands from both female and male mice were enlarged and contained discrete islands of tyrosine hydroxylase–positive chromaffin cells diffusely distributed in the medulla. Using MRI, we measured a 1.7-fold increase in the volume of GR-C3 fetal adrenal glands (Supplemental Fig. 2). To assess the function of the fetal adrenal cortex, we measured corticosterone (the major glucocorticoid in rodents) in serum from the E18.5 embryos. Corticosterone was elevated 1.4-fold in female embryos and 1.6-fold in male GR-C3 embryos compared with WT littermates (Fig. 2D).

TABLE 1.

Expected and observed genotypes from intercross of heterozygous GR-C3 knockin mice

Parameter	WT/WT (%)	WT/C3 (%)	C3/C3 (%)
Expected (%)	25	50	25
Birth	4 (12.5)	16 (50)	12 (37.5)
Postnatal d 21	80 (29.9)	172 (64.2)	16 (6.0)

Figure 2.

Homozygous GR-C3 knockin mice die shortly after birth due to respiratory distress. A) Representative images of hematoxylin and eosin–stained sections from E18.5 embryos show impaired lung development in homozygous GR-C3 knockin embryos. Scale bar, 2 mm. B) Representative images of lungs, liver, and adrenal glands from E18.5 embryos. Myeloid hematopoietic cells surrounding portal veins and central veins in the liver of homozygous GR-C3 knockin mice are indicated (black arrows). Original magnification, ×40. C) Representative immunohistochemistry images of tyrosine hydroxylase staining of left adrenal gland from female and male WT and homozygous GR-C3 knockin E18.5 embryos. Scale bars, 150 μm. D) Corticosterone levels in serum from female and male WT and homozygous GR-C3 knockin E18.5 embryos. Data are means ± sem (n = 5–11 mice/group). E) RT-PCR analysis of ENaC mRNA in lungs from neonatal WT and homozygous GR-C3 knockin mice. Data are means ± sem (n = 3 neonatal mice/group). A 1-way ANOVA was performed to determine significance. *P < 0.05, **P < 0.01, ***P < 0.001 for C3/C3 vs. WT/WT.

As the lung matures, the production of surfactant proteins, the proliferation of alveolar type II epithelial cells, and the continued thinning of the lining of the air spaces are observed. Due to the histopathology in the homozygous GR-C3 knockin embryos, we evaluated the expression of the 3 surfactant proteins (SFTPA, SFTPB, and SFTPC) and the amiloride-sensitive sodium channel (ENaC) in lungs from newborn pups. Homozygous GR-C3 mice displayed no significant alteration in the expression of the Sftpa, Sftpb, and Sftpc genes (Supplemental Fig. 3). However, ENaC mRNA expression was significantly reduced (Fig. 2E). Because ENaC expression is required for removal of fluid from the lung at birth (11), decreased levels of ENaC would result in fluid accumulation in the lung and atelectasis. Collectively, these data suggest that respiratory distress is the principle pathology responsible for the neonatal death of the homozygous GR-C3 knockin mice. These results also demonstrate that expression of the GR-C3 isoform alone in mice is not sufficient to support normal mouse development and viability.

GR-C3 MEFs have a unique glucocorticoid-dependent transcriptome

Because the homozygous GR-C3 knockin mice die soon after birth, we generated MEFs from E12.5 embryos to characterize the gene regulatory activity of the GR-C3 isoform. MEFs prepared from the WT embryos predominantly express the GR-A isoform, and MEFs prepared from homozygous GR-C3 embryos exclusively express GR-C3 (Fig. 3A). The total amount of GR-C3 protein was ∼40% of the total amount of GR in WT MEFs (Fig. 3A). Immunocytochemistry was performed to determine the subcellular distribution of GR-C3. GR in the WT MEFs was predominantly cytoplasmic in the absence of hormone and translocated into the nucleus after addition of the synthetic glucocorticoid Dex (Fig. 3B). A similar distribution was observed for GR-C3, which resided in the cytoplasm of cells in the absence of glucocorticoids and translocated into the nucleus after Dex treatment (Fig. 3B). The transcriptional activity of the GR-C3 isoform was evaluated on 2 classic glucocorticoid-responsive genes: glucocorticoid-induced leucine zipper (Gilz) and tristetraprolin (Ttp). As expected, both genes were up-regulated in WT MEFs in response to Dex and to the natural glucocorticoid corticosterone (Fig. 3C, D). Dex treatment of the GR-C3 MEFs also resulted in the induction of GILZ and TTP; however, the magnitude of GILZ induction was significantly less than observed in WT MEFs (Fig. 3C). An even greater reduction in the transcriptional activity of GR-C3 was observed on both the Gilz and Ttp genes in response to corticosterone (Fig. 3D). These data indicate that the GR-C3 isoform expressed in MEFs derived from GR-C3 knockin mice is functional and that GR-C3 has an altered ability to regulate the transcription of certain glucocorticoid-responsive target genes.

Figure 3.

Expression and function of GR-C3 in MEFs from knockin mice. A) Representative immunoblot and quantitation of GR expression levels in WT, heterozygous GR-C3, and homozygous GR-C3 MEFs. Data are means ± sem from 3 independent experiments. A 1-way ANOVA was performed to determine significance. **P < 0.01 for WT/C3 vs. WT/WT and for C3/C3 vs. WT/WT. B) WT and GR-C3 MEFs were treated with vehicle or 100 nM Dex for 1 h and processed for immunocytochemistry. Representative confocal microscopic images show the distribution of WT GR (upper panel) and GR-C3 (lower panel). Cell nuclei are stained with DAPI. Scale bars, 50 μm. C, D) RT-PCR analysis of Gilz and Ttp genes in WT and GR-C3 MEFs that were treated with vehicle, 100 nM Dex, or 500 nM corticosterone (Cort) for 6 h. Data are means ± sem from 3 to 4 independent experiments. **P < 0.01, ***P < 0.001 for Dex vs. vehicle and Cort vs. vehicle; ^##P < 0.01, ^###P < 0.001 for C3 Dex vs. WT Dex and C3 Cort vs. WT Cort (1-way ANOVA).

The glucocorticoid-regulated transcriptome for individual GR isoforms expressed at endogenous levels has not been determined. Therefore, we performed a genome-wide microarray on WT and GR-C3 MEFs treated with vehicle or Dex for 6 h. Glucocorticoid treatment resulted in the regulation of 1352 transcripts in the WT MEFs and 1052 transcripts in GR-C3 MEFs (Fig. 4A). The percentage of genes induced and repressed by Dex was similar in the WT MEFs (37.4% induced and 62.6% repressed) and GR-C3 MEFs (34.8% induced and 65.2% repressed) (Fig. 4A). A comparison of the 2 sets of glucocorticoid-regulated genes revealed only a small amount of overlap (Fig. 4B). Most of the regulated genes were unique to the WT MEFs (1062/1352 genes, 78.6%) or GR-C3 MEFs (762/1052 genes, 72.4%). To validate the unique gene-regulatory actions of GR-C3, we performed RT-PCR on RNA isolated from an independent set of WT and GR-C3 MEFs. In agreement with the microarray results, the adenylate cyclase activating polypeptide 1 receptor 1 gene (Adcyap1r1) was induced by Dex in the GR-C3 MEFs but not in the WT MEFs (Fig. 4C). In addition, the TNF-α–induced protein 3 gene (Tnfaip3) was repressed by glucocorticoids in the GR-C3 MEFs but not in the WT MEFs (Fig. 4C). These results demonstrate that the GR-C3 translational isoform has a gene-regulatory profile distinct from GR in WT MEFs and that the cellular complement of receptor isoforms can lead to a dramatic change in the glucocorticoid-regulated transcriptome.

Figure 4.

GR-C3 has a unique glucocorticoid-regulated transcriptome. Microarray analysis was performed on RNA isolated from WT and GR-C3 MEFs that were treated with vehicle or Dex for 6 h. A) Total number of genes in WT and GR-C3 MEFs that were regulated by Dex (ANOVA, P < 0.01) and total number of genes exhibiting increased or decreased expression in response to Dex treatment. B) Comparison analysis of the glucocorticoid-regulated genes in WT and GR-C3 MEFs. The Venn diagram shows the number of genes uniquely regulated by glucocorticoids in WT MEFs, commonly regulated by glucocorticoids in both WT and GR-C3 MEFs, and uniquely regulated by glucocorticoids in GR-C3 MEFs. C) Total RNA was isolated from an independent set of WT and GR-C3 MEFs that had been treated with vehicle or Dex for 6 h. RT-PCR was performed to measure the expression of ADCYAP1R1 and TNFAIP3 mRNA. Data are means ± sem for 3 independent experiments. **P < 0.01 for Dex vs. control mice (Student’s t test).

Literature-based IPA was used to gain insight into the diseases and biological functions most significantly associated with the genes regulated by glucocorticoids in the WT and GR-C3 MEFs. Among the top 10 gene ontology annotations, only 2 were shared across both sets of regulated genes: Organismal Injury and Abnormalities and Cancer (Supplemental Fig. 4). One of the annotations displaying the greatest divergence between WT and GR-C3 gene sets was Respiratory Disease. Respiratory Disease was poorly associated with the glucocorticoid-regulated genes in WT MEFs (rank, 72 out of 74 significant annotations) but was strongly associated with the glucocorticoid-regulated genes in GR-C3 MEFs (rank, 11 out of 77 significant annotations). A total of 22 genes associated with Respiratory Disease were altered in the WT MEFs, whereas 205 genes associated with Respiratory Disease were altered in the GR-C3 MEFs (Supplemental Fig. 5). This finding is particularly noteworthy because the GR-C3 mice die shortly after birth due to deficient lung maturation and respiratory distress. We also used IPA to evaluate the canonical signaling pathways that were most significantly associated with the glucocorticoid-regulated genes in the WT and GR-C3 MEFs (Table 2). No overlap was observed among the top 10 signaling pathways for the WT and GR-C3 MEFs. Consistent with the well-known actions of glucocorticoids to modulate the immune response, 4 of the top 10 signaling pathways for the WT MEFs have important roles in inflammation and immunity: Role of JAK1, JAK2, and TYK2 in IFN Signaling; IFN Signaling; Role of IL-17F in Allergic Inflammatory Airway Diseases; and STAT3 Pathway. These and other immune-related signaling pathways were not strongly associated with the glucocorticoid-regulated genes in the GR-C3 MEFs. Glucocorticoids are also known to regulate the core circadian clock genes and influence circadian rhythm (12). Circadian Rhythm Signaling was the top ranked pathway in the GR-C3 MEFs but was not significantly associated with the regulated genes in the WT MEFs (Table 2). Collectively, these gene enrichment predictions suggest that the GR translational isoforms affect distinct signaling pathways and biological processes by regulating unique sets of genes.

TABLE 2.

Top 10 canonical pathways most significantly associated with the glucocorticoid regulated genes in the WT and GR-C3 MEFs

MEFs	Rank	Canonical pathways	−Log (P)	Z score	Genes
WT	1	Huntington’s disease signaling	2.71	1.387	30
	2	PKA signaling	2.69	−0.539	43
	3	Role of JAK1, JAK2, and TYK2 in IFN signaling	2.47	0	6
	4	Polyamine regulation in colon cancer	2.36	N/A	6
	5	IFN signaling	2.26	−0.378	7
	6	Glioblastoma multiforme signaling	1.94	0.5	20
	7	Type I diabetes mellitus signaling	1.84	−0.905	14
	8	Role of IL-17F in allergic inflammatory airway diseases	1.75	−0.816	7
	9	STAT3 pathway	1.7	0.302	11
	10	HIPPO signaling	1.68	−0.333	12
C3	1	Circadian rhythm signaling	3.46	N/A	8
	2	Guanosine nucleotides degradation III	2.43	N/A	4
	3	Urate biosynthesis/inosine 5′-phosphate degradation	2.29	N/A	4
	4	Adenosine nucleotides degradation II	2.17	N/A	4
	5	Superpathway of melatonin degradation	2.09	N/A	8
	6	Prostanoid biosynthesis	1.91	N/A	3
	7	Coagulation system	1.87	−1.633	6
	8	Purine nucleotides degradation II (aerobic)	1.85	N/A	4
	9	Role of tissue factor in cancer	1.81	N/A	13
	10	Melatonin degradation II	1.74	N/A	2

Microarray analysis was performed on RNA isolated from WT and GR-C3 MEFs that were treated with vehicle or Dex for 6 h. Genes significantly regulated by Dex (ANOVA, P < 0.01) in WT and GR-C3 MEFs were analyzed by IPA software for association with canonical signaling pathways. N/A, not applicable.

GR-C3 knockin mice are rescued by antenatal glucocorticoid administration

Homozygous GR-C3 knockin mice exhibit abnormal lung development and die soon after birth due to respiratory distress. Synthetic glucocorticoids are known to accelerate pulmonary maturation and are used clinically to prevent respiratory distress syndrome in premature infants (13). Therefore, we attempted to rescue the GR-C3 mice with antenatal Dex administration on E15.5 and E16.5. All the GR-C3 mice born to mothers that received Dex survived (Table 3). Antenatal glucocorticoid administration resulted in the lungs nearly filling the thoracic cavity and in the complete restoration of saccular airspace and a reduction in the thickness of the saccular walls (Fig. 5A, B). ENaC is known to be a glucocorticoid-regulated gene (11), and antenatal Dex administration increased ENaC expression to approximately normal levels in the GR-C3 knockin mice (Fig. 5C). After birth, the rescued GR-C3 mice exhibited normal postnatal development. These findings show that the GR-C3 isoform has the capacity to support normal lung development independent of the other translational isoforms when activated in utero by pharmacological levels of glucocorticoids.

TABLE 3.

Expected and observed genotypes from intercross of heterozygous GR-C3 knockin mice in which pregnant dams received antenatal Dex

Parameter	WT/WT (%)	WT/C3 (%)	C3/C3 (%)
Expected (%)	25	50	25
Birth	11 (20.4)	32 (59.2)	11 (20.4)
Postnatal d 21	10 (19.2)	31 (59.6)	11 (21.1)

Figure 5.

GR-C3 knockin mice are rescued from neonatal death by antenatal glucocorticoids. A) Representative images of hematoxylin and eosin–stained sections from E18.5 embryos show improved lung development in homozygous GR-C3 knockin embryos treated with antenatal Dex. Scale bars, 2 mm. B) Representative images of lungs from E18.5 embryos showing improved morphology consistent with mature lung development in homozygous GR-C3 knockin embryos treated with antenatal Dex. Original magnification, ×40. C) RT-PCR analysis of ENaC mRNA in lungs from neonatal mice in which the pregnant female mouse was treated with PBS or Dex. Data are means ± sem (n = 3 neonatal mice/group). **P < 0.01 for C3/C3 Antenatal Dex vs. C3/C3 PBS (1-way ANOVA).

Body composition of rescued GR-C3 knockin mice

The rescued GR-C3 knockin mice provide a novel model system for investigating whether individual GR translation isoforms can regulate distinct biological processes in vivo. Because glucocorticoids have important roles in metabolism, we examined the body composition of rescued GR-C3 knockin mice that were 3 mo old using dual-energy X-ray absorptiometry. Female and male GR-C3 mice did not differ from sex-matched WT mice in any of the measured parameters (total body weight, lean body weight, body fat weight, and body fat percentage) (Supplemental Fig. 6A–D). In addition, differences in total body weight and lean body weight measured across male and female WT mice were preserved in the GR-C3 mice. The male GR-C3 mice were found to have increased body fat weight compared with female GR-C3 mice, but no difference was observed in body fat percentage. These findings suggest that the GR-C3 isoform has minimal effects on body composition and is sufficient to support normal growth and metabolism in rescued mice maintained on a standard chow diet.

Rescued female GR-C3 knockin mice have altered circadian rhythm

The glucocorticoid-regulated genes in the GR-C3 MEFs exhibited a strong association with circadian rhythm (Table 2); therefore, we next examined circadian behavior in rescued GR-C3 knockin mice. Mice sleep during the day and are active at night. The ability to consolidate activity to either the light or dark portion of the day is referred to as circadian photoentrainment and requires light input to the circadian clock (12). Activity of mice at night is particularly robust in the presence of a running wheel, and measuring this behavior is a minimally invasive method that can be used to evaluate the functionality of the circadian system. Actograms from running wheel experiments performed on female and male WT mice show that the vast majority of the wheel running activity was confined to the dark cycle and minimal activity was observed in the light cycle. Thus, circadian photoentrainment was maintained in WT mice (Fig. 6A–C). For female GR-C3 knockin mice, wheel running activity was significantly increased in the light phase, and a decrease was observed in the dark phase that showed a strong trend toward significance (P = 0.0642) (Fig. 6A, C). Wheel running activity was not significantly altered in male GR-C3 mice (Fig. 6B, C). These results demonstrate that female GR-C3 knockin mice have alterations in their circadian photoentrainment, suggesting that the GR-C3 isoform can affect circadian function in a sex-specific manner.

Figure 6.

Rescued female GR-C3 knockin mice exhibit alterations in circadian function. A) Representative running wheel actograms of female WT and rescued GR-C3 knockin mice. Time periods of light (shaded) and dark (unshaded) are indicated on the x axis, and running wheel rotations are depicted on the y axis. B) Representative running wheel actograms of male WT and rescued GR-C3 knockin mice. Time periods of light (shaded) and dark (unshaded) are indicated on the x axis, and running wheel rotations are depicted on the y axis. C) Mean daily running wheel activity during light and dark cycles for female and male WT and rescued GR-C3 knockin mice. Data are means ± sem (n = 10 mice/group). *P < 0.05 for C3/C3 female vs. WT/WT female (repeated-measures ANOVA).

Rescued GR-C3 knockin mice have increased sensitivity to LPS-induced endotoxemia

Compared with WT MEFs, immune and inflammatory response signaling pathways were poorly associated with the glucocorticoid-regulated transcriptome in the GR-C3 MEFs (Table 2). Therefore, we investigated whether the rescued GR-C3 knockin mice were deficient in their ability to control inflammation when challenged with LPS. Endotoxemia produced by LPS elicits a strong inflammatory response that requires the action of endogenous glucocorticoids for its resolution (14). Male WT and rescued GR-C3 knockin mice were injected with a sublethal dose of LPS and monitored for survival for 48 h. WT mice treated with LPS survived the 48-h study period (Fig. 7A). In marked contrast, LPS treatment induced premature death in GR-C3 knockin mice. This result suggests that the GR-C3 isoform alone is insufficient to mount an effective anti-inflammatory response in mice challenged with a systemic infection.

Figure 7.

Rescued GR-C3 knockin mice are deficient in their ability to resolve LPS-induced inflammation. A) WT and rescued GR-C3 knockin mice were treated with LPS (10 μg/g body weight) and monitored for moribundity over the next 48 h. Shown are Kaplan-Meier survival curves for WT (n = 10) and rescued GR-C3 knockin mice (n = 10). B) Nanostring analysis was performed on RNA isolated from spleens from WT and rescued GR-C3 knockin mice treated for 3 and 24 h with LPS (10 μg/g body weight). Shown is a heat map cluster of significantly regulated genes (ANOVA, P < 0.05). C) Comparison analysis of the total number of genes significantly regulated in the WT and GR-C3 knockin spleens after 3-h (left panel) and 24-h (right panel) treatment with LPS. The Venn diagram shows the number of genes uniquely regulated in the WT spleens, commonly regulated in both the WT and GR-C3 knockin spleens, and uniquely regulated in the GR-C3 knockin spleens at the 3- and 24-h time points. D) Nanostring RNA counts for Il1b and Tnf genes in the WT and GR-C3 knockin spleens. E) Nanostring RNA counts for Notch1, Irf1, Il12rb2, Stat2, Tnfrsf9, and Litaf genes in the WT and GR-C3 knockin spleens. Data shown in D and E are the raw RNA counts normalized to the housekeeping genes and represent the means ± sem (n = 3–4 spleens/group). *P < 0.05, **P < 0.01, ***P < 0.001 for LPS vs. con; ^##P < 0.01, ^###P < 0.001 for GR-C3 LPS-24 vs. WT LPS-24 (1-way ANOVA followed by Tukey’s post hoc test).

The major mechanism by which glucocorticoids protect against LPS-induced inflammation is at the level of gene transcription (15). Therefore, we evaluated gene expression in spleens from WT and rescued GR-C3 knockin mice that had been treated with vehicle (PBS) or LPS for 3 and 24 h. For this experiment, we used nanostring technology to measure the expression of 547 immunology-related genes. A heat map of the significantly regulated genes in WT and GR-C3 spleens shows a very similar profile after the 3-h treatment with LPS (Fig. 7B). A total of 262 and 291 genes were altered by LPS in the WT and GR-C3 knockin spleens, respectively, and the majority of these genes (239/314 genes, 76.1%) were commonly regulated (Fig. 7C). A considerably different profile of gene regulation emerged after 24 h of LPS exposure (Fig. 7B). At this later time point, a total of 218 and 294 genes were regulated in the spleens from WT and GR-C3 knockin mice, respectively (Fig. 7C). However, less than half of these genes (160/352, 45.5%) were commonly regulated, as evidenced by a marked expansion in the number of genes uniquely altered in the WT and GR-C3 knockin spleens.

Many of the inflammatory genes commonly induced by LPS at the early 3-h time point in both the WT and GR-C3 knockin spleens returned to baseline expression levels after 24 h of LPS treatment. The proinflammatory cytokines IL1B and TNF-α are representative members of this group. Both genes were strongly up-regulated after 3 h of LPS treatment, and both genes declined back to baseline by 24 h (Fig. 7D). Glucocorticoids have been shown to inhibit Il1b and Tnf gene expression (16, 17). Therefore, the reduction in these inflammatory genes at the later 24-h time point is likely mediated by endogenous glucocorticoids released in response to LPS. GR-C3 appears equivalent to WT GR in its ability to suppress the expression of these inflammatory genes because they both return to basal levels in the WT and GR-C3 spleens. Some genes increased by LPS at the early 3-h time point, however, did not return to baseline levels in spleens from GR-C3 knockin mice. The immune and inflammatory response genes Notch1, Irf1, Il12rb2, Stat2, Tnfrsf9, and Litaf are representative members of this group. All 6 genes were strongly induced by 3 h of LPS treatment in both WT and GR-C3 spleens (Fig. 7E). However, although these genes returned to baseline levels in the WT spleens treated for 24 h with LPS, they remained significantly elevated in the GR-C3 knockin spleens. Many of these genes and associated signaling pathways are known to be repressed by glucocorticoids (18–21). Therefore, the inability of GR-C3 to fully counteract the LPS-mediated rise in their expression suggests that additional GR translational isoforms are required for effective inhibition and resolution of inflammation. Collectively, these data suggest that the rescued GR-C3 knockin mice exhibited enhanced sensitivity to LPS-induced endotoxemia due to a deficiency in the ability of GR-C3 to repress a large cohort of immune and inflammatory response genes.

DISCUSSION

We have generated knockin mice that exclusively express the GR-C3 translational isoform. These mice died shortly after birth due to respiratory distress. Microarray analysis of MEFs prepared from WT and GR-C3 embryos demonstrated that most of the genes regulated by GR-C3 in response to Dex differed from those regulated by WT GR. The genes regulated by GR-C3 were strongly associated with circadian rhythm signaling pathways but weakly associated with immune and inflammatory response signaling pathways. Antenatal glucocorticoid therapy completely rescued the GR-C3 mice from neonatal death. Rescued GR-C3 mice showed no major differences in body composition but exhibited alterations in circadian behavior in a sex-specific manner. In addition, rescued GR-C3 knockin mice were deficient in their ability to counteract systemic inflammation induced by LPS. These findings demonstrate that specific GR translational isoforms can influence unique biological functions through the regulation of distinct sets of genes.

The GR-C3 knockin embryos exhibited several developmental defects. The adrenal glands were enlarged, and serum corticosterone levels were elevated. A rise in circulating glucocorticoids would be expected to activate GR signaling in the brain to inhibit the hypothalamic-pituitary-adrenal axis in a negative feedback manner. GR-C3 may be impaired in its ability to regulate genes involved in this negative feedback function, resulting in elevated hypothalamic-pituitary-adrenal axis activity and the consequent increase in adrenal size. The GR-C3 adrenals also exhibited an alteration in chromaffin cell distribution. These cells did not coalesce into a central mass but rather were observed in discrete islands within the medulla. Whereas studies in GR null mice have concluded that GR is largely unnecessary for chromaffin cell development (22), our findings in the GR-C3 knockin adrenals suggest that chromaffin cell development is subject to modulation by GR isoform–specific signaling.

Lungs from GR-C3 knockin embryos were immature, resulting in respiratory distress and neonatal death. Knockout studies have demonstrated a critical role for GR signaling in fetal lung maturation (23). Results from our knockin mice indicate the GR-C3 isoform alone is insufficient to support fetal lung maturation, even in the presence of elevated levels of corticosterone. After antenatal Dex administration, GR-C3 gained the ability to mediate proper maturation of the lung, resulting in the complete rescue of GR-C3 mice from neonatal death. These findings suggest that corticosterone-activated GR-C3 and Dex-activated GR-C3 differ in their ability to regulate key target genes, such as ENaC, that are necessary for the maturation of the lung. Once past this critical developmental window for lung maturation, the GR-C3 isoform alone appears to be sufficient to support normal postnatal development in mice in the presence of endogenous glucocorticoids.

The ability of GR-C3 to regulate genes distinct from WT GR suggests that sequences within the NTD play a critical role determining the gene regulatory profile of GR. Residues 98–115 of the NTD were recently shown to be important for the heightened transcriptional activity of GR-C3 via recruitment of specific coregulators (24). The NTD of GR is also subject to a variety of posttranslational modifications. GR is phosphorylated on multiple serine/threonine residues located in the NTD (25), and the pattern and extent of receptor phosphorylation has a major influence on the GR transcriptome via changes in cofactor recruitment and chromatin occupancy (26, 27). The GR-C3 isoform lacks 1 putative phosphorylation site (Ser-46), and its shortened NTD may influence the efficiency of phosphorylation at remaining sites, such as Ser-220 and Ser-234. Indeed, compared with WT GR, we found a small but significant increase in the Dex-dependent phosphorylation of GR-C3 at Ser-220 but not at Ser-234 (Supplemental Fig. 7), suggesting that alterations in the phosphorylation status of GR-C3 may contribute to its unique gene-regulatory profile.

Gene enrichment predictions from the glucocorticoid-regulated transcriptome in WT and GR-C3 MEFs suggested that the rescued GR-C3 knockin mice may differ from WT mice in their circadian function. Using voluntary wheel running activity as a measure of circadian behavior, we found that female GR-C3 knockin mice were more active than female WT mice in the light portion of the day. Male GR-C3 knockin mice exhibited a normal circadian behavior, indicating that the effect on circadian function was sexually dimorphic. The circadian clock and stress response system are closely interconnected (28). Glucocorticoids are released in a circadian rhythm and in turn regulate clock genes and synchronize peripheral and central circadian oscillators. One of the circadian rhythm genes identified in our microarray to be uniquely regulated by GR-C3 is Adcyap1r1. The ligand for this receptor, pituitary adenylate cyclase-activating polypeptide, is involved in phase resetting in response to light, and its dysregulation has been associated with posttraumatic stress disorder selectively in female mice (29, 30). Studies in animal models have suggested a link between circadian dysregulation and vulnerability to posttraumatic stress disorder (28). Cellular alterations in the relative levels of the GR translational isoforms to favor more GR-C3 expression may therefore underlie perturbations in circadian function and contribute to pathologies with known circadian and stress components, including metabolic syndrome, cardiovascular disease, psychiatric disorders, inflammatory/autoimmune disease, and sleep disorders (31). The mechanisms underlying the sex-specific effects of GR-C3 on circadian behavior are unclear but may involve cross-talk with estrogen, which has been shown to alter GR recruitment to target genes (32).

Gene ontology analysis of the glucocorticoid-regulated genes in the WT and GR-C3 MEFs also suggested a deficiency in the ability of GR-C3 to antagonize immune and inflammatory response signaling pathways. This appears to be the case because the rescued GR-C3 knockin mice were more sensitive to LPS-induced endotoxemia than WT mice and died prematurely. At a molecular level, GR-C3 effectively repressed many of the proinflammatory genes that were induced by LPS. However, a large cohort of other immune and inflammatory response genes increased by LPS were not efficiently suppressed by GR-C3. Additional GR translational isoforms and their unique gene regulatory profiles appear necessary for glucocorticoids to effectively combat a strong systemic inflammatory response. These findings may have significant implications for the treatment of sepsis. Sepsis is an acute systemic inflammatory disease with high morbidity and mortality rates (14). Glucocorticoids are widely used for treating patients with sepsis, but controversy exists over their potential benefits and harms (33). A major limitation to the effective use of glucocorticoids for sepsis is the development of glucocorticoid resistance in certain patient groups. Results from our GR-C3 knockin mice suggest that patients with elevated expression of GR-C3 will exhibit an impaired response to glucocorticoids that is insufficient to resolve the systemic inflammation. Critical goals of future studies will be to measure GR-C3 isoform levels in patients with sepsis and other inflammatory diseases and to determine what factors control GR-C3 expression.

In summary, we have generated knockin mice that exclusively express a single GR translational isoform. The GR-C3 knockin mice exhibit a distinct gene regulatory profile and unique developmental and physiologic outcomes in response to glucocorticoids. These findings suggest that the complement of GR translational isoforms will be a major factor determining how glucocorticoids influence the biological function and activity of specific cells and tissues. The identification of isoform-specific signaling pathways may provide insight not only into the pathogenesis of multiple diseases but also into the molecular mechanisms underlying changes in glucocorticoid responsiveness. Moreover, a determination of the GR isoform repertoire in diseased tissues may aid in the development of treatment strategies that maximize the therapeutic benefit of glucocorticoid therapy.

Supplementary Material

This article includes supplemental data. Please visit http://www.fasebj.org to obtain this information.

Glossary

Dex

dexamethasone

ENaC

amiloride-sensitive sodium channel

ES

embryonic stem

GILZ

glucocorticoid-induced leucine zipper

GR

glucocorticoid receptor

IPA

ingenuity pathway analysis

MEF

mouse embryonic fibroblast

NTD

N-terminal domain

TTP

tristetraprolin

WT

wild type

AUTHOR CONTRIBUTIONS

R. H. Oakley, S. Ramamoorthy, J. F. Foley, J. T. Busada, N. Z. Lu, and J. A. Cidlowski designed the research; R. H. Oakley, S. Ramamoorthy, J. F. Foley, and J. T. Busada performed the research; R. H. Oakley, S. Ramamoorthy, J. F. Foley, J. T. Busada, N. Z. Lu, and J. A. Cidlowski analyzed the data; and R. H. Oakley, S. Ramamoorthy, and J. A. Cidlowski wrote the paper.