Inflammatory Cytokines Regulate Glycoprotein Subunit β5 of Thyrostimulin through Nuclear Factor-κB

Thyrostimulin is a heterodimeric hormone comprised of two glycoprotein hormone subunits, namely glycoprotein hormone subunit α2 and glycoprotein hormone subunit β5 (GPB5). Immunological studies have revealed that both subunits colocalize in human pituitary corticotroph cells. Although recombinant thyrostimulin protein selectively activates the TSH receptor and has thyrotropic activity in rats, its biological functions have not been clarified. To explore the physiological regulators for the GPB5, the 5′-flanking region of the GPB5 coding sequence up to 3-kb upstream was analyzed by luciferase reporter assays. We found that nuclear factor-κB (NF-κB) markedly activated GPB5 transcription. Disruption of the putative NF-κB-binding motifs in the GPB5 5′-flanking region silenced the GPB5 activation by p65. Chromatin immunoprecipitation assays revealed that recombinant p65 bound to the predicted NF-κB-binding sites. Because NF-κB is known to associate with acute phase inflammatory cytokines, we examined whether TNFα or IL-1β could regulate GPB5. Both these cytokines activated GPB5 transcription by 2- to 3-fold, and their effects were abolished by the addition of MG132, a NF-κB inhibitor. Our results suggest that inflammatory cytokines positively regulate thyrostimulin through NF-κB activation.

Two novel glycoprotein hormone subunits were recently identified by database mining, and designated glycoprotein hormone subunit α2 (GPA2) and glycoprotein subunit β5 (GPB5) from their structural similarities to other glycoprotein hormone subunits (1). The amino acid sequence of GPA2 exhibits 35% identity to the common glycoprotein α-subunit (GSUα), whereas GPB5 shows approximately 30% identity to each of the glycoprotein β-subunits of TSH, LH, FSH, and chorionic gonadotropin (TSHβ, LHβ, FSHβ, and CGβ, respectively). The deduced amino acid sequences of GPA2 and GPB5 have conserved cysteine residues that form cysteine-knot motifs, similar to the case for other glycoprotein hormone subunits. Recombinant GPA2 and GPB5 form a heterodimeric hormone, designated thyrostimulin, which activates the TSH receptor (TSHR) both in vivo and in vitro (2, 3, 4). The carbohydrate chains on GPA2 are crucial for binding to TSHR (5). When administered to the peripheral circulation in rats, recombinant thyrostimulin stimulates T₄ secretion from the thyroid.

We previously analyzed the sites of GPA2 and GPB5 expression in rat and human tissues using quantitative PCR, and found discrepancies between the two subunits (2). The GPA2 levels were more than 100-fold higher than the GPB5 levels in all the tissues examined, and the highest expressions were observed in the pituitary and pancreas. GPA2 expression was also found in extrathyroidal tissues, including the retina, skin, and testis. Recently, both GPA2 and GPB5 were found to colocalize in the corticotroph of the human anterior pituitary (3), suggesting that thyrostimulin is a novel member of the anterior pituitary hormones.

Genetic intervention studies have revealed limited information regarding the biological functions of thyrostimulin. Overexpression of GPB5 caused thyrotrophic effects of thyrostimulin with elevated serum T₄ levels, proptosis, and reduced body weights in mice (3, 6). On the other hand, GPB5-knockout mice showed no overt phenotypes. Regarding GPA2, neither overexpression nor disruption of the gene caused any significant phenotypes. These data indicate that thyrostimulin is not critical for embryogenesis, development, or reproduction, and suggest the presence of alternative bioactive reagents. The physiological and pathophysiological roles of thyrostimulin remain to be clarified.

The regulation of human GPB5 gene expression is poorly understood. To explore the biological functions of thyrostimulin by studying its transcriptional regulation, we examined the promoter activity of the 5′-flanking region of the human GPB5 gene, and focused on its putative binding sites for the transcription factor nuclear factor-κB (NF-κB). In the present study, we investigated the regulation of human GPB5 gene transcription by cytokines via transfection experiments in a mouse corticotroph cell line, AtT20. By deletion analyses, site-directed mutagenesis, and chromatin immunoprecipitation (ChIP) assays, we found that NF-κB positively regulates GPB5 expression. In addition, the acute phase inflammatory cytokines IL-1β and TNFα regulate GPB5 through NF-κB activation.

Materials and Methods

Cell culture

AtT20 mouse pituitary corticotroph cell line cells were maintained in DMEM (Invitrogen Corp., Carlsbad, CA) supplemented with 10% fetal bovine serum (Invitrogen) and antibiotics (50 U/ml penicillin and 50 mg/ml streptomycin; Invitrogen) under a 5% CO₂/95% air atmosphere at 37 C. The culture media were exchanged for fresh media three times a week, and the cells were subcultured once a week. For some experiments, cells were exposed to TNFα, IL-1β, or MG132, a NF-κB inhibitor. TNFα and IL-1β were purchased from PeproTech EC Ltd. (London, UK), and MG132 was purchased from Sigma-Aldrich Corp. (St. Louis, MO).

Plasmid construction and site-directed mutagenesis

Based on the published sequence of chromosome 14, −3, −2, and −1-kb fragments (relative to the first nucleotide of the ATG start codon) upstream of the GPB5 gene were obtained from human genomic DNA by PCR cloning using pcDNA3.1/V5-His-TOPO (Invitrogen). Each of the fragments was subcloned into a reporter plasmid, pGL3-Basic (Promega Corp., Madison, WI), for fusion to an upstream position of the luciferase gene (−3, −2, −1k-GPB5-Luc). The −1k-GPB5-Luc construct was then used as a template to generate deletion mutant reporter constructs of various lengths (−656/−93 fragment: −656-GPB5-Luc; −412/−93 fragment: −412-GPB5-Luc; −181/−93 fragment: −181-GPB5-Luc). The putative NF-κB-binding sites in the −656-GPB5-Luc construct were disrupted using a QuikChange XL Site-Directed Mutagenesis Kit (Stratagene, Palo Alto, CA) according to the manufacturer’s instructions. The primers used for these constructs are listed in Table 1. The entire length of the PCR-cloned insert and mutant plasmids were confirmed by DNA sequencing.

TABLE 1.

Primers used for deletion constructs, substitution mutants, and ChIP assays

Name	Sequence
Primers for deletion constructs
−3k (sense)	5′-GCCTAAGATCTACCCAGGATGAAGCTGATG-3′
−2k (sense)	5′-GTGGTAAGATCTAATTGATGGACTCACTGGG-3′
−1K (sense)	5′-GCTAGATCTTTCTATTCCATGGCTTTGATAA-3′
−656 (sense)	5′-GTAGATGCTGTTGCAGTGGAAGC-3′
−412 (sense)	5′-CCCAAATACCCTGGACCTGAATA-3′
−182 (sense)	5′-CAGTTACAGGAATCTCTGGATGCC-3′
Common antisense	5′-GACCTAAGCTTTGCTCCTAGCATTTCAGGGA-3′
Primers for site-directed mutagenesis
P1m (sense)	5′-GTAACGATGCATATAAGTAAGACTATTATTAAT-3′
P1m (antisense)	5′-ATTAATAATAGTCTTACTTATATGCATCGTTAC-3′
P2m (sense)	5′-GGCATCTAGAGAAGTCCAGGGCATCCACACT-3′
P2m (antisense)	5′-AGTGTGGATGCCCTGGACTTCTCTAGATGCC-3′
P3m (sense)	5′-GATGCCTGTGTTGGAGTTTGTTGACATGTACAATTTCTGG-GCTCATTT-3′
P3m (antisense)	5′-AAATGAGCCCAGAAATTGTACATGTCAACAAACTCCAAC-ACAGGCATC-3′
Primers for ChIP assay
ChIP-B5 (sense)	5′-CCATGTTTAGTGGTGAGTTGGTTTAG-3′
ChIP-B5 (antisense)	5′-GGAAAATGAGCCCAGAAATTGTA-3′

A full-length p65 cDNA was purchased from CLONTECH (Palo Alto, CA), and subcloned into the pRcRSV mammalian cell expression vector (Invitrogen) to obtain p65/pRcRSV. For immunoprecipitation experiments, the FLAG-epitope was fused to the N terminal of p65 by oligo-directed insertion and subcloned into the pcDNA3.1/V5-His-TOPO plasmid. All the constructs were confirmed by a DNA sequencing service (Macrogen, Seoul, Korea).

Transient transfection and reporter gene assays

At 24 h before transfection, AtT20 cells were seeded in 24-well plates at approximately 70% confluence. The cells were then transfected with each of the reporter constructs containing a GPB5 upstream region (1 μg) or pGL3-Basic (1 μg). Each of the reporter constructs was cotransfected with the p65/pRcRSV vector (1 μg) or mock pRcRSV vector (1 μg). The pGL3-Basic vector lacking promoter/enhancer elements was used as a negative control. Cells were transfected using Lipofectamine (Invitrogen) and Plus Reagent (Invitrogen) according to the manufacturer’s instructions. At 48 h after the transfection, the cells were harvested, and their luciferase activities were determined as previously described (7). The pNF-κB-Luc plasmid in PathDetect in Vivo Signal Transduction Pathway cis-Reporting Systems (Stratagene) having 5× repeat of NF-κB element just upstream to the TATA box was also examined with other GPB5 upstream reporter constructs.

Western blotting

Western blotting was used to detect p65 expression in AtT20 cells. Briefly, AtT20 cells transiently expressing p65-pRcRSV or FLAG-p65-pcDNA3.1 were lysed at 48 h after transfection. After boiling for 5 min, the samples were subjected to 10% SDS-PAGE and then electrotransferred to Immobilon-P membranes (Millipore, Bedford, MA). The membranes were incubated with anti-p65 (Chemicon International, Inc., Temecula, CA) or anti-FLAG (M2) (Sigma-Aldrich) antibodies. As a secondary antibody, antirabbit IgG conjugated with a horseradish peroxidase (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) was used, and antibody-bound bands were detected by chemiluminescence using ECL-Plus (Amersham Pharmacia Biotech, Buckinghamshire, UK).

ChIP assays

ChIP assays were performed as described previously (8). Briefly, −656-GPB5-Luc and various mutant constructs were cotransfected with FLAG-p65-pcDNA3.1 into AtT20 cells. After an overnight incubation, the AtT20 cells were cross-linked by incubation with formaldehyde at a final concentration of 1% for 10 min at 37 C. Lysates of the cells were sonicated 25 times for 5 sec each on ice to shear the genomic DNA to lengths of 0.2–1 kb. The chromatin solutions were precleared with salmon sperm DNA-protein A agarose (Upstate Biotechnology, Charlottesville, VA). An aliquot (200 μl) of each supernatant was removed and retained as an input control. Anti-FLAG M2 Affinity Gel for mice (Sigma-Aldrich) was added to the remaining supernatant fractions and incubated for 2 h at 4 C with rotation. In addition, polyclonal antibody for NFκB p50 (Santa Cruz Biotechnology) was added to the remaining supernatant and incubated 1 h at 4 C with rotation, and followed by the addition of Protein A/G PLUS-Agarose Immunoprecipitation Reagent (Santa Cruz Biotechnology) for another 2 h. As nonantibody controls, Protein A/G PLUS-Agarose was used. After incubation of the antibody conjugated gel with the precleared supernatants, the gel were washed four times using wash buffer. The beads were then incubated in 200 ml elution buffer containing 5 mg proteinase K at 65 C overnight to reverse the DNA-protein cross-links. The DNA samples eluted from the beads were purified using a PCR purification kit (QIAGEN, Inc., Valencia, CA). The purified DNA samples were eluted into 40 ml Buffer EB, and 4 ml was used for each PCR amplification. The PCR conditions were one cycle of 95 C for 3 min, followed by 25 cycles of 95 C for 30 sec, 55 C for 30 sec, and 72 C for 20 sec. The obtained PCR products were electrophoresed in a 2% agarose gel and visualized by ethidium bromide staining. The PCR primers for the ChIP assays are listed in Table 1. These primers amplified the region containing the NF-κB-binding sites to produce fragments of 433 bp.

Statistical analyses

Most of the experiments were performed more than three times, and representative sets of data are presented. The samples in each group of experiments were examined in triplicate or quadruplicate. All data were expressed as the mean ± se. For statistical analyses, the data were compared by one-way ANOVA with Fisher’s multiple range test, and values of P < 0.05 were considered significant.

Results

Cloning and analyses of the 5′ upstream region of the human GPB5 gene

To characterize the regulatory sequences involved in human GPB5 gene expression, an approximate 3-kb 5′ upstream region of the GPB5 gene was cloned from human genomic DNA, and the sequence was confirmed to be identical with the sequence in GenBank (NW_925561.1). Promoter analyses using the TRANSFAC (BIOBASE GmbH, Wolfenbuettel, Germany) database suggested several transcription start sites within −300 to −600 nt relative to the start codon of hGPB5 (Fig. 1). It also found various binding sites for the transcription factors, including NF-κB, Pit1, stimulatory protein 1, and glucocorticoid receptor in the adjacent region. Of these, three putative NF-κB elements were indicated.

Fig. 1.

Nucleotide sequences of the 5′ upstream region of the human GPB5 gene (GenBank NW_925561.1) were analyzed to search for transcription consensus motifs (TRANSFAC database). Sequences homologous to TATA box (TFIID binding site) and elements for transcription factors are underlined. The translation start site is shown by an asterisk and set as +1. Three putative NF-κB-binding sites are labeled: P1, P2, and P3.

Identification of NF-κB-binding sites and deletion analyses of the human GPB5 promoter

In estimation the promoter locates relatively proximal region, up to 3 kb upstream of hGPB5 gene was analyzed by luciferase reporter assay in AtT20 cells. The assay showed stepwise increase by the truncation of upstream regions (Fig. 2A). To remove multiple NF-κB-binding sites one by one, a series of reporter constructs with stepwise deletions of the 5′ promoter region ligated upstream of the luciferase gene were generated. It is known that the active forms of NF-κB are heterodimers or homodimers of p65 (RelA) and p50 subunit proteins (9). AtT20 cells were transfected with the deletion mutants with or without a p65 expression vector, and subjected to reporter assays at 48 h after transfection (Fig. 2, B and C). In the presence of the longest construct, −1k-GPB5-Luc, p65 increased the promoter activity by 400-fold of the mock cotransfection. Deletion of approximately 400 bp (−656-GPB5-luc), which kept the three putative NF-κB-binding motifs intact, had no significant difference with −1k-GPB5-Luc. Step-by-step deletion of the putative NF-κB binding sites correspondingly decreased the enhancement by p65. These data suggest that p65 of NF-κB promotes GPB5 transcription by binding to each of the three putative NF-κB-binding sites located within −656 to −412, −412 to −181, and −181 to −93.

Fig. 2.

Analyses of GPB5 promoter activity. A, AtT20 cells expressing reporter plasmid having various lengths of upstream regions (−656 to −3 kb) were analyzed by luciferase assay. The values are expressed as the mean ± se. *, P < 0.05. B, Schematic representation of the human GPB5 proximal upstream region. Closed circles, Locations of the NF-κB-binding motifs. Numbers are relative to the translation start site of the GPB5 gene. C, Promoter activities of the 5′-deleted promoter-luciferase constructs. AtT20 cells were transfected with constructs of various lengths or pGL3-Basic together with p65-pRcRSV, and subjected to luciferase assays. D, Site-specific disruption analyses of putative NF-κB-binding sites. White crosses represent the disrupted NF-κB-binding sites in the constructs. Luciferase reporter plasmids containing the wild-type or mutant constructs were transfected into AtT20 cells, and the responses to cotransfection with p65-pRcRSV were examined. The results of the independent determinations are expressed as the fold stimulation by cotransfection of p65-pRcRSV relative to that obtained by cotransfection of the mock plasmid (pRcRSV). All the experiments were performed in quadruplicate, and the values are expressed as the mean ± se. *, P < 0.05 vs. wild-type. n.s., Not significant.

Mutation analyses of the human GPB5 promoter

To elucidate the functions of the putative NF-κB-binding sites for human GPB5 promoter activity, the three sites were disrupted in all combinations by oligo-directed mutagenesis (Fig. 2D). AtT20 cells were transfected with the wild-type 656-GPB5-luc, or various mutant constructs with the p65 expression vector or mock vector. For the wild-type construct, cotransfection of the p65 vector markedly increased the promoter activity compared with mock cotransfection. Each of the single site-disrupted mutants decreased the magnitude of enhancement to 30% of the wild type. Triple site-disrupted mutants further reduced to approximately 10% of the wild type (Fig. 2D). These data indicate that each of the putative NF-κB-binding sites is pivotal for NF-κB-induced expression of GPB5.

Binding of NF-κB to the three consensus elements of the human GPB5 promoter

To investigate the molecular mechanisms by which NF-κB protein binds to the GPB5 promoter region in chromatin, we performed ChIP assays. First, we performed Western blotting analyses using anti-FLAG and anti-p65 antibodies to confirm that recombinant p65 derived from the FLAG epitope-tagged p65 expression vector (FLAG-p65/pcDNA3.1) was expressed in AtT20 cells (data not shown). For ChIP assays, AtT20 cells were cotransfected with wild-type or site-disrupted reporter constructs and FLAG-p65/pcDNA3.1, and then treated with formaldehyde to cross-link the DNA-protein complexes. After sonication, immunoprecipitation, and reversal of the cross-links, the sheared DNA was purified and used as a template for PCR amplification with primers encompassing the region from −545 to −113 bp that contains the three putative NF-κB-binding sites (Fig. 3A). Both anti-FLAG and anti-p50 antibody effectively immunoprecipitated the wild-type GPB5 promoter region, as revealed by specific amplification of the NF-κB-binding region (Fig. 3). As a negative control, we performed the ChIP procedure without an antibody, and confirmed that no target sequences were amplified. In both p50 antibody and FLAG antibody, combined disruptions of two of the three NF-κB-binding sites did not interfere with the amplification. No bands were amplified after disruption of all three sites. These findings illustrate that each of the three sites are responsible for NF-κB binding in vivo in AtT20 cells.

Fig. 3.

ChIP assays were performed to confirm the binding of p65 to the GPB5 promoter in vivo. A, Schematic representation showing the putative NF-κB-binding sites. The two arrows indicate the primers used to amplify the region from −545 to −113 spanning the putative NF-κB-binding sites. B, The protein-DNA complexes were analyzed by PCR with the indicated primers. Aliquots of the cell lysates before immunoprecipitation were used as input controls (first panel). As a negative control, the protein-DNA complexes were incubated with a column without the antibody (second panel). The protein-DNA complexes were incubated with an anti-FLAG antibody (third panel) or p50 antibody (fourth panel) conjugated column.

Cytokine-stimulated activities of the isolated promoter

To determine whether the constitutively functional GPB5 promoter could be induced by exposure to cytokines, AtT20 cells transfected with −1k-GPB5-Luc were stimulated with different doses of TNFα and IL-1β (Fig. 4, A and B). After 6 h TNFα and IL-1β treatments, the luciferase productions were significantly increased by up to 2-fold in dose-dependent manners. Furthermore, TNFα and IL-1β showed time-dependent enhancements with the maximum responses at 12 h after the cytokine addition (Fig. 4, C and D). The effect of IL-1β was further examined using reporter constructs with various lengths of upstream region ranging from −656 bp to −3 kb (Fig. 4E). On −656, −1 kb, and −2k-GPB5-Luc, the magnitude of the IL-1β induction was approximately 2-fold. On the positive control vector having 5× NF-κB binding site (pNF-kB/Luc), IL-1β increased the activity at the same level. No effect was found on −3kb-GPB5-Luc.

Fig. 4.

Effects of proinflammatory cytokines on GPB5 promoter activity in AtT20 cells. A and B, Dose-response effects. Cells transiently expressing −1k-GPB5-Luc were treated with various doses of TNFα (0–200 mg/ml; C) or IL-1β (0–50 U/ml; D) for 6 h. C and D, Time course effects. Cells transiently expressing −1k-GPB5-Luc were treated with TNFα (200 μg/ml; A) or IL-1β) (50 U/ml; B) for up to 48 h. *, P < 0.05 vs. vehicle-treated group (A and B) or the initial point of the course (C and D). E, The IL-1β effects on various lengths of upstream regions. AtT20 cells expressing reporter plasmid having −656 to −3 kb of GPB5 upstream region or 5× tandem repeat of NF-κB element were treated with (closed bar) or without (open bar) 50 U/ml IL-1β for 6 h. Fold induction by IL-1β for each of the reporter constructs is expressed. *, P < 0.05 vs. each of IL-1β (−) group.

To clarify whether the endogenous NF-κB pathway is involved in the cytokine-mediated induction of GPB5, cells were pretreated with the proteasome inhibitor MG132, which blocks NF-κB activation. MG132 abolished the cytokine-induced GPB5 gene activation (Fig. 5A). When all three NF-κB-binding sites were disrupted, neither TNFα nor IL-1β induced any significant transcription (Fig. 5B).

Fig. 5.

Involvement of the NF-κB pathway in cytokine-induced GPB5 expression. A, MG132 (10 mg/ml) was added to the media at 30 min before treatment with TNFα (200 μg/ml) or IL-1β (50 U/ml). ‡, P < 0.05 vs. the vehicle-treated group; §, P < 0.05 vs. the IL-1β-treated group; #, P < 0.05 vs. the TNFα-treated group. B, AtT20 cells transfected with wild-type (−656-GPB5-Luc) or mutant (P1m2m3m) plasmids are incubated in the presence or absence of TNFα (0–200 μg/ml) or IL-1β (0–50 U/ml) for 6 h. All the experiments were performed in quadruplicate, and for each construct, the values (mean ± se) are expressed as the fold induction relative to the luciferase activity with vehicle treatment (open bar). *, P < 0.05 vs. vehicle treated group.

Discussion

The physiological functions of thyrostimulin are currently poorly understood. One of the reasons for this lack of knowledge is its species differences because thyrostimulin is scarcely expressed in rodents commonly used for scientific research (2) but expressed in the corticotroph of the human pituitary (3), which is one of the most difficult tissues to investigate directly. In an attempt to study thyrostimulin from the aspect of transcriptional regulation, we previously showed that the GPA2 subunit is positively regulated by the LIM domain homeobox gene isl-1 (10). Isl-1 is related to the development or regulation of various endocrine tissues (11, 12, 13, 14), suggesting that GPA2 is associated with endocrine systems. In addition, the fact that GPA2 expression is independent of TRH and T₃ suggests that GPA2 may be regulated differently from GSUα or TSHβ, referred to as the hypothalamus-pituitary-thyroid (HPT) axis (10).

In the present study, we explored the regulatory systems for GPB5 by examining the 5′-flanking region of the GPB5 gene extending from −3 kb to −93 nt relative to the ATG initiation codon. It requires the identification of the transcription start site to locate the promoter region of the gene, and there is a fear that the true promoter locates in further upstream region. However, reporter analyses within −3-kb regions illustrated more potent activities in proximal regions within −1 kb by itself or with cytokine treatments. Combined with the database analysis showing putative TATA boxes found side by side with regulatory elements, it is possible that the functional promoter might locate within the −1-kb upstream region of the human GPB5 gene.

We focused on the three putative NF-κB binding sites that are most frequently identified in the region. Overexpression of the p65 subunit of NF-κB extremely induced the transcriptional activity of GPB5 by several hundreds-fold. Site-disruption studies illustrate that each of the three elements is required for the full activation by p65. ChIP assay showed that not only the exogenous FLAG-p65 but also endogenous p50 of NF-κB binds to each of the three elements in AtT20 cells. This result correlates with the previous findings that the NF-κB active form is usually a hetero- or homodimer of p50 or p65.

It is known that the NF-κB pathway is associated with the cytokine system. Our data clearly showed that the proinflammatory cytokines TNFα and IL1β increased the transcriptional activity of GPB5 through the NF-κB pathway in the corticotroph cell line, AtT20. This cell line has expressed receptors for cytokines, including IL-1β and TNFα endogenously (15, 16, 17), and addition of these cytokines induces proopiomelanocortin gene expression through the NF-κB pathway (7, 17). Cytokines lose positive effect when all of the three elements are disrupted (Fig. 5B), which suggests that NF-κB is unique in the cytokine signaling system in AtT20. This result also coincides with the ChIP data that overexpressed p65 does not bind to P1m2m3m. However, the reporter construct with P1m2m3m showed approximately 30-fold activation by p65, which might suggest another NF-κB element in the region. It would be possible that ChIP assays are less sensitive than luciferase assay for the p65 activation. The presumable NF-κB element would bind to p65 only weakly, and not associate with cytokine signaling.

Binding sites for NF-κB are present in the promoter regions of many proinflammatory cytokines and immunoregulatory mediators important for the induction of acute inflammatory responses associated with critical illnesses (18). Among the many transcriptional regulatory proteins identified, NF-κB has particular importance in modulating the expressions of immunoregulatory genes relevant to critical illnesses, including infections, blood loss, and ischemia-reperfusion injury (19). Various cytokines generated during infectious stress induce ACTH expression directly or via hypothalamic CRH, thereby establishing the concept of immune-neuroendocrine interactions (20, 21). In the present study, we have revealed that GPB5 is also positively regulated by IL-1β and TNFα. Combined with the fact that GPB5 colocalizes with ACTH in the corticotroph of the human pituitary, it is likely that neuroimmunoregulatory factors, including proinflammatory cytokines, induce thyrostimulin in the human corticotroph.

In an animal model of infectious stress, bacterial lipopolysaccharide was found to suppress the HPT axis (22, 23), a disorder commonly referred to as nonthyroidal illness syndrome in humans (24). In this syndrome, decreases in the circulating TSH and thyroid hormone levels are accompanied by a reduction in TRH gene expression in hypophysiotropic neurons of the hypothalamic paraventricular nucleus (22), suggesting that infection alters the set point for negative feedback control of the HPT axis by thyroid hormone. However, in the acute phase, the circulating T₄ level becomes significantly increased within 3 h lipopolysaccharide administration (25). Therefore, it is possible that thyrostimulin is involved in acute control of the T₄ level, and that suppression of the HPT axis gradually becomes dominant after several hours.

On the other hand, thyroid hormone does not affect the GPB5 promoter activity in AtT20 cells (data not shown), unlike the TSH composing subunit GSUα and TSHβ, which are negatively regulated by T₄ (26, 27). Moreover, TRH does not affect thyrostimulin expression in primary cultures of rat pituitaries (2). These results indicate that thyrostimulin has different roles from those of TRH and TSH in the HPT axis.

It remains unknown whether thyrostimulin acts in an endocrine or paracrine manner. Thyrotrope or folliculo-stellate cells in the anterior pituitary express TSHR (28, 29), indicating that thyrostimulin secreted by the corticotroph may act as a paracrine reagent. Expression of TSH and TSHR in a variety of tissues outside of the HPT axes is also well documented (30, 31, 32). These results suggest that thyrostimulin and TSH have additional roles other than their functions in the thyroid.

In summary, proinflammatory cytokines regulate GPB5 transcription through the NF-κB pathway in the AtT20 corticotroph cell line. These findings suggest a stress-related nature for the functions of thyrostimulin. It may be useful to measure the circulating levels of thyrostimulin in patients with severe illnesses, including sepsis, acute respiratory distress syndrome, and multiple organ dysfunction syndrome. Further studies will be necessary to elucidate the physiological roles of thyrostimulin.

Acknowledgments

We are grateful for the excellent technical assistance of Ms. Michiko Yamada, and the technical advice of Dr. Akihiro Tomita regarding the chromatin immunoprecipitation assays.

This paper was supported in part by a Grant-in-Aid from the Ministry of Education, Culture, Science and Technology of Japan (No. 19590245-00).

Disclosure Summary: The authors have nothing to declare.

First Published Online December 18, 2008

Abbreviations

 
• ChIP,

  Chromatin immunoprecipitation;

 
• GPA2,

  glycoprotein hormone subunit α2;

 
• GPB5,

  glycoprotein subunit β5;

 
• GSUα,

  glycoprotein α-subunit;

 
• HPT,

  hypothalamus-pituitary-thyroid;

 
• NF-κB,

  nuclear factor-κB;

 
• TSHR,

  TSH receptor.