Skip to main content
PMC home page
Endocrinology logoLink to Endocrinology
. 2008 Feb 14;149(5):2230–2240. doi: 10.1210/en.2007-1344

Growth Hormone Regulation of Insulin-Like Growth Factor-I Gene Expression May Be Mediated by Multiple Distal Signal Transducer and Activator of Transcription 5 Binding Sites

Satyanaryana Eleswarapu 1, Zhiliang Gu 1, Honglin Jiang 1
PMCID: PMC2329286  PMID: 18276757

Abstract

The transcription factor signal transducer and activator of transcription (STAT)-5 mediates GH stimulation of IGF-I gene expression in the liver. Previous studies suggested that STAT5 might exert this effect by binding to an IGF-I intron 2 region and a distal 5′-flanking region each containing two STAT5 binding sites. Here we report the identification of three additional chromosomal regions containing a total of five putative STAT5 binding sites that may mediate GH-induced STAT5 activation of IGF-I gene expression in the mouse liver. By comparing an 170-kb mouse genomic DNA containing the IGF-I gene with the corresponding human sequence, we identified 19 putative STAT5 binding sites that bear the consensus sequence of STAT5 binding site and are conserved across the two species. Chromatin immunoprecipitation assays indicated that five chromosomal regions containing a total of nine of the 19 putative STAT5 binding sites were bound by STAT5 in the mouse liver in response to GH administration and that these bindings preceded or coincided with GH-increased IGF-I gene transcription. Two of the five chromosomal regions correspond to those previously identified in other species, and the three new chromosomal regions that contain a total of five putative STAT5 binding sites are IGF-I intron 3 regions located at least 26 kb from the transcription start site. Gel-shift assays confirmed the binding of the five new putative STAT5 binding sites as well as three of the four previously identified STAT5 binding sites to GH-activated STAT5 from the mouse liver. Cotransfection analyses indicated that, although each of the five chromosomal regions was able to mediate STAT5 activation of reporter gene expression, together they mediated greater STAT5 activation of reporter gene expression in response to GH. Overall, these results suggest that GH-induced STAT5 activation of IGF-I gene expression in the mouse liver might be collectively mediated by at least eight STAT5 binding sites located in distal intronic and 5′-flanking regions of the IGF-I gene.

IGF-I, A POLYPEPTIDE hormone, is essential for normal development and growth in mammals (1). IGF-I gene- or IGF-I receptor gene-deleted mice die at or shortly after birth, and those that survive are 70% smaller than wild-type mice; have defects in brain, bone, muscle, and lung; and are infertile (2,3). IGF-I may also play a role in tumorigenesis and aging because high concentrations of circulating IGF-I are associated with increased risk of developing breast and prostate cancers (4,5), and low IGF-I levels are linked to extended longevity (6,7). Although the IGF-I gene is widely expressed in the body (8), most (∼75%) of the circulating IGF-I comes from its gene expression in the liver (9). Liver expression of the IGF-I gene is mainly controlled at the transcriptional level by GH from the pituitary (8). Therefore, understanding the mechanism by which GH regulates IGF-I gene expression in the liver is important to understanding how circulating IGF-I concentration is regulated.

The mechanism by which GH regulates IGF-I gene expression in the liver is only beginning to be understood. Studies on signal transducer and activator of transcription (STAT)-5 knockout mice (10,11) and on rats overexpressing a dominant-negative STAT5b mutant in the liver (12) demonstrate that GH-increased IGF-I gene expression in the rodent liver is primarily mediated by STAT5. The same transcription factor also appears to mediate GH regulation of IGF-I gene expression in the human liver because a missense mutation in the STAT5b gene was associated with reduced serum IGF-I concentration in a man (13). Through mapping GH-induced deoxyribonuclease I hypersensitive sites, two STAT5 binding sites in intron 2 of the rat IGF-I gene were indicated to mediate STAT5 activation of IGF-I gene expression in the liver (14). Through mapping STAT5 binding enhancers, two functional STAT5 binding sites were identified approximately 70 kb upstream from the human IGF-I gene (15), and these two STAT5 binding sites were later found to also mediate GH activation of IGF-I gene expression in the rat liver (16). The identification of four STAT5 binding sites upstream and downstream from the IGF-I promoter prompted us to conduct this study to determine whether there are additional STAT5 binding sites mediating GH regulation of IGF-I gene expression in the liver. Our results suggest that eight distantly located STAT5 binding sites might mediate GH-induced STAT5 activation of IGF-I gene expression in the mouse liver.

Materials and Methods

Reagents

Oligonucleotides were synthesized by Operon Technologies, Inc. (Alameda, CA). Cell culture media, fetal bovine serum, and other cell culture reagents were purchased from Sigma (St. Louis, MO). Restriction enzymes, T4 DNA ligase, T4 polynucleotide kinase, and other molecular biology enzymes were purchased from New England BioLabs (Beverly, MA) or Promega (Madison, WI). All other reagents were purchased from Sigma or Fisher Scientific (Pittsburgh, PA) unless otherwise indicated.

Promoter-reporter and cDNA constructs

Fourteen mouse IGF-I gene and 5′-flanking regions (designated S1-S14, 200–400 bp each) containing putative STAT5 binding sites were amplified from mouse (C57BL/6) genomic DNA by PCR using 2× PCR master mix (Promega) and sequence-specific primers containing SmaI and KpnI restriction sites at their 5′ ends (Table 1). The PCR products were digested with restriction enzymes SmaI and KpnI, gel purified using a QIAquick gel extraction kit (QIAGEN Inc., Valencia, CA), and cloned into pGL2TK, a minimal thymidine kinase (TK) promoter-luciferase reporter vector constructed previously (15). The resulting constructs were named pGL2TK-S1, -S2, -S3, -S4, -S5, -S6, -S7, -S8, -S9, -S10, -S11, -S12, -S13, and -S14, respectively. A mouse cytokine inducible SH2-containing protein gene (Cis) promoter region containing four STAT5 binding sites (17,18) was similarly amplified and cloned to generate the construct pGL2TK-Cis. The IGF-I DNA regions S5, S9, S10, and S11 were ligated and cloned into pGL2TK-S1 at the BamHI and SalI sites to generate the construct pGL2TK-S1S5S9S10S11. An approximately 3.2-kb mouse IGF-I promoter was amplified by PCR from the mouse genomic DNA using gene-specific primers (Table 1) and used to replace the TK promoter in pGL2TK-S1 and pGL2TK-S1S5S9S10S11 to generate the constructs pGL2IP1-S1 and pGL2IP1-S1S5S9S10S11, respectively. These promoter-reporter gene constructs were used in the transfection analyses described below.

Table 1.

PCR primers used in this study

Name Sequencea (5′–3′) Gene and location Amplicon size (bp) Application
S1F gaaagtgggtttggcttgg IGF-I 5′ flanking 301 Cloning and ChIP
S1R tgtgcaaaccaaccagtcat
S2F tctgttctgggcaaggtcat IGF-I 5′ flanking 252 Cloning and ChIP
S2R ggtttggaaccaaggacaga
S3F ggcaattttccaagagtcca IGF-I 5′ flanking 239 Cloning and ChIP
S3R gctcttctagactcccaagtgc
S4F cctagccccagcaaaggtat IGF-I 5′ flanking 230 Cloning and ChIP
S4R tgggggaaagcaatgaatag
S5F gggtggctcacctcatactc IGF-I intron 2 245 Cloning and ChIP
S5R gccgatggttagtagccaaa
S6F gagcaaaggtgaaaagggaat IGF-I intron 3 227 Cloning and ChIP
S6R accatcaccttctgccaaac
S7F gtagaaggcgaggcagtagc IGF-I intron 3 249 Cloning and ChIP
S7R acagcactgttgctgggtta
S8F tccaccatcccttgagtagg IGF-I intron 3 251 Cloning and ChIP
S8R tctgtttgagtgtagacattctgct
S9F aaggtggaggtggcctttag IGF-I intron 3 232 Cloning and ChIP
S9R gcctgagaatgacctttgga
S10F attcctcccagctgtgtgtc IGF-I intron 3 199 Cloning and ChIP
S10R ggacttggtctgaggcaatg
S11F aaaggaaggctgggtggtag IGF-I intron 3 278 Cloning and ChIP
S11R gtcctgcatgtctgtggaag
S12F gcctcgctcctaaagagtca IGF-I intron 5 299 Cloning and ChIP
S12R attgacaggtggcacagaca
S13F caacacagtaaaaggagaaagcaa IGF-I exon 6 251 Cloning and ChIP
S13R aaagaaaccaggactcccaaa
S14F ccaccccacacacacctatt IGF-I exon 6 260 Cloning and ChIP
S14R agctggccaaacagtaaagg
CisPF gtccagcgatacgattggtc Cis promoter 264 Cloning and ChIP
CisPR gaacagcttggaaggacgag
AlbPF gcaaacatacgcaagggatt A1b promoter 178 ChIP
AlbPR tggggttgataggaaaggtg
IGFIE2F tgtaaacgacccggacctac IGF-I mRNA 171 mRNA qPCR
IGFIE3R cacgaactgaagagcatcca
IGFIPE2F actacgcgtcctgtggcaaa IGF-I promoter 3,248 Cloning
IGFIPE2R tgcctcgagtttggtaggtc
IGFII3F catgggaaggagacagagga IGF-I intron 3 252 Pre-mRNA qPCR
IGFII3R ggggttcactgaggtgattt
A1bE7F aattggcaacagacctgacc Albumin mRNA 237 mRNA qPCR and cloning
A1bE8R cctcaacaaaatcagcagca
A1bI3F tgggagcttgacagtgacag Albumin intron 3 166 Pre-mRNA qPCR
A1bI3R gggatgaccattggtattgg
IGFIE3F cttgctcaccttcaccagct IGF-I mRNA 210 Cloning
IGFIE4R tacatctccagtctcctcag
GapdhF acccagaagactgtggatgg GAPDH mRNA 81 Cloning
GapdhR ggatgcagggatgatgttct
GhrhrE2F tgagcttgcatgtcttcagg GHRH receptor exon 2 and intron 3 482 Genotyping
GhrhrI3R ggtgaagtggacgatgaggt
Open in a new tab
a

Top sequence, forward primer; bottom sequence, reverse primer. 

qPCR, Quantitative PCR. 

A 210-bp mouse IGF-I, a 237-bp mouse albumin, and an 81-bp mouse glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA fragments were amplified by standard RT-PCR from mouse liver total RNA using the ImProm-II reverse transcription system (Promega), 2× PCR master mix, and sequence-specific primers (Table 1). These cDNA fragments were cloned into pGEM-T Easy vector (Promega). The resulting cDNA plasmids served as templates for syntheses of antisense probes for ribonuclease protection assays (RPAs) described below. The inserts in all recombinant plasmids were verified by sequencing.

Animal experiments

Male C57BL/6 mice hypophysectomized by a parapharyngeal approach at age of 4 wk were purchased from Charles River Laboratories (Wilmington, MA). The mice were housed on automatically timed 12-h dark, 12-h light cycles and had free access to standard rodent diet (Harlan Teklad, Madison, WI) and 5% sucrose water. The mice were weighed weekly for the first 2 wk, and those that did not continue to grow were considered completely hypophysectomized. The confirmed hypophysectomized mice at 8 wk old were administered sc with 2 μg/g body weight of recombinant bovine GH (National Hormone and Peptide Program, Torrance, CA) or equal amount of PBS. At this or similar supraphysiological dosages, bovine GH has been shown to be effective in restoring normal gene expression or normal growth in GH-deficient mice (11,19,20,21). The mice were killed 6 h later and the livers collected, immediately used for nuclei isolation, or frozen for RNA extraction. Based on previous studies (22,23), expression of IGF-I mRNA in the liver reached maximum by 6 h after a single GH injection.

Breeding pairs of C57BL/6J-Ghrhrlit mice, which contained a mutation in the GHRH receptor (Ghrhr) gene (24,25), were purchased from the Jackson Laboratory (Bar Harbor, ME). The mice were maintained as described above and bred to generate GH-deficient homozygous lit/lit offspring. The lit/lit mice were identified based on smaller size, compared with their littermates, and confirmed by genotyping. For genotyping, genomic DNA was extracted from tail clips using a DNeasy tissue kit (QIAGEN) and the Ghrhr DNA region containing the suspected mutation was amplified by PCR using gene-specific primers (Table 1). The amplified DNA was gel purified and sequenced. Lit/lit male mice 4–5 months old were administered sc with 2 μg/g body weight of recombinant bovine GH as described above and were killed for liver collection at 5, 15, and 30 min and 2, 6, and 24 h after the administration. Lit/lit male mice administered with PBS served as controls. The animal-related procedures were approved by the Virginia Tech Animal Care Committee.

RNA extraction

Total and nuclear RNA were extracted from liver tissue and liver nuclei, respectively. To isolate liver nuclei, an approximately 200-mg fresh liver sample was homogenized in ice-cold PBS at a low speed. The homogenate was centrifuged at 3000 × g for 10 min at 4 C, and the pellet was washed quickly in 10 ml of ice-cold hypotonic buffer composed of 10 mm HEPES (pH 7.9), 1.5 mm MgCl2, 10 mm KCl, 0.2 mm phenylmethylsulfonyl fluoride (PMSF), and 0.5 mm dithiothreitol. The pellet was resuspended in 10 ml of the same buffer and homogenized in a glass Dounce homogenizer. The nuclei were collected by centrifugation as described above and frozen for later RNA extraction. All RNA extractions were done using Tri reagent (Molecular Research Center, Cincinnati, OH), according to the manufacturer’s protocol. RNA concentration and quality were determined by spectrophotometry and RNA gel electrophoresis, respectively.

RPAs

An RPA was used to analyze liver IGF-I, albumin, and GAPDH mRNAs in the hypophysectomized mice. 32P-labeled riboprobes for RPA were synthesized from linearized IGF-I, albumin, and GAPDH cDNA plasmids by standard in vitro transcription in the presence of [α-32P]CTP (PerkinElmer Life and Analytical Sciences, Wellesley, MA). The probes were purified to remove free [α-32P]CTP using G-50 Sephadex columns (Roche Diagnostics Corp., Indianapolis, IN). About 1 × 105 dpm of IGF-I or albumin and 1 × 105 dpm of GAPDH riboprobes were hybridized with 15 μg of total RNA in 20 μl of hybridization buffer for about 16 h at 42 C and then digested with ribonucleases A and T1 at 37 C for 30 min. The digestion was stopped and the undigested RNA fragments precipitated by adding inactivation/precipitation buffer. These steps were performed using a RPAII kit (Ambion, Inc., Austin, TX). The precipitated RNA fragments were resolved in 6% polyacrylamide gels containing 7 m urea. After electrophoresis, the gels were dried, exposed to phosphor screens, and scanned on a Molecular Imager FX System (Bio-Rad, Hercules, CA).

Real-time RT-PCR

Real-time RT-PCR was used to quantify IGF-I and GAPDH mRNAs and IGF-I and albumin pre-mRNAs in the lit/lit mice. Two micrograms of total RNA or nuclear RNA were reverse transcribed to cDNA in a total volume of 20 μl using TaqMan reverse transcribing reagents (Applied Biosystems, Foster City, CA), according to the manufacturer’s protocol. Real-time PCRs of IGF-I, GAPDH, and albumin cDNAs were performed on 0.2 μg of cDNA product in a total volume of 25 μl, using Power Sybr Green PCR master mix (Applied Biosystems) on an Applied Biosystems 7300 real-time PCR system. The primers for these PCRs were presented in Table 1 and the conditions for these PCRs were 40 cycles of 95 C for 15 sec and 60 C for 1 min. For RT-PCR of nuclear RNA, a reverse transcription that did not contain the reverse transcriptase was included for each sample to control for genomic DNA contamination. The RT-PCR of each sample was performed in duplicate. The real-time PCR data were analyzed using the cycle threshold (2−ΔΔCT) method, as recommended by Applied Biosystems.

Chromatin immunoprecipitation assays (ChIP)

A fresh liver sample (∼200 mg) was homogenized at a low speed in 6 ml of PBS supplemented with 1% formaldehyde, a protease inhibitor cocktail tablet (Roche), 0.5 mm sodium orthovanadate, 10 mm sodium β-glycerophosphate, 50 mm sodium fluoride, and 5 mm sodium pyrophosphate. The homogenate was incubated with gentle rocking at room temperature for 15 min, followed by addition of 125 mm glycine to stop the cross-linking. The homogenate was centrifuged at 3000 × g for 10 min at 4 C, and the pellet was washed once in 10 ml of ice-cold PBS. The pellet was resuspended in 1 ml of lysis buffer, 5 μl of protease inhibitors, and 5 μl of PMSF, all from a ChIP-IT kit (Active Motif, Carlsbad, CA). The resuspension was incubated on ice for 30 min, followed by homogenization in a glass Dounce homogenizer. The homogenate was centrifuged as described above, and the pelleted nuclei were resuspended in 1 ml of shearing buffer, 5 μl of protease inhibitors, and 5 μl of PMSF from the ChIP-IT kit. The nuclei were subsequently sheared on ice with 10 pulses of 20 sec sonication using a sonic dismembrator model 100 at setting 3 (Fisher Scientific). Under these conditions, the chromatin was sheared to fragments 200–500 bp long. Immunoprecipitation of the sheared chromatin was performed as described (26). Briefly, 100 μl of protein G-Dynal beads (Invitrogen, Carlsbad, CA) were conjugated to 3 μg of STAT5 antibody (sc-835; Santa Cruz Biotechnology, Inc., Santa Cruz, CA) by overnight incubation at 4 C with gentle rocking. The STAT5 antibody-conjugated beads were collected in 150 μl of ChIP immunoprecipitation buffer and 1 μl of protease inhibitors from the ChIP-IT kit and were incubated with 50 μl of sonicated chromatin at 4 C overnight. The beads were washed and the chromatin was eluted using the reagents from the ChIP-IT kit, essentially following the manufacturer’s instructions. The eluted chromatin fragments were reverse cross-linked by overnight incubation at 65 C, and the DNA was extracted using the columns from the ChIP-IT kit, according to the manufacturer’s instructions.

The relative abundance of a STAT5 binding site-containing DNA fragment in STAT5 antibody-precipitated chromatin was determined by PCR using 2× PCR master mix (Promega) and sequence-specific primers (Table 1). All primers were tested for specificity on sonicated chromatin, i.e. input DNA. The PCR conditions were 30 cycles of 94 C for 30 sec, 60 C for 1 min, and 72 C for 2 min. In each PCR, an albumin promoter region was also amplified. The albumin promoter region was not expected to be bound by STAT5 because it does not contain a putative STAT5 binding site and because it is not regulated by GH (see Results for this). The presence of the albumin promoter DNA in the STAT5 antibody-precipitated chromatin reflected nonspecific binding of the antibody to chromatin and/or incomplete removal of the unbound chromatin during the ChIP procedure. The PCR products were resolved through standard agarose gels, and the intensities of DNA bands were measured using the National Institutes of Health ImageJ software (Bethesda, MD). The relative abundance of a putative STAT5 binding site-containing chromosomal region in a ChIP DNA sample was obtained by normalizing the intensity of the PCR product for this DNA region to that for the albumin promoter region. This normalization was expected to control for tube-to-tube variation in DNA loading and PCR efficiency in addition to nonspecific antibody-chromatin binding and incomplete removal of nonbound DNA.

EMSAs

Liver nuclei were isolated from lit/lit mice 30 min after GH or PBS injection as described above for ChIP. The nuclei were resuspended in low-salt buffer [20 mm HEPES (pH 7.9), 25% glycerol, 1.5 mm MgCl2, 0.02 m KCl, 0.2 mm EDTA] supplemented with protease and phophatase inhibitors as described above. The nuclei were lysed by addition of an equal volume of high-salt buffer [20 mm HEPES (pH 7.9), 25% glycerol, 1.5 mm MgCl2, 1.2 m KCl, 0.2 mm EDTA]. The lyses were pelleted by centrifugation at 14,500 rpm for 30 min. The supernatant was collected and dialyzed in a dialysis buffer [20 mm HEPES (pH 7.9), 25% glycerol, 100 mm KCl, 0.2 mm EDTA] for 2 h at 4 C.

Double-stranded oligonucleotides (Table 2) corresponding to the 19 putative STAT5 binding sites were end labeled with 32P using T4 polynucleotide kinase and [γ-32P]ATP (PerkinElmer). Ten micrograms of liver nuclear proteins were incubated with 1 × 105 dpm of 32P-labeled oligonucleotide in a reaction buffer containing 20% glycerol, 20 mm Tris-HCl (pH 7.5), 100 mm KCl, 1 mm dithiothreitol, 1 mm EDTA, and 2 μg of poly d(IC) for 1 h at 4 C. For the supershift assay, the nuclear proteins were incubated with 2 μg of STAT5 antibody (Santa Cruz Biotechnology) or 2 μg of rabbit preimmune serum for 1 h at 4 C before being incubated with the labeled oligonucleotide. After the incubation, the DNA-protein mixtures were resolved by electrophoresis through 6% polyacrylamide gels in 0.5× Tris-borate-EDTA buffer. The gels were analyzed by phosphor imaging as described above.

Table 2.

Oligonucleotides used in the gel-shift assays

STAT5 binding site Sequencea (5′–3′) Chromosomal locationsb
S1ac aaattctaagaaact chr10:87250569–87250583
S1bc tttttcttagaagta chr10:87250802–87250816
S2 tccttccttgaaact chr10:87258619–87258633
S3 gagttctgggaatgt chr10:87261780–87261794
S4 ttattcatagaatga chr10:87279703–87279717
S5ac gccttcctggaagaa chr10:87325725–87325739
S5bc tgcttcttagaatga chr10:87325801–87325815
S6 gtcttccatgaagaa chr10:87339430–87339444
S7 attttctgtgaacta chr10:87340981–87340995
S8 tgtttcagggaaaaa chr10:87346761–87346775
S9 tctttcagggaaatc chr10:87348654–87348668
S10a cagttctcagaaagg chr10:87364876–87364890
S10b aaattcgcagaagtg chr10:87364891–87364905
S11a tgattcctagaagag chr10:87369686–87369700
S11b tagttcacagaaaaa chr10:87369821–87369835
S12 gaattccttgaagtc chr10:87388381–87388395
S13 gccttccaagaagaa chr10:87394585–87394599
S14a catttctttgaaagt chr10:87396823–87396837
S14b tccttctttgaatgt chr10:87396881–87396895
Open in a new tab

chr, Chromosome. 

a

The sequence is for the sense strand; the core sequences of the STAT5 binding sites are indicated in bold. 

b

These correspond to the locations in the mouse July 2007 (mm9) assembly at the University of California, Santa Cruz, genome browser (http://genome.ucsc.edu). 

c

These correspond to the previously identified STAT5 binding sites in the rat and human IGF-I loci (14–16). 

Cell culture, transfection, and luciferase assays

The CHO cells were grown in MEM supplemented with 10% fetal bovine serum, 1 mm glutamine, 100 U/ml penicillin, and 100 μg/ml streptomycin under 5% CO2 at 37 C. The cells at approximately 50% confluency in 24-well plates were transfected with 200 ng of promoter-reporter gene plasmid, 200 ng of bovine GH receptor (GHR) expression plasmid (15), 200 ng of mouse STAT5b expression plasmid (27), and 1 ng of transfection efficiency control plasmid pRL-CMV (Promega) per well, using FuGENE6 (Roche). Twenty-four hours after the transfection, the medium was replaced with serum-free MEM, and the cells were further cultured for 16 h. The cells were subsequently treated with 500 ng/ml of recombinant bovine GH or PBS for 8 h before being lysed for luciferase assay. Lysis of the cells and luciferase assay were carried out using dual-luciferase reporter assay system (Promega), according to the manufacturer’s instructions. The firefly luciferase activity encoded by the promoter-reporter construct was normalized to the renilla luciferase activity encoded by pRL-CMV to control for variation in transfection efficiency.

Statistical analyses

Comparisons between two means were done using t test. Comparisons between multiple means were done using ANOVA followed by the Tukey test. These statistical analyses were performed using the general linear model of SAS (SAS Institute, Inc., Cary, NC). All data are expressed as mean ± sem.

Results

Nineteen of 90 STAT5 binding consensus sequences in approximately 170 kb of mouse IGF-I gene and flanking regions were conserved in the corresponding human sequence

We started this study by identifying consensus sequence of STAT5 binding site, TTCNNNGAA, where N is any nucleotide (28,29), which was conserved between the mouse and human chromosomal regions surrounding the IGF-I gene because regulatory elements are often evolutionarily conserved (30,31). A search of a 170 kb of mouse genomic sequence (GenBank accession no. AC125082 and AC139754) consisting of approximately 78 kb IGF-I gene (exon 1 to exon 6), approximately 80 kb 5′-flanking region and approximately 10 kb 3′-flanking region revealed 90 stretches of TTCNNNGAA. An alignment of the 170-kb mouse DNA sequence with the corresponding human sequence (GenBank accession no. AC010202) using the VISTA program (32) revealed that 19 of the 90 9-bp STAT5 consensus sequences were conserved in highly similar (>70% identical) DNA regions (>100 bp) between the two species (Fig. 1). Among the 19 conserved STAT5 consensus sequences, five (designated S1a, S1b, S2, S3, and S4) were located in the distal 5′-flanking region; two (S5a, S5b) in intron 2; eight (S6, S7, S8, S9, S10a, S10b, S11a, and S11b) in intron 3; one (S12) in intron 5; and three (S13, S14a, and S14b) in exon 6 of the IGF-I gene (Fig. 1). S1a and S1b, S5a and S5b, S10a and S10b, S11a and S11b, and S14a and S14b were less than 250 bp apart in the genome (Fig. 1). S1a, S1b, S5a, and S5b corresponded to the previously identified STAT5 binding sites in the rat and human IGF-I loci (14,15,16).

Figure 1.

Figure 1

Open in a new tab

Nineteen STAT5 binding consensus sequences are conserved across the mouse and human IGF-I gene and flanking regions. The conserved STAT5 binding consensus sequences (TTCNNNGAA) were identified by comparative sequence analysis using VISTA. The highly similar regions (>70% identical over > 100 bp) between the mouse and human IGF-I gene (∼80 kb) and flanking (∼80 kb 5′ flanking and ∼10 kb 3′ flanking) sequences are indicated by peaks. The IGF-I exons are indicated by rectangles labeled with E1 (exon 1), E2, E3, E4, E5, and E6. The locations of the 19 conserved STAT5 binding consensus sequences are marked by asterisks. The 200- to 300-bp DNA regions containing these conserved STAT5 consensus sequences are designated S1, S2 … S14.

GH stimulated STAT5 binding to five putative STAT5 binding site-containing chromosomal regions in the mouse liver

We next determined which of the 19 conserved STAT5 consensus sequences, or putative STAT5 binding sites, were bound by STAT5 in the mouse liver in response to GH treatment. We injected the hypophysectomized mice with GH or PBS and used the ChIP assay to measure STAT5 binding to the 14 chromosomal regions (S1–S14) containing one or two of the 19 putative STAT5 binding sites at 6 h after the injection. We also measured STAT5 binding to a Cis gene promoter region that contains four STAT5 binding sites (17,18) as a positive control. Before the ChIP assays, we confirmed the GH responsiveness of the mice by measuring changes in liver IGF-I mRNA. As shown in Fig. 2A, GH administration increased liver abundance of IGF-I mRNA in the hypophysectomized mice and had no effects on liver expression of albumin or GAPDH mRNA. These results are consistent with previous studies (11,33). Figure 2A also shows that some expression of IGF-I mRNA remained in three of the four hypophysectomized mice in the absence of exogenous GH. This continued expression of IGF-I mRNA may be due to incomplete hypophysectomies or suggests regulation of IGF-I mRNA expression by factors in addition to GH.

Figure 2.

Figure 2

Open in a new tab

GH stimulated STAT5 binding to five putative STAT5 binding site-containing IGF-I chromosomal regions and IGF-I mRNA expression in the mouse liver. A, RPAs of IGF-I, albumin (alb), and GAPDH mRNAs in the livers of four hypophysectomized mice injected with PBS (hypox) and four hypox mice injected with GH (hypox + GH). The arrows point to the ribonuclease-protected IGF-I, albumin, and GAPDH mRNA bands. Y. RNA, Yeast RNA, serving as a negative control in the assay. B, Chromatin immunoprecipitation assays of STAT5 binding to the putative STAT5 binding site-containing chromosomal regions in the mouse liver. Liver chromatin from hypophysectomized mice receiving GH (+GH) or PBS (−GH) was immunoprecipitated with an anti-STAT5 antibody, and the abundance of 14 putative STAT5 binding site-containing IGF-I DNA regions (S1 to S14) and a Cis promoter region (as a positive control that was known to bind to GH-activated STAT5) as well as an albumin promoter region (as a negative control that was not expected to bind to GH-activated STAT5) in the immunoprecipitated chromatin was quantified by semiquantitative PCR. Shown are representative agarose gel images of PCR analysis of STAT5 antibody-precipitated chromatin. One of the two DNA bands in each lane corresponds to the albumin promoter and the other to S1, S2 … S14 or the Cis promoter. C, Relative abundance of S1, S2 … S14 or the Cis promoter DNA in STAT5 antibody-precipitated liver chromatin from four +GH and four −GH hypophysectomized mice. These data were obtained from densitometric analyses of DNA bands like those in B. Relative abundance on the y-axis represents the ratio of the abundance of S1, S2 … S14 or the Cis promoter DNA to that of the albumin promoter in the same chromatin sample. * and #, P < 0.05 and P < 0.1, respectively, +GH vs. −GH for the same DNA region.

The ChIP assays showed that the STAT5 antibody precipitated more of the Cis promoter DNA from the GH-treated liver than from the control liver (P < 0.05), indicating that GH increased binding of STAT5 to the Cis promoter (Fig. 2, B and C). This result is as expected because the Cis promoter contains four putative STAT5 binding sites that have been previously shown to bind to STAT5 (17,18). The STAT5 antibody immunoprecipitated significantly more (P < 0.05) S1, S5, S9, S10, and S11 DNA from the GH-treated liver than from the control liver (Fig. 2, B and C), whereas it did not enrich the remaining nine IGF-I DNA regions (i.e. S2, S3, S4, S6, S7, S8, S12, S13, and S14) that corresponded to the other 10 putative STAT5 binding sites (Fig. 2, B and C), suggesting that GH increased STAT5 binding to the IGF-I DNA regions S1, S5, S9, S10, and S11 in the liver. Based on Fig. 2B, there appeared to be more albumin promoter DNA than the IGF-I chromosomal regions S6, S7, and S13 in the immunoprecipitated chromatin from both the GH-treated and the control livers. These differences probably do not indicate STAT5 binding to the albumin promoter because it was not more abundant than most of the other IGF-I DNA regions and the Cis promoter in the same chromatin samples (Fig. 2B). Lack of a putative STAT5 binding site and not being regulated by GH (Fig. 2A) also suggest that the albumin promoter is not bound by STAT5. The differences were perhaps caused by more efficient amplification of the albumin promoter than S6, S7, and S13 in the PCRs.

Because the GH-increased binding of STAT5 to S1, S5, S9, S10, and S11 was associated with GH-increased IGF-I mRNA, the nine putative STAT5 binding sites in these chromosomal regions may mediate GH-induced STAT5 activation of IGF-I gene expression in the liver.

GH-stimulated binding of STAT5 to S1, S5, S9, S10, and S11 preceded or coincided with GH-stimulated IGF-I gene transcription in the mouse liver

We next conducted a time-course experiment to determine whether GH-stimulated STAT5 binding to the chromosomal regions S1, S5, S9, S10, and S11 preceded GH-stimulated IGF-I gene transcription in the mouse liver. In this experiment, we treated genetically GH-deficient lit/lit mice with GH, collected livers at different times after GH injection, and analyzed STAT5 binding to S1, S5, S9, S10, and S11 as well as other putative STAT5 binding site-containing chromosomal regions using ChIP. We also measured GH-stimulated changes in IGF-I pre-mRNA and mRNA. As shown in Fig. 3A, liver expression of IGF-I pre-mRNA was significantly higher at 30 min after than before GH administration (P < 0.05) and continued to rise by 2 h after the administration; it started to decline at 6 h after GH administration and was near its basal level by 24 h after the administration. As also shown in Fig. 3A, GH administration caused a similar but delayed time-dependent increase in liver IGF-I mRNA: liver IGF-I mRNA expression was significantly higher at 6 h after than before GH administration (P < 0.05), and it returned to its basal level by 24 h after the administration (Fig. 3A). The delayed response of IGF-I mRNA to GH obviously reflected the fact that IGF-I mRNA is processed from IGF-I pre-mRNA. These time-dependent GH responses of IGF-I mRNA and pre-mRNA in mice were similar to those in rats (23,34). As expected, GH had no effect on liver expression of albumin pre-mRNA or GAPDH mRNA (Fig. 3A). It should be noted that IGF-I mRNA was not completely absent in the liver of the lit/lit mouse (Fig. 3A). This observation is consistent with the studies that indicate the presence of some GH and IGF-I in the serum of the lit/lit mouse (35,36).

Figure 3.

Figure 3

Open in a new tab

GH-stimulated binding of STAT5 to chromosomal regions S1, S5, S9, S10, and S11 preceded or coincided with GH-increased IGF-I transcription in the mouse liver. A, Time courses of GH-stimulated expression of IGF-I mRNA and pre-mRNA. Total RNA and nuclear RNA from GH-deficient lit/lit mice (n = 3) at different times after GH administration were analyzed for the abundance of IGF-I mRNA, IGF-I pre-mRNA, GAPDH mRNA (as a total RNA loading control), and albumin pre-mRNA (as a nuclear RNA loading control) by quantitative real-time RT-PCR. Relative mRNA or pre-mRNA abundance at different times after GH administration is presented as changes over that in lit/lit mice not administered with GH (indicated as time 0 in this figure). Note that the x-axis is not drawn to scale. *, P < 0.05, compared with the zero time value. B, Time courses of GH-stimulated binding of STAT5 to S1, S5, S9, S10, S11, S14 (as a negative control), the Cis promoter (as a positive control), and the albumin promoter (as a DNA loading control) in the livers of lit/lit mice. The ChIP assays were carried out as described for Fig. 2. Note the time-dependent changes in the abundance of the PCR products from amplifying S1, S5, S9, S10, S11, and Cis but not from S14 or Alb. C, Relative abundance of S1, S5, S9, S10, S11, S14, and Cis promoter DNA in STAT5 antibody-precipitated chromatin from lit/lit mice at different times after GH administration (n = 3 for each time point). These data were obtained by densitometric analyses of DNA bands like those in B. Relative abundance on the y-axis represents the ratio of the abundance of S1, S5, S9, S10, S11, S14, or the Cis promoter DNA at a later time to that at the zero time. Note that the x-axis is not drawn to scale.

The ChIP assays indicated that GH administration increased STAT5 binding to S1, S5, S9, S10, and S11 in a time-dependent manner. STAT5 binding to S10 and S11 started to increase 5 min after GH administration and reached the highest level at 2 h; thereafter it declined and returned to its basal level by 24 h after GH administration (Fig. 3, B and C). GH-increased binding of STAT5 to S1, S5 and S9 appeared to be slightly delayed, compared with that to S10 and S11, with significant increases in STAT5 binding to S9 occurring 15 min after GH administration and significant increases in STAT5 binding to S1 and S5 occurring 30 min after GH administration (Fig. 3, B and C). Like STAT5 binding to S10 and S11, that to S1, S5, and S9 also peaked at 2 h after GH administration (Fig. 3, B and C). As expected, GH increased STAT5 binding to the Cis promoter and had no effect on STAT5 binding to the albumin promoter (Fig. 3, B and C). Interestingly, the GH-increased binding of STAT5 to the Cis promoter peaked at both 15 min and 6 h after GH administration (Fig. 3, B and C).

A comparison of the time courses of STAT5 binding to S1, S5, S9, S10, and S11 with those of IGF-I pre-mRNA expression indicated that the increases in STAT5 binding to these chromosomal regions preceded or coincided with GH-stimulated increases in IGF-I pre-mRNA expression (Fig. 3), further suggesting that the STAT5 binding sites in S1, S5, S9, S10, and S11 may mediate GH-induced STAT5 activation of IGF-I gene transcription in the liver.

The putative STAT5 binding sites in S1, S5, S9, S10, and S11 were able to bind to STAT5 in vitro

We next determined whether the 19 putative STAT5 binding sites can directly bind to STAT5 in vitro. We incubated a 15-bp oligonucleotide corresponding to each of the 19 putative STAT5 binding sites with liver nuclear proteins from lit/lit mice treated with GH or PBS. As shown in Fig. 4A, 15 of the 19 oligonucleotides, including S1a, S1b, S2, S3, S5a, S7, S8, S9, S10a, S10b, S11a, S11b, S12, S13, and S14a, formed DNA-protein complexes, which appeared to have the same mobility with liver nuclear proteins from the GH-injected mice. The same complexes were not formed between the oligonucleotides and liver nuclear proteins from PBS-treated lit/lit mice. To confirm the presence of STAT5 in and the specificity of these DNA-protein complexes, the oligonucleotide S1a was further analyzed in supershift and competitive gel-shift assays. As shown in Fig. 4B, the DNA-protein complexes formed between the oligonucleotide S1a and the GH-treated liver nuclear proteins were partially supershifted by a STAT5 antibody. The same complexes were completely competed away by a molar excess (10- or 100-fold) of unlabeled oligonucleotide S1a but were barely affected by the same molar excess of an oliogonucleotide that did not contain a STAT5 consensus binding site (Fig. 4C).

Figure 4.

Figure 4

Open in a new tab

Eight of the nine putative STAT5 binding sites in S1, S5, S9, S10, and S11 sequences bound to STAT5 in vitro. A, Electrophoretic mobility shift assays of the nineteen putative STAT5 binding sites. 32P-labeled double-stranded oligonucleotides (15 bp) corresponding to the putative STAT5 binding sites (i.e., S1a, S1b … S14b) were incubated with liver nuclear proteins from the lit/lit mice injected with GH (+) or PBS (−), followed by polyacrylamide gel electrophoresis. Arrow “B” points to DNA-protein complexes formed specifically between some of the oligonucleotides and the GH-treated liver nuclear proteins. B, Supershift assay of the oligonucleotide S1a. In this assay, the 32P-labeled oligonucleotide S1a was incubated with GH-treated liver nuclear proteins in the presence of anti-STAT5 antibody (anti-STAT5) or rabbit preimmune serum. Arrow “S” indicates a partial supershift of the DNA-protein complex pointed by arrow “B”. Due to extended electrophoresis, the DNA-protein complex in panel A was separated into two bands, which probably reflect phosphorylation of STAT5 at two distinct sites (38). C, Competitive gel-shift assay of the oligonucleotide S1a. In this assay, the 32P-labeled oligonucleotide S1a was incubated with GH-treated liver nuclear proteins in the presence of 1×, 10×, or 100× molar excess of unlabeled S1a or unlabeled oligonucleotide H that did not contain a STAT5 binding site.

Based on these gel-shift assays, all of the putative STAT5 binding sites in S1, S5, S9, S10, and S11 except the 3′ end STAT5 binding site in S5 were able to bind to GH-activated liver STAT5 (Fig. 4A). These data strongly suggest that the GH-stimulated binding of STAT5 to these chromosomal regions in vivo (Figs. 2 and 3) is due to the presence of STAT5 binding sites in them. Based on the gel-shift assays, the putative STAT5 binding sites in S3, S7, S8, S12, S13, and S14 were also able to bind to STAT5 in vitro (Fig. 4A), but GH did not stimulate STAT5 binding to S3, S7, S8, S12, S13, and S14 in the liver (Figs. 2 and 3), indicating that transcription factor binding in cells may be chromatin or sequence context dependent.

S1, S5, S9, S10, and S11 were each able to mediate GH-induced STAT5 activation of reporter gene expression

We next determined whether S1, S5, S9, S10, and S11 as well as other putative STAT5 binding site-containing IGF-I DNA regions can mediate GH-induced STAT5 activation of gene expression, using cotransfection assays. In these assays, a luciferase reporter gene construct bearing a minimal TK promoter and one of the 14 STAT5 binding site-containing IGF-I DNA regions (i.e. S1, S2 … S14) was cotransfected with a GHR expression plasmid and a STAT5b expression plasmid into CHO cells, and reporter gene response to GH was measured. As shown in Fig. 5, GH increased reporter gene expression from the constructs containing S1, S5, S9, S10, and S11 by 50–220%, compared with PBS (P < 0.05), and GH had no effect on reporter gene expression from the promoter-only plasmid. These data indicate that S1, S5, S9, S10, and S11 can mediate GH-induced STAT5 activation of gene expression.

Figure 5.

Figure 5

Open in a new tab

The IGF-I DNA regions S1, S5, S9, S10, and S11 mediated GH-induced STAT5 activation of gene expression. The CHO cells were transfected with a GHR expression plasmid, a STAT5b expression plasmid, a renilla luciferase plasmid (as a transfection efficiency control), and a pGL2TK-based firefly luciferase reporter plasmid that contained S1, S2 … S14, or Cis. The CHO cells were treated with PBS or GH for 8 h before dual-luciferase assay. This experiment was repeated four times. Fold activation on the y-axis represents the ratio of reporter gene expression in the presence of GH to that in the presence of PBS. * and #, P < 0.05 and P < 0.1, respectively, compared with pGL2TK.

GH also stimulated reporter gene expression from the constructs containing S2, S7, and S8, but it had no effect on reporter gene expression from the constructs containing S4 and S6 (Fig. 5). These transfection data were consistent with the gel-shift data that the putative STAT5 binding sites in S2, S7, and S8 were able to bind to STAT5, whereas those in S4 and S6 were not (Fig. 4A). However, GH did not stimulate reporter gene expression from the constructs containing S12, S13, and S14, despite the fact that the putative STAT5 binding sites in these sequences were able to bind to STAT5 in vitro (Fig. 4A), indicating that STAT5 binding does not always lead to transcriptional activation.

Multiple STAT5 binding sites together mediated greater GH-induced STAT5 activation of reporter gene expression

One possible functional advantage of the presence of multiple STAT5 binding sites in the IGF-I gene is that multiple binding sites can cooperate to mediate greater STAT5 activation of IGF-I gene expression. To test this possibility, GH-induced STAT5 activation of reporter gene expression was compared between a plasmid that contained the S1 region as the enhancer and a plasmid that contained S5, S9, S10, and S11, in addition to S1, as the enhancers. As shown in Fig. 6, GH-stimulated luciferase activity from the plasmid containing S1, S5, S9, S10, and S11 was 40–50% greater (P < 0.05) than that from the plasmid containing S1 in CHO cells, no matter whether the promoter of the plasmid was a TK promoter or an IGF-I promoter. These data suggest that the STAT5 binding sites in S1, S5, S9, S10, and S11 working together may mediate greater STAT5 activation of IGF-I gene expression in vivo.

Figure 6.

Figure 6

Open in a new tab

S1, S5, S9, S10, and S11 together mediated greater STAT5 activation of gene expression than did S1. The plasmids pGL2TK-S1 contained the TK promoter as the promoter and S1 as the enhancer. The plasmid pGL2IP1-S1 contained a 3.2-kb mouse IGF-I promoter and S1. The plasmids pGL2TK-S1S5S9S10S11 and pGL2IP1-S1S5S9S10S11 contained S5, S9, S10, and S11 in addition to S1 as the enhancers. The cotransfection analyses to compare GH-activated luciferase expression from these plasmids were conducted essentially as described for Fig. 5. *, P < 0.05, compared with pGL2TK-S1 or pGL2IP1-S1.

Discussion

The major results of this study are: 1) GH increased STAT5 binding to four distal intronic regions and one distal 5′-flanking region of the IGF-I gene in the mouse liver that together contain nine putative STAT5 binding sites; 2) GH-increased binding of STAT5 to these DNA regions preceded or coincided with GH-increased IGF-I gene transcription in the liver; 3) eight of the nine putative STAT5 binding sites in these DNA regions were found to bind to STAT5 in vitro; 4) each of the five DNA regions was able to mediate GH-induced STAT5 activation of reporter gene expression in cultured cells; and 5) the five DNA regions together mediated a greater GH-induced STAT5 activation of reporter gene expression than did a single DNA region containing two STAT5 binding sites. These results indicate that GH-induced STAT5 activation of IGF-I gene expression in the mouse liver may involve eight STAT5 binding sites located distantly from the IGF-I promoter.

The presence of eight GH-responsive STAT5 binding sites controlling IGF-I gene expression is a stark contrast to the identification of one or two STAT5 binding sites in most STAT5-regulated genes (29,37). A question therefore arises: why are so many STAT5 binding sites involved in GH regulation of IGF-I gene expression? Our cotransfection analyses have shown that, although each of the five STAT5 binding site-containing IGF-I DNA regions can mediate GH-induced STAT5 activation of gene expression, the magnitude of the activation is relatively small. Our cotransfection analyses also have shown that the five STAT5 binding site-containing regions together mediate greater GH-induced STAT5 activation of reporter gene expression than a single STAT5 binding site-containing region. Therefore, a possible reason for the existence of multiple STAT5 binding sites for GH regulation of IGF-I gene expression is that each of these STAT5 binding sites is transactivationally weak, and a combined action of them is needed to have a significant effect on IGF-I gene transcription. This possibility seems to be supported by our observation that GH-induced maximum binding of STAT5 to the five chromosomal regions containing multiple STAT5 binding sites coincides with GH-induced maximum transcription of the IGF-I gene in the liver. It should be noted that the dosage of bovine GH used in this study was supraphysiological (38); the STAT5 binding sites identified in this study therefore may, or may not, correspond to the STAT5 binding sites at physiological GH concentrations. It is possible that not all of the eight STAT5 binding sites identified in this study are activated under physiological concentrations of GH, and it is also possible that the STAT5 binding sites identified in this study may be selectively activated in the liver, depending on the physiological condition of the animal.

Two of the STAT5 binding sites identified in this study lie more than 70 kb 5′ from the transcription start site of the IGF-I gene, and five of the remaining six are located at least 26 kb downstream from the transcription start site. The distal locations of these GH-responsive STAT5 binding sites seem to explain previous failure to identify GH-responsive elements in the IGF-I promoter or proximal intron in cell culture and transgenic mice (39,40,41) and also seem to be in line with the increasing identification of long-range enhancers (42,43,44,45). But how do these STAT5 binding sites mediate STAT5 action to the IGF-I promoter over seemingly very long distances? Two major models have been proposed for distant enhancer-promoter interaction (42,43). One is the looping model, in which the long intervening DNA loops out to allow direct contact between the transcription factor bound at the distant enhancer and the general transcription machinery or sequence-specific transcription factors bound at the promoter; the other model is the tracking model, and based on this model, the transcription factor and perhaps other associated proteins bound at the distant enhancer track along the intervening DNA to the promoter. Increasing evidence favors the looping model (43). For example, it was recently reported that an interchromosomal enhancer regulates expression of olfactory receptor genes through direct physical contact with the olfactory receptor gene promoters (45). We speculate that the eight distal STAT5 binding sites in the IGF-I gene affect its transcription also through physical association with the IGF-I promoter. Such a physical association would not only allow direct STAT5 interaction with the general transcription machinery and/or sequence-specific transcription factors bound at the promoter but also bring all the distal STAT5 binding sites together for potential STAT5-STAT5 interaction, which has been shown to benefit STAT5 binding and STAT5 transactivation (17,46,47).

Our in vitro and in vivo DNA-protein binding analyses indicate that several STAT5 binding sites (S2, S3, S7, S8, S12, S13, and S14) can bind to STAT5 in vitro, but their corresponding chromosomal regions do not bind to GH-activated STAT5 in the liver. Whereas these data underscore the importance of validating the physiological relevance of in vitro observations, they also make one wonder why the STAT5 binding sites in S1, S5, S9, S10, and S11 can bind to STAT5 both in vitro and in vivo. As pointed out earlier, S1, S5, S10, and S11 each contain a pair of STAT5 binding sites. Although S9 contains only one conserved STAT5 binding site, there is a nonconserved STAT5 consensus site 240 bp upstream from the conserved STAT5 binding site. Two adjacent STAT5 binding sites located on the same side of the DNA helix can stabilize each other’s binding to STAT5 through formation of STAT5 tetramers (46). It is therefore tempting to speculate that the paired STAT5 binding sites in S1, S5, S9, S10, and S11 might allow for the formation of STAT5 tetramers or some type of STAT5-STAT5 interaction, thereby strengthening their binding to STAT5 in vivo. Alternatively, S1, S5, S9, S10, and S11 might contain binding sites for other transcription factors, and the binding of these other transcription factors either makes the neighboring chromatin more accessible to STAT5 or makes STAT5 binding to the chromatin more stable.

Have all the STAT5 binding sites involved in GH stimulation of IGF-I gene expression been identified? In this study, the identification of the eight STAT5 binding sites from a list of evolutionarily conserved STAT5 consensus sequences was initially based on the observation that the chromosomal regions containing them were bound by STAT5 in the liver at 6 h after GH treatment. It has been shown that STAT5 activation and nuclear translocation can occur within as short as 15 min of GH administration (38), and our time course experiment in this study also indicated that maximal STAT5 binding to chromatin in the lit/lit mouse liver may occur long before 6 h after GH injection. Therefore, some of the evolutionarily conserved STAT5 consensus sequences that bind to STAT5 only in the early stage of GH action on IGF-I gene expression might have been missed in this study. As just mentioned, the eight STAT5 binding sites were identified from a list of evolutionarily conserved putative STAT5 binding sites on the basis of the widely accepted hypothesis that evolutionarily conserved noncoding sequences have great potential to function as regulatory sequences (30,31). However, a recent study comparing transcription factor binding sites between three related yeast species demonstrated that transcription factor binding sites are not necessarily conserved across species (48). Thus, it is possible that some of the unconserved STAT5 consensus sequences that were not tested in this study might be bona fide STAT5 binding sites. Whereas most of the natural STAT5 binding sites conform to the consensus sequence TTCNNNGAA, sequence variants that differ from this consensus sequence by one or two nucleotides have been shown to bind to STAT5, too, at least under in vitro situations (29). The IGF-I locus contains hundreds of such variants of TTCNNNGAA, and some of them might be true STAT5 binding sites.

It is likely that GH regulation of IGF-I gene expression involves transcription factors in addition to STAT5. The human IGF-I promoter contains binding sites for liver-enriched transcription factor hepatocyte nuclear factor (HNF)-3β (49,50). Mice bearing a truncated GHR have reduced expression of HNF-3β protein in the liver (51). The mouse IGF-I promoter contains three putative HNF-3β binding sites, and overexpression of HNF-3β significantly enhances reporter gene expression from it in CHO cells (Eleswarapu, S., and H. Jiang, unpublished data). These observations together suggest that GH may induce IGF-I mRNA expression in the liver of the mouse by stimulating HNF-3β protein expression. Besides STAT5 and HNF-3β, many other transcription factors, such as STAT1, STAT3, HNF-1, HNF-4, and HNF-6 are activated or regulated by GH (52,53). Therefore, it would not be surprising if some of these transcription factors are found to participate in GH regulation of IGF-I gene expression.

In summary, the results of this study suggest that GH-induced STAT5 activation of IGF-I gene expression in the mouse liver might be mediated by at least eight STAT5 binding sites that are all located distantly from the IGF-I promoter. The identification of multiple distal STAT5 binding sites underscores the complexity of the mechanism that mediates GH regulation of IGF-I gene expression.

Acknowledgments

We thank B. Heid for technical assistance with the animal experiments and Y. Zhou for performing some of the transfection assays. We are grateful to Dr. R. H. Costa (University of Illinois at Chicago) for providing the HNF-3b expression plasmid.

Footnotes

This work was supported by National Institutes of Health Grant DK67961 (to H.J.).

Disclosure Statement: The authors have nothing to disclose.

First Published Online February 14, 2008

Abbreviations: ChIP, Chromatin immunoprecipitation assay; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GHR, GH receptor; HNF, hepatocyte nuclear factor; PMSF, phenylmethylsulfonyl fluoride; RPA, ribonuclease protection assay; STAT, signal transducer and activator of transcription; TK, thymidine kinase.

References

  1. Stewart CE, Rotwein P 1996 Growth, differentiation, and survival: multiple physiological functions for insulin-like growth factors. Physiol Rev 76:1005–1026 [DOI] [PubMed] [Google Scholar]
  2. Baker J, Liu JP, Robertson EJ, Efstratiadis A 1993 Role of insulin-like growth factors in embryonic and postnatal growth. Cell 75:73–82 [PubMed] [Google Scholar]
  3. Liu JP, Baker J, Perkins AS, Robertson EJ, Efstratiadis A 1993 Mice carrying null mutations of the genes encoding insulin-like growth factor I (Igf-1) and type 1 IGF receptor (Igf1r). Cell 75:59–72 [PubMed] [Google Scholar]
  4. Pollak MN, Schernhammer ES, Hankinson SE 2004 Insulin-like growth factors and neoplasia. Nat Rev Cancer 4:505–518 [DOI] [PubMed] [Google Scholar]
  5. Samani AA, Yakar S, Leroith D, Brodt P 2007 The role of the IGF system in cancer growth and metastasis: overview and recent insights. Endocr Rev 28:20–47 [DOI] [PubMed] [Google Scholar]
  6. Kenyon C 2005 The plasticity of aging: insights from long-lived mutants. Cell 120:449–460 [DOI] [PubMed] [Google Scholar]
  7. Bartke A 2005 Minireview: role of the growth hormone/insulin-like growth factor system in mammalian aging. Endocrinology 146:3718–3723 [DOI] [PubMed] [Google Scholar]
  8. Daughaday WH, Rotwein P 1989 Insulin-like growth factors I and II. Peptide, messenger ribonucleic acid and gene structures, serum, and tissue concentrations. Endocr Rev 10:68–91 [DOI] [PubMed] [Google Scholar]
  9. Yakar S, Liu JL, Stannard B, Butler A, Accili D, Sauer B, LeRoith D 1999 Normal growth and development in the absence of hepatic insulin-like growth factor I. Pro Nat Acad Sci USA 96:7324–7329 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Udy GB, Towers RP, Snell RG, Wilkins RJ, Park SH, Ram PA, Waxman DJ, Davey HW 1997 Requirement of STAT5b for sexual dimorphism of body growth rates and liver gene expression. Proc Natl Acad Sci USA 94:7239–7244 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Davey HW, Xie T, McLachlan MJ, Wilkins RJ, Waxman DJ, Grattan DR 2001 STAT5b is required for GH-induced liver IGF-I gene expression. Endocrinology 142:3836–3841 [DOI] [PubMed] [Google Scholar]
  12. Woelfle J, Billiard J, Rotwein P 2003 Acute control of insulin-like growth factor-I gene transcription by growth hormone through Stat5b. J Biol Chem 278:22696–22702 [DOI] [PubMed] [Google Scholar]
  13. Kofoed EM, Hwa V, Little B, Woods KA, Buckway CK, Tsubaki J, Pratt KL, Bezrodnik L, Jasper H, Tepper A, Heinrich JJ, Rosenfeld RG 2003 Growth hormone insensitivity associated with a STAT5b mutation. N Engl J Med 349:1139–1147 [DOI] [PubMed] [Google Scholar]
  14. Woelfle J, Chia DJ, Rotwein P 2003 Mechanisms of growth hormone (GH) action. Identification of conserved Stat5 binding sites that mediate GH-induced insulin-like growth factor-I gene activation. J Biol Chem 278:51261–51266 [DOI] [PubMed] [Google Scholar]
  15. Wang Y, Jiang H 2005 Identification of a distal STAT5-binding DNA region that may mediate growth hormone regulation of insulin-like growth factor-I gene expression. J Biol Chem 280:10955–10963 [DOI] [PubMed] [Google Scholar]
  16. Chia DJ, Ono M, Woelfle J, Schlesinger-Massart M, Jiang H, Rotwein P 2006 Characterization of distinct Stat5b binding sites that mediate growth hormone-stimulated IGF-I gene transcription. J Biol Chem 281:3190–3197 [DOI] [PubMed] [Google Scholar]
  17. Verdier F, Rabionet R, Gouilleux F, Beisenherz-Huss C, Varlet P, Muller O, Mayeux P, Lacombe C, Gisselbrecht S, Chretien S 1998 A sequence of the CIS gene promoter interacts preferentially with two associated STAT5A dimers: a distinct biochemical difference between STAT5A and STAT5B. Mol Cell Biol 18:5852–5860 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Matsumoto A, Masuhara M, Mitsui K, Yokouchi M, Ohtsubo M, Misawa H, Miyajima A, Yoshimura A 1997 CIS, a cytokine inducible SH2 protein, is a target of the JAK-STAT5 pathway and modulates STAT5 activation. Blood 89:3148–3154 [PubMed] [Google Scholar]
  19. Noshiro M, Negishi M 1986 Pretranslational regulation of sex-dependent testosterone hydroxylases by growth hormone in mouse liver. J Biol Chem 261:15923–15927 [PubMed] [Google Scholar]
  20. Pell JM, Bates PC 1992 Differential actions of growth hormone and insulin-like growth factor-I on tissue protein metabolism in dwarf mice. Endocrinology 130:1942–1950 [DOI] [PubMed] [Google Scholar]
  21. Davey HW, Park SH, Grattan DR, McLachlan MJ, Waxman DJ 1999 STAT5b-deficient mice are growth hormone pulse-resistant—role of STAT5b in sex-specific liver P450 expression. J Biol Chem 274:35331–35336 [DOI] [PubMed] [Google Scholar]
  22. Bichell DP, Kikuchi K, Rotwein P 1992 Growth hormone rapidly activates insulin-like growth factor I gene transcription in vivo. Mol Endocrinol 6:1899–1908 [DOI] [PubMed] [Google Scholar]
  23. Carlsson L, Nilsson I, Oscarsson J 1998 Hormonal regulation of liver fatty acid-binding protein in vivo and in vitro: effects of growth hormone and insulin. Endocrinology 139:2699–2709 [DOI] [PubMed] [Google Scholar]
  24. Lin SC, Lin CR, Gukovsky I, Lusis AJ, Sawchenko PE, Rosenfeld MG 1993 Molecular basis of the little mouse phenotype and implications for cell type-specific growth. Nature 364:208–213 [DOI] [PubMed] [Google Scholar]
  25. Godfrey P, Rahal JO, Beamer WG, Copeland NG, Jenkins NA, Mayo KE 1993 GHRH receptor of little mice contains a missense mutation in the extracellular domain that disrupts receptor function. Nat Genet 4:227–232 [DOI] [PubMed] [Google Scholar]
  26. Odom DT, Zizlsperger N, Gordon DB, Bell GW, Rinaldi NJ, Murray HL, Volkert TL, Schreiber J, Rolfe PA, Gifford DK, Fraenkel E, Bell GI, Young RA 2004 Control of pancreas and liver gene expression by HNF transcription factors. Science 303:1378–1381 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Ariyoshi K, Nosaka T, Yamada K, Onishi M, Oka Y, Miyajima A, Kitamura T 2000 Constitutive activation of STAT5 by a point mutation in the SH2 domain. J Biol Chem 275:24407–24413 [DOI] [PubMed] [Google Scholar]
  28. Darnell Jr JE 1997 STATs and gene regulation. Science 277:1630–1635 [DOI] [PubMed] [Google Scholar]
  29. Ehret GB, Reichenbach P, Schindler U, Horvath CM, Fritz S, Nabholz M, Bucher P 2001 DNA binding specificity of different STAT proteins. Comparison of in vitro specificity with natural target sites. J Biol Chem 276:6675–6688 [DOI] [PubMed] [Google Scholar]
  30. Boffelli D, Nobrega MA, Rubin EM 2004 Comparative genomics at the vertebrate extremes. Nat Rev Genet 5:456–465 [DOI] [PubMed] [Google Scholar]
  31. Prakash A, Tompa M 2005 Discovery of regulatory elements in vertebrates through comparative genomics. Nat Biotechnol 23:1249–1256 [DOI] [PubMed] [Google Scholar]
  32. Frazer KA, Pachter L, Poliakov A, Rubin EM, Dubchak I 2004 VISTA: computational tools for comparative genomics. Nucleic Acids Res 32(Web Server issue):W273–W279 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Norstedt G, Palmiter R 1984 Secretory rhythm of growth hormone regulates sexual differentiation of mouse liver. Cell 36:805–812 [DOI] [PubMed] [Google Scholar]
  34. Gronowski AM, Rotwein P 1995 Rapid changes in gene expression after in vivo growth hormone treatment. Endocrinology 136:4741–4748 [DOI] [PubMed] [Google Scholar]
  35. Cheng TC, Beamer WG, Phillips 3rd JA, Bartke A, Mallonee RL, Dowling C 1983 Etiology of growth hormone deficiency in little, Ames, and Snell dwarf mice. Endocrinology 113:1669–1678 [DOI] [PubMed] [Google Scholar]
  36. Donahue LR, Beamer WG 1993 Growth hormone deficiency in ‘little’ mice results in aberrant body composition, reduced insulin-like growth factor-I and insulin-like growth factor-binding protein-3 (IGFBP-3), but does not affect IGFBP-2, -1 or -4. J Endocrinol 136:91–104 [DOI] [PubMed] [Google Scholar]
  37. Grimley PM, Dong F, Rui H 1999 Stat5a and Stat5b: fraternal twins of signal transduction and transcriptional activation. Cytokine Growth Factor Rev 10:131–157 [DOI] [PubMed] [Google Scholar]
  38. Ram PA, Park SH, Choi HK, Waxman DJ 1996 Growth hormone activation of Stat 1, Stat 3, and Stat 5 in rat liver. Differential kinetics of hormone desensitization and growth hormone stimulation of both tyrosine phosphorylation and serine/threonine phosphorylation. J Biol Chem 271:5929–5940 [DOI] [PubMed] [Google Scholar]
  39. Thomas MJ, Kikuchi K, Bichell DP, Rotwein P 1994 Rapid activation of rat insulin-like growth factor-I gene transcription by growth hormone reveals no alterations in deoxyribonucleic acid-protein interactions within the major promoter. Endocrinology 135:1584–1592 [DOI] [PubMed] [Google Scholar]
  40. Le Stunff C, Thomas MJ, Rotwein P 1995 Rapid activation of rat insulin-like growth factor-I gene transcription by growth hormone reveals no changes in deoxyribonucleic acid-protein interactions within the second promoter. Endocrinology 136:2230–2237 [DOI] [PubMed] [Google Scholar]
  41. Ye P, Umayahara Y, Ritter D, Bunting T, Auman H, Rotwein P, D’Ercole AJ 1997 Regulation of insulin-like growth factor I (IGF-I) gene expression in brain of transgenic mice expressing an IGF-I-luciferase fusion gene. Endocrinology 138:5466–5475 [DOI] [PubMed] [Google Scholar]
  42. Kleinjan DA, van Heyningen V 2005 Long-range control of gene expression: emerging mechanisms and disruption in disease. Am J Hum Genet 76:8–32 [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Dean A 2006 On a chromosome far, far away: LCRs and gene expression. Trends Genet 22:38–45 [DOI] [PubMed] [Google Scholar]
  44. Carroll JS, Liu XS, Brodsky AS, Li W, Meyer CA, Szary AJ, Eeckhoute J, Shao W, Hestermann EV, Geistlinger TR, Fox EA, Silver PA, Brown M 2005 Chromosome-wide mapping of estrogen receptor binding reveals long-range regulation requiring the forkhead protein FoxA1. Cell 122:33–43 [DOI] [PubMed] [Google Scholar]
  45. Lomvardas S, Barnea G, Pisapia DJ, Mendelsohn M, Kirkland J, Axel R 2006 Interchromosomal interactions and olfactory receptor choice. Cell 126:403–413 [DOI] [PubMed] [Google Scholar]
  46. Meyer WK, Reichenbach P, Schindler U, Soldaini E, Nabholz M 1997 Interaction of STAT5 dimers on two low affinity binding sites mediates interleukin 2 (IL-2) stimulation of IL-2 receptor α gene transcription. J Biol Chem 272:31821–31828 [DOI] [PubMed] [Google Scholar]
  47. John S, Vinkemeier U, Soldaini E, Darnell Jr JE, Leonard WJ 1999 The significance of tetramerization in promoter recruitment by Stat5. Mol Cell Biol 19:1910–1918 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Borneman AR, Gianoulis TA, Zhang ZD, Yu H, Rozowsky J, Seringhaus MR, Wang LY, Gerstein M, Snyder M 2007 Divergence of transcription factor binding sites across related yeast species. Science 317:815–819 [DOI] [PubMed] [Google Scholar]
  49. Nolten LA, Steenbergh PH, Sussenbach JS 1995 Hepatocyte nuclear factor 1 alpha activates promoter 1 of the human insulin-like growth factor I gene via two distinct binding sites. Mol Endocrinol 9:1488–1499 [DOI] [PubMed] [Google Scholar]
  50. Nolten LA, Steenbergh PH, Sussenbach JS 1996 The hepatocyte nuclear factor 3β stimulates the transcription of the human insulin-like growth factor I gene in a direct and indirect manner. J Biol Chem 271:31846–31854 [DOI] [PubMed] [Google Scholar]
  51. Rowland JE, Lichanska AM, Kerr LM, White M, d’Aniello EM, Maher SL, Brown R, Teasdale RD, Noakes PG, Waters MJ 2005 In vivo analysis of growth hormone receptor signaling domains and their associated transcripts. Mol Cell Biol 25:66–77 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Rastegar M, Lemaigre FP, Rousseau GG 2000 Control of gene expression by growth hormone in liver: key role of a network of transcription factors. Mol Cell Endocrinol 164:1–4 [DOI] [PubMed] [Google Scholar]
  53. Ceseña TI, Cui TX, Piwien-Pilipuk G, Kaplani J, Calinescu AA, Huo JS, Iñiguez-Lluhí JA, Kwok R, Schwartz J 2007 Multiple mechanisms of growth hormone-regulated gene transcription. Mol Genet Metab 90:126–133 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Endocrinology are provided here courtesy of The Endocrine Society

RESOURCES