help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhu, J.-L.
Right arrow Articles by Phillips, L. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhu, J.-L.
Right arrow Articles by Phillips, L. S.
Endocrinology Vol. 140, No. 10 4761-4771
Copyright © 1999 by The Endocrine Society

ARTICLES

Identification of Core Sequences Involved in Metabolism-Dependent Nuclear Protein Binding to the Rat Insulin-Like Growth Factor I Gene1

Juan-Li Zhu, Ching-I Pao, Edward Hunter, Jr., Kai-wei M. Lin, Guang-jer Wu and Lawrence S. Phillips

Departments of Medicine (J.-L.Z., C.-I.P., K.-w.M.L., L.S.P.) and Microbiology-Immunology (G.-j.W.), Emory University School of Medicine, Department of Biology, Morris Brown College (E.H.), Atlanta, Georgia 30322

Address all correspondence and requests for reprints to: Lawrence S. Phillips, M.D., Division of Endocrinology and Metabolism, Department of Medicine, Emory University School of Medicine, 1639 Pierce Drive, Woodruff Memorial Research Building, Room 1301, Atlanta, Georgia 30322. E-mail: medlsp{at}emory.edu


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
In the liver, most insulin-like growth factor I (IGF-I) transcripts originate in exon 1, where important cis-regulatory regions are located downstream from the major transcription initiation sites. Within these regions, we have attempted to identify sequences which are involved in the decrease in IGF-I gene transcription associated with diabetes mellitus. The function of different genomic templates was assessed by in vitro transcription, which revealed a consistent 50–80% decrease in the activity of nuclear extracts from streptozotocin-diabetic as compared with normal rats. The disparity in transcriptional activity between normal and diabetic nuclear extracts was reduced with templates containing 11-bp mutations within DNase I protected regions III or V (+42 and +129 bp, respectively, from the major transcription initiation site), but a mutation between regions IV and V had little effect. Within region III, gel mobility shift analysis and methylation interference studies indicated that DNA-protein interactions involve a GCGC core sequence. In region V, gel mobility shift studies and uracil interference analysis revealed interactions involving a TTAT core. While gel mobility shift analysis and transient transfection studies indicate that the GCGC core sequence in region III recognizes C/EBP, the AT-rich sequence in region V is likely to recognize a protein with homeodomain characteristics. Identification of the nuclear factor(s) interacting with regions III and V, downstream from exon 1 initiation sites, will be important for understanding the mechanism of reduced IGF-I gene transcription due to diabetes mellitus.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
INSULIN-LIKE GROWTH FACTOR I (IGF-I) is a critical cellular regulatory factor with amino acid sequence, tertiary structure, and biological actions similar to those of insulin (1). IGF-I is important for both fetal and postnatal growth and development, and plays an anabolic role in metabolic regulation as well (2, 3, 4). IGF-I was first thought to be regulated largely by GH. However, nutrition, local cellular factors, and other hormones also modulate IGF-I production (5, 6, 7). IGF-I messenger RNA (mRNA) has been found in many sites, but expression is particularly high in the liver, where IGF-I appears to be produced and released for action in an endocrine rather than a paracrine or autocrine mode (2, 3, 4, 8).

The IGF-I gene extends over 80 kb, and contains at least six exons (9). Although IGF-I mRNA species approximately 7.5, 2, and 1.2 kb in size are found in many tissues, such heterogeneity appears to be due largely to differences in polyadenylation (10). While abundance of the 7.5-kb species may be determined in part by changes in mRNA stability (11), strong correlations between IGF-I expression and rates of transcription in nuclear run-on assays indicate a major role for regulation at the level of gene transcription (12). IGF-I transcripts originate at multiple initiation sites in both exon 1 and exon 2, with splicing to common regions in exons 3 and 4; alternative splicing in exon 6 leads to IGF-Ia and IGF-Ib species (13). Under a variety of metabolic conditions, the majority of IGF-I transcripts appear to be initiated in exon 1 (14). Because the IGF-I gene is complex and poorly expressed in many immortal cell lines (15, 16), there is relatively little understanding of the IGF-I promoter region. Our laboratory and others have reported that transcription in exon 1 is initiated mainly at two sites, approximately 100 bp apart (12, 17). Relative to the major downstream initiation site, Hall et al. (18) have reported that reduction of 5' sequence from approximately 1.0 to 0.5 kb is attended by an 80% decrease in expression in SK-N-MC neuroblastoma cells, and we (19) have found that reduction from -300 to -54 bp results in a 75% decrease in expression in an in vitro transcription system, although Lowe et al. (20, 21) have reported that 5' deletions may increase expression in rat fibroblasts and C6 glioma cells. Low expression in constructs with 5' deletions may reflect reduction in binding of transcription factors such as HNF-1 and C/EBP, which Nolten et al. (22, 23) have shown to promote transcription of IGF-I constructs in Hep3B cells. Reduction in 3' sequence appears to reduce expression as well, both in SK-N-MC cells (18) and in C6 glioma cells (21, 24), although not in rat fibroblasts (21). We have also found that 3' regions are important for expression, both when constructs are transfected into rat hepatocytes in primary culture (25) and when rat liver nuclear extracts are examined by in vitro transcription (26).

Although HNF-1, HNF-3, and C/EBP bind to 5' regions of the IGF-I promoter in exon 1 (23, 27), very little is known about DNA-protein interactions in the 3' region. On the basis of DNase I footprinting, results from our laboratory (26) and that of Thomas et al. (28) indicate several areas of DNA-protein interaction within the 3' region (Fig. 1Go); we have found diabetes-sensitive binding in downstream regions III and V (26)—+42 and +129 bp, respectively, from the major transcription initiation site (14). In the present studies, we have used in vitro transcription, gel mobility shift and methylation/uracil interference analysis to identify core binding sequences within regions III and V. A GCGC sequence in region III and an AT-rich sequence in region V appear to be particularly important for metabolically regulated nuclear factor binding to the IGF-I gene in diabetes mellitus.



View larger version (6K):
[in this window]
[in a new window]
  		Figure 1. Diagram of exon 1 of the rat IGF-I gene, showing transcription initiation sites and DNase I footprints in the downstream region.

 

   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Chemicals
Restriction endonucleases and DNA modifying enzymes were obtained from New England Biolabs (Beverly, MA). RNase-free DNase I and SuperScript reverse transcriptase were ordered from Life Technologies, Inc. (Gaithersburg, MD). Nucleotides [{gamma}-32P] (ATP) (6000 Ci/mmol), [{alpha}-32P]deoxy (ATP, GTP, TTP, and CTP) (800 Ci/mmol) were purchased from Amersham Pharmacia Biotech (Arlington Heights, IL). Oligonucleotides were obtained from Life Technologies, Inc. Proteinase inhibitors were ordered from Roche Molecular Biochemicals (Indianapolis, IN). Nonfat dry milk was obtained from Bio-Rad Laboratories, Inc. (Hercules, CA), and streptozotocin was from Pfanstiehl (Waukegan, IL). All other molecular biology grade chemicals were purchased from Sigma (St. Louis, MO).

Animals
Male Sprague Dawley rats (Charles River Laboratories, Inc., Lexington, MA), weighing 160–180 g, were fed ad libitum. Diabetes was produced through tail vein injection of streptozotocin at 250 mg/kg. Animals were killed by cervical dislocation, and livers were used for nuclear extract preparation immediately.

Liver nuclear extract preparation
Nuclear extracts were prepared according to the methods described previously (26), except that 1% nonfat dry milk was included in the homogenization buffer (29). The homogenate was layered onto a 2 M sucrose cushion and centrifuged at 27,000 rpm in an SW28 rotor for 1 h at 2 C. Nuclei were lysed by adding 1/10 volume of 4 M ammonium sulfate, nuclear proteins were concentrated by (NH4)2SO4 precipitation (0.33 g/ml), and finally dialyzed against buffer containing 25 mM HEPES, pH 7.6, 100 mM KCl, 0.1 mM EDTA, 1 mM DTT, and 10% glycerol for 4 h at 4 C, and frozen in aliquots at -70 C.

Gel mobility shift assay
Oligonucleotides used for gel mobility shift assays are summarized in Table 1Go. Additional oligonucleotides used in competition studies include: C/EBP, 5'-TGCAGATTGCGCAATCTGCA-3'; AP1, 5'-CGCTTGATGACTCACCGGAA-3'; AP2, 5'-GATCGAACTGACCGCCCGC-GGCCCGT-3'; and TFIID, 5'-GCAGAGCATATAAAATGAGGTAG-GA-3'. Oligonucleotides were annealed, end labeled, then gel purified as described (26). The binding reaction was carried out with 8 µg of extract in 10 mM Tris, pH 7.5, 100 mM NaCl, 1 mM EDTA, 0.5 mM DTT, 0.2% NP-40, 1 mg/ml of BSA, and 20 µg/ml of salmon sperm DNA. For competition studies, extract was first incubated on ice with 25- to 200-fold molar excess of competitor before addition of probe. DNA/protein complexes were resolved on a 6% polyacrylamide gel at 11 v/cm in 1 x TGE (25 mM Tris, pH 8.0, 190 mM glycine, and 1 mM EDTA) for 2 h at 4 C, and visualized by autoradiography. Supershift analysis used 1 µg of C/EBP antiserum and preimmune serum from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA).


View this table:
[in this window]
[in a new window]
  		Table 1. Oligonucleotides used as competitors in gel mobility shift analyses

 
Methylation interference assay
A 272-bp AccI/BglII (-54/+219) fragment was end labeled, partially methylated with dimethylsulfate, and used as a probe for gel mobility shift analysis. Bound and free DNA were recovered, purified, and finally cleaved with piperidine at 95 C for 30 min as described (30). The sample was electrophoresed on an 8 M urea, 6% polyacrylamide gel, and visualized by autoradiography.

Uracil interference assay
To study DNA/protein interactions involving region V, a 77-bp probe (+103/+179) was synthesized by PCR. Oligonucleotides 5'-TCCCATCTCTCTGGA-3' and 5'-CGCTTCTGAAGTACAAAG-3' were end labeled and used as primers. The reactions were performed with 15 cycles of amplification in the presence of 1 x PCR buffer (Perkin-Elmer Corp.), 100 ng 272-bp genomic DNA fragment (-54/+219 bp), 200 µM dGTP, dCTP, dATP, and UTP, 20 pmol of both labeled and unlabeled primers, and 5 U of Taq polymerase. The 77-bp uracil-substituted fragment was gel purified and used as a probe in a gel mobility shift assay. The bound and unbound DNA were repurified and digested with 1 U of uracil-N-glycosylase. Finally, samples were digested with piperidine, electrophoresed on an 8 M urea, 10% polyacrylamide gel, and visualized by autoradiography.

Construction of site specific mutants
For in vitro transcription studies, all of the mutants were constructed by PCR as described previously (26) according to the methods of Higuchi et al. (31) or Picard et al. (32). A 5-kb HindIII genomic DNA fragment in pGEM4Z was used as a wild-type template. Primers used are shown in Table 2Go. In each case, a -309/+373 bp PCR product with site-specific mutations was digested with BanII and BglII, gel purified, then subcloned into a wild-type template (-471/+240 bp). For preparation of IGF-I basal promoter constructs, a 240-bp product containing wild-type IGF-I downstream sequences and the native IGF-I promoter was amplified using forward (-88/-69, 5'-AGAGAGAAGGCGAATGTTCC-3') and reverse (+129/+152, 5'-ATTGGTGGGCAGGAATAATGAGGC-3') primers. The amplified product was inserted into pCRII (a TA cloning vector), and the KpnI/XhoI fragment containing the insert was subcloned into pGL3-basic (Promega Corp.) to generate p(-88/+152)LUC. To generate a similar construct containing a mutation of the GCGC element, a two-step, megaprimer approach was used (33). To amplify the megaprimer, the first set of PCR reactions contained 5 ng of pIGF-I (-471/+240) template DNA, 1 µM reverse primer (+129/+152), and 1 µM mutagenic primer (+42/+64, 5'-AGATAGAGCCTatatAATGCAAA-3'); the underlined bases (atat) in the mutagenic primer were substituted for the wild-type nucleotides (GCGC). The resulting 110-bp product (megaprimer) was purified from agarose. The second set of PCR reactions contained 5 ng of pIGF-I (-471/+240) template DNA and 5 µg of megaprimer DNA, with five initial cycles at 94 C for 1 min and 72 C for 3 min. During the final 72 C incubation, 1 µg of forward primer (-88/-69) was added and reactions were continued for 25 cycles at 94 C for 30 sec, 55 C for 30 sec, and 72 C for 1 min. The KpnI/XhoI fragment was subcloned into pGL3-basic to create p(-88/+152){Delta}GCGC LUC. All mutants were sequenced.


View this table:
[in this window]
[in a new window]
  		Table 2. Oligonucleotides used to construct mutants for in vitro transcription assays

 
In vitro transcription assay
As described previously (26), transcription reactions (30 µl) contained 1.25 µg of linear DNA template. The wild-type template extended from -471 to +240 bp relative to the major transcription initiation site in exon 1 (14), and mutations within this template are shown in Figs. 1Go and 2Go. The reactions contained 30 µg of liver extract, 66 mM KCl, 6 mM MgCl2, 0.5 mM ATP, CTP, GTP, and UTP, 10% glycerol, 1 U/µl of RNasin, 0.05 mM EDTA, and 1 mM DTT. The DNA template and extract were incubated on ice for 20 min. Transcription was then initiated by addition of nucleotides, and carried out at 30 C for 45 min. The DNA template was degraded by 250 U of RNase-free DNase I. The repurified in vitro transcripts were quantitated by primer extension, autoradiography, and laser densitometry (26). In separate experiments, normal and diabetic extracts were also evaluated by transcription supported by a -471/+3 bp IGF-I template and an adenovirus major late promoter AdMLP template, both linked to (C2AT) "G-free cassette" reporters, as described previously (26).



View larger version (29K):
[in this window]
[in a new window]
  		Figure 2. In vitro transcription assay with wild-type templates (-471/+240) and templates bearing site-specific mutations in DNase I protected regions III and V within downstream region of exon 1. A, Wild-type and mutant regions III and V, with mutated bases indicated by lower case. B, Representative experiment showing transcriptional activity of hepatic nuclear extracts from normal (lanes 1, 3, and 5) and streptozotocin-diabetic rats (lanes 2, 4 and 6) with both wild-type DNA templates (lanes 1 and 2) and templates with 11-bp mutations in regions III (lanes 3 and 4) and V (lanes 5 and 6). In vitro transcripts were quantitated by primer extension as described in Materials and Methods, with rat liver RNA (10 µg) as control (lane 7). C, Representative experiment showing 32P-labeled G-free reporter cassette transcripts from in vitro transcription with normal and diabetic extracts, an adenovirus major late promoter (AdMLP) template (190-bp reporter), and a -471/+3 bp IGF-I template (373-bp reporter). D, Representative experiment showing transcriptional activity of hepatic nuclear extracts from normal (lanes 2 and 4) and streptozotocin-diabetic rats (lanes 3 and 5), with both wild-type DNA templates (lanes 2 and 3) and templates with mutations in a 10-bp region between footprints IV and V (lanes 4 and 5). E, In vitro transcriptional activity of normal and diabetic nuclear extracts with wild-type IGF-I templates, templates with 11-bp mutations in regions III or V, a template with a 10-bp mutation between regions IV and V, an IGF-I template containing only upstream sequence, and the AdML template. With each template, the transcriptional activity of the normal extract was defined as 1.00, and transcriptional activity of diabetic extracts was expressed relative to this internal standard. Mean ± SEM for at least three separate experiments with at least two pairs of normal and diabetic extracts.

 
Primer extension
An end-labeled oligonucleotide complementary to the sequence from +79/+101 was used as probe to quantitate in vitro transcripts. Probe (1 x 106 cpm) was annealed to 10 µg of liver RNA, or in vitro transcripts at 42 C for 8–10 h. The complementary DNA (cDNA) was synthesized as described by Boorstein and Craig (34) at 42 C for 1 h, ethanol precipitated, and resolved on an 8 M urea-6% polyacrylamide gel.

Transient transfection
A CCAAT/enhancer binding protein (C/EBP) expression vector (pMSV-C/EBP) was kindly provided by Dr. Steven McKnight (35); this encodes C/EBP-{alpha} (McKnight, S., Department of Molecular Genetics, University of Texas Southwestern Medical Center, personal communication). The day before transfection, CHO cells were plated at a density of 5 x 105 cells/well in 6-well plates in Ham’s F12 medium containing 10% FBS. The medium was changed to Ham’s F12 containing 3% FBS for 3 h before transfection, and transfection was usually performed at 70–80% confluence using a calcium phosphate coprecipitation method. Precipitates contained 2 µg of reporter constructs, with or without 0.5 µg of the C/EBP expression vector. The amount of DNA per transfection was kept constant by the addition of carrier DNA (pUC19). Four hours after transfection, cells were placed in fresh medium containing 10% FBS. Cells were harvested 48 h later and extracts used in luciferase assays, performed according to the Promega Corp. protocol. Each transfection included triplicate wells, and luciferase activity was expressed relative to total protein in cell extracts.

Statistics
Results are expressed as mean ± SEM, and differences evaluated by paired and unpaired t tests, as appropriate. Significance was taken as P < 0.05.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Transcriptional activity of region III and V mutated constructs
The functional importance of downstream DNase I footprint regions III and V in metabolic regulation was evaluated according to the in vitro transcriptional activity of site-directed mutant templates, as shown in Fig. 2Go. Consistent with our previous report (26), a wild-type rIGF-I template containing both upstream and downstream sequence exhibited reduced activity with nuclear extracts from the livers of streptozotocin-diabetic as compared with normal rats (Fig. 2BGo, lane 2 vs. 1). Similar content of transcriptional machinery in different extracts was established by transcription from a control IGF-I template containing only upstream sequence, and transcription from a control adenovirus major late promoter (AdMLP) template (Ref. 26 and Fig. 2CGo); with these templates, the transcriptional activity of the diabetic extract was comparable to or even greater than that of the normal extract. Mutation of 11 bp within DNase I footprint region III resulted in a slight decrease in transcriptional activity with normal extracts (lane 3 vs. 1), but a substantial increase in transcriptional activity with diabetic extracts (lane 4 vs. 2). Mutations in footprint region V had less effect on the transcriptional activity of the normal extracts (lane 5 vs. 1) but restored the activity of diabetic extracts close to that of normal extracts with wild-type templates (lane 6 vs. 1). In contrast, mutations in a 10-bp region between footprints IV and V had no effect on the relative transcriptional activity of normal and diabetic nuclear extracts (Fig. 2DGo). After quantitation by laser densitometry (Fig. 2EGo), the IGF-I transcriptional activity of diabetic extracts was 46 ± 6% that of normal extracts with wild-type templates, and 45 ± 12% with region IV/V mutant templates (both P < 0.02, diabetic vs. normal), but rose to 77 ± 21% with mutation of region III, and 96 ± 12% with mutation of region V (both p = NS, diabetic vs. normal). In contrast, the transcriptional activity of diabetic extracts compared with normal extracts was 164 ± 6% with the AdML template, and 245 ± 25% with a control IGF-I template containing only upstream sequence, confirming the transcriptional integrity of the diabetic extracts. In combination, these results indicate that downstream regions III and V [+42 and +129 bp, respectively, from the major transcription initiation site in exon 1 (14)] may both be important in mediating depressed IGF-I gene transcription in conditions of diabetes mellitus.

Localization of core binding sites in regions III and V
Transcription factor binding to DNase I footprint region III was examined by methylation interference analysis, as shown in Fig. 3Go. Binding of nuclear factors to a 272-bp (-54/+219) fragment produced two major protein/DNA complexes (B1 and B2), as described previously (26). Methylation interference was observed only with complex "B1" (lanes 3 and 7). Methylation on guanidine residues in a core GCGC sequence (+53/+56) inhibited binding on both coding and noncoding strands (Fig. 3Go, A and B). Methylation interference analysis was less informative for DNase I footprint region V (+125/+159) because portions of region V are relatively AT rich. Therefore, uracil interference analysis was used to evaluate binding. As shown in Fig. 4Go, substitution of thymidine with uracil altered binding in a core "TTATT" region from +135 to +139 bp.



View larger version (52K):
[in this window]
[in a new window]
  		Figure 3. Methylation interference assay. A, An end-labeled 272-bp AccI/BglII (-54/+219) fragment was partially methylated, then incubated with liver nuclear extract. The protein/DNA complexes (B1 and B2) and free probe were separated on a 5% gel, and processed as described in Materials and Methods, and samples finally applied to an 8 M urea-6% polyacrylamide gel. Bases essential for binding are indicated with by •, and bases that become hypersensitive to piperidine digestion are indicated bu {blacktriangleup}. The sequence of region III determined with the Maxam-Gilbert method (lanes 1 and 5) is shown. The coding strand is shown left, and the noncoding strand right. B, Expanded view of autoradiogram in Panel A (noncoding strand) demonstrating in relevant of G residues within the core binding region.

 


View larger version (58K):
[in this window]
[in a new window]
  		Figure 4. Uracil interference assay. A, Labeled, uracil-substituted 77-bp (+103/+179) fragment was prepared by PCR, then incubated with normal liver nuclear extract. Protein/DNA complexes were separated on a 5% gel, and processed as described in Materials and Methods. Samples were separated on an 8 M urea-6% polyacrylamide gel. Bases essential for binding are indicated by •, and bases that become hypersensitive to piperidine digestion are indicated by {blacktriangleup}. The sequence of region V determined with the Maxam-Gilbert method (lanes 1, 2 and 6, 7) is also indicated. The coding strand is shown left, and the noncoding strand right. B, Expanded view of the autoradiogram in Panel A (coding strand) demonstrating involvement of T residues within the core binding region.

 
To confirm the putative core binding sites identified on the basis of interference analysis, DNA/protein interactions were examined with gel mobility shift assays, as shown in Fig. 5Go. With the wild-type region III oligonucleotide (Table 1Go), binding of nuclear proteins resulted in three protein/DNA complexes: B1, B2, and N (nonspecific). In competition assays, complexes B1 and B2 could be competed readily with a wild-type oligonucleotide (lane 3), but not with an oligonucleotide mutated in the core binding region (mIII, lane 4). Complex N could be competed with both wild-type and mutant oligonucleotides. These studies confirmed the importance of the "GCGC" sequence for nuclear factor binding in region III.



View larger version (36K):
[in this window]
[in a new window]
  		Figure 5. Gel mobility shift analysis. A, Assays were performed as described in Materials and Methods. Double-stranded oligonucleotide III was used as a probe. The mutants used in competition assays are shown in Table 1Go. Competing oligonucleotides were incubated with nuclear protein for 20 min before addition of probe, and binding was performed at 25 C for 20 min. The protein/DNA complexes (B1, B2, and N) were separated from free probe (F) on a 5% polyacrylamide gel. B, Supershift analysis. Using a similar design, binding of nuclear factors to a region III probe was tested with and without addition of anti-C/EBP antibody or preimmune serum. C, Transient transfection. CHO cells were transfected as described with IGF-I basal promoter constructs with or without a 4-bp mutation in the GCGC core region, and with or without a C/EBP expression vector. Mean ± SEM for three separate experiments.

 
Comparison with the Genetics Computer Group (GCG) database of transcription factor consensus binding sites revealed potential overlap with critical sequences in regions III and V (Table 3Go). Competition analyses for region III binding were performed with oligonucleotides containing binding sites for AP-1, AP-2, C/EBP and TFIID. As shown in Fig. 5AGo, competition was undetectable with AP2 and TFIID, and weak with AP1, but strong with C/EBP. In separate studies, interactions of C/EBP with region III were confirmed by the presence of a supershift with an antibody which recognizes C/EBP-{alpha}, -ß, and -{delta} (Fig. 5BGo); no supershift was seen with preimmune serum. A C/EBP expression construct also provided 10.7-fold stimulation of a wild-type IGF-I basal promoter construct, but 3.6-fold stimulation (comparable to 2.8-fold stimulation of the reporter vector) of a construct with a 4-bp mutation in the core binding region (Fig. 5CGo), demonstrating the importance of the GCGC element for C/EBP action in vivo.


View this table:
[in this window]
[in a new window]
  		Table 3. Comparison of rat IGF-I genomic DNA sequence in DNase I protection regions III and V with the consensus binding sequences of known transcription factors

 
Similar studies were also performed with oligonucleotides reflecting portions of region V (Figs. 6Go and 7Go). Four protein/DNA complexes (B1/B2, B3, and N) were seen with a wild-type oligonucleotide (lanes 2 and 11). Complex N appeared to be nonspecific, but B1/B2 and B3 could be competed with a wild-type oligonucleotide. To determine the core binding site, a series of mutant oligonucleotides were used in competition studies (Fig. 6Go). Competition of complexes B1/B2 with a wild-type probe was least effective with the M1 mutant, next least effective with the M2 mutant and effective (at least at low concentration), with the M3 mutant. Decreased ability to compete suggested that the mutated bases involved might be important for binding, and directed subsequent studies of the M1/M2 junction. The hypothesis of the importance of this area was then confirmed by poor competition with the M4, M5, M6 series of mutants centered on the M1/M2 junction, as shown in Fig. 6Go. These results indicate that DNA-protein interactions involve the junction of the 5' end and middle portion of region V, centered on the "TTAT" region identified as important on the basis of uracil interference analysis. Complex N was occasionally competed with mutants, but the observation was not consistent. Lack of competition with a "core" mutant (Fig. 7Go, lane 4) provides strong evidence that the TTAT sequence is important for nuclear factor binding. No competition was obtained with an AP1 oligonucleotide, but formation of complexes B1/B2 and B3 was slightly competed with AP2, C/EBP and TFIID oligonucleotides even though region V has little homology with these consensus binding sequences.



View larger version (104K):
[in this window]
[in a new window]
  		Figure 6. Gel mobility shift analysis. Assays were performed as described in Materials and Methods. Double-stranded oligonucleotide V was used as probe. The mutants used in competition assays are shown in Table 1Go. Competitors were incubated with nuclear extracts for 20 min before addition of probe, and binding was performed at 25 C for 20 min. The protein/DNA complexes (B1, B2, and B3, and N) were separated from free probe (F) on a 5% polyacrylamide gel.

 


View larger version (63K):
[in this window]
[in a new window]
  		Figure 7. Gel mobility shift analysis. Assays were performed as described in Materials and Methods. Double-stranded oligonucleotide V was used as probe. Competitors were incubated with nuclear extracts for 20 min before addition of probe, and binding was performed at 25 C for 20 min. Protein/DNA complexes (B1/B2, B3, and N) were separated from free probe (F) on a 5% polyacrylamide gel.

 

   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
In the present studies, we have used in vitro transcription to show that sequences downstream from the IGF-I major transcription initiation sites in exon 1 are important for metabolic responsiveness. Wild-type templates supported a diabetes-sensitive decrease in in vitro transcription that was similar to the diabetes-induced decrease in the rate of IGF-I gene transcription measured by nuclear runon assays (12). However, extensive mutation in either DNase I protected region III or V reduced the difference in transcriptional activity with hepatic nuclear extracts from normal as compared with streptozotocin-diabetic rats. In contrast, control mutation of the bases between regions IV and V did not alter the decreased IGF-I transcriptional activity seen with hepatic nuclear extracts from diabetic as compared with normal rats, indicating that these bases are not important in mediating the decreased IGF-I transcriptional activity associated with diabetes mellitus. Within region III, methylation interference analysis indicated involvement of a GCGC core sequence, and the importance of this sequence for nuclear factor binding was confirmed by gel mobility shift analysis. Within region V, uracil interference analysis indicated that the target site for nuclear factor binding involves a TTAT core sequence.

Regulation of gene expression depends on the assembly of factors involved in transcription initiation, and an interplay of both activator and repressor proteins bound to the DNA upstream or downstream from the promotor. Such cis-acting elements may be dispersed throughout the gene (e.g. ß-globin (36), or located within introns (37) or in the 5' or 3' flanking regions (38, 39). Tissue-specific enhancers or silencers have been identified in the downstream region of the human placental lactogen gene (40), the gene coding for von Willebrand factor in endothelial cells (41), the human osteocalcin gene (42), and the human mdr1 gene (39). In addition to regulating promoter function, such downstream elements may be necessary to direct accurate transcription initiation, and/or to influence mRNA stability, processing, and elongation.

The biological importance of downstream sequences has previously been suggested for the IGF-I gene. The presence of downstream sequences confers increased expression with IGF-I constructs as examined by transient transfection into SK-N-MC neuroblastoma cells by Hall et al. (18), and C6 rat glioma cells as studied by Lowe et al. (20, 21) and Wang et al. (24). We also found that downstream sequences were involved in metabolically regulated IGF-I gene expression in studies with both rat liver and rat hepatocytes in primary culture (26). In the present studies, the metabolic significance of downstream sequences was confirmed by in vitro transcription assays with templates containing extensive mutations in regions III or V (Fig. 2Go). Thus, our data provide additional evidence that the presence of downstream sequences containing regions III and V may be important for metabolic regulation of IGF-I gene transcription. Further studies will be required to determine the specific contributions of alterations in different metabolic fuels and hormones (glucose, FFA, insulin, glucagon, etc.) to the decrease in IGF-I gene transcription seen in diabetic animals.

The present findings also demonstrate that GCGC and TTAT are core binding sequences in regions III and V, respectively. Search of a transcription factor consensus binding sites database using GCG software revealed homology in region III with transcription factor(s) such as C/EBP, and c-fos, and which share similar GCGC motifs in their binding sequences. Our results indicate that C/EBP may be one of the factors binding to region III, based on evidence from gel shift competition, antibody supershift, and transfection studies. Thomas et al. (43) have reported that a CGCAATCG element in region III can mediate C/EBP involvement in the cAMP-induced activation of IGF-I constructs in osteoblasts, and Umayahara et al. (44) found that bases CGCAATCG (in boldface) are particularly important for PGE2-induced nuclear factor binding in COS-7 cells. However, our present gel shift and methylation interference studies indicate that binding of hepatic nuclear extracts involves a GCGC core in region III, slightly 5' to the region reported by Umayahara et al. (44). Such differences may reflect cAMP signal transduction pathways that are relatively tissue specific, as we have been unable to obtain stimulation of our constructs by forskolin or a PKA expression vector in CHO cells (not shown). Despite some homology in both regions to consensus sequences binding jun-fos heterodimers or jun-jun homodimers (45), we found that oligonucleotides for AP-1 or AP-2 did not compete well for factor binding to either region III or V. Neither region appeared to bind TFIID, despite some homology to region V. Within region V, the tetranucleotide ATTA is the core motif for binding of homeodomain proteins encoded by a family of developmentally regulated genes (46). Because IGF-1 gene expression is also developmentally regulated (4), nuclear factor(s) binding onto the TTAT core in region V may involve a homeodomain protein.

It is possible that regulatory proteins interacting with the two regions modulate IGF-I gene transcription via different mechanisms, conceivably involving protein-protein as well protein/DNA interactions. The homeodomain protein Msx-1 represses activity of the SV-40 promoter without direct DNA contact (47), and maximal activity of the CytR repressor depends on the formation of nucleoprotein complexes more than target sequence recognition (48). Moreover, the regulatory function of the GCGC and TTAT core elements may also depend on the presence of neighboring sequences which are not part of the core binding regions but contribute to either the binding or the activity of factors bound to these sites.

It is likely that IGF-I gene transcription depends on combinatorial interactions between ubiquitous, cell/tissue type specific, and signal-specific transcription factors, as outlined by Hill and Treisman (49). For example, different transcription factors modulate transferrin expression in liver vs. Sertoli cells (50), both HNF-4 and C/EBPß are required for tissue specificity of the ornithine transcarbamylase enhancer (51), and simultaneous binding of HNF-3ß or -{gamma} along with pancreas transcription factor 1 are required for {alpha}-amylase expression (52). Thus, factors involved with core sites in regions V and III may be of particular importance for the metabolic regulation of IGF-I expression, whereas factors binding to other regions may play a supportive role. Identification of nuclear factors involved through DNA/protein or protein-protein interactions with regions III and V should help to elucidate the mechanisms which underlie the decrease in IGF-I gene transcription associated with diabetes mellitus.\.


   Acknowledgments
 
We thank Sharon DePeaza, Belinda Richardson, and Mary Lou Mojonnier for assistance in the preparation of this manuscript.\.


   Footnotes
 
1 This work was presented in part at The Endocrine Society Annual Meetings in 1995 and 1996. This work was supported in part by research awards from the Emory Medical Care Foundation (to C.-I.P.), the American Cancer Society Cancer Center Seed Money Fund and Emory University Research Committee (to G.-J.W.), and NIH Grant GM-37261 (G.-J.W.); and the National Institutes of Health Grant DK-33475 (to L.S.P., C.-I.P., and E.H.). Back

Received February 18, 1999.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Blundell TL, Bedarkar S, Humbel RE 1983 Tertiary structures, receptor binding, and antigenicity of insulin like growth factors. Fed Proc 42:2592–2597[Medline]
  2. Daughaday WH, and 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[Abstract/Free Full Text]
  3. Rotwein P 1991 Mini-review: Structure, evolution, expression and regulation of insulin-like growth factors I and II. Growth Factors 5:3–18[Medline]
  4. Adamo ML 1995 Regulation of insulin-like growth factor I gene expression. Diabetes Rev 3:2–35
  5. Phillips LS, Goldstein S, Pao C-I 1991 Nutrition and somatomedin XXVI. Molecular regulation of insulin-like growth factor-1 by insulin in cultured rat hepatocytes. Diabetes 40:1525–1530[Abstract]
  6. Barreca A, Voci A, Minuto F, deMarchis M, Cecchelli E, Fugassa E, Giordano G, Gallo G 1992 Effect of epidermal growth factor on insulin-like growth factor-I (IGF-1) and IGF-binding protein synthesis by adult hepatocytes. Mol Cell Endocrinol 84:119–126[CrossRef][Medline]
  7. Umayahara Y, Kawamori R, Watada H, Imano E, Iwama N, Morishima T, Yamasaki Y, Kajimoto Y, Kamada T 1994 Estrogen regulation of the insulin-like growth factor I gene transcription involves an AP-1 enhancer. J Biol Chem 269:16433–16442[Abstract/Free Full Text]
  8. Murphy LJ, Bell GI, Friesen HG 1987 Tissue distribution of insulin-like growth factor I and II messenger ribonucleic acid in the adult rat. Endocrinology 120:1279–1282[Abstract/Free Full Text]
  9. Shimatsu A, and Rotwein P 1987 Mosaic evolution of the insulin-like growth factors. Organization, sequence, and expression of the rat insulin-like growth factor I gene. J Biol Chem 262:7894–7900[Abstract/Free Full Text]
  10. Lund PK, Hepler JE, Hoyt EC, Simmons JG 1991 Physiological relevance of IGF-I mRNA heterogeneity. In: Spencer EM (ed) Modern Concepts of Insulin-Like Growth Factors. Elsevier Science Publishing Co Inc 111–120
  11. Steenbergh PH, Koonen-Reemst AMCB, Cleujens CBJM, Sussenbach JS 1991 Complete nucleotide sequence of the high molecular weight human IGF-I mRNA. Biochem Biophys Res Commun 175:507–514
  12. Pao C-I, Farmer PK, Begovic S, Goldstein S, Wu G-J, Phillips LS 1992 Expression of hepatic insulin-like growth factor-I and insulin-like growth factor-binding protein-1 genes is transcriptionally regulated in streptozotocin-diabetic rats. Mol Endocrinol 6:969–977[Abstract/Free Full Text]
  13. Lowe Jr WL, Roberts Jr CT, Lasky SR, LeRoith D 1987 Differential expression of alternative 5' untranslated regions in mRNAs encoding rat insulin-like growth factor-I. Proc Natl Acad Sci USA 84:8946–8950[Abstract/Free Full Text]
  14. Adamo M, Ben-Hur H, Roberts CT, Jr, LeRoith D 1991 Regulation of start site usage in the leader exons of the rat insulin-like growth factor-I gene by development, fasting and diabetes. Mol Endocrinol 5:1677–1686[Abstract/Free Full Text]
  15. Su T-S, Liu W-Y, Han S-H, Jansen M, Yang-Fen TL 1989 Transcripts of the insulin-like growth factors I and II in human hepatoma. Cancer Res 49:1773–1777[Abstract/Free Full Text]
  16. Conover CA, Baker BK, Bale LK, Clarkson JT, Liu F, Hintz RL 1993 Human hepatoma cells synthesize and secrete insulin-like growth factor Ia prohormone under growth hormone control. Regulatory Peptides 48:1–8[CrossRef]
  17. Adamo ML, Ben-Hur H, LeRoith D, Roberts CT, Jr 1991 Transcription initiation in the two leader exons of the rat IGF- 1 gene occurs from disperse versus localized sites. Biochem Biophys Res Commun 176:887–893[CrossRef][Medline]
  18. Hall LJ, Kajimoto Y, Bichell D, Kim S-W, James PL, Counts D, Nixon LJ, Tobin G, Rotwein P 1992 Functional analysis of the rat insulin-like growth factor I gene and identification of an IGF-I gene promoter. DNA Cell Biol 11:301–313[Medline]
  19. Pao C-I, Lin KWM, Zhu JL, Wu GJ, Farmer PK, Phillips LS 1996 In vitro transcription of the rat insulin-like growth factor-I gene. J Biol Chem 271:8667–8674[Abstract/Free Full Text]
  20. Lowe WL Jr, and Teasdale RM 1992 Characterization of a rat insulin-like growth factor I gene promoter. Biochem Biophys Res Commun 189:972–978[CrossRef][Medline]
  21. An MR, and Lowe WL, Jr 1995 The major promoter of the rat insulin-like growth factor-I gene binds a protein complex that is required for basal expression. Mol Cell Endocrinol 114:77–89[CrossRef][Medline]
  22. Nolten LA, van Schaik FMA, Steenbergh PH, Sussenbach JS 1994 Expression of the insulin-like growth factor I gene is stimulated by the liver-enriched transcription factors C/EBPa and LAP. Mol Endocrinol 8:1636–1645[Abstract/Free Full Text]
  23. 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[Abstract/Free Full Text]
  24. Wang X, Yang Y, Adamo ML 1997 Characterization of the rat insulin-like growth factor I gene promoters and identification of a minimal exon 2 promoters. Endocrinology 138:1528–1536[Abstract/Free Full Text]
  25. Huang S, Thulé PM, Phillips LS 1995 Identification of novel promoter and repressor elements in the 5'-flanking regions of the rat insulin-like growth factor-I gene. Biochem Biophys Res Commun 206:279–286[CrossRef][Medline]
  26. Pao C-I, Zhu J-L, Robertson DG, Lin KWM, Farmer PK, Begovic S, Wu G-J, Phillips LS 1995 Transcriptional regulation of the rat insulin-like growth factor-I gene involves metabolism-dependent binding of nuclear proteins to a downstream region. J Biol Chem 270:24917–24923[Abstract/Free Full Text]
  27. Kulik VP, Kavsan VM, van Schaik FMA, Nolten LA, Steenbergh PH, Sussenbach JS 1995 The promoter of the salmon insulin-like growth factor I gene is activated by hepatocyte nuclear factor 1. J Biol Chem 270:1068–1073[Abstract/Free Full Text]
  28. 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[Abstract]
  29. Sierra F 1990 In: Azzi A, Polak JM, Saluz HP (eds) A Laboratory Guide to In Vitro Transcription. Basel, Boston, Berlin, Birkhäuser, Verlag, pp 34–42
  30. Kerrigan LA, Kadonaga JT 1994 DNA-protein interactions. In: Ausubel FM, Brent R, Kington RE, Moore DD, Seidmann JG, Smith JA, Struhl K (eds) Current Protocols in Molecular Biology. Wiley Interscience [12], pp 12.0.1–12.10.18
  31. Higuchi R, Krummel B, Saiki RK 1988 A general method of in vitro preparation and specific mutagenesis of DNA fragments: study of protein and DNA interactions. Nucleic Acids Res 16:7351–7367[Abstract/Free Full Text]
  32. Picard V, Ersdal-Badju E, Lu A, Bock SC 1994 A rapid and efficient one-tube PCR-based mutagenesis technique using Pfu DNA polymerase. Nucleic Acids Res 22:2587–2591[Abstract/Free Full Text]
  33. Smith AM, and Klugman KP 1997 "Megaprimer" tethod of PCR-based mutagenesis: the concentration of megaprimer is a critical factor. Biotechniques 22:438–442[Medline]
  34. Boorstein WR, and Craig EA 1989 Primer extension analysis of RNA. Methods Enzymol 101:347–368
  35. Friedman AD, Landschulz WH, McKnight SL 1989 CCAAT/enhancer binding protein activates the promoter of the serum albumin gene in cultured hepatoma cells. Genes Dev 3:1314–1322[Abstract/Free Full Text]
  36. Collis P, Antoniou M, Grosveld F 1990 Definition of the minimal requirements within the human b-globin gene and the dominant control region for high level expression. EMBO J 9:233–240[Medline]
  37. Chung S, and Perry P 1989 Importance of introns for expression of mouse ribosomal protein gene rpL32. Mol Cell Biol 9:2075–2082[Abstract/Free Full Text]
  38. Emerson BM, Nickol JM, Jackson PD, Felsenfeld G 1987 Analysis of the tissue-specific enhancer at the 3' end of the chicken adult b-globin gene. Proc Natl Acad Sci USA 84:4786–4790[Abstract/Free Full Text]
  39. Madden MJ, Morrow CS, Nakagawa M, Goldsmith ME, Fairchild CR, Cowan KH 1993 Identification of 5' and 3' sequences involved in the regulation of transcription of the human mdr 1 gene in vivo. J Biol Chem 268:8290–8297[Abstract/Free Full Text]
  40. Jacquemin P, Oury C, Peers B, Morin A, Belayew A, Martial JA 1994 Characterization of a single strong tissue-specific enhancer downstream from the three human genes encoding placental lactogen. Mol Cell Biol 14:93–103[Abstract/Free Full Text]
  41. Jahroudi N, and Lynch DC 1994 Endothelial-cell-specific regulation of von Willebrand factor gene expression. Mol Cell Biol 14:999–1008[Abstract/Free Full Text]
  42. Li YP, Chen W, Stashenko P 1995 Characterization of a silencer element in the first exon of the human osteocalcin gene. Nucleic Acids Res 23:5064–5072[Abstract/Free Full Text]
  43. Thomas MJ, Umayahara Y, Shu H, Centrella M, Rotwein P, McCarthy TL 1996 Identification of the cAMP response element that controls transcriptional activation of the insulin-like growth factor-I gene by prostaglandin E(2) in osteoblasts. J Biol Chem 271:21835–21841[Abstract/Free Full Text]
  44. Umayahara Y, Ji C, Centrella M, Rotwein P, McCarthy TL 1997 CCAAT/enhancer-binding protein d activates insulin-like growth factor-I gene transcription in osteoblasts. J Biol Chem 272:31793–31800[Abstract/Free Full Text]
  45. Pfahl M 1993 Nuclear receptor/AP-1 interaction. Endocr Rev 14:651–658[Abstract/Free Full Text]
  46. Gehring WJ, Affolter M, Bürglin T 1994 Homeodomain proteins. Annu Rev Biochem 63:487–526[CrossRef][Medline]
  47. Catron KM, Zhang H, Marshall SC, Inostroza JA, Wilson JM, Abate C 1995 Transcriptional repression by Msx-1 does not require homeodomain DNA-binding sites. Mol Cell Biol 15:861–871[Abstract]
  48. Sogaard-Andersen L, Valentin-Hansen P 1993 Protein-protein interactions in gene regulation: The cAMP-CRP complex sets the specificity of a second DNA-binding protein, the CytR repressor. Cell 75:557–566[CrossRef][Medline]
  49. Hill CS, and Treisman R 1995 Transcriptional regulation by extracellular signals: mechanisms and specificity. Cell 80:199–211[CrossRef][Medline]
  50. Schaeffer E, Guillou F, Part D, Zakin MM 1993 A different combination of transcription factors modulates the expression of the human transferrin promoter in liver and Sertoli cells. J Biol Chem 268:23399–23408[Abstract/Free Full Text]
  51. Nishiyori A, Tashiro H, Kimura A, Akagi K, Yamamura K, Mori M, Takiguchi M 1994 Determination of tissue specificity of the enhancer by combinatorial operation of tissue-enriched transcription factors. J Biol Chem 269:1323–1331[Abstract/Free Full Text]
  52. Cockell M, Stolarczyk D, Frutiger S, Hughes GJ, Hagenbüchle O, Wellauer PK 1995 Binding sites for hepatocyte nuclear factor 3ß or 3{gamma} and pancreas transcription factor 1 are required for efficient expression of the gene encoding pancreatic {alpha}-amylase. Mol Cell Biol 15:1933–1941[Abstract]



This article has been cited by other articles:


Home page
 EndocrinologyHome page
J. Chahal, C.-C. Chen, M. J. Rane, J. P. Moore, M. T. Barati, Y. Song, and B. C. Villafuerte
Regulation of Insulin-Response Element Binding Protein-1 in Obesity and Diabetes: Potential Role in Impaired Insulin-Induced Gene Transcription
Endocrinology, October 1, 2008; 149(10): 4829 - 4836.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
B. C. Villafuerte and E. N. Kaytor
An Insulin-response Element-binding Protein That Ameliorates Hyperglycemia in Diabetes
J. Biol. Chem., May 20, 2005; 280(20): 20010 - 20020.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
B. C. Villafuerte, L. S. Phillips, M. J. Rane, and W. Zhao
Insulin-response Element-binding Protein 1: A NOVEL Akt SUBSTRATE INVOLVED IN TRANSCRIPTIONAL ACTION OF INSULIN
J. Biol. Chem., August 27, 2004; 279(35): 36650 - 36659.
[Abstract] [Full Text] [PDF]

Home page
 EndocrinologyHome page
L. N. N. Shoba, M. Newman, W. Liu, and W. L. Lowe Jr.
LY 294002, an Inhibitor of Phosphatidylinositol 3-Kinase, Inhibits GH-Mediated Expression of the IGF-I Gene in Rat Hepatocytes
Endocrinology, September 1, 2001; 142(9): 3980 - 3986.
[Abstract] [Full Text] [PDF]

Home page
 EndocrinologyHome page
E. N. Kaytor, J. L. Zhu, C.-I Pao, and L. S. Phillips
Physiological Concentrations of Insulin Promote Binding of Nuclear Proteins to the Insulin-Like Growth Factor I Gene
Endocrinology, March 1, 2001; 142(3): 1041 - 1049.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Zhu, J.-L.
Right arrow Articles by Phillips, L. S.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Zhu, J.-L.
Right arrow Articles by Phillips, L. S.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology Endocrine Reviews J. Clin. End. & Metab.
Molecular Endocrinology Recent Prog. Horm. Res. All Endocrine Journals