Acute control of insulin-like growth factor-I gene transcription by growth hormone through Stat5b.

Many of the effects of growth hormone (GH) are mediated by insulin-like growth factor-I (IGF-I), a secreted peptide whose gene transcription is induced by GH by unknown mechanisms. Recent studies in mice have implicated Stat5b as part of a GH-regulated somatic growth pathway, because mice lacking this transcription factor show diminished growth rates and a decline in serum IGF-I levels. To test the role of Stat5b in GH-stimulated IGF-I gene expression, we have delivered modified versions of the protein to pituitary-deficient male rats by quantitative adenovirus-mediated gene transfer. In pilot studies in cell culture, both constitutively active and dominant-negative Stat5b appropriately regulated transcription from a GH-responsive Stat5-dependent reporter gene. After in vivo expression, neither protein impaired GH-induced activation of cytoplasmic signaling pathways or blocked nuclear accumulation of Stats 1 and 3 in the liver, the major site of IGF-I production. Dominant-negative Stat5b completely prevented GH-stimulated IGF-I gene transcription, whereas constitutively active Stat5b led to robust IGF-I gene expression in the absence of hormone. An adenovirus encoding enhanced green fluorescent protein was without effect. Similar results were seen with the GH-responsive Stat5b-dependent Spi 2.1 gene, whereas GH-stimulated c-fos transcription was minimally altered. These results establish Stat5b as a key component of GH-stimulated IGF-I gene transcription, and they demonstrate the feasibility of using in vivo gene transfer to target distinct components of hormone-activated signaling pathways.

Growth hormone (GH) 1 exerts broad effects on growth, differentiation, and intermediary metabolism through mechanisms dependent on activation of the transmembrane GH re-ceptor and its associated tyrosine-protein kinase, Jak2 (1,2). Among the intracellular mediators of GH action are several members of the signal transduction and activator of transcription (Stat) family, including Stats 1, 3, 5a, and 5b (2)(3)(4)(5). Gene deletion studies in mice have implicated Stat5b and to a lesser extent Stat5a as components of the GH signaling pathways that are responsible for normal postnatal somatic growth. The absence of Stat5b appears to phenocopy GH receptor deficiency in male mice but not in females, leading to animals up to 30% smaller than heterozygous or wild-type littermates, whereas in both female and male mice the double knockout of Stat5a and Stat5b caused a decline in growth rates (6,7).
Insulin-like growth factor-I (IGF-I), a highly conserved 70residue secreted protein, also plays an essential role in regulating somatic growth (8,9). In growing animals the concentration of IGF-I in the circulation is GH-dependent (8). In GH-deficient hypophysectomized rats IGF-I serum levels are reduced to less than 5% of normal, and hormone treatment restores IGF-I toward the normal range within several hours (8,10). GH also rapidly induces IGF-I mRNA expression in multiple tissues in vivo through acute stimulation of IGF-I gene transcription by mechanisms that have not been fully characterized but appear to be independent of new protein synthesis (11)(12)(13)(14).
In male mice lacking Stat5b and in female and male mice lacking both Stat5a and b, serum IGF-I concentrations are reduced only to 50 -70% of normal (6,7), significantly higher than values observed in pituitary-deficient rodents, thus arguing indirectly against a primary role for Stat5b in mediating GH-dependent IGF-I gene expression. By contrast, Davey et al. (15) have shown that in Stat5b-deficient hypophysectomized male mice a week of GH treatment failed to induce either IGF-I mRNA or protein expression, potentially implicating Stat5b as being important for this chronic response to hormone treatment but without uncovering any mechanisms. We now have used replication-defective adenoviruses encoding either a constitutively active or dominant-negative version of rat Stat5b to directly assess the role of this transcription factor in the acute regulation of IGF-I gene expression by GH in vivo. Our results demonstrate that Stat5b is an essential intermediate in GHstimulated IGF-I gene transcription and establish the feasibility of selectively targeting components of complex signal transduction pathways in experimental animals.

EXPERIMENTAL PROCEDURES
Materials-Antibodies to Akt and Erks were from Cell Signaling Technology (Beverly, MA). Anti-Stat1 (anti-ISGF3 p91/p84) was from Transduction Laboratories (Lexington, KY), anti-Stat3 was from Upstate Biotechnology (Lake Placid, NY), anti-SP1 was from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA), and anti-FLAG (M2) was from Sigma (St. Louis, MO). Alkaline phosphatase-conjugated goat antirabbit and anti-mouse IgG were from Southern Biotechnology Associates (Birmingham, AL). Goat anti-mouse-Alexa 488 conjugate was from * This work was supported by National Institutes of Health Grants 5RO1DK37449 (to P. R.) and 5F32DK09802 (to J. B.) and by a Research Fellowship of the European Society for Pediatric Endocrinology, sponsored by Novo Nordisk and the Eli Lilly International Foundation (to J. W.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18  Generation of Recombinant Adenoviruses-Adenoviruses encoding EGFP, Stat5b, and mutant Stat5bs were prepared following procedures described by the provider of pAdEasy. The DNA inserts were checked by PCR and by restriction enzyme mapping, and the recombinant viruses were amplified in HEK293 cells. All viruses were purified over a discontinuous CsCl gradient and titered by optical density.
Reporter Gene Assays-COS-7 cells were transfected with an expression plasmid for the mouse GH receptor and 8xGHRE-TK Luc reporter gene using Fugene-6 and a protocol from the supplier. After incubation for 16 h cells were infected with either Ad-EGFP, or different Ad-Stat5b variants, and Ad-tTA, each at a multiplicity of infection of 100. Ad-tTA encodes a tetracycline-inhibited transcriptional activator that binds to response elements adjacent to the promoter-regulating gene encoding either EGFP or Stat5b in each recombinant adenovirus and is required for gene expression. 8 h later, after addition of serum-free media containing 1% bovine serum albumin, and recombinant rat GH (1 g/ml final concentration) or vehicle, cells were incubated for 16 h, harvested, and cell lysates used for luciferase assays. For co-infection experiments, COS-7 cells were transfected as described above, followed by infection with Ad-Stat5b WT , Ad-tTA, and either Ad-␤ galactosidase or Ad-Stat5b DN at a ratio of 1:10. This corresponded to a multiplicity of infection of 10 for Ad-Stat5b WT , 90 for Ad-␤ galactosidase or Ad-Stat5b DN , and 100 for Ad-tTA. Subsequent steps were as outlined above.
Animal Studies-Male Sprague-Dawley rats, hypophysectomized by a transauricular route at age 7 weeks, were obtained from Harlan Sprague-Dawley (Indianapolis, IN). Animals were housed at the OHSU Animal Care Facility on a 12-h light/dark schedule with free access to food and water, and received care according to National Institutes of Health guidelines. All procedures were pre-approved by the OHSU Animal Care and Use Committee. Glucocorticoids (cortisol phosphate, 400 g/kg/day) and thyroxine (10 g/kg/day) were replaced by daily subcutaneous injection, and GH deficiency was confirmed by failure to grow during a 2-week observation period. Following this interval, rats were injected via tail vein with 0.2 ml of sterile phosphate-buffered saline containing 2 ϫ 10 10 plaque forming units (pfu) of Ad-EGFP, Ad-Stat5b DN , or Ad-Stat5b CA plus 2 ϫ 10 9 pfu of helper virus Ad-tTA. In pilot experiments, the lowest dose of viruses that gave nearly complete liver cell infection in vivo was determined by analysis of primary hepatocytes isolated using the collagenase method (16). A similar approach was used to document quantitative infection with Ad-Stat5b DN and Ad-Stat5 CA , except that isolated hepatocytes were cultured for 24 h, fixed in 4% paraformaldehyde, and immunostained using antibody to the FLAG epitope tag at a 1:1000 dilution (see Fig. 2). For studies of GH action, rats were injected intraperitoneally with either vehicle (saline) or 1.5 g/g recombinant rat GH at 48 h after infection. Following an incubation period of 30 -120 min, animals were anesthetized with pentobarbital (50 mg/kg intraperitoneally) and sacrificed. Liver proteins and RNA were isolated as outlined below.
DNA-Protein Binding Studies-Electrophoretic mobility shift assays were performed as described previously (17) with 10 or 2 g of rat hepatic nuclear proteins and a 5Ј-fluorescein-labeled double-stranded oligonucleotide from the rat ␤-casein gene (top strand, GAACTTCTTG-GAATTAAG; bottom strand, CTTAATTCCAAGAAGTTC) (18,19). After incubation of proteins and DNA for 30 min at 4°C, products were separated by electrophoresis through non-denaturing, 4 -12% polyacrylamide gels in 0.5ϫ TBE (45 mM Tris, 45 mM boric acid, and 1 mM EDTA, pH 8.3) at 120 V for 2 h at 20°C. Results were detected using a Molecular Imager FX and Quantity One software.
RNA Isolation and Analysis-Total liver and hepatic nuclear RNA was isolated as described previously (13). RNA concentration was determined spectrophotometrically at 260 nm, and its quality was assessed by agarose gel electrophoresis. RNA (5 g) was reverse-transcribed in a final volume of 20 l using an RT-PCR kit (Invitrogen, Carlsbad, CA), with either oligo-dT primers (for total RNA) or random hexamers (for nuclear RNA). Each PCR reaction contained 0.5 l of cDNA. Primer sequences are listed in Table I. The linear range of product amplification was established in pilot studies for each primer pair, and the cycle number that reflected the approximate midpoint was used in final experiments. This varied from 18 to 25 cycles for total RNA

FIG. 1. Regulation of gene expression by recombinant adenoviruses encoding wild-type, dominant-negative, and constitutively active Stat 5b.
A, transcriptional activity of wild-type Stat 5b and variants in COS-7 cells transfected with the mouse GH receptor, and the 8xGHRE-TK Luc reporter gene. Cells were infected with different recombinant adenoviruses as indicated, and incubated with recombinant rat GH (1 g/ml) or vehicle (10 mM NaHCO 3 ) for 16 h before analysis of luciferase activity as described under "Experimental Procedures." B, immunoblot using an antibody to the FLAG epitope tag showing equivalent expression of different Stat5b proteins after transduction of recombinant adenoviruses into COS-7 cells. C, dominantnegative Stat5b blocks gene activation by wild-type Stat5b. COS-7 cells were transfected as in A. Cells were infected with adenoviruses encoding wild-type Stat5b and ␤-galactosidase at a ratio of 1:10 (WT/Mock) or with wild-type Stat5b and dominant-negative Stat 5b at a ratio of 1:10 (WT/DN). After incubation with rat GH as in B, cells were harvested, and luciferase activity measured. Results for A and C represent the means Ϯ S.E. of three independent experiments, each performed in duplicate, and were normalized for protein concentration. and from 25 to 30 cycles for nuclear RNA. Results were quantified by densitometry after electrophoresis through 1.5% agarose gels.

Development of Recombinant Adenoviruses Encoding Rat
Stat5b and Variants-To determine if Stat5b plays a role in the transcriptional regulation of IGF-I gene expression, we first constructed a series of recombinant adenoviruses encoding wild-type, dominant-negative, and constitutively active versions of the protein. Dominant-negative rat Stat5b (Ad-Stat5b DN ) contained a substitution mutation of phenylalanine for tyrosine at amino acid 699. The mutation prevents phosphorylation at this residue, thus blocking the inducible dimerization of two Stat5b molecules and the subsequent nuclear import that occurs normally after hormone treatment (20,21). Constitutively active Stat5b (Ad-Stat5b CA ) contains a point mutation of histidine for asparagine at residue 642 that was modeled after an activating mutation in the analogous location in Stat5a (22). To test the function of recombinant adenoviruses encoding these proteins, they were introduced by infection into COS-7 cells that had been transfected a day earlier with expression plasmids for the mouse GH receptor and the hormone-responsive reporter gene, 8xGHRE-TK Luc. GH treatment caused a 7.5-fold increase in luciferase activity in cells infected with Ad-Stat5b WT . No increase in gene expression was observed after GH in cells infected with Ad-Stat5b DN or control virus (Ad-EGFP), and luciferase activity remained at basal levels. In cells receiving Ad-Stat5b CA , reporter gene activity was enhanced in the absence of GH to levels equivalent to those seen after hormone treatment of cells infected with Ad-Stat5b WT , and GH caused a small further rise in luciferase (Fig. 1A). As seen in Fig. 1B, Stat5b protein expression from all three recombinant adenoviruses was similar. To test the func-tion of Ad-Stat5b DN , cells were co-infected with Ad-Stat5b WT , and either Ad-Stat5b DN or an adenovirus encoding ␤-galactosidase (Ad-␤-galactosidase) at a ratio of 1:10 as indicated under "Experimental Procedures." Ad-Stat5b DN completely prevented induction of reporter gene expression by GH, whereas Ad-␤galactosidase was ineffective (Fig. 1C). Based on these results, we conclude that both Ad-Stat5b CA and Ad-Stat5b DN function appropriately.
Quantitative Liver Expression after in Vivo Infection with Ad-Stat5b Variants-Traditional analyses of gene activation and studies in transgenic mice have not yielded the molecular mechanisms by which GH regulates IGF-I gene expression (23)(24)(25), even though in vivo studies have demonstrated that GH exerts a rapid and potent stimulatory effect on IGF-I transcription in the liver (13). Thus to assess any potential role of Stat5b in GH-induced IGF-I gene expression, it was necessary to establish conditions in which quantitative in vivo hepatic infection with recombinant adenoviruses could be achieved. At optimal doses of viruses (2 ϫ 10 10 pfu of Ad-Stat5b and 2 ϫ 10 9 pfu of Ad-tTA), we found that 80 -90% of liver cells were infected in vivo, as demonstrated by immunocytochemistry of subsequently isolated hepatocytes ( Fig. 2A). In the absence of co-infection with Ad-tTA, no Stat5b was produced (data not shown). As seen in Fig. 2B, hypophysectomized rats treated without GH or with hormone for up to 120 min expressed relatively comparable levels of FLAG-tagged Stat5b DN or Stat5b CA , which was absent in rats infected with Ad-EGFP. We were unable to determine the effects of infection with either Ad-Stat5b DN or Ad-Stat5b CA on endogenous Stat5b, because both native and recombinant proteins reacted with Stat5b antibodies. Fig. 3 shows results of time-course experiments designed to evaluate intracellular signal transduction pathways that are usually activated by GH (1). As seen in Fig. 3A, Erk phosphorylation is acutely stimulated by in vivo GH treatment, as evidenced in rats infected with Ad-EGFP by detection of phosphorylated Erks 1 and 2 beginning at 30 min after hormone injection and continuing for up to 120 min, with a peak response of 3-fold above baseline at 30 min. An analogous time course of induction of Erk phosphorylation was observed in rats receiving either Ad-Stat5b DN or Ad-Stat5b CA (2-fold above time 0 at 30 min for Ad-Stat5b DN and 3-fold for Ad-Stat5b CA ). Akt also becomes rapidly phosphorylated after GH (1), as seen in Fig.  3B, and was induced for up to 120 min after GH in hypophysectomized rats infected with any of the three recombinant viruses, although a slightly higher level of basal phosphorylation was noted in animals expressing Ad-Stat5b CA . This latter result is consistent with a prior observation in cultured cells overexpressing Stat5a CA (26). At 30 min after GH in rats infected with any of the recombinant adenoviruses, levels of pAkt were nearly 7-fold above baseline values.

Assessing Signal Transduction Pathways Regulated by GH in Hypophysectomized Rats Expressing Stat5b Variants-
In addition to Stat5, the nuclear expression and activity of Stats 1 and 3 is acutely stimulated by in vivo GH treatment (3-5). As shown in Fig. 3C, nuclear accumulation of these latter transcription factors was rapidly induced by GH in rats expressing each recombinant virus, although in animals infected with Ad-Stat5b DN , basal nuclear expression of each protein was increased slightly in the absence of GH. Nevertheless, at 30 min after GH, nuclear Stat1 was 2-fold above basal levels in rats infected with each of the recombinant adenoviruses; nuclear Stat3 was 3-fold higher than baseline, except in animals infected with Ad-Stat5b DN , where it was 2-fold higher. No change in abundance was seen for the constitutively nuclear transcription factor SP1, confirming that equal amounts of nuclear protein were examined at each time point. Taken together, results from these experiments show that forced expression of recombinant adenoviruses for Stat5b or EGFP did not impede the acute biological effects of GH at the levels of cytoplasmic or nuclear signaling.
The activity of Stat5 was assessed in rats infected with recombinant adenoviruses by gel mobility shift assays, using hepatic nuclear protein extracts isolated before or after GH treatment, and a double-stranded DNA oligonucleotide containing the GH/prolactin response element from the ␤-casein gene. As seen in Fig. 3D, in Ad-EGFP-infected pituitary-deficient rats, GH stimulated sustained nuclear expression and DNA binding activity of Stat5. GH-induced binding toward the ␤-casein GH/prolactin DNA response element was markedly reduced after forced expression of Stat5b DN with a detectable gel shift being seen only at a single time point, 30 min after GH, where Stat5b DN protein expression was slightly diminished compared with other animals (see Fig. 2B). In contrast, DNA binding activity was greatly enhanced even in the absence of GH in rats infected with Ad-Stat5b CA . Taken together with the results in Fig. 3 (A-C), the observations in Fig. 3D demonstrate the effectiveness and specificity of in vivo delivered Stat5b DN and Stat5b CA , respectively, in repressing or activating Stat5 function in the liver.
Hormonal Regulation of Gene Expression-We next examined the effects of Stat5b DN and Stat5b CA on GH-regulated hepatic gene expression. RNA was isolated at various times after GH injection and used as template for measurement of mRNA abundance by semi-quantitative RT-PCR (for primers see Table I). Spi 2.1 is a GH-stimulated liver-specific gene whose expression is dependent on active Stat5 (16,18). As shown in Fig. 4A, GH induced Spi 2.1 mRNA within 120 min of hormone treatment in Ad-EGFP-infected rats but not in Ad-  (27) but by mechanisms that are dependent in part on Stats 1 and 3 (28). As seen in Fig. 4A, c-fos gene expression was increased with similarly rapid kinetics after GH administration in all three groups of rats, although a greater induction was observed in animals infected with Ad-Stat5b DN . Control mRNAs encoding GAPDH and ribosomal protein S17 were unaltered by GH under all experimental conditions (Fig. 4A). IGF-I gene transcription is regulated in the liver by two adjacent promoters, termed P1 and P2, respectively, which control expression of IGF-I mRNAs containing distinct 5Ј exons, 1 and 2, that are individually spliced to a common exon 3 (29). Regulation by GH of transcripts directed by each promoter was assessed by semi-quantitative RT-PCR using the primers listed in Table I. In hypophysectomized rats infected with Ad-EGFP, GH induced an increase in expression of exon 1-containing mRNAs within 60 min of hormone treatment, and exon 2-containing transcripts within 120 min. Induction of IGF-I mRNAs was blocked in rats infected with Ad-Stat5b DN and was enhanced in animals expressing Ad-Stat5b CA , even in the absence of GH (Fig. 4B).
Hormonal Control of Gene Transcription-Based on the results seen in Fig. 4, it appears that Stat5b was both necessary and sufficient for GH-dependent gene expression of Spi 2.1 and IGF-I but was not required for c-fos. To extend these observations in a more mechanistic direction, we next assessed hormone-regulated gene transcription in nuclear RNA by use of a semi-quantitative RT-RCR assay that measures accumulation of nascent transcripts for each gene. A diagram showing the locations of gene-specific primers is found in Fig. 5A, and their DNA sequences are listed in Table I. In pituitary-deficient rats infected with Ad-EGFP, GH caused a rapid and progressive increase in Spi 2.1 transcription beginning at 30 min after hormone injection. Ad-Stat5b DN completely blocked gene activation even at the 30-min time point where nuclear Stat5b was measurable (see Fig. 3D). In contrast, Ad-Stat5b CA stimulated accumulation of nascent Spi 2.1 transcripts even in the absence of GH (Fig. 5B). Results with c-fos were different. GH led to a similarly transient increase in nascent c-fos mRNA in all three groups of rats, with peak expression at the earliest time point examined, 30 min, followed by a subsequent decline at 60 min. In animals infected with Ad-EGFP or Ad-Stat5b CA , a small rise was observed at 120 min, which was not seen in rats infected with Ad-Stat5b DN (Fig. 5B). Transcription of the S17 gene was not regulated by GH (Fig. 5B).
IGF-I gene transcription from promoter 1 was also rapidly and progressively stimulated within 30 min after systemic GH injection in rats infected with Ad-EGFP, was completely inhibited after Ad-Stat5b DN infection, and was markedly enhanced even in the absence of GH in Ad-Stat5b CA rats (Fig. 5C). Nearly identical results were observed with primers designed to measure promoter 2-mediated IGF-I gene transcription. However, these latter primers also detect nascent transcripts directed by promoter 1, which resides 5Ј to promoter 2 (see Fig. 5A). In conjunction with the observations in Fig. 4B, the results in Fig.  5C indicate that GH-stimulated transcription from both promoters of the IGF-I gene is acutely dependent on the availability of functional Stat5b. perimental animals (in this case activation of Stat5b by GH) through use of an adenovirus-mediated gene delivery system. This approach offers a potentially rapid and simple alternative to technically more complicated and slower transgenic strategies to knock out or overexpress specific genes in selected tissues in adult rodents.
Our observations surprisingly identify Stat5b as a key agent regulating the acute actions of GH in stimulating IGF-I gene transcription in the liver, the major site of growth factor production in vivo (8). We find that dominant-negative Stat5b completely prevents GH-stimulated IGF-I gene transcription, whereas active Stat5b effectively substitutes for GH, leading to robust IGF-I gene expression even in the absence of hormone. Although our results establish that Stat5b is an essential component of an acutely regulated signal transduction pathway that extends from the transmembrane GH receptor at the cell surface to the IGF-I gene in the nucleus, they do not show that this transcription factor is critical for normal somatic growth. In fact, published studies demonstrating the modest growth impairment and only ϳ50% decrease in serum IGF-I levels in male mice lacking Stat5b (6,7) indicate that other factors in addition to Stat5b may play collaborative roles in regulating normal somatic growth and IGF-I gene and protein expression in response to GH. In mice without Stat5b since conception, these factors may compensate for the defect in signaling by this transcription factor, such that little growth impairment is seen in female Stat5b-deficient mice and modest defects observed in males.
The IGF-I gene is more complicated than would have been predicted from its simple protein sequence. In rodents the 6 exon gene is transcribed from tandem promoters, each with its own unique leader exon, into several RNA precursor molecules that undergo both alternative processing and differential polyadenylation to yield a multiplicity of mature mRNA species (29). A similarly complex pattern of mRNA production has been observed for other mammalian IGF-I genes, including the human gene (29). Previous studies have demonstrated that GH exerts its primary effects on IGF-I gene transcription but have not established molecular mechanisms or identified the contributing transcription factors. In pituitary-deficient rats, GH acutely induces IGF-I mRNAs regulated by both promoters, leading to a sustained accumulation of multiple IGF-I transcripts (13). As documented by run-on assays using rat hepatic nuclei, GH can stimulate a robust increase in IGF-I gene transcription within 30 min of systemic administration of hormone that can last for at least 6 h (13). Other studies have demonstrated that the induction of IGF-I gene transcription by GH is part of a primary hormonal response that does not require continuous protein synthesis (14).
Despite accumulating evidence showing that the activity of both IGF-I gene promoters is stimulated by GH, little understanding of regulation of promoter function has been achieved to date. Although basal control regions have been identified for each promoter in several mammalian species (25), no GHresponse elements have been mapped. The only physiologically significant regulatory domain that has been characterized has been a binding site in IGF-I promoter 1 for the transcription factor C/EBP␦ that appears responsible for regulation through parathyroid hormone, cAMP, and protein kinase A in osteoblasts (30). Studies using cultured cells generally have been unsuccessful in even showing a significant response of transfected IGF-I promoter-reporter genes to GH (23), and a single experiment using transgenic mice failed to even detect IGF-I transgene expression in the liver, although it did show regulation by GH in the brain (24). Although our results do not identify a GH-response region within the IGF-I gene, they may be interpreted to indicate that this putative region is likely to contain a binding site for Stat5b, and thus offer a rational framework for defining such an element. Alternatively, our data are also consistent with a model in which Stat5b acutely activates another transcription factor that is responsible for induction of IGF-I gene expression by GH. Further studies will be needed to distinguish between these alternatives and to determine the biochemical mechanisms by which Stat5b mediates GH-stimulated IGF-I gene transcription.
Stat5b is normally activated by direct phosphorylation by the GH receptor-associated tyrosine-protein kinase, Jak2 (1, 2), leading to its dimerization and subsequent nuclear translocation (20,21). Many studies have established that these events occur rapidly in the liver and in liver-derived cell lines, as well as in other cell types (5,14,20). Nuclear accumulation of Stat5b has been detected within 5-15 min after a single in vivo hormonal pulse, with specific DNA binding activity being measurable soon thereafter (5). The activation of Stats 1 and 3 by GH occurs with similar kinetics (5). Sustained GH treatment causes the initiation of counter-regulatory measures that dampen subsequent biological responses, and in terms of Stat5b trigger its dephosphorylation by nuclear phosphatases and subsequent export to the cytoplasm (31).
As documented in this manuscript, the genes for both IGF-I and Spi 2.1 comprise part of an acute response to GH that is dependent on Stat5b. GH also rapidly activates other genes in vivo through additional signal transduction pathways. As seen in this report, the early-response gene, c-fos, is induced with kinetics similar to those of IGF-I, but as expected is unmodified by dominant-negative or constitutively active Stat5b, because previous studies have demonstrated that c-fos gene transcription is stimulated by GH through pathways involving Stats 1 and 3, and the Erk-activated Elk-1 transcription factor (28,32,33).
In addition to altering gene activity through acute regulation of several signal transduction pathways, the pattern of GH secretion and action in vivo also influences gene expression. In male rodents, GH is secreted in a pulsatile fashion, whereas in females it is more continuous (34). These sexually dimorphic patterns of GH secretion reversibly modify the transcriptional activity of hepatic genes encoding several cytochrome P450 2C steroid hydroxylases (35) and rodent major urinary proteins (36). To date the molecular mechanisms underlying patternspecific gene regulation by GH remain unknown, and the role of Stat5b has been undefined, although its activity in the nucleus is lower in models of continuous GH exposure than pulsatile treatment (37).
IGF-I plays a multifaceted role in human physiology. It is essential for normal development and somatic growth during childhood (38) and is a key agent in maintaining tissue integrity during aging (39). However, IGF-I also has been linked by epidemiological studies in humans and by experimental manipulations in animals to several cancers (40). Thus, full understanding of the molecular mechanisms by which GH and other trophic agents control IGF-I gene expression is essential for implementing rational therapeutic decisions designed to regulate this critical growth factor.