Growth Hormone Rapidly Activates Insulin-Like Growth Factor I Gene Transcription in Vivo
David
P. Bichell,
Departments a&f Surgery Washington Saint Louis,
Kiyoshi
Kikuchi,
and Peter
of Medicine and Biochemistrv (D.P.B.) University School of Medicine Missouri 63110
Rotwein
, and Molecular
Many of the growth-promoting properties of GH are mediated by insulin-like growth factor I (IGF-I), a highly conserved circulating 70-amino acid peptide. Recent studies have shown that multiple mechanisms influence IGF-I gene expression, including transcription from two promoters, alternative RNA splicing, and variable polyadenylation. In order to determine how GH regulates IGF-I gene expression we have analyzed the response of hypophysectomized rats to a single ip injection of recombinant GH. A rise in hepatic IGF-I mRNA was detected within 2 h of GH treatment, with peak values of more than 15-fold above untreated animals by 4 h, and a decline by 16 h. A coordinate increase was seen in all IGF-I mRNA splicing and polyadenylation variants, indicating that neither alternative RNA processing nor differential poly A addition were altered by GH. Transcription run-on experiments using isolated hepatic nuclei and direct analysis of nuclear RNA demonstrated a rise in nascent IGF-I mRNA within 30 min of GH treatment, with peak levels reaching more than lo-fold above background by 2 h and declining by 6 h. As determined by RNase protection assays, transcripts directed by each promoter were coordinately and equivalently activated after GH. A single GH-responsive DNase I hypersensitive site was mapped in chromatin to the second IGF-I intron. This site exhibited rapid kinetics of induction which mirrored the pattern of transcriptional stimulation after GH treatment. These experiments show that GH enhances IGF-I expression in viva by predominantly transcriptional mechanisms. The rapid kinetics of IGF-I gene activation and the temporally associated chromatin changes demonstrate a direct link between a GH-dependent signal transduction pathway and nuclear events. (Molecular Endocrinology 6: 1899-1908, 1992)
Biophysics
and molecular mechanisms by which GH initiates these biological effects are not understood (1). An initial step in GH action involves binding to a specific membraneassociated receptor, and a putative receptor has been cloned and characterized (2-6). Subsequent steps in GH-stimulated signal transduction remain poorly defined, despite recent progress in identifying early events after GH binding (7-l 3). Insulin-like growth factor I (IGF-I) is the major mediator of GH’s actions on somatic growth (14-I 6). IGF-I is a highly conserved 70-residue secreted peptide whose appearance in the circulation is GH-dependent (14-l 7). In GH-deficient hypophysectomized (hypox) rats IGF-I values are less than 5% of normal, and GH treatment leads to a rise in serum IGF-I within 12 h (17, 18). GH therapy also increases hepatic IGF-I mRNA levels (1821) and chronic treatment has a modest (-3-fold) potentiating effect on IGF-I gene transcription (22). Recent studies have shown that the organization of the IGF-I gene is more complicated than would have been predicted from its simple protein structure (2330). In rats the six-exon gene is transcribed by two promoters into RNA precursors that undergo both alternative splicing and differential polyadenylation to produce multiple mature mRNA species (25-30). As part of a long-term goal to determine how GH promotes growth, we have investigated the mechanisms involved in acute regulation of IGF-I gene expression in viva after GH treatment. We find that GH induces IGF-I gene transcription in the liver within 30 min of systemic injection, and that gene activation is associated with a transiently appearing DNase I hypersensitive site (HS) in the second IGF-I intron. The subsequent rise by 2 h in abundance of all IGF-I mRNA species readily explains the later increase in serum IGF-I (17, 18). Rapid activation of IGF-I gene expression after GH treatment suggests direct modification of preexisting transcription factor(s) by a GH-regulated signaling pathway.
INTRODUCTION RESULTS GH has a broad range of actions on growth, tion, and intermediary metabolism, although 0888-8809/92/l 899-l 908$03 00/O Molecular Endocmology CopyrIght 0 1992 by The Endocme Socety
differentiathe cellular
GH Treatment Transcription
Rapidly
Activates
IGF-I Gene
In order to determine how GH regulates IGF-I expression in vivo we first examined the response of male
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
MOL 1900
ENDO.
1992
Vo16No.11
hypox rats to a single injection of 1.5 pg/g recombinant GH. GH treatment caused a rapid and transient rise in the abundance of hepatic IGF-I mRNA. As shown in Fig. 1 A, a nearly lo-fold increase in the major 7.5kilobase (kb) IGF-I transcript was seen by 2 h after GH treatment, with peak levels of greater than 30-fold above hypox values observed by 4 h, followed by a decline toward baseline by 16 h. In other experiments, peak expression of IGF-I mRNA occurred by either 4 or 6 h after GH, ranged from 15- to 30-fold above baseline (data not shown), and reached 40-100% of that seen in control RNA samples (see also Figs. 2 and 4-6). Longer radiographic exposures demonstrated that minor IGF-I mRNAs of approximately l-l .8 kb followed a pattern of induction similar to the 7.8kb band (data not shown). These smaller IGF-I mRNAs represent transcripts that use more proximal polyadenylation sites (25, 28). Thus GH does not alter differential polyadenylation of IGF-I mRNA. Similar results
g
A. IGF-I
f
I
GH-treated 2
4
(hrs) 8
I6
I
22
were obtained with injection of either human or bovine GH (bGH), indicating that the rapid induction of IGF-I mRNA is a consequence of activation of somatogenic receptors (31). In contrast to the rapid change in IGF-I mRNA content seen after GH treatment, levels of albumin mRNA remained relatively constant (Fig. 1 B). Having established that GH uniformly stimulated the accumulation of IGF-I mRNAs with different 3’-ends, we next sought to determine whether GH acutely modified alternative RNA splicing, since published studies had demonstrated that several days of GH treatment caused a preferential increase in IGF-I transcripts lacking differentially processed exon 5 (32). The rat IGF-I gene can undergo alternative RNA splicing in two locations, leading to mRNAs with or without a 186nucleotide (nt) segment of exon 1 (26) and containing or lacking the entire 52-nt exon 5 (32) as diagrammed in Fig. 28. As demonstrated in the autoradiograph of Fig. 2A each class of transcripts was coordinately upregulated by GH. The proportion of IGF-I mRNAs containing the 186-nt fragment of exon 1 (90-95%) was the same after GH treatment as in control rat liver RNA.
s A. -9.5 * -7.5
Exon I, 3 Probe
Kb
Exon 4,5,6 Probe i
GH treated (hrs)
2 2 I
5 $
&E=24816g
-28s /
-2.4 -18s - 1.4
B. Albumin
-28s B.
-2.4 -18s 1.4
5’ r
-
Fig. 1. Induction of IGF-I mRNA after GH Treatment The autoradiographs illustrate results of Northern blots using 15 fig/lane total liver RNA isolated from hypox rats before or 2-16 h after a single ip injection of human GH and from pituitary-intact untreated controls. A, Results of hybridization with a 32P-labeled single-stranded antisense RNA probe derived from a rlGF-I cDNA. B, The same filter after rehybridization with a 32P-labeled antisense rat albumin probe. Major bands are indicated by arrows. Molecular size markers are derived from an RNA ladder. Autoradiographic exposures were for 16 h (A) and 3 h (B).
--_-
-&an1
Fig. 2. GH Treatment IGF-I Transcripts
Ia~--lo~+--r3
Does
Not Regulate
Differential
Splicing
of
A, Autoradiograph depicting results of RNase protection experiments performed with 32P-labeled antisense RNA probes derived from the rlGF-I cDNA diagrammed in B. The 5’-probe (exons 1 and 3) spans a region that is alternatively spliced within exon 1 (26), and the 3’-probe (exons 4, 5, and 6) includes alternatively processed exon 5 (32). Probes and protected bands are outlined below the map. A DNA sequence ladder was included as a size marker, and calibrations at 50nt intervals are indicated. Autoradiographic exposure was for 15 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
GH Stimulates
IGF-I
Gene Transcription
in Vivo
1901
Similar results were seen with IGF-I mRNAs lacking exon 5. Thus acute GH treatment does not modify alternative splicing of the primary IGF-I transcript. We next investigated the effect of GH treatment on IGF-I gene transcription. As illustrated in Fig. 3, GH caused a rapid rise in IGF-I gene transcriptional elongation, as demonstrated by a nuclear run-on assay. Induction of transcription was detected within 30 min of GH injection, reached peak values of lo-fold above baseline by 2 h, and declined by 6 h. By contrast, albumin gene transcription varied by approximately 2fold over the same interval (33). GH also rapidly stimulated the appearance of nascent IGF-I RNA. As determined by an RNase protection
assay using a 345nt probe containing the 182-nt exon 4 and a portion of its 5’-intron (Fig. 4), a rise in both precursor and processed IGF-I mRNA was seen in nuclear RNA by 30 min after GH, with peak values of 9- to 1 O-fold over baseline attained by 2 h. The increase in nuclear IGF-I mRNA preceded its appearance in the cytoplasm by more than 1 h. A similar result was seen with an exon 3 probe (data not shown). GH thus rapidly activated IGF-I gene transcription. Transcriptional Activation Occurs Coordinately through Both IGF-I Gene Promoters The rat IGF-I (rlGF-I) gene contains each capable of directing transcription Whole Cell RNA
A.
GH treated (hrs) I 2
B f’0.5
two promoters, at multiple initi-
Nuclear RNA
6’ exon
*mlr-
I
2 exon 4.5 albumin exon
*am*-
-cIccm-* -
-
is%
1*
Pre-mRNA-
PUC
8. 5kb 5+-j--+ 12
EXOIl
Plasmid
5Okbfl; 3
BXOn 1 exon2
targetS
-3, 45
6
era
5
mRNAC.
HYPOX
1
0.5 TIME
(MS)
AFTER
2
6
GH
Fig. 3. GH Treatment Enhances IGF-I Gene Transcription A, Results of a transcription run-on experiment performed as described in Materials and Methods using rat liver nuclei isolated from hypox animals before and 0.5, 1, 2, and 6 h after a single ip injection of recombinant GH. Autoradiographic exposure was for 72 h. B, Diagram of the rlGF-1 gene, showing the plasmid targets used in A. The distance between the start of exons 1 and 4 is 57 kb. C, Summary of seven similar experiments with results expressed as the relative increase in specific hybridization (mean + SE) after GH treatment.
Fig. 4. Rapid Accumulation of Hepatic Nuclear IGF-I mRNA after GH Treatment The autoradiograph shows results of an RNase protection experiment using 10 pg total rat liver RNA (lert panel), or nuclear RNA (right pane/), hybridized to the 3*P-labeled antisense RNA probe depicted in the lower part of the figure. The sizes of the protected bands were determined from a DNA sequencing ladder electrophoresed in adjacent lanes (not shown). The protected band of 345 nt, seen with increasing abundance in nuclear RNA after GH treatment, is consistent in length with an mRNA precursor, and the 182-nt fragment is the length of exon 4, as diagrammed in the lower part of the figure. Autoradiographic exposure was for 3 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
MOL 1902
ENDO.
1992
Vo16No.11
ation sites within unique leader exons (29, 30). We next asked whether GH injection acutely altered promoter use, since published studies suggested that prolonged GH treatment to hypox rats preferentially stimulated transcripts containing exon 2 (and thus regulated by promoter 2) (34). As demonstrated by RNase protection assay using a single-stranded probe derived from exons 2 and 3 (Fig. 5) the abundance of IGF-I mRNAs either GH-
containing or lacking exon 2 increased coordinately (by 9- to 1Cfold in three independent experiments) after GH injection. Levels of IGF-I mRNAs using each of the three major transcription start sites within exon 2 rose in parallel after GH treatment, as did IGF-I mRNAs using the multiple initiation sites directed by promoter 1 (Fig. 6). At all times after GH injection the proportion of transcripts containing each exon remained constant (70-75% exon 1, 25-30% exon 2). In other experiments we have demonstrated that nuclear transcripts initiating within exons 1 and 2 appeared within 30 min
a 3 +rrr$d $E-124816GATCu
2 a-
s “,I
GH-(trr?.$ed 248
z E SGATC
Undigested Probe -
243 nt-
329
nt -
w 2911
210-213202-
__--------
241 nt
Fig. 5. GH Treatment Coordinately Induces IGF-I mRNAs with Different 5’-Ends The autoradiograph illustrates results of an RNase protection experiment using 10 Kg/lane total rat liver RNA hybridized to a 32P-labeled single-stranded antisense probe from exons 2 and 3 of the rlGF-I gene, as diagrammed in the lower panel. Protected bands include exon P-derived transcripts of 202243 nt (major transcription start sites within exon 2 are indicated by arrows in the lower panel) and a 143-nt fragment representing exon l-containing IGF-I mRNAs. Autoradiographic exposure was for 48 h.
Fig. 6. GH Treatment Stimulates Accumulation of Exon lContaining IGF-I mRNAs with Different 5’-Ends The autoradiograph depicts results of an RNase protection experiment performed with 12.5 pg/lane total rat liver RNA and a 32P-labeled single-stranded antisense RNA probe derived from the 5’-untranslated region of rlGF-I exon 1, as shown in the lower panel. Protected bands representing IGF-I mRNAs originating from both major and minor transcription start sites (29, 30) are indicated by long and short arrows, respectively. Control RNA is from normally growing rat liver. A DNA sequencing ladder using a specific primer derived from the 3’-end of the probe provides localization of protected bands to previously described transcription initiation sites (30). Autoradiographic exposure was for 12 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
GH Stimulates
IGF-I Gene
Transcription
in Viva
of GH treatment (data not shown). GH thus stimulated IGF-I gene transcription through both promoters equivalently.
equivalently, leading to the coordinate appearance of multiple IGF-I mRNA species, including all splicing and polyadenylation variants. Activation of IGF-I gene transcription by GH is accompanied by an equally prompt alteration in chromatin structure, manifested by a shortlived DNase I HS in the second intron. Taken together with published observations on another GH-responsive gene, serum protease inhibitor (Spi) 2.1 (35, 36), our studies show that GH initiates a signaling pathway that culminates in a rapid and powerful transcriptional stimulus. Our results extend on a mechanistic level several previously published experiments that showed a rise in hepatic IGF-I mRNA content after a single GH injection to hypox rats. In three studies the peak response of 4to 6-fold above baseline was attained at 6 h after GH administration (18, 20, 21) while Hynes et al. (19) found an 8- to lo-fold increase by 4 h. Rapid responses to GH also have been reported for the Ob1771 adipocyte cell line (37) and for primary rat hepatocyte cultures (38, 39). In addition, in Obl771 cells 3 days of incubation in GH-containing medium also stimulated IGF-I gene transcription (37). We cannot explain why the magnitude of the GH response was greater in our experiments than in other studies, except to note that the IGF-I mRNA values in our hypox rats approached the limits of detectability, even for the highly sensitive RNase protection assay (see Figs. 2, 4, 5, and 6). The acute effects of GH treatment on IGF-I gene expression that we observed differ in several ways from what has been seen after several days of intermittent or continuous in vivo GH therapy. Daily injections of GH led to a greater response of transcripts containing exon 2 (directed by promoter 2) than mRNAs containing exon 1 (34), while a single GH injection caused a rapid and coordinate increase in all mRNAs transcribed by both promoters (Fig. 5). Four days of daily intermittent injections of GH stimulated a greater rise in IGF-I mRNAs containing alternatively spliced exon 5 than in mRNAs lacking the 52-nt exon (32) while a single GH injection
Identification of a GH-Responsive DNase I Hypersensitive Site in the IGF-I Gene We examined chromatin around the rat IGF-I locus for structural alterations potentially regulated by GH. Figure 7 shows a map of the IGF-I gene, illustrating the 15 DNase I HSs that were identified by digestion of nuclei from normally growing control rats (our manuscript in preparation). DNase I HSs were found throughout the gene, with eight sites clustered within a lo-kb region that included promoters 1 and 2, exons 1-3, and adjacent introns. When nuclei from hypox rats were incubated with DNase I, and the genomic DNA was examined by Southern blotting, 14 of 1.5 hypersensitive sites were readily detected (Figs. 8 and 9, and data not shown). One site, HS 7, located in the second intron, was consistently absent in hypox rats but appeared after GH injection (Fig. 8, top panel). Kinetic studies showed that HS 7 appeared within 30 min of GH treatment, reached maximal intensity by 1 h, and declined by 6 h, while an adjacent site, HS 6, remained unaltered after GH (Fig. 8, bottom panel). In contrast to results observed with IGF-I, DNase I HSs 5’ to the albumin gene did not change after GH injection (Fig. 10).
DISCUSSION The observations presented in this paper demonstrate that in vivo GH treatment rapidly and transiently stimulates IGF-I gene transcription in the liver, the major site of growth factor production in the rat (15-l 7). The induction of IGF-I gene transcription by GH occurs at the level of mRNA initiation and involves both promoters
EcoRl Hindlll BamHl
-30
-20
I
I
-10
0 I
I
I
11, /,I
-f -
I HS site
/A\\-
12345678
:-
+30
+40
+50
+60
I
I
I
I
II1111
----.-+-+-,,+ d-2
DNAse
+20
+10 I
6
1
tttl+tilit tt /I I
-tttt--SltttltH 1I
/I L
IU
t St -
A
i
--:
+70
1I I”
4
9
10
11 12
+80
+90
kb
I
I
it
-1
II
L
i’ \ i
13
14
15
Fig. 7. Map Depicting DNase I HSs in and around the rlGF-I Gene The locations of DNase I HSs 1-15 surrounding the rlGF-I gene are indicated by small vertical lines. Sites were detected Southern blotting using DNA derived from DNase l-treated rat liver nuclei, as described in Materials and Methods, and probes 11, which are illustrated above the map. Restriction sites for EcoRI, Hindlll, and BarnHI are shown near the top of the figure.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
by l-
MOL 1904
ENDO.
’ I
1992
Normal 2 3
Vo16No.11
4”5
Hypox +GH ’ 9 IO II’
HYPOX 6 7 8’
HS Site I-. 23456*7a--
HYP~X 2 3 4
‘I
Kb -23.1 -9.4 -6.6
Hypox 5’
‘6
7
+ 8
GH 9
IO’ -2K3ql -9.4 -6.6
HS 14’
-2.3 -2.0
0
2 5 IO 0 [DRAse
2
5
IO
I lug/ml)] 2
+GH +GH +GH Hypox (30 min) (60 min) (360min) ‘I 2 3”4 5 6 7”8 9101”‘1213 14 15’
2>; Kb -23.1-
2 NOIlllOl
5
HS l5-
17’18 19 20 211&
-94-
-66-44-
-
56.78-
HS S,te -5 -6 -7. -0
-2 3 -20 0 2 to 0 2 5 IO 0 2 5 10 0 2 5 10
-0
2 5 IO -
[DNAS~ I (ug/ml)]
Fig. 8. GH Treatment Stimulates the Appearance of a DNase I HS in the rlGF-I Gene The autoradiographs in the top pane/ illustrate Southern blots of DNA digested with BarnHI after hybridization to probe 2 of Fig. 7, following the protocol described in Materials and Methods. The locations of eight DNase I HSs visible in control rat DNA (lanes l-4) are indicated. Concentrations of DNase I are listed below each lane. Autoradiographic exposure was for 12 days. The bottom panel shows Southern blots of DNA digested with Hindlll and hybridized to probe 2 of Fig. 7. DNase I HSs 5-8 are indicated. Autoradiographic exposure was for 14 days. Lanes 17 and 22 represent DNA isolated from whole liver. Molecular size standards are shown for both panels. GH-dependent HS 7 is indicated by an asterisk.
caused a coordinate induction in both classes of transcripts over the next 4 h (Fig. 2). The reasons for these differences are not known but may be a function of differential RNA stability. It has been reported that GH reduces the half-life of IGF-I mRNAs in primary hepatocyte culture (33) but this observation has not been confirmed (38). Our results on the transcriptional effects of GH also are at variance with the single in vivo study that has been published. A 7-day infusion of GH into /it/ lir mice (an inbred strain with isolated GH deficiency) has been shown to cause only a 3-fold rise in IGF-I gene transcription (22) while a single injection leads to a peak lo-fold induction by 2 h (Fig. 3). It is possible that the transcriptional response becomes attenuated after chronic GH infusion because of down-regulation of one or more components of the signaling pathway. Alternatively, as has been described for several other genes that are modulated by GH, a pulsatile male pattern of GH secretion may be a more effective stimulus to the IGF-I gene than the more continuous female pattern (40-43).
0
I
2
5
[DNAse
7
01257
I (ug/ml)]
Fig. 9. DNase I HSs at the 3’-End of the rlGF-I Gene Are Not Altered by GH Treatment The autoradiographs show results of Southern blots of DNA digested with BarnHI and hybridrzed to probes 9 (upper pane/s), and 10 (lower pane/s) of Fig. 7. DNase I HSs 14 and 15 are indicated. Autoradiographic exposure was for 14 days. Molecular size standards are to the right of each pair of figures.
Based on observations reported here and on other publishedstudies, it appears that GH stimulatesboth rapid and slow genomic effects. For some genes, exemplifiedby IGF-I and Spi 2.1 (35, 36) the actions of GH are rapid. In our studies IGF-I gene transcription was stimulated within 30 min of ip injection. At this earliest time point nearly full-length transcripts were already present in the nucleus and could be detected with an exon 4 probe (Fig. 4). Assumingan elongation rate for RNA polymeraseof up to 1.5 kb/min (44) these resultssuggestthat activation of the GH receptor leads to an almost instantaneoustranscriptional signal. The rapid kinetics of appearance of the GH-dependent DNaseI HS identifiedwithin intron 2 are alsoconsistent with this hypothesis. Rapid actions of GH also have been reported for the Spi 2.1 gene (35) and Yoon et al. (36) have mappeda GH-induced DNaseI HS to the Spi 2.1 gene promoter. They also have detected GHdependent DNA binding activity directed toward this promoter fragment in liver nuclear extracts within 1 h of GH injection to hypox rats (36). The effects of GH seenin vivo on IGF-I and Spi 2.1 genetranscription are nearly as rapid as actionsobserved in vitro on the early responsegenes fos and jun (IO). For other genes, such as some members of the sexually dimorphic P450 2C steroid hydroxylases (41, 42), and the major urinary protein genes (40, 43) GH
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
GH Stimulates
IGF-I Gene Transcription
in Vivo
1905
2 D +GH HYPOX ’ I 2 3
4”5
(30
mm) 6
!‘6
7
+GH (60mrd 9
IO
II
+GH (360 mm) II
12
13
; I
5 14
Kb -23.1 -9.4 -6.6
HS Site I-
-4.4
$
-2.3 -2.0
0
2
5
IV
z
3
IU
[DNAse
ON&e I HS szte:
z
3
IU
L
3
IU
I (up/ml)]
1
2
3
Fig. 10. DNase I HSs in the Rat Albumin Promoter Are Not Altered by GH Treatment The autoradiograph illustrates results of Southern blots of DNA digested with Sstl and hybridized to the probe indicated on the map in the lower part of the figure. DNase I HSs l-3 are indicated by arrows, and their locations are illustrated on the albumin gene map. In lane 14 an Sstl site polymorphism is seen. Autoradiographic exposure was for 12 days. Molecular size markers are indicated to the right of the figure. Lane 14 represents DNA isolated from whole liver.
has a more gradual effect on transcription. Depending on the gene, one or more days of either continuous or intermittent GH treatment to hypox rats are required to demonstrate stimulation in a nuclear run-on assay (41, 42). It thus seems likely that distinct pathways of gene activation are linked to the hepatic GH receptor. Despite the characterization and molecular cloning of a GH receptor from several species (2-6) the mechanisms of action of GH are poorly understood. The identification of rapidly induced nuclear targets could provide a critical impetus for unraveling a signaling pathway that is essential for normal growth and development.
MATERIALS
AND METHODS
Materials Restriction endonucleases, DNA and RNA polymerases, and other enzymes were purchased from commercial suppliers (United States Biochemical Corp., Cleveland, OH; Perkin-Elmer Cetus, Norwalk, CT; Stratagene Cloning Systems, La Jolla, CA; Promega Biotec, Madison, WI; New England Biolabs, Boston, MA: Bethesda Research Laboratories, Gaithersburg,
MD; and Sigma Chemical Co., St. Louis, MO). Ribonucleotide triphosphates were obtained from Pharmacia LKB Biotechnology (Piscataway, NJ) and radionuclides ([~u-~‘P]CTP, [N-~‘P] UTP, [a-32P]dATP, [a-32P]dCTP, [g-3’P]ATP, and [a-35S]dATP) from DuPont-New England Nuclear (Boston, MA) and Amersham Corp. (Arlington Heights, IL). Nitrocellulose filters were purchased from Schleicher & Schuell (Keene, NH) and Immobilon-N transfer membranes from Millipore Corp. (Bedford, MA). Plasmids pGEM3Z. pBluescript SK, and pUC18 were obtained from Promega Biotec, Stratagene, and Bethesda Research Labs, respectively. Animal
Studies
Male Sprague-Dawley rats, hypox by a transauricular route at age 7 weeks, were purchased from Harlan Sprague-Dawley (Indianapolis, IN). Completeness of hypophysectomy was confirmed by lack of weight gain over the next lo-14 days and by low serum IGF-I levels (
David
P. Bichell,
Departments a&f Surgery Washington Saint Louis,
Kiyoshi
Kikuchi,
and Peter
of Medicine and Biochemistrv (D.P.B.) University School of Medicine Missouri 63110
Rotwein
, and Molecular
Many of the growth-promoting properties of GH are mediated by insulin-like growth factor I (IGF-I), a highly conserved circulating 70-amino acid peptide. Recent studies have shown that multiple mechanisms influence IGF-I gene expression, including transcription from two promoters, alternative RNA splicing, and variable polyadenylation. In order to determine how GH regulates IGF-I gene expression we have analyzed the response of hypophysectomized rats to a single ip injection of recombinant GH. A rise in hepatic IGF-I mRNA was detected within 2 h of GH treatment, with peak values of more than 15-fold above untreated animals by 4 h, and a decline by 16 h. A coordinate increase was seen in all IGF-I mRNA splicing and polyadenylation variants, indicating that neither alternative RNA processing nor differential poly A addition were altered by GH. Transcription run-on experiments using isolated hepatic nuclei and direct analysis of nuclear RNA demonstrated a rise in nascent IGF-I mRNA within 30 min of GH treatment, with peak levels reaching more than lo-fold above background by 2 h and declining by 6 h. As determined by RNase protection assays, transcripts directed by each promoter were coordinately and equivalently activated after GH. A single GH-responsive DNase I hypersensitive site was mapped in chromatin to the second IGF-I intron. This site exhibited rapid kinetics of induction which mirrored the pattern of transcriptional stimulation after GH treatment. These experiments show that GH enhances IGF-I expression in viva by predominantly transcriptional mechanisms. The rapid kinetics of IGF-I gene activation and the temporally associated chromatin changes demonstrate a direct link between a GH-dependent signal transduction pathway and nuclear events. (Molecular Endocrinology 6: 1899-1908, 1992)
Biophysics
and molecular mechanisms by which GH initiates these biological effects are not understood (1). An initial step in GH action involves binding to a specific membraneassociated receptor, and a putative receptor has been cloned and characterized (2-6). Subsequent steps in GH-stimulated signal transduction remain poorly defined, despite recent progress in identifying early events after GH binding (7-l 3). Insulin-like growth factor I (IGF-I) is the major mediator of GH’s actions on somatic growth (14-I 6). IGF-I is a highly conserved 70-residue secreted peptide whose appearance in the circulation is GH-dependent (14-l 7). In GH-deficient hypophysectomized (hypox) rats IGF-I values are less than 5% of normal, and GH treatment leads to a rise in serum IGF-I within 12 h (17, 18). GH therapy also increases hepatic IGF-I mRNA levels (1821) and chronic treatment has a modest (-3-fold) potentiating effect on IGF-I gene transcription (22). Recent studies have shown that the organization of the IGF-I gene is more complicated than would have been predicted from its simple protein structure (2330). In rats the six-exon gene is transcribed by two promoters into RNA precursors that undergo both alternative splicing and differential polyadenylation to produce multiple mature mRNA species (25-30). As part of a long-term goal to determine how GH promotes growth, we have investigated the mechanisms involved in acute regulation of IGF-I gene expression in viva after GH treatment. We find that GH induces IGF-I gene transcription in the liver within 30 min of systemic injection, and that gene activation is associated with a transiently appearing DNase I hypersensitive site (HS) in the second IGF-I intron. The subsequent rise by 2 h in abundance of all IGF-I mRNA species readily explains the later increase in serum IGF-I (17, 18). Rapid activation of IGF-I gene expression after GH treatment suggests direct modification of preexisting transcription factor(s) by a GH-regulated signaling pathway.
INTRODUCTION RESULTS GH has a broad range of actions on growth, tion, and intermediary metabolism, although 0888-8809/92/l 899-l 908$03 00/O Molecular Endocmology CopyrIght 0 1992 by The Endocme Socety
differentiathe cellular
GH Treatment Transcription
Rapidly
Activates
IGF-I Gene
In order to determine how GH regulates IGF-I expression in vivo we first examined the response of male
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
MOL 1900
ENDO.
1992
Vo16No.11
hypox rats to a single injection of 1.5 pg/g recombinant GH. GH treatment caused a rapid and transient rise in the abundance of hepatic IGF-I mRNA. As shown in Fig. 1 A, a nearly lo-fold increase in the major 7.5kilobase (kb) IGF-I transcript was seen by 2 h after GH treatment, with peak levels of greater than 30-fold above hypox values observed by 4 h, followed by a decline toward baseline by 16 h. In other experiments, peak expression of IGF-I mRNA occurred by either 4 or 6 h after GH, ranged from 15- to 30-fold above baseline (data not shown), and reached 40-100% of that seen in control RNA samples (see also Figs. 2 and 4-6). Longer radiographic exposures demonstrated that minor IGF-I mRNAs of approximately l-l .8 kb followed a pattern of induction similar to the 7.8kb band (data not shown). These smaller IGF-I mRNAs represent transcripts that use more proximal polyadenylation sites (25, 28). Thus GH does not alter differential polyadenylation of IGF-I mRNA. Similar results
g
A. IGF-I
f
I
GH-treated 2
4
(hrs) 8
I6
I
22
were obtained with injection of either human or bovine GH (bGH), indicating that the rapid induction of IGF-I mRNA is a consequence of activation of somatogenic receptors (31). In contrast to the rapid change in IGF-I mRNA content seen after GH treatment, levels of albumin mRNA remained relatively constant (Fig. 1 B). Having established that GH uniformly stimulated the accumulation of IGF-I mRNAs with different 3’-ends, we next sought to determine whether GH acutely modified alternative RNA splicing, since published studies had demonstrated that several days of GH treatment caused a preferential increase in IGF-I transcripts lacking differentially processed exon 5 (32). The rat IGF-I gene can undergo alternative RNA splicing in two locations, leading to mRNAs with or without a 186nucleotide (nt) segment of exon 1 (26) and containing or lacking the entire 52-nt exon 5 (32) as diagrammed in Fig. 28. As demonstrated in the autoradiograph of Fig. 2A each class of transcripts was coordinately upregulated by GH. The proportion of IGF-I mRNAs containing the 186-nt fragment of exon 1 (90-95%) was the same after GH treatment as in control rat liver RNA.
s A. -9.5 * -7.5
Exon I, 3 Probe
Kb
Exon 4,5,6 Probe i
GH treated (hrs)
2 2 I
5 $
&E=24816g
-28s /
-2.4 -18s - 1.4
B. Albumin
-28s B.
-2.4 -18s 1.4
5’ r
-
Fig. 1. Induction of IGF-I mRNA after GH Treatment The autoradiographs illustrate results of Northern blots using 15 fig/lane total liver RNA isolated from hypox rats before or 2-16 h after a single ip injection of human GH and from pituitary-intact untreated controls. A, Results of hybridization with a 32P-labeled single-stranded antisense RNA probe derived from a rlGF-I cDNA. B, The same filter after rehybridization with a 32P-labeled antisense rat albumin probe. Major bands are indicated by arrows. Molecular size markers are derived from an RNA ladder. Autoradiographic exposures were for 16 h (A) and 3 h (B).
--_-
-&an1
Fig. 2. GH Treatment IGF-I Transcripts
Ia~--lo~+--r3
Does
Not Regulate
Differential
Splicing
of
A, Autoradiograph depicting results of RNase protection experiments performed with 32P-labeled antisense RNA probes derived from the rlGF-I cDNA diagrammed in B. The 5’-probe (exons 1 and 3) spans a region that is alternatively spliced within exon 1 (26), and the 3’-probe (exons 4, 5, and 6) includes alternatively processed exon 5 (32). Probes and protected bands are outlined below the map. A DNA sequence ladder was included as a size marker, and calibrations at 50nt intervals are indicated. Autoradiographic exposure was for 15 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
GH Stimulates
IGF-I
Gene Transcription
in Vivo
1901
Similar results were seen with IGF-I mRNAs lacking exon 5. Thus acute GH treatment does not modify alternative splicing of the primary IGF-I transcript. We next investigated the effect of GH treatment on IGF-I gene transcription. As illustrated in Fig. 3, GH caused a rapid rise in IGF-I gene transcriptional elongation, as demonstrated by a nuclear run-on assay. Induction of transcription was detected within 30 min of GH injection, reached peak values of lo-fold above baseline by 2 h, and declined by 6 h. By contrast, albumin gene transcription varied by approximately 2fold over the same interval (33). GH also rapidly stimulated the appearance of nascent IGF-I RNA. As determined by an RNase protection
assay using a 345nt probe containing the 182-nt exon 4 and a portion of its 5’-intron (Fig. 4), a rise in both precursor and processed IGF-I mRNA was seen in nuclear RNA by 30 min after GH, with peak values of 9- to 1 O-fold over baseline attained by 2 h. The increase in nuclear IGF-I mRNA preceded its appearance in the cytoplasm by more than 1 h. A similar result was seen with an exon 3 probe (data not shown). GH thus rapidly activated IGF-I gene transcription. Transcriptional Activation Occurs Coordinately through Both IGF-I Gene Promoters The rat IGF-I (rlGF-I) gene contains each capable of directing transcription Whole Cell RNA
A.
GH treated (hrs) I 2
B f’0.5
two promoters, at multiple initi-
Nuclear RNA
6’ exon
*mlr-
I
2 exon 4.5 albumin exon
*am*-
-cIccm-* -
-
is%
1*
Pre-mRNA-
PUC
8. 5kb 5+-j--+ 12
EXOIl
Plasmid
5Okbfl; 3
BXOn 1 exon2
targetS
-3, 45
6
era
5
mRNAC.
HYPOX
1
0.5 TIME
(MS)
AFTER
2
6
GH
Fig. 3. GH Treatment Enhances IGF-I Gene Transcription A, Results of a transcription run-on experiment performed as described in Materials and Methods using rat liver nuclei isolated from hypox animals before and 0.5, 1, 2, and 6 h after a single ip injection of recombinant GH. Autoradiographic exposure was for 72 h. B, Diagram of the rlGF-1 gene, showing the plasmid targets used in A. The distance between the start of exons 1 and 4 is 57 kb. C, Summary of seven similar experiments with results expressed as the relative increase in specific hybridization (mean + SE) after GH treatment.
Fig. 4. Rapid Accumulation of Hepatic Nuclear IGF-I mRNA after GH Treatment The autoradiograph shows results of an RNase protection experiment using 10 pg total rat liver RNA (lert panel), or nuclear RNA (right pane/), hybridized to the 3*P-labeled antisense RNA probe depicted in the lower part of the figure. The sizes of the protected bands were determined from a DNA sequencing ladder electrophoresed in adjacent lanes (not shown). The protected band of 345 nt, seen with increasing abundance in nuclear RNA after GH treatment, is consistent in length with an mRNA precursor, and the 182-nt fragment is the length of exon 4, as diagrammed in the lower part of the figure. Autoradiographic exposure was for 3 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
MOL 1902
ENDO.
1992
Vo16No.11
ation sites within unique leader exons (29, 30). We next asked whether GH injection acutely altered promoter use, since published studies suggested that prolonged GH treatment to hypox rats preferentially stimulated transcripts containing exon 2 (and thus regulated by promoter 2) (34). As demonstrated by RNase protection assay using a single-stranded probe derived from exons 2 and 3 (Fig. 5) the abundance of IGF-I mRNAs either GH-
containing or lacking exon 2 increased coordinately (by 9- to 1Cfold in three independent experiments) after GH injection. Levels of IGF-I mRNAs using each of the three major transcription start sites within exon 2 rose in parallel after GH treatment, as did IGF-I mRNAs using the multiple initiation sites directed by promoter 1 (Fig. 6). At all times after GH injection the proportion of transcripts containing each exon remained constant (70-75% exon 1, 25-30% exon 2). In other experiments we have demonstrated that nuclear transcripts initiating within exons 1 and 2 appeared within 30 min
a 3 +rrr$d $E-124816GATCu
2 a-
s “,I
GH-(trr?.$ed 248
z E SGATC
Undigested Probe -
243 nt-
329
nt -
w 2911
210-213202-
__--------
241 nt
Fig. 5. GH Treatment Coordinately Induces IGF-I mRNAs with Different 5’-Ends The autoradiograph illustrates results of an RNase protection experiment using 10 Kg/lane total rat liver RNA hybridized to a 32P-labeled single-stranded antisense probe from exons 2 and 3 of the rlGF-I gene, as diagrammed in the lower panel. Protected bands include exon P-derived transcripts of 202243 nt (major transcription start sites within exon 2 are indicated by arrows in the lower panel) and a 143-nt fragment representing exon l-containing IGF-I mRNAs. Autoradiographic exposure was for 48 h.
Fig. 6. GH Treatment Stimulates Accumulation of Exon lContaining IGF-I mRNAs with Different 5’-Ends The autoradiograph depicts results of an RNase protection experiment performed with 12.5 pg/lane total rat liver RNA and a 32P-labeled single-stranded antisense RNA probe derived from the 5’-untranslated region of rlGF-I exon 1, as shown in the lower panel. Protected bands representing IGF-I mRNAs originating from both major and minor transcription start sites (29, 30) are indicated by long and short arrows, respectively. Control RNA is from normally growing rat liver. A DNA sequencing ladder using a specific primer derived from the 3’-end of the probe provides localization of protected bands to previously described transcription initiation sites (30). Autoradiographic exposure was for 12 h.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
GH Stimulates
IGF-I Gene
Transcription
in Viva
of GH treatment (data not shown). GH thus stimulated IGF-I gene transcription through both promoters equivalently.
equivalently, leading to the coordinate appearance of multiple IGF-I mRNA species, including all splicing and polyadenylation variants. Activation of IGF-I gene transcription by GH is accompanied by an equally prompt alteration in chromatin structure, manifested by a shortlived DNase I HS in the second intron. Taken together with published observations on another GH-responsive gene, serum protease inhibitor (Spi) 2.1 (35, 36), our studies show that GH initiates a signaling pathway that culminates in a rapid and powerful transcriptional stimulus. Our results extend on a mechanistic level several previously published experiments that showed a rise in hepatic IGF-I mRNA content after a single GH injection to hypox rats. In three studies the peak response of 4to 6-fold above baseline was attained at 6 h after GH administration (18, 20, 21) while Hynes et al. (19) found an 8- to lo-fold increase by 4 h. Rapid responses to GH also have been reported for the Ob1771 adipocyte cell line (37) and for primary rat hepatocyte cultures (38, 39). In addition, in Obl771 cells 3 days of incubation in GH-containing medium also stimulated IGF-I gene transcription (37). We cannot explain why the magnitude of the GH response was greater in our experiments than in other studies, except to note that the IGF-I mRNA values in our hypox rats approached the limits of detectability, even for the highly sensitive RNase protection assay (see Figs. 2, 4, 5, and 6). The acute effects of GH treatment on IGF-I gene expression that we observed differ in several ways from what has been seen after several days of intermittent or continuous in vivo GH therapy. Daily injections of GH led to a greater response of transcripts containing exon 2 (directed by promoter 2) than mRNAs containing exon 1 (34), while a single GH injection caused a rapid and coordinate increase in all mRNAs transcribed by both promoters (Fig. 5). Four days of daily intermittent injections of GH stimulated a greater rise in IGF-I mRNAs containing alternatively spliced exon 5 than in mRNAs lacking the 52-nt exon (32) while a single GH injection
Identification of a GH-Responsive DNase I Hypersensitive Site in the IGF-I Gene We examined chromatin around the rat IGF-I locus for structural alterations potentially regulated by GH. Figure 7 shows a map of the IGF-I gene, illustrating the 15 DNase I HSs that were identified by digestion of nuclei from normally growing control rats (our manuscript in preparation). DNase I HSs were found throughout the gene, with eight sites clustered within a lo-kb region that included promoters 1 and 2, exons 1-3, and adjacent introns. When nuclei from hypox rats were incubated with DNase I, and the genomic DNA was examined by Southern blotting, 14 of 1.5 hypersensitive sites were readily detected (Figs. 8 and 9, and data not shown). One site, HS 7, located in the second intron, was consistently absent in hypox rats but appeared after GH injection (Fig. 8, top panel). Kinetic studies showed that HS 7 appeared within 30 min of GH treatment, reached maximal intensity by 1 h, and declined by 6 h, while an adjacent site, HS 6, remained unaltered after GH (Fig. 8, bottom panel). In contrast to results observed with IGF-I, DNase I HSs 5’ to the albumin gene did not change after GH injection (Fig. 10).
DISCUSSION The observations presented in this paper demonstrate that in vivo GH treatment rapidly and transiently stimulates IGF-I gene transcription in the liver, the major site of growth factor production in the rat (15-l 7). The induction of IGF-I gene transcription by GH occurs at the level of mRNA initiation and involves both promoters
EcoRl Hindlll BamHl
-30
-20
I
I
-10
0 I
I
I
11, /,I
-f -
I HS site
/A\\-
12345678
:-
+30
+40
+50
+60
I
I
I
I
II1111
----.-+-+-,,+ d-2
DNAse
+20
+10 I
6
1
tttl+tilit tt /I I
-tttt--SltttltH 1I
/I L
IU
t St -
A
i
--:
+70
1I I”
4
9
10
11 12
+80
+90
kb
I
I
it
-1
II
L
i’ \ i
13
14
15
Fig. 7. Map Depicting DNase I HSs in and around the rlGF-I Gene The locations of DNase I HSs 1-15 surrounding the rlGF-I gene are indicated by small vertical lines. Sites were detected Southern blotting using DNA derived from DNase l-treated rat liver nuclei, as described in Materials and Methods, and probes 11, which are illustrated above the map. Restriction sites for EcoRI, Hindlll, and BarnHI are shown near the top of the figure.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
by l-
MOL 1904
ENDO.
’ I
1992
Normal 2 3
Vo16No.11
4”5
Hypox +GH ’ 9 IO II’
HYPOX 6 7 8’
HS Site I-. 23456*7a--
HYP~X 2 3 4
‘I
Kb -23.1 -9.4 -6.6
Hypox 5’
‘6
7
+ 8
GH 9
IO’ -2K3ql -9.4 -6.6
HS 14’
-2.3 -2.0
0
2 5 IO 0 [DRAse
2
5
IO
I lug/ml)] 2
+GH +GH +GH Hypox (30 min) (60 min) (360min) ‘I 2 3”4 5 6 7”8 9101”‘1213 14 15’
2>; Kb -23.1-
2 NOIlllOl
5
HS l5-
17’18 19 20 211&
-94-
-66-44-
-
56.78-
HS S,te -5 -6 -7. -0
-2 3 -20 0 2 to 0 2 5 IO 0 2 5 10 0 2 5 10
-0
2 5 IO -
[DNAS~ I (ug/ml)]
Fig. 8. GH Treatment Stimulates the Appearance of a DNase I HS in the rlGF-I Gene The autoradiographs in the top pane/ illustrate Southern blots of DNA digested with BarnHI after hybridization to probe 2 of Fig. 7, following the protocol described in Materials and Methods. The locations of eight DNase I HSs visible in control rat DNA (lanes l-4) are indicated. Concentrations of DNase I are listed below each lane. Autoradiographic exposure was for 12 days. The bottom panel shows Southern blots of DNA digested with Hindlll and hybridized to probe 2 of Fig. 7. DNase I HSs 5-8 are indicated. Autoradiographic exposure was for 14 days. Lanes 17 and 22 represent DNA isolated from whole liver. Molecular size standards are shown for both panels. GH-dependent HS 7 is indicated by an asterisk.
caused a coordinate induction in both classes of transcripts over the next 4 h (Fig. 2). The reasons for these differences are not known but may be a function of differential RNA stability. It has been reported that GH reduces the half-life of IGF-I mRNAs in primary hepatocyte culture (33) but this observation has not been confirmed (38). Our results on the transcriptional effects of GH also are at variance with the single in vivo study that has been published. A 7-day infusion of GH into /it/ lir mice (an inbred strain with isolated GH deficiency) has been shown to cause only a 3-fold rise in IGF-I gene transcription (22) while a single injection leads to a peak lo-fold induction by 2 h (Fig. 3). It is possible that the transcriptional response becomes attenuated after chronic GH infusion because of down-regulation of one or more components of the signaling pathway. Alternatively, as has been described for several other genes that are modulated by GH, a pulsatile male pattern of GH secretion may be a more effective stimulus to the IGF-I gene than the more continuous female pattern (40-43).
0
I
2
5
[DNAse
7
01257
I (ug/ml)]
Fig. 9. DNase I HSs at the 3’-End of the rlGF-I Gene Are Not Altered by GH Treatment The autoradiographs show results of Southern blots of DNA digested with BarnHI and hybridrzed to probes 9 (upper pane/s), and 10 (lower pane/s) of Fig. 7. DNase I HSs 14 and 15 are indicated. Autoradiographic exposure was for 14 days. Molecular size standards are to the right of each pair of figures.
Based on observations reported here and on other publishedstudies, it appears that GH stimulatesboth rapid and slow genomic effects. For some genes, exemplifiedby IGF-I and Spi 2.1 (35, 36) the actions of GH are rapid. In our studies IGF-I gene transcription was stimulated within 30 min of ip injection. At this earliest time point nearly full-length transcripts were already present in the nucleus and could be detected with an exon 4 probe (Fig. 4). Assumingan elongation rate for RNA polymeraseof up to 1.5 kb/min (44) these resultssuggestthat activation of the GH receptor leads to an almost instantaneoustranscriptional signal. The rapid kinetics of appearance of the GH-dependent DNaseI HS identifiedwithin intron 2 are alsoconsistent with this hypothesis. Rapid actions of GH also have been reported for the Spi 2.1 gene (35) and Yoon et al. (36) have mappeda GH-induced DNaseI HS to the Spi 2.1 gene promoter. They also have detected GHdependent DNA binding activity directed toward this promoter fragment in liver nuclear extracts within 1 h of GH injection to hypox rats (36). The effects of GH seenin vivo on IGF-I and Spi 2.1 genetranscription are nearly as rapid as actionsobserved in vitro on the early responsegenes fos and jun (IO). For other genes, such as some members of the sexually dimorphic P450 2C steroid hydroxylases (41, 42), and the major urinary protein genes (40, 43) GH
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 17 November 2015. at 14:58 For personal use only. No other uses without permission. . All rights reserved.
GH Stimulates
IGF-I Gene Transcription
in Vivo
1905
2 D +GH HYPOX ’ I 2 3
4”5
(30
mm) 6
!‘6
7
+GH (60mrd 9
IO
II
+GH (360 mm) II
12
13
; I
5 14
Kb -23.1 -9.4 -6.6
HS Site I-
-4.4
$
-2.3 -2.0
0
2
5
IV
z
3
IU
[DNAse
ON&e I HS szte:
z
3
IU
L
3
IU
I (up/ml)]
1
2
3
Fig. 10. DNase I HSs in the Rat Albumin Promoter Are Not Altered by GH Treatment The autoradiograph illustrates results of Southern blots of DNA digested with Sstl and hybridized to the probe indicated on the map in the lower part of the figure. DNase I HSs l-3 are indicated by arrows, and their locations are illustrated on the albumin gene map. In lane 14 an Sstl site polymorphism is seen. Autoradiographic exposure was for 12 days. Molecular size markers are indicated to the right of the figure. Lane 14 represents DNA isolated from whole liver.
has a more gradual effect on transcription. Depending on the gene, one or more days of either continuous or intermittent GH treatment to hypox rats are required to demonstrate stimulation in a nuclear run-on assay (41, 42). It thus seems likely that distinct pathways of gene activation are linked to the hepatic GH receptor. Despite the characterization and molecular cloning of a GH receptor from several species (2-6) the mechanisms of action of GH are poorly understood. The identification of rapidly induced nuclear targets could provide a critical impetus for unraveling a signaling pathway that is essential for normal growth and development.
MATERIALS
AND METHODS
Materials Restriction endonucleases, DNA and RNA polymerases, and other enzymes were purchased from commercial suppliers (United States Biochemical Corp., Cleveland, OH; Perkin-Elmer Cetus, Norwalk, CT; Stratagene Cloning Systems, La Jolla, CA; Promega Biotec, Madison, WI; New England Biolabs, Boston, MA: Bethesda Research Laboratories, Gaithersburg,
MD; and Sigma Chemical Co., St. Louis, MO). Ribonucleotide triphosphates were obtained from Pharmacia LKB Biotechnology (Piscataway, NJ) and radionuclides ([~u-~‘P]CTP, [N-~‘P] UTP, [a-32P]dATP, [a-32P]dCTP, [g-3’P]ATP, and [a-35S]dATP) from DuPont-New England Nuclear (Boston, MA) and Amersham Corp. (Arlington Heights, IL). Nitrocellulose filters were purchased from Schleicher & Schuell (Keene, NH) and Immobilon-N transfer membranes from Millipore Corp. (Bedford, MA). Plasmids pGEM3Z. pBluescript SK, and pUC18 were obtained from Promega Biotec, Stratagene, and Bethesda Research Labs, respectively. Animal
Studies
Male Sprague-Dawley rats, hypox by a transauricular route at age 7 weeks, were purchased from Harlan Sprague-Dawley (Indianapolis, IN). Completeness of hypophysectomy was confirmed by lack of weight gain over the next lo-14 days and by low serum IGF-I levels (
