| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Studies |
Department of Medicine, Clinical Sciences Center, Sheffield University, Sheffield, United Kingdom S5 7AU; the Department of Medicine (R.I.G.H., J.P.M.), Institute of Liver Studies (B.P.), King’s College Hospital, London, United Kingdom SE5 9PJ; and INSERM U-344, Endocrinologie Moleculaire, Faculte de Medecine Necker (M.-C.P.-V.), 75730 Paris, France
Address all correspondence and requests for reprints to: Dr. R. J. M. Ross, Department of Medicine, Clinical Sciences, Northern General Hospital, Sheffield, United Kingdom S5 7AU. E-mail: r.j.ross{at}sheffield.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We recently demonstrated that cirrhotic patients treated with GH for 7 days showed a 2.5-fold increase in their IGF-I levels (9). This significant increase in IGF-I levels suggests that these patients either retain some hepatic GHR on their liver or that IGF-I is being generated from nonhepatic tissues. In addition, we studied the expression of IGF-I and IGF-binding proteins (IGFBPs) in end-stage cirrhotic liver (10, 11). Cirrhotic liver remains transcriptionally active, and in regenerating hepatic nodules, a high level of expression of IGFBP-1 and IGF-I is also detected. In addition to the full- length GHR (GHRfl) messenger ribonucleic acid (mRNA), another transcript corresponding to a truncated receptor is present in normal human liver (12). This truncated receptor lacks 97% of the cytoplasmic domain of the GHR, has a high level of expression on the cell surface, is able to generate large amounts of GH-binding protein (GHBP) (12, 13), and inhibits receptor signalling (12, 14). We now report a detailed analysis of the forms of GHR expression in cirrhotic liver. The results demonstrate that both the full-length receptor and the truncated receptor are expressed in cirrhotic liver, but at a reduced level.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Liver tissue was taken from two groups of patients: normal liver
from liver transplant donors and cirrhotic liver removed at the time of
liver transplant for end-stage liver disease (see Table 1
for clinical details). Before organ
donation, the liver donors had spent a mean time of 24 h (range,
6–72 h) in the intensive care unit, during which time fluid
replacement was given as a 5% dextrose infusion. All donors were
diagnosed as brain dead and had normal liver function throughout their
clinical courses. The other patients had been fasted for 12 h
before surgery. Tissue samples were taken when the liver was mobilized,
placed in liquid nitrogen immediately, and stored at -70 C until
analysis. The study had the approval of the local research ethics
committee.
|
[125I]Human (h) GH was prepared using
chloramine-T, as previously described (15). Iodinated GH had specific
activity of about 30 µCi/µg. The microsomal liver membrane fraction
was prepared by differential centrifugation of liver homogenate in 0.25
mol/L sucrose (15). Protein concentrations were determined by the
method of Lowry et al. (16), using BSA as the standard.
Three hundred micrograms of liver membrane protein were incubated with
125I-labeled GH (
6 x 104 cpm) in 300
µL 25 mmol/L Tris-HCl, 20 mmol/L MgCl2 (pH 7.4), and
0.1% BSA. Parallel incubations were performed in the absence or
presence of various amounts of cold GH (0–3000 ng/mL) in duplicate or
triplicate (depending on sample size) for 4 h at room temperature.
At the end of the incubation, bound and free hormones were separated by
centrifugation (1.3 x 104 x g) for 30 min
at 4 C. The supernatants were discarded, and the membrane pellets were
counted. GH specific binding = (sample counts - nonspecific
binding)/total radioactivity x 100%. An excess of cold GH (3000
ng/mL) was used to determine nonspecific binding. GH association
constants (Ka) and binding capacities
(Bmax) were calculated by Scatchard analysis from
competition experiments using the EBDA-LIGAND program.
Total RNA isolation
Total RNA were extracted from liver tissues using TRI-zol reagent (Molecular Research Center, Cincinnati, OH). The integrity of the RNA and the accuracy of the spectrophotometric determinations were assessed and normalized by densitometry analysis of the ethidium bromide-stained 28S and 18S ribosomal RNA bands after electrophoresis through 1.2% agarose-formaldehyde gels.
RT-PCR and primers
Complementary DNA (cDNA) was made using 5 µg total RNA with 200 U Moloney murine leukemia virus reverse transcriptase (Life Technologies, Gaithersburg, MD), 5 µg random hexamer primers (Boehringer Mannheim, Indianapolis, IN), and 200 µmol/L (final concentration) deoxy-NTP in a 50-µL reaction. The reaction was performed at 24 C for 10 min, at 37 C for 60 min, and at 92 C for 10 min. PCR amplification of GHR transcripts was performed using the GHR709 primer in exon 7 (5'-GGATAAGGAATATGAAGTGC-3') and the GHR1161 primer in exon 10 (5'-GATTTCTCATGGTCACTGC-3'; manufactured by Genosys, Cambridge, UK). The PCR product for GHRfl was predicted to have a size of 453 bp. Five microliters of cDNA (equivalent to 0.5 µg total RNA) were added in a 50-µL volume to 200 µmol/L (final concentration) deoxy-NTP, 2.5 mmol/L (final concentration) MgCl2, 2 µmol/L (final concentration) of each primer, and 1 U Taq DNA polymerase. After heating to 94 C for 3 min, 30 cycles were performed at 94 C for 30 s, 56 C for 1 min, and 72 C for 2 min before a final step of 72 C for 10 min. All PCR reactions included the following controls: 1) the RT was performed with the total RNA without reverse transcriptase to detect possible contamination in RNA samples by previously amplified cDNA or genomic DNA; 2) total RNA was omitted in the RT reaction; and 3) cDNA product was omitted in the PCR reaction. These confirmed that no contamination had occurred during the course of the RT-PCR procedure. The experiment was only considered useful if no band was observed in the negative control lanes. PCR products were separated by gel electrophoresis in a 6% polyacrylamide gel using the Miniprotean apparatus according to the manufacturer’s protocol (Bio-Rad, Hemel Hempstead, UK). Southern blotting was performed as previously described (12).
Polyadenylated [poly(A)+] RNA purification and Northern blotting
Poly(A)+ mRNA from 500 µg total RNA from liver and
IM-9 cells (control for GHR as they express high level of GHR) were
purified with oligo(deoxythymidine)25 Dynabeads (Dynal,
Bromborough, UK). Poly(A)+ mRNA were electrophoresed
through a 1.2% (wt/vol) agarose-formaldehyde gel, transferred to a
Hybond N+ membrane (Amersham International, Aylesbury, UK),
and UV cross-linked to the membrane by autocross-link. The GHR cDNA
(exons 7–10) was labeled with [
-32P]deoxy-CTP (3000
Ci/mmol; Amersham International) using the Primer a Gene kit (Promega,
Madison, WI) and hybridized to the membrane in buffer containing 50%
deionized formamide, 5 x SSPE, 5 x Denhardt’s reagent, and
200 µg/mL fragmented salmon sperm DNA at 42 C for 18 h. The
membrane was washed with 2 x SSPE and 0.2% SDS for 15 min at
room temperature and with 0.2 x SSPE and 0.2% SDS at 60 C for 15
min before exposure to Kodak Omat AR film (Cambridge, UK) at -70
C.
In situ hybridization for GHR
Antisense and sense digoxigenin-labeled probes were made by in vitro transcription with SP6 or T7 RNA polymerase after digestion with SmaI or EcoRV, respectively, of the plasmid pCRII (Invitrogen, San Diego, CA) containing GHR709–1161 nucleotides (17). An aliquot of all digoxigenin-labeled probes was run on an agarose-formaldehyde gel, transferred to Hybond N+, and detected using the manufacturer’s protocol (antidigoxigenin-alkaline phosphatase Fab fragments; Boehringer Mannheim) to confirm probe size and quantity. In situ hybridization was performed as previously described (11). For each experiment, negative controls were sense RNA probe and tissue sections treated prehybridization with 200 mU ribonuclease 1 (RNase-1).
Immunohistochemistry
Tissue was fixed for 24 h in 10% (vol/vol) neutral buffered formalin before dehydration and embedding in paraffin. Four-micron sections were cut using a rotary microtome and mounted on glass microscope slides. Paraffin was removed with xylene, and immunohistochemical staining was performed using the avidin-biotin-peroxide methods (Dako, High Wycombe, UK). Sections were incubated for 2 h at 20 C with the mouse monoclonal anti-GHR antibody, Mab 263 (Biogenesis, Poole, UK), used at a concentration of 10 µg/mL. This antibody was raised against rat liver membrane, but also recognizes the hGHR. The sections were then incubated with goat antimouse biotinylated IgG (Dako) diluted 1:400 in phosphate-buffered saline-1% (wt/vol) BSA for 2 h at 20 C. Positive staining was visualized using 0.075% (wt/vol) 3,3'-diaminobenzidine (Sigma, Dorset, UK) diluted in Tris buffer, and color was developed by the addition of 0.02% (wt/vol) hydrogen peroxide. The primary and secondary incubations were performed for 2 h at room temperature. The avidin-biotin-peroxidase complex/streptavidin incubation was performed for 30 min at room temperature. The 3,3'-diaminobenzidine incubation was performed for 10 min at room temperature. Sections were counterstained with Carazzi’s hematoxylin, dehydrated with ethanol, and mounted, then examined by light microscopy.
The specificity of staining was confirmed by 1) omission of primary antibody, 2) omission of secondary antibody, and 3) preabsorption of primary antibodies with GHR-rich rat liver membrane preparations.
For image analysis, two samples of cirrhotic and normal livers were immunostained for GHR and photographed under identical conditions. The prints were scanned and analyzed using NIH image.
RNase protection assay (RPA; hepatic GHR mRNA quantification)
Hepatic GHR mRNA was measured by RPA. The GHR riboprobe was
against the truncated GHR alternative splice, as previously described
(12). The expected protected fragments were 296 bp for the truncated
GHR and 217 bp for the GHRfl. An antisense probe (385 bp) with high
specific activity was produced by in vitro transcription
using T7 RNA polymerase with 3.7 megabecquerels
[-32P]UTP (Amersham International) and 0.575 µmol/L
unlabeled UTP (Pharmacia, Uppsala, Sweden). Freshly synthesized probe
was gel purified before use in solution hybridization. Expression of
the housekeeping gene, human glyceraldehyde 3-phosphate dehydrogenase
(hGAPDH), was measured simultaneously with GHR mRNA in each sample as a
control for gel loading. An antisense hGAPDH riboprobe (257 bp) with
low specific activity was synthesized by T7 RNA polymerase in
vitro transcription. It protected a 108-bp nucleotide fragment of
hGAPDH mRNA.
In preliminary studies, we observed that RNase protection assay using 25–50 µg total RNA produced a linear signal when hybridized with GHR (2 x 105 cpm/assay) and hGAPDH (1 x 105 cpm/assay) (17A ). Therefore, in the present studies, 25 µg total RNA from normal (n = 4) and cirrhotic (n = 5) human livers were hybridized with the same amounts of the GHR and hGAPDH probes. The hybridization was performed for 16 h at 45 C in 80% (vol/vol) formamide, 40 mmol/L piperazine-N,N'-bis(2-ethanesulfonic acid (pH 6.4), 400 mmol/L sodium chloride, and 1 mmol/L ethylenediamine tetraacetate. After hybridization, the samples were treated with 6.7 µg/mL RNase A and 0.33 µg/mL RNase T1 (Sigma) for 1 h at 30 C. Protected hybrids were isolated by ethanol precipitation and separated on a 8% polyacrylamide/7 mol/L urea denaturing sequencing gel. The dried gel was exposed to Kodak Biomax film (-70 C) for 16 h (normal livers) or 4 days (cirrhotic livers) for the autoradiograph and overnight on a phosphorimager (GS 670, Bio-Rad) to determine the optical densities. The GHR mRNA levels were analyzed by molecular analysis software (Bio-Rad) using the data from the phosphorimager and normalized by the level of hGAPDH mRNA.
Statistical analyses
Results were expressed as the mean ± SEM. Comparison of GHR expression between groups was determined by Mann-Whitney U test, and significance accepted as P < 0.05.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
GH binding assay was carried out using liver membranes from normal
and cirrhotic patients (Table 1
). Specific binding for
[125I]hGH was detected in all cirrhotic samples from
patient with various causes of cirrhosis. GH binding was variable in
both normal and cirrhotic livers, but was generally reduced in
cirrhotic compared to normal liver samples.
There was insufficient liver sample to perform Scatchard analysis on
all liver samples. GH binding affinities (Ka) and
binding capacities (Bmax), however, were
calculated from one normal and three cirrhotic livers (Fig. 1). The GH binding affinity for the
cirrhotic liver samples (Ka, 0.93 ±
0.1 x 109 mol/L-1; mean ±
SEM) was similar to that in the normal liver
(Ka, 1.29 x 109
mol/L-1). The GH binding capacity was 7-fold reduced in
cirrhotic livers (59.6 ± 16.2 fmol/mg protein; mean ±
SEM) compared to that in normal liver (424 fmol/mg
protein). Scatchard plot analysis (Fig. 1) was linear for both normal
and cirrhotic liver samples, indicating the presence of a single class
of binding site for hGH on liver membranes.
|
A single transcript for GHR at 4.8 kb was identified by Northern
blotting using poly(A)+ mRNA from normal and cirrhotic
livers and IM-9 cells (Fig. 2a). Northern
blotting was not used to quantify GHR mRNA levels, as this has been
performed by RPA using a larger number of liver samples. RT-PCR of
normal and cirrhotic livers revealed two products at 453 and 427 bp
(Fig. 2b). As previously reported, the 453-bp product represents the
GHRfl, and the 427-bp product represents a truncated isoform that lacks
97% of the cytoplasmic domain of the receptor (12, 13). Thus, it
appears that, as in normal liver, there is expression of the truncated
receptor in cirrhotic liver.
|
In situ hybridization was performed to confirm and
localize GHR gene expression in the cirrhotic lever. A hematoxylin- and
eosin-stained section showed the typical pattern for cirrhotic liver
(Fig. 3a): hepatocytes surrounded by
fibrous tissue. When the antisense probe was used, intense staining was
detected in hepatocytes and in a few cells in the fibrous tissue (Fig. 3b). However, this staining was not seen with the GHR sense probe (Fig. 3c) and was completely abolished by RNase pretreatment (data not
shown). GHR expression in cirrhotic liver was confirmed by
immunohistochemistry (Fig. 4). GHR
expression was predominantly in regenerating hepatocytes, although
there were individual cells in fibrotic tissue that showed a high level
of expression on both in situ hybridization and
immunohistochemistry. The examples shown are typical of the 10 normal
and 6 cirrhotic liver samples that were analyzed. The examples shown in
Fig. 4 represent the extremes in immunostaining of the cirrhotic liver,
with weak staining in one sample (Fig. 4e) and intense staining in the
other (Fig. 4f). In these two cirrhotic liver samples, processed under
identical conditions, the level of immunostaining was reduced in one
cirrhotic liver by 30% compared to that in the normal liver sample
(Fig. 4, e vs. c), and the other cirrhotic liver showed an
identical level of immunostaining as normal liver (Fig. 4, f
vs. d).
|
|
RNase protection assay was performed using a probe complementary
to the alternatively spliced mRNA for the truncated GHR (12). As
expected, the truncated GHR, in addition to the GHRfl, was detected by
RPA in both normal (n = 4) and cirrhotic (n = 5) livers (Fig. 5). The level of total GHR mRNA
expression (truncated GHR plus GHRfl) was reduced 4.4-fold in cirrhotic
liver compared to that in normal liver (Fig. 6). The reduction in the truncated GHR
mRNA level was significantly greater (5.9-fold) than that in
GHRfl mRNA (4.0-fold; P < 0.05; Fig. 6). The
proportion of the alternative splice GHR mRNA to the total GHR mRNA
decreased 1.3-fold in cirrhotic livers compared to that in control
livers (P < 0.05).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Our results are in contrast to those of the single previous report on the measurement of GH binding in cirrhotic liver presented by Chang et al. (8). Their study only examined one aspect of the GHR, GH binding. There are a number of important methodological differences between Chang’s and the present studies. The sensitivity of the assay used by Chang may have been considerably lower than that reported here. The level of GH binding they found in normal livers was variable and was 3- to 36-fold less than we and others have reported (15). Thus, Chang may have missed low level GH binding in their cirrhotic tissue. GH binding assays are dependent on the method of labeling GH (15), and this may explain the difference between the two studies. In addition, Chang only studied cirrhotic liver that was adjacent to hepatoma tissue, and it is possible the tumor may have influenced their results, as even normal liver tissue adjacent to a hepatoma showed no GH binding. In this study the reduction in binding of labeled GH in cirrhotic liver was similar to that seen in another patient group with acquired GH resistance, thallassemic patients (20).
Whether the reduction of GHR expression and binding of GH is the predominant cause of GH resistance in cirrhosis is not established. The extracellular domain of the GHR circulates as a GHBP. A number of studies suggest that GHBP levels reflect GHR protein expression (21). GHBP levels are low in most situations of acquired GH resistance, including renal failure (22), cirrhosis, malnutrition (23, 24, 25), and critical illness (21). However, in many of these conditions there are changes in other aspects of the GH/IGF-I axis (7). In cirrhosis, these include changes in the pattern of GH secretion (18) and low levels of IGFBP-3 (3). These changes could be secondary to a reduction in GHR signaling. In cirrhosis there are also high levels of IGFBP-1 (9, 19), which inhibits the biological actions of IGF-I, and this is unlikely to be related to a direct effect of reduced GHR signaling. In many conditions of GH resistance, a protease appears in the circulation that reduces the affinity of IGF-I for IGFBP-3 (26). It is of interest that increased protease activity is not a feature of patients with cirrhosis (11, 19). The reduced expression of GHR mRNA in cirrhotic liver was of a similar order of magnitude as the reduction in GH binding. This suggests that the reduced GH binding may be secondary to reduced GHR gene expression and is unlikely to be due to changes in the receptor, as the affinities of the receptor were similar in normal and cirrhotic liver. Immunohistochemistry confirmed expression of the GHR on hepatocytes from cirrhotic liver, and in the two samples analyzed under identical conditions as those used for normal liver, the level of immunostaining for the GHR was reduced in one cirrhotic liver and was similar to that in normal liver in the other cirrhotic liver sample. Overall, it is likely that the reduced level of GH binding in cirrhotic liver is due to a combination of reduced expression of the GHR and reduced hepatocyte number, which may vary between cirrhotic liver samples.
Recently, truncated forms of the GHR have been cloned from normal human liver (12, 13). These truncated receptors, which lack the majority of the cytoplasmic portion of the GHR, result from alternative splicing of exon 9. An identical truncated receptor has been cloned from a permanent cell line (13), and its expression has been confirmed in a number of human tissues (13). The physiological role of these receptors is not fully established. However, in vitro the truncated receptor shows a high level of expression on the cell surface and produces a large amount of GHBP, suggesting that it provides a mechanism for the production of the GHBP (12, 13). Thus, the finding of reduced expression of the truncated GHR in cirrhotic liver may be a further explanation for the reduced GHBP levels found in cirrhosis (21) together with the low level of membrane GHR that can be proteolysed to GHBP.
The truncated receptor heterodimerizes with the full-length receptor and acts as a dominant negative inhibitor of receptor signaling (12, 14). The greater reduction in expression of the truncated receptor compared to that of the full-length receptor in cirrhotic liver may be a compensatory mechanism in cirrhosis. Thus, where there is a low level of expression of the receptor, there is reduced inhibition of the receptor.
It is now clear that treatment with pharmacological doses of GH (10 times the normal replacement dose for adults) can augment IGF-I levels in patients with cirrhosis (2, 9, 19). The rise in IGF-I levels is considerably less than that in normal subjects given a similar dose of GH (9). Thus, these patients remain GH resistant despite GH therapy. This may reflect the low level of GHR expression. The majority of GHR expression in cirrhotic liver appears to be in regenerating hepatocytes, but a low level of expression was also identified in the fibrotic tissue. In normal individuals, the majority of circulating IGF-I is derived from the liver. IGF-I expression is found in most tissues, and it is possible that the rise in IGF-I levels seen in cirrhotic patients during GH treatment is related to tissues other than the liver.
The present study has identified GH binding and GHR expression in cirrhotic liver from patients with end-stage liver disease. Binding and expression were reduced compared to those in normal liver. These results are compatible with the observation that patients with cirrhosis have GH resistance but respond to pharmacological doses of GH.
|
Footnotes |
|---|
Received October 21, 1997.
Revised January 29, 1998.
Revised March 26, 1998.
Accepted April 8, 1998.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  A. Hosui, A. Kimura, D. Yamaji, B.-m. Zhu, R. Na, and L. Hennighausen Loss of STAT5 causes liver fibrosis and cancer development through increased TGF-{beta} and STAT3 activation J. Exp. Med., April 13, 2009; 206(4): 819 - 831. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. A. Held, W. Cosme-Blanco, L. M. Difedele, E. L. Bonkowski, R. K. Menon, and L. A. Denson Alterations in growth hormone receptor abundance regulate growth hormone signaling in murine obstructive cholestasis Am J Physiol Gastrointest Liver Physiol, May 1, 2005; 288(5): G986 - G993. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. D. Wallace, W. J. Abbott-Johnson, D. H. G. Crawford, R. Barnard, J. M. Potter, and R. C. Cuneo GH Treatment in Adults with Chronic Liver Disease: A Randomized, Double-Blind, Placebo-Controlled, Cross-Over Study J. Clin. Endocrinol. Metab., June 1, 2002; 87(6): 2751 - 2759. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Beuschlein, C. J. Strasburger, V. Siegerstetter, D. Moradpour, P. Lichter, M. Bidlingmaier, H. E. Blum, and M. Reincke Acromegaly Caused by Secretion of Growth Hormone by a Non-Hodgkin's Lymphoma N. Engl. J. Med., June 22, 2000; 342(25): 1871 - 1876. [Full Text] [PDF]
|
|
|
|
|
|
T. Amit, M. B. H. Youdim, and Z.'e. Hochberg Does Serum Growth Hormone (GH) Binding Protein Reflect Human GH Receptor Function? J. Clin. Endocrinol. Metab., March 1, 2000; 85(3): 927 - 932. [Abstract] [Full Text]
|
|
|
|
|
|
R. I. G. Holt, J. S. Jones, A. J. Baker, C. R. Buchanan, and J. P. Miell The Effect of Short Stature, Portal Hypertension, and Cholestasis on Growth Hormone Resistance in Children with Liver Disease J. Clin. Endocrinol. Metab., September 1, 1999; 84(9): 3277 - 3282. [Abstract] [Full Text]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| 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 |
