| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institutes of Molecular Animal Breeding/Gene Center and
* Veterinary Pathology, Ludwig-Maximilian University, Munich, Germany;
Laboratory of Endocrine Research, Medical Department II Köln-Merheim, University of Cologne, Germany; and
Laboratory of Endocrine Research, University Children’s Hospital, Tuebingen, Germany
1Correspondence: Institute of Molecular Animal Breeding/Gene Center, Ludwig-Maximilian University, Feodor-Lynen-Str. 25, 81377 Munich, Germany. E-mail: ewolf{at}lmb.uni-muenchen.de
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: adrenal gland • hypertrophy • hyperplasia • growth hormone • IGF binding protein 2 • transgenic mouse • corticosterone
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Both local and systemic controls determine organ size, but their relative importance can vary greatly (reviewed in ref 2
). Growth hormone (GH) and insulin-like growth factor I (IGF-I) are important stimulators of cell growth (cell mass increase), survival and proliferation, depending on tissue type and physiological/pathological status (3
4
5
6)
. IGF binding proteins (IGFBPs) are capable of modulating IGF effects via different mechanisms (7)
, but their role in regulating cell size, survival and proliferation in vivo is largely unclear.
To address this point, we used GH-overexpressing transgenic mice, which exhibit excessive total cell mass in many organs (8)
, as a model. We investigated the adrenal glands, which have been shown to be enlarged as a consequence of GH overexpression. Since adrenal enlargement in this model is associated with corticosterone hypersecretion (9
, 10)
, studies of the regulation of size and number of adrenocortical cells can easily be combined with a functional readout, i.e., plasma corticosterone concentration.
Development and function of the adrenal glands are markedly influenced by components of the IGF system (11
12
13
14
15)
. Besides IGF-I and -II, IGF-I receptors (IGF-IR), and IGF-II/mannose 6-phosphate receptors (IGF-II/M6PR) (16
, 17)
, messenger RNA expression for all six high-affinity IGF binding proteins (IGFBPs) has been detected in fetal (18)
and adult human adrenocortical tissue (18
, 19)
. Conditioned media of adrenal cells from various species were found to contain several IGFBPs (19
20
21
22)
.
Found to account for 
12% of the IGF binding capacity secreted by cultured human adult adrenocortical cells (19)
, IGFBP-2 attracted particular attention in the context of adrenal physiology and pathology since 1) IGFBP-2 is coregulated with IGF-II, an important factor inducing proliferation of adrenocortical cells and steroidogenesis (23)
; 2) overexpression of IGFBP-2 and secretion into the circulation are characteristic for adrenocortical carcinomas (24
25
26
; reviewed in ref 27
); and 3) overexpression of IGFBP-2 in mouse adrenocortical tumor cells (Y-1) increased their tumorigenic potential, i.e., stimulation of anchorage-dependent and -independent proliferation and changes in cellular morphology (28
; reviewed in ref 27
). Therefore, IGFBP-2 is a good candidate to modulate GH/IGF actions on the adrenal gland or to act by GH/IGF-independent mechanisms.
To characterize effects of IGFBP-2 on normal and GH-stimulated growth and function on the adrenal gland, we crossed transgenic mice overexpressing IGFBP-2 under the control of the CMV promoter (29)
with PEPCK-bGH transgenic mice (30)
and performed quantitative morphological and functional studies of the adrenal glands of the resulting groups of mice. We asked 1) whether adrenal enlargement of GH transgenic mice is due to hypertrophy, hyperplasia, or both; 2) how corticosterone hypersecretion is linked with these changes; and 3) whether distinct components of GH/IGF-I-induced adrenal enlargement and hyperfunction are affected by IGFBP-2.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. A PEPCK-bGH transgenic founder mouse was originally generated on a C57BL/6 x SJL genetic background (31)
. The transgene was then transferred to an NMRI outbred background (12 generations). CMV-IGFBP-2 transgenic mice originated from a B6D2F2 founder mouse (29)
after three generations backcrossing to C57BL/6 mice. To ensure a similar genetic background of all mice investigated, female hemizygous CMV-IGFBP-2 transgenic mice were mated with male hemizygous PEPCK-bGH transgenic mice, producing all experimental groups within the same litters: mice carrying both transgenes (GB), a single GH (G) or IGFBP-2 transgene (B), and nontransgenic controls (C).
Male offspring were analyzed for adrenal growth and corticosterone secretion to avoid bias of the results by the juxtamedullary X zone, which degenerates after birth in male but persists in female adrenal glands with highly variable size (32
, 33)
. All mice were maintained under standard (nonbarrier) conditions and had free access to a standard diet (V1534; Ssniff, Soest, Germany) and tap water.
Serum levels of GH, IGFBP-2, and IGF-I in the various experimental groups have been reported elsewhere (30)
. Serum concentrations of GH were in the range of 2 µg/mL both in G and GB mice. The endogenous GH in C and B mice was not detected by the ELISA used in our study (34)
. Serum IGFBP-2 levels were four- to ninefold increased both in B and GB mice vs. C and G mice. Circulating IGF-I levels were two- to threefold increased in the GH-overexpressing groups. Serum GH, IGFBP-2, and IGF-I levels of mice investigated in the present study are summarized in Table 1
.
|
Detection of IGFBP-2 in adrenal tissue
IGFBP-2 in protein extracts of adrenal glands from the different genetic groups was detected by Western immunoblot (WIB) analysis as described previously (29
, 30)
. Tissue samples were homogenized in extraction buffer [10 mM Na2HPO4, pH 7.0; 0.2% (w/v) sodium dodecyl sulfate (SDS); 10% (w/v) glycerin] using a cell homogenizer (ART, Mühlheim, Germany); 5 µg of protein was boiled (7 min) and separated under reducing conditions on a 5% stacking/12% separating SDS-polyacrylamide gel using the Mini Protean IITM system (Bio-Rad, Munich, Germany). Separated proteins were transferred to a PVDF membrane (ImmobilonTM-P; Millipore, Eschborn, Germany). The blots were blocked with 3% bovine serum albumin and IGFBP-2 was identified using a specific antiserum (diluted 1/800) raised in rabbit (29)
. Bound antibodies were detected with peroxidase-coupled antibodies (diluted 1/5000) against rabbit IgG (Dianova, Hamburg, Germany) using an ECL detection kit (Amersham Pharmacia Biotech, Freiburg, Germany).
Biological activity of IGFBP-2 was shown by Western ligand blot (WLB) analysis (35)
using [125I]IGF-II as tracer. Blots were prepared as for WIB except that proteins were separated under nonreducing conditions and membranes were blocked in 1% gelatin. All incubation and washing steps were performed at 4°C. Bound [125I]IGF-II was visualized on Phosphor-Imager Storm (Molecular Dynamics, Krefeld, Germany) (29)
. Two mice per genetic group were analyzed at the age of 5 wk.
Body weight and weight of the adrenal glands
Body weight of all mice was recorded in weekly intervals to the nearest 0.1 g (30)
. At an age of 5 wk and 4 months, respectively, male mice from all genetic groups were ether anesthetized and killed by bleeding from the retro-orbital sinus. The abdominal cavity was opened, and all internal organs were removed. Adrenal glands were carefully isolated from adhering adipose tissue under a dissecting microscope and weighed immediately to the nearest 0.1 mg.
Stereological investigations
Stereological analysis of the adrenal glands was carried out on 12 male mice (3 per genetic group) at the age of 11 wk. Animals were anesthetized by ether inhalation and fixed by orthograde perfusion via the left heart ventricle. The vasculature was preflushed for 20 s with phosphate-buffered saline (PBS, pH 7.4) and perfused with a 3% glutaraldehyde-PBS (pH 7.4) solution for 5 min under a constant pressure of 100 mm Hg. An incision of the inferior vena cava served as an outlet for the perfusate. After perfusion fixation, internal organs were removed and the adrenal glands were postfixed in situ for an additional 48 h in the same fixative. Adrenal glands were subsequently excised and carefully prepared free from adjacent tissues under a dissecting microscope. The left adrenal gland was weighed to the nearest 0.1 mg, embedded in paraffin wax following standard procedures, and exhaustively sectioned at a nominal thickness of 3 µm on a microtome equipped with a section counter. Every 20th section of the series was saved, stained with hematoxylin and eosin, and used for morphometric evaluation carried out on a Videoplan image analysis system (Zeiss-Kontron, Germany) coupled to a light microscope via a color video camera. The total number of paraffin sections sampled per adrenal gland ranged from 14 to 28. On these sections, the cross-sectional areas of cortex and medulla were planimetrically determined; cross-sectional areas of the adrenocortical zones were estimated by point counting at a 340 x final magnification provided by a 10 x objective. For calibration, an object micrometer (Zeiss, Germany) was used. Since in mice the zona fasciculata and reticularis are not distinguishable at the light microscopic level (36)
, these zones were calculated jointly and for simplicity will henceforth be denoted zona fasciculata only. Volume fractions of the cortex VV(cortex/adrenal gland) and the medulla VV(medulla/adrenal gland) in the adrenal gland were determined following the principle of Delesse and calculated as the sum of cross-sectional areas of cortical and medullary tissue, respectively, divided by the sum of cross-sectional areas of the whole adrenal gland. The fractional volume of the zona fasciculata in the adrenal cortex (VV(zona fasciculata/cortex)) was estimated from the corresponding point fraction (37)
. The absolute volume of the cortex and the medulla was obtained as the product of the corresponding volume fractions (VV(cortex/adrenal gland) and VV(medulla/adrenal gland), respectively) and the volume of the adrenal gland. The latter was calculated by dividing the weight of the adrenal gland by the specific weight, which is reported to correspond to 1.039 mg/mm3 for adrenal glands (38)
. The volume of the zona fasciculata was obtained as the product of VV(zona fasciculata/cortex) and the volume of the cortex.
Another aim of this study was to analyze in quantitative terms changes in the zona fasciculata at a cellular level. For this analysis, the disector method was applied, which allows unbiased assumption-free counting and sizing of particles (39)
. Based on pilot studies, we used the physical disector in combination with systematic point counting to estimate the numerical density and the mean volume of zona fasciculata cells. Paraffin sections are not useful for counting cells with physical disectors (40)
. Therefore, the right adrenal gland was embedded in Epon after being washed several times in 0.1 M phosphate buffer and after subsequent postfixing in 1% osmium tetroxide and dehydration through a series of acetone solutions. After trimming of the blocks, at least eight serial semi-thin sections (0.5 µm) including cortex and medulla were cut from each adrenal gland with a Reichert-Jung "Ultracut E" microtome (Leica, Germany), mounted on consecutively numbered glass slides, and stained with toluidine blue and safranine. From the stack of serially cut semi-thin sections, one was drawn at random by means of a random number (R) between 2 and 8 (the baseline section was not used for sampling) as a reference section in a disector. The second section (look-up section) was sampled among no. 2–8 by means of R ± 3, i.e., the disector height was equivalent to the thickness of three semi-thin sections (1.5 µm). Five fields were systematically sampled at random in the zona fasciculata of the reference section and the corresponding fields were identified in the look-up section. Light microscopic images of the selected fields were acquired with a color video camera using a 63x oil immersion objective and color prints were prepared at the same final magnification (2300x). A plastic transparency with equally spaced test points (n=70) and an unbiased counting frame representing an area of 8800 µm2 was superimposed on the printed images. All profiles of zona fasciculata cell nuclei sampled in the frame in the reference section not present in the look-up section were counted (Q-). On reference sections, the number of points hitting zona fasciculata cells was counted as well as the points hitting the zona fasciculata. The operation of counting of Q- (cell nuclei) was then repeated by interchanging the roles of the look-up section and the reference section, thereby increasing the efficiency by a factor of two. Given that zona fasciculata cells have only one nucleus, which is documented in the literature (36)
and also observed in this study, no cell is counted twice with this procedure.
On average, 61 nuclei (range: 45–83) were counted with the five disector pairs per adrenal gland. The numerical density of the epithelial cells in the zona fasciculata was calculated by dividing the total number of cells counted in all disectors in an adrenal gland by the cumulative volume of the disectors (area of the unbiased counting framexdisector heightxnumber of disectors) sampled in the adrenal gland. Assuming the same numerical density of epithelial cells in the zona fasciculata in both (right and left) adrenal glands and neglecting the small amount of tissue shrinkage when using Epon-embedding (37)
, the total number of zona fasciculata cells per adrenal gland was calculated as the product of the numerical density of epithelial cells in the zona fasciculata and the volume of the zona fasciculata. The mean volume of zona fasciculata cells was obtained by dividing the volume density by the numerical density of the epithelial cells in the zona fasciculata. To control the nominal thickness of semi-thin sections (0.5 µm), a resectioning technique was used. Five semi-thin sections that were not used for sampling were selected from different section series and re-embedded in Epon. Ultra-thin sections (70 nm) of the semi-thin sections were stained with uranyl acetate and lead citrate and examined with a Zeiss EM 10 electron microscope. Electron micrographs (final magnification 10,500x) were taken from segments of the primary section showing every evidence of having been resectioned normally to their surface. A waffle grating replica having 2160 lines/mm (Plano, Germany) was used for calibration. The measured mean thickness of the semi-thin sections was 0.497 ± 0.020 µm. Thus, the precision of the ultramicrotome was confirmed.
Adrenal function tests
Adult male mice (4 months) were housed under controlled illumination (lights on from 0700 to 1900) and temperature (23°C) with free access to food and tap water. Animals were housed singly in opaque cages for 2 wk before measurement of basal corticosterone levels. Blood samples were taken between 1600 and 1800 h. The animals were anesthetized individually in a glass jar containing saturated ether vapor and retro-orbital blood was collected within 30 s of the initial disturbance from the cage (23)
. To measure stimulated corticosterone levels, the anesthetized animals were treated with 1 IU/100 g body weight ACTH (1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24)
(Synacthen®, Ciba-Geigy, Basel, Switzerland) intraperitoneally and a second blood sample was obtained 60 min later. Blood was collected in ice-chilled, EDTA-coated Eppendorf tubes containing 200 IU aprotinin (Trasylol®, Bayer AG, Leverkusen, Germany). Plasma samples were stored at -80°C until analysis by radioimmunoassay. For the corticosterone assay (ICN Biomedicals, Costa Mesa, CA), inter- and intra-assay coefficients of variance were 7.2% and 6.9%, respectively, with a detection limit of 25 ng/mL.
To clarify whether differences in plasma corticosterone levels between genetic groups are due to altered basal plasma ACTH, blood samples from another set of animals (4–7 months old) were taken between 0800 and 0900 h and between 1600 and 1700 h, then collected in EDTA-coated tubes as described above. For measurements of plasma ACTH levels, we obtained blood samples by making a small incision into the tail vein of conscious animals. In our experience, this is the least stressful method of blood sampling if only a small amount of blood is needed and the animals are used to being handled. To exclude strain-specific effects we also determined plasma ACTH levels in a different transgenic mouse strain expressing bGH under the control of the mouse metallothionein I (MT) promoter (41)
. Blood samples from male MTbGH transgenic mice and littermate controls were taken in the morning. The inter- and intra-assay coefficients of variance for the ACTH assay (Diagnostic Systems Laboratories, Webster, MA) were below 10% and 7%, respectively, with a detection limit of 25 pg/mL.
RT real-time PCR analysis of IGFBP-2, IGF-I, IGF-IR, and GHR mRNA expression
Individual adrenal glands were homogenized in guanidinium thiocyanate and total RNA was pelleted by CsCl gradient ultracentrifugation as described before (29)
. The pellet was dissolved in 200 µL water, ethanol-precipitated, and dissolved in water. RNA concentration was determined by measuring absorbency at 260 nm. Total RNA (2 µg) was DNase digested using 2 µL DNase (10 IU/µl) (Roche, Mannheim, Germany) at 37°C for 15 min to remove genomic DNA contamination. After enzyme inactivation at 65°C for 10 min, the absence of genomic DNA was controlled by PCR using ß-actin-specific primers (Table 2
).
|
For mRNA quantification, 1 µg of DNase-digested total RNA was incubated for 5 min at 94°C, vortexed for 5 s, and chilled on ice/water. RT was performed in a total volume of 20 µL containing 50 mM Tris HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, 10 mM DTT, 1 mM each dGTP, dATP, dTTP, dCTP (MBI Fermentas, St. Leon-Rot, Germany), 600 ng/mL random hexamer primers (pN6) (Roche), and 20 IU murine leukemia virus (MuLV) reverse transcriptase (Gibco BRL, Grand Island, NY). The RT reaction was carried out at 4°C for 1 h, terminated by 5 min at 95°C and placed on ice/water.
Quantification of mRNA abundance was performed by real-time PCR detection using an ABI PRISM 7700 Sequence Detector (PE Biosystems, Weiterstadt, Germany) and Sybr®Green as a dsDNA-specific fluorescent dye (42)
. Amplification mixes (25 µL) contained 2.5 µL cDNA solution, 2 x Sybr®Green PCR Master Mix, 12.5 pM of each primer, and 0.25 IU AmpErase uracil N-glycosylase (UNG, PE Biosystems). Amplification primers (Table 2)
were designed using Primer Express software (PE Biosystems).
PCR was started with 2 min at 50°C for AmpErase activation and 10 min at 95°C for denaturation. The program continued with 40 cycles of 15 s at 95°C and 60 s at 62°C. Each assay included triplicates of cDNA for the gene of interest, a no-template control, and four dilutions of cDNA pooled from all genetic groups for the gene of interest and for the reference gene ß-actin to calculate the corresponding amplification efficiency (E=10-(1/b)-1; b = regression coefficient). The parameter CT (threshold cycle) is defined as the cycle number at which fluorescence intensity exceeds a fixed threshold. Relative mRNA expression for each gene of interest (I) was calculated using the formula:
 |
Statistical analysis
Data were analyzed by ANOVA taking the effect of genetic group into account. Means were compared by using LSD post hoc tests (SPSS program package; SPSS, Chicago, IL). Data are presented as means and standard deviations (SD) throughout the study.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). As observed before (29)
, the apparent size of IGFBP-2 was different in WIB (34 kDa) and WLB (32 kDa), presumably because the former was performed under reducing, the latter under nonreducing, conditions.
|
Body weight and weight of the adrenal glands
Body weight was markedly stimulated by expression of the PEPCK-bGH transgene in G mice. This became significant from an age of 5 wk, when the body weight of G mice was increased by 34% compared with C mice. At the age of 4 months, the body weight of G vs. C mice was almost twofold increased (Fig. 2
a). Overexpression of IGFBP-2 in GB mice reduced the effect of GH overexpression to control levels in 5-wk-old animals (P<0.001) and by 17% (P<0.001 vs. G mice) at the age of 4 months. The body weight of 4-month-old B vs. C mice was significantly (13%; P<0.05) reduced.
|
Absolute weight of the adrenal glands was significantly (P<0.05) increased in 5-wk- and 4-month-old G mice vs. C and B mice. This effect of GH excess was eliminated by IGFBP-2 overexpression in GB mice, which exhibited adrenal weights similar to C and B mice (Fig. 2b
). In 4-month-old animals, IGFBP-2 overexpression in B mice significantly (P<0.05) reduced the weight of their adrenal glands compared with C mice. The effect of IGFBP-2 to reduce adrenal weights was greater (P<0.05) in GH excess and elevated levels of IGF-I than in mice with normal GH and IGF-I expression. Adrenal weights were reduced by 14% in 4-month-old B vs. C mice, but by 37% in 4-month-old GB vs. G mice, a greater reduction than observed for any other organ (30)
. Thus, in the context of GH excess, an important role of the IGF system in the adrenal gland can be assumed.
Relative adrenal weights were not different between the genetic groups at 5 wk of age but were reduced (P<0.05) in 4-month-old G and GB vs. C and B mice (Fig. 2c
).
Volumes of adrenal cortex and medulla
To evaluate the contributions of adrenal cortex and medulla to the differences in adrenal weights between the genetic groups (age=11 wk), the volume fractions of cortex and medulla were determined and their absolute volumes were calculated. The volume fractions VV(cortex/adrenal gland) and VV(medulla/adrenal gland) ranged from 81 to 85% and from 15 to 19%, respectively, and were not significantly different between the genetic groups (data not shown).
However, the absolute volume of the adrenal cortex of G mice was twofold (P<0.001) increased vs. C mice (Fig. 3
a). Coexpression of the IGFBP-2 transgene in GB mice caused a significant (P<0.001) reduction of the adrenocortical volume vs. G mice. The volume of the adrenal cortex was smaller in B than in C mice (P<0.01).
|
The volume of the adrenal medulla was 1.7-fold (P<0.01) larger in G than in C mice, whereas in B and GB mice the means for this parameter were not significantly different from that of controls (Fig. 3b
).
These data show that GH excess stimulates growth of the adrenal cortex and medulla and that both compartments are affected by elevated IGFBP-2. Representative sections of adrenal glands from the different groups are shown in Fig. 3c
.
Size and number of zona fasciculata cells
To clarify whether adrenocortical enlargement of GH-overexpressing mice is due to hypertrophy or due to hyperplasia, we determined the volume and number of zona fasciculata cells using stereological methods. The mean volume of zona fasciculata cells was 40–50% (P<0.05) increased in G vs. C and B mice. Strikingly, this increase was completely abolished by overexpression of IGFBP-2 in GB mice whose mean volume of zona fasciculata cells was not different from those of C and B mice (Fig. 4
a). In contrast, the total number of zona fasciculata cells was not different in GB and B mice, i.e., GH-induced hyperplasia of adrenocortical cells (50%) was not affected by elevated IGFBP-2 levels (Fig. 4b
). Thus, a clear-cut effect of IGFBP-2 to inhibit hypertrophy, but not hyperplasia, of zona fasciculata cells can be demonstrated in this model. Representative segments of adrenal cortex from the groups of mice investigated are shown in Fig. 4c
.
|
Adrenal function tests
Basal plasma corticosterone levels of 4-month-old G mice were > threefold (P<0.001) increased if compared with C mice. In contrast, plasma corticosterone was elevated by only 40% (P=0.117) in GB vs. C mice but was 47% (P<0.001) reduced in GB vs. G mice, indicating a marked inhibitory effect of elevated IGFBP-2 on corticosterone secretion of giant GH transgenic mice. IGFBP-2 slightly (26%) reduced plasma corticosterone levels in B vs. C mice, but the difference between these groups was not statistically significant (Fig. 5
).
|
Basal plasma corticosterone levels of C mice in the present study were higher than observed in an earlier study evaluating effects of IGF-II overexpression on adrenal growth and function (23)
. This is probably because blood samples of the present study were taken in the afternoon, the last third of the light period where the highest levels of circulating corticosterone are measured in mice (43)
, and the mice investigated had a different genetic background.
Stimulation of corticosterone secretion by injection of ACTH resulted in an threefold increase of plasma corticosterone levels in all groups. Again, overexpression of IGFBP-2 caused significant (P<0.001) reduction of corticosterone secretion in GB vs. G mice whereas plasma corticosterone concentrations of B vs. C mice were only slightly reduced (Fig. 5)
.
Plasma ACTH levels were not significantly different in the genetic groups when investigated in the afternoon and were slightly increased in G and GB vs. C and B mice (Fig. 6
a). Only differences between GB vs. C and B mice reached a level of statistical significance (P<0.05). Plasma ACTH levels of C and B mice were significantly (P<0.05) higher in the afternoon than in the morning. This physiological diurnal variation of ACTH secretion was not seen in G and GB mice.
|
To further characterize effects of GH excess on plasma ACTH levels, we extended our studies to transgenic mice carrying mouse MT promoter bGH fusion genes. Plasma ACTH levels of MTbGH transgenic males did not differ from those of nontransgenic littermate controls (Fig. 6b
). We had investigated plasma ACTH levels previously in PEPCK-bGH transgenic mice maintained on an NMRI outbred background using a different method of blood collection (puncture of the retro-orbital sinus in ether-anesthetized mice). Morning plasma ACTH levels were slightly, not significantly, increased in PEPCK-bGH transgenic mice vs. nontransgenic controls (233±43 vs. 204±39 pg/mL; mean±SE; n=11 per group) (44)
.
Adrenal IGFBP-2, IGF-I, IGF-IR, and GHR mRNA expression
RT real-time PCR analysis of total RNA extracted from adrenal glands of the different genetic groups confirmed marked overexpression of IGFBP-2 in B and GB mice (Table 3
). GH excess in G and GB mice resulted in a 1.3-fold (P<0.05) and 1.7-fold (P<0.01) increase in adrenal IGF-I mRNA abundance compared with C mice. In contrast, no significant differences between the genetic groups were found for the mRNA levels of IGF-IR and GHR (Table 3)
.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
PEPCK-bGH transgenic mice, an established model for studying long-term consequences of elevated GH, exhibited markedly increased adrenal weights, which is in line with previous findings (9)
. The decrease of relative adrenal weights of 4-month-old G mice is due to a disproportionate increase of several other organs, mainly of liver, skin, pancreas, and spleen (30)
.
To further characterize the differences in adrenal weights between the genetic groups, we estimated the volume fractions of adrenal cortex and medulla by stereological methods and calculated their volumes. Cortex and medulla increased in volume, which is consistent with the presence of IGF-IR in both zones (16
, 17)
.
Using state-of-the-art stereological methods, we demonstrated for the first time that adrenocortical enlargement of GH transgenic mice involves both hypertrophy (44% increase in mean cell volume) and hyperplasia (50% increase in number) of zona fasciculata cells. In contrast, the adrenal glands of genetically GH-deficient Snell dwarf mice are several times smaller than those of normal mice (45)
.
Adrenocortical enlargement and increased corticosterone secretion in GH-overexpressing mice are likely to be caused by increased levels of IGF-I for several reasons. 1) In cultured adult adrenocortical cells, supplementation of recombinant human GH in the culture medium did not stimulate basal or ACTH-stimulated cortisol secretion (46)
. In contrast, many studies showed mitogenic and steroidogenic actions of IGFs on fetal and adult adrenocortical cells from different species (14
, 15
, 21
, 47
48
49)
. 2) In vivo, infusion of IGF-I into guinea pigs caused an increase in the fractional weight of the adrenal glands, and this effect was more pronounced with LongR3 
IGF-I, an analog with reduced affinity for IGFBPs (50)
. 3) Overexpression of IGF-II in transgenic mice caused adrenal enlargement and enhanced steroidogenesis (23)
. 4) Both adrenal IGF-I mRNA expression and circulating levels of IGF-I were significantly elevated in PEPCK-bGH transgenic (G and GB) mice. 5) The elevation in plasma corticosterone in different strains of GH transgenic mice correlated largely with serum IGF-I levels rather than with circulating GH (9)
. 6) IGFBP-2, a presumed inhibitor of IGF actions in various human and rodent cell culture models (51
, 52)
and in vivo (29
, 30
, 53
, 54)
, strongly reduced adrenal growth and corticosterone secretion in the context of high GH and IGF-I levels in GB mice (see below).
To evaluate effects of IGFBP-2 overexpression on normal and GH-stimulated adrenal growth and function in vivo, we investigated transgenic mice harboring IGFBP-2 and/or GH expression vectors. IGFBP-2 overexpression in GB mice markedly reduced the effect of GH excess on adrenal weight and volume. We had previously observed that the growth-inhibiting effect of elevated IGFBP-2 is greater in mice with high GH and IGF-I levels than in normal mice. This effect was most marked for spleen, pancreas, and kidney in 5-wk-old and for the carcass and kidney in 4-month-old mice (30)
. However, a reduction in weight of the adrenals by 37% as observed in 4-month-old GB vs. G mice was the greatest effect observed for any organ. The fact that even the relative adrenal weight tended to be smaller in GB than in G mice points to a specific effect of IGFBP-2 on adrenal growth, especially with regard to GH excess.
Surprisingly, IGFBP-2 inhibited only hypertrophic action of GH/IGF-I excess on zona fasciculata cells and hyperplasia was unaffected. Our data show for the first time that even with saturating amounts of GH and IGF-I, IGFBP-2 is able to limit cell growth and size. For understanding how total cell mass of an organ is determined, the mechanisms that stop cell growth are as important as those that stimulate it (2)
. For IGFBP-2, the molecular mechanisms underlying the specific hypertrophy-inhibiting effect remain unclear. If this effect were independent of GH or IGF-I, one would also expect a decrease in cell volume in B vs. C mice, which was not the case. On the other hand, IGFBP-2 could specifically block an increase in cell size rather than decrease normal-sized cells. If the effect of IGFBP-2 were GH/IGF-I dependent, it remains unclear why only the hypertrophic actions of GH/IGF-I excess were blocked. One could speculate that IGFBP-2 targets specific signal transduction cascades downstream of the IGF-I receptor. These pathways may be tissue specific, since in cultured mesangial cells IGF-I has been shown to induce hypertrophy via a calcineurin-dependent pathway (55)
whereas IGF-I-induced skeletal myotube hypertrophy is mediated through calcineurin-independent PI(3)K/Akt/mTOR and PI(3)K/Akt/GSK3 pathways (56)
. Future studies should address potential effects of IGFBP-2 on specific IGF-I signaling pathways in adrenocortical cells by, for example, evaluating the phosphorylation status of key molecules such as PI(3)K/Akt and extracellular signal-regulated kinase (ERK) as well as major regulators of protein translation such as the FRAP/p70s6k pathway.
The number of adrenocortical cells was not affected by IGFBP-2 overexpression in either the absence or presence of GH excess and elevated IGF-I levels. This is in contrast to findings in transfected Y-1 adrenocortical carcinoma cells where IGFBP-2 overexpression stimulated cell proliferation (28)
. Thus, IGFBP-2 is likely to exhibit specific effects in a malignant context. Potential mechanisms have been reviewed recently (27)
.
Analysis of plasma corticosterone levels in the groups of mice investigated in the present study revealed that changes in adrenal weight and volume were paralleled by their capacity to secrete glucocorticoids. Basal and ACTH-induced plasma corticosterone levels were both markedly increased in G mice, and this effect was significantly reduced by coexpression of the CMV-IGFBP-2 transgene in GB mice. Since the number of adrenocortical cells was not different between G and GB mice, corticosterone hypersecretion seems to depend on cell size rather than cell number.
In contrast to a previous report in which threefold increased morning plasma ACTH levels were measured in male mice of one PEPCK-bGH transgenic strain (10)
, plasma ACTH levels were normal or only slightly elevated in several GH transgenic collectives investigated in our study. Therefore, it is unlikely that the trophic effects of GH on the adrenals gland are mediated via increased ACTH levels, although we cannot exclude that GH might increase the responsiveness of the adrenal cortex to ACTH. Such a mechanism has been observed in children with Turner syndrome after long-term treatment with high doses of GH (57)
. The interesting observation that plasma ACTH levels in GH transgenic mice were not reduced despite markedly increased plasma corticosterone concentrations suggests disturbed feedback regulation of ACTH secretion (58)
.
The ability of IGFBP-2 overexpression to antagonize adrenal enlargement and corticosterone hypersecretion in giant GH transgenic mice may be IGF-dependent or IGF-independent. Since IGFs have been shown to be potent stimulators of adrenal growth and steroidogenesis (see above) and IGFBP-2 can block IGF effects in vitro (51)
and in vivo (29
, 30
, 53)
, an IGF-dependent mechanism is likely to underlie the consequences of elevated IGFBP-2 for adrenal size and corticosterone secretion as observed in the present study. This assumption is in line with previous findings that IGF-I analogs with reduced affinity for IGFBPs are more potent stimulators of adrenal growth and steroidogenesis than IGF-I (14
, 50)
. Nevertheless, an influence of elevated IGFBP-2 on direct GH effects cannot be excluded. For instance, IGFBP-2 has been shown to increase GHR mRNA expression and binding of GH to its receptor in rat osteosarcoma cells (59)
.
To further clarify the mechanisms underlying differences in adrenal weight and steroid production observed between the different genetic groups, we performed RT real-time PCR assays to determine levels of mRNA expression for IGF-I, IGF-IR, and GHR. Consistent with previous observations in other tissues (60
, 61)
, IGF-I mRNA levels in the adrenal glands were significantly increased by GH excess, although the level of increase was not dramatic. In contrast, no significant differences were observed in levels of IGF-IR and GHR mRNA abundance.
In summary, our data show that 1) GH excess in transgenic mice results in hypertrophy and hyperplasia of the adrenal cortex; 2) corticosterone hypersecretion in this model is a consequence of adrenocortical hypertrophy and not affected by hyperplasia; and 3) IGFBP-2 is a strong inhibitor of GH/IGF-I-induced adrenocortical hypertrophy without any effect on hyperplasia. This selective inhibition of cell size increase is an important new facet in the spectrum of mechanisms by which IGFBP-2 may affect cell growth and differentiation (reviewed in ref 27
). In view of the clear-cut effects regarding hypertrophy and hyperplasia, the adrenal cortex from the panel of mice generated in this study is an attractive model for systematic analyses of mechanisms involved in organ size control.
|
ACKNOWLEDGMENTS |
|---|
Note added in proof: In mice carrying hypomorphic PDK1 alleles, reduced cell size of adrenal zona fasciculata cells was observed without an alteration of cell number. Lawlor, M. A., Mora, A., Ashby, P. R., Williams, M. R., Murray-Tait, V., Malone, L., Prescott, A. R., Lucocq, J. M., Alessi, D. R. (2002) Essential role of PDK1 in regulating cell size and development in mice. EMBO J. 21, 3728–3738.
Received for publication April 23, 2002. Accepted for publication June 19, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
