| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Endocrinology and Metabolism Unit and Institute of Human Nutrition (A.A.J.), Division of the Fetal Origins of Adult Disease, School of Medicine, Southampton General Hospital, Southampton, United Kingdom SO16 6YD
Address all correspondence and requests for reprints to: Dr. Caroline Bertram, Centre for FOAD, Mailpoint 887, Princess Anne Hospital, Coxford Road, Southampton, United Kingdom SO16 5YD. E-mail: c.bertram{at}soton.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
To determine the molecular mechanisms underlying the role of
alterations in glucocorticoid hormone action in the prenatal
programming of hypertension in MLP offspring, tissues were analyzed for
expression of the glucocorticoid receptor (GR), mineralocorticoid
receptor (MR), 11ßHSD1, 11ßHSD2, and corticosteroid-responsive
Na/K-adenosine triphosphatase 
1 and ß1. GR protein (95 kDa) and
messenger RNA (mRNA) expression in kidney, liver, lung, and brain was
more than 2-fold greater in MLP vs. control offspring
during fetal and neonatal life and was more than 3-fold higher during
subsequent juvenile and adult life (P < 0.01).
This was associated with increased levels of Na/K-adenosine
triphosphatase 1- and ß1-subunit mRNA expression. Levels of MR
gene expression remained unchanged. Exposure to the MLP diet also
resulted in markedly reduced levels of 11ßHSD2 expression in the MLP
placenta on days 14 and 20 of gestation (P <
0.001), underpinning similar effects on 11ßHSD2 enzyme activity that
we reported previously. Levels were also markedly reduced in the kidney
and adrenal of MLP offspring during fetal and postnatal life
(P < 0.001). This programmed decline in 11ßHSD2
probably contributes to marked increases in glucocorticoid hormone
action in these tissues and potentiates both GR- and MR-mediated
induction of raised blood pressure. In contrast, levels of 11ßHSD1
mRNA expression in offspring central and peripheral tissues remained
unchanged.
In conclusion, we have demonstrated that mild protein restriction during pregnancy programs tissue-specific increases in glucocorticoid hormone action that are mediated by persistently elevated expression of GR and decreased expression of 11ßHSD2 during adult life. As glucocorticoids are potent regulators not only of fetal growth but also of blood pressure, our data suggest important potential molecular mechanisms contributing to the prenatal programming of hypertension by maternal undernutrition in the rat.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
These epidemiological data are strongly supported by experimental animal studies. In rodents both severe nutrient restriction to the fetus [through either uterine artery ligation (7) or major calorific restriction (8)] and mild undernutrition arising from a maternal low protein (MLP) diet during part or all of gestation result in offspring with low body weight and/or disproportionate body size at birth that have elevated blood pressure (8, 9, 10) and dysregulation of glucose metabolism in later life (11). Similarly, in the sheep, early to midgestation maternal nutrient restriction program raised blood pressure in the offspring (12).
The precise molecular mechanisms underlying the programming of adult disease by maternal undernutrition are unknown. However, recent experimental studies strongly suggest that fetal overexposure to maternal glucocorticoids (10, 13, 14, 15, 16) triggers programming events in utero that establish persistent increases in glucocorticoid hormone action throughout life (14, 16). Glucocorticoids are potent regulators of fetal growth and development (17). They also promote increased blood pressure by potentiating tissue sensitivity to vasoactive hormones (18), promote gluconeogenesis (19), and antagonize the metabolic actions of insulin (20). Indeed, glucocorticoid excess results in a metabolic syndrome-like phenotype (21) similar to that programmed in human populations (22) and rat models (9, 10, 11, 12, 13, 14). Programmed alterations in glucocorticoid action are likely, therefore, to play a key role linking intrauterine nutrient availability, fetal growth, and CVD risk.
Glucocorticoid hormone action within the cell is regulated by expression of the glucocorticoid receptor (GR) and isoforms of 11ß-hydroxysteroid dehydrogenase (11ßHSD1 and 11ßHSD2) at the level of gene transcription (23, 24)., Most of the classical effects of glucocorticoid are therefore mediated by the GR (23), with levels of glucocorticoid binding largely dictated by levels of GR mRNA expression (23). In addition, 11ßHSD1 behaves predominantly as an 11-oxo-reductase, catalyzing the conversion of cortisone to bioactive cortisol [11-dehydrocorticosterone (A) to corticosterone (B) in the rat], and acts as an intracellular amplifier of glucocorticoid access to the GR (23, 24). Conversely, 11ßHSD2 behaves as an 11-dehydrogenase, catalyzing the inactivation of cortisol to cortisone (B to A in the rat). It maintains the aldosterone specificity of the mineralocorticoid receptor (MR) (24) and may also potentiate glucocorticoid hormone action through the GR (25).
11ßHSD2 is also expressed at high levels in feto-placental tissues,
where it is thought to protect the fetus from overexposure to
glucocorticoid (13, 15, 24, 26, 27). In the rat the potent
effects of the MLP diet on fetal growth and programming of hypertension
and dysregulation of glucose metabolism are thought to be mediated by
inhibition of placental 11ßHSD2 activity (10, 13). In
support of this hypothesis, maternal treatment during pregnancy
with the 11ßHSD inhibitor carbenoxolone mimics the effects of the
MLP diet on offspring birth weight and later blood pressure
(15). In another pharmacological model of programmed
hypertension in which the pregnant dam is treated with
dexamethasone, the offspring exhibit persistently elevated GR
expression in the liver (16) and attenuated GR expression
in the hippocampus (28). The molecular mechanisms
underlying the intrauterine programming of hypertension by modest
variations in the maternal diet during pregnancy (9),
however, are unknown. In the present study we describe, for the first
time, the prenatal programming effects of a maternal isocaloric low
protein-high carbohydrate (MLP) diet on the expression of GR, MR,
11ßHSD1, and 11ßHSD2 and also the expression of
corticosteroid-responsive Na/K-adenosine triphosphatase
(Na/K-ATPase) 1- and ß1-subunits (29) in
feto-placental and postnatal central and peripheral tissues in the
rat.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Animals
All animal experiments were performed in accordance with the
provisions of Home Office Project Licenses PPL30/1522 and PPL30/1523
granted under the Animal Procedures Act 1986.
Virgin female Wistar rats (230–250 g; Harlan UK Ltd., Bicester, UK) were maintained under controlled lighting (lights on at 0700 h and off at 1900 h) and temperature (22 C) conditions and allowed ad libitum access to standard rat chow (56.3% carbohydrate, 18.3% protein, and 0.7% NaCl) and tap water for at least 1 week before the experimental protocol.
Experimental animal protocol
Virgin female Wistar rats were mated using one of four adult
male Wistar rats (Harlan UK Ltd., Bicester, UK), with conception
defined by the presence of a vaginal plug. Thereafter, the pregnant
female rats were housed in pairs and randomly allocated to receive a
control diet comprised of 18% casein or a low protein diet comprised
of 9% casein as previously described (9) throughout
pregnancy (term = 22 days). Pregnant rats were weighed at 4- to
5-day intervals and on day 18 were housed individually. A proportion of
pregnancies (n = 10/diet group) were terminated on days 14 and 20
of gestation by euthanasia of the pregnant dam with barbiturate (100
mg/kg pentobarbital-sodium, i.e. Euthatal, PMB Animal
Health, UK). Placentas and fetuses were excised and weighed. Fetal
tissues were dissected and along with the placentas were snap-frozen in
liquid nitrogen and stored at -80 C for
molecular analyses. At term, offspring were also weighed within 6
h of delivery, sexed, and culled to eight (n = 4 of each sex) per
litter. To further standardize the postnatal environment, the
experimental diet was replaced with standard chow ad
libitum. Culled offspring were dissected for excision of neonatal
tissues, which were also snap-frozen and stored at -80
C.
After blood pressure measurements (described below), offspring were weighed, and two from each litter were randomly killed by CO2 asphyxiation at 4, 8, 12, and 16 weeks of age. In additional litters offspring were randomly killed from each litter at 2, 6, 10, and 20 weeks of age. Organs were excised and stored at -80 C for molecular analyses. This protocol enabled relatively continuous analyses of offspring gene expression from birth to 5 months of age in which variation within each diet group at each time point reflected interlitter variability between pregnancies rather than intralitter variability from the same pregnancy.
Blood pressure measurements
Systolic blood pressure was determined by tail cuff
plethysmography, which we described in detail previously
(9). An IITC model 229 blood pressure recorder linked to a
computer software package was used to determine blood pressure using a
preset algorithm (Linton Instrumentation, Diss, UK). Tail cuffs were
selected according to the size of each rat and were inflated to 300 mm
Hg to occlude the tail arterial pulse. Deflation at approximately 3 mm
Hg/sec allowed accurate determination of systolic pressure as the pulse
returned. Each animal was acclimatized to this procedure before
consecutive measurements (n = 4–5) on each rat over a period of
less than 5 min between 1100–1300 h. The mean value was recorded. All
blood pressure measurements were determined repetitively in the same
animal to assess the reproducibility of recordings. A single operator
blinded to the prenatal experience of the rats was employed in blood
pressure measurements, resulting in intra- and interassay coefficients
of variation of less than 5% and less than 8% respectively. This
method has previously been validated against direct arterial
cannulation measurements taken in conscious, unrestrained, animals from
which a correlation coefficient of 0.974 was established between the
two methods (30).
Molecular analyses
RNA and protein isolation from tissues. Total RNA and
protein were isolated in parallel from the same tissue aliquot using
Tri-Reagent (Sigma-Aldrich Corp., Poole, UK). Total RNA
was also isolated using RNAzol B (Biogenesis, UK)
as previously described (29, 31). Both procedures are
modifications of the single step acidified phenol-chloroform extraction
method. The integrity and quantification of total RNA, pooled from
three separate aliquots of tissue from the same tissue sample, were
assessed by comparison with RNA mol wt markers (Amersham Pharmacia Biotech-Pharmacia, Little Chalfont, UK)
coelectrophoresed in an ethidium bromide-stained agarose gel and also
by UV spectrophotometric absorbance at 260 nM. Protein was
also isolated from the same sample by precipitation with propan-2-ol,
centrifugation (10,000 x g, 20 min, 4 C), 90% ethanol
wash, and resuspension in sterile Milli-Q water comprising 0.1% SDS
and 1 mM phenylmethylsulfonylfluoride and was
stored at -80 C. Protein concentrations were
quantified by the Bradford method using a commercially available kit
(Bio-Rad Laboratories, Inc., Hemel Hempstead, UK)
Northern blot analysis
Northern blot analyses were performed as previously described
(25, 29). Briefly, total RNA (30 µg) was electrophoresed
through an agarose (1.5%)/formaldehyde
(15%)/3-[N-morpholino]propanesulfonic acid (MOPS) gel in
MOPS buffer. RNA was transferred overnight onto Hybond
N+ nylon membrane (Amersham Pharmacia Biotech-Pharmacia) by capillary action facilitated by 20 x
SSC (standard saline citrate) and then UV cross-linked (CL-1000 UV
cross-linker, UVP, Wolf Laboratories, York, UK).
For complementary DNA (cDNA) probe hybridizations (11ßHSD1,
11ßHSD2, and Na/K-ATPase 1- and ß1-subunits), each membrane was
prehybridized at 65 C in hybridization buffer (0.77 M
sodium phosphate, 5 mM EDTA, and 7% SDS, pH 7.2)
containing 100 µg denatured salmon sperm (ss) DNA. Probe was added,
and hybridization was performed in the same buffer for 16 h at 65
C. For complementary RNA (cRNA) probe hybridizations (GR and MR),
membranes were prehybridized at 42 C in a rotary incubator (Hybaid,
Ashford, MIDDX, UK) in hybridization buffer as described
previously (25). Membranes were hybridized at 60 C for
16 h. cDNA- and cRNA-probed membranes were then washed in 2
x SSC/1% SDS for 10 min at room temperature at progressively higher
wash stringencies to a maximum of 0.1 x SSC/0.1% SDS at
42-68 C for 30 min
depending on the probe, followed by washes in 2 x SSC/0.1% SDS
(10 min, room temperature) and 0.2 x SSC/0.1% SDS (20 min, 68 C)
as previously described (33, 37). After phosphorimage
analysis (Storm 850 Phosphor-Imager, Molecular Dynamics, Inc., Sunnyvale, CA), membranes were also subjected to
autoradiography between two intensifying screens at -80
C for up to 10 days. The relative abundance of
specific mRNA species in each tissue sample was quantified from either
the phoshorimaged membrane or the autoradiograph using Phoretix Gel
Analysis Software (NonLinear Dynamics, Newcastle upon Tyne, UK) within
the linear range of the image or autoradiographic film (DuPont-Cronex).
mRNA abundance was expressed as a fraction of the relative abundance of
18S ribosomal RNA (rRNA) to correct for variations in gel loading and
efficiency of RNA transfer.
Probes
cDNAs encoding rat GR (1150-bp fragment in pBluescript, pGR14X)
(32), 11ßHSD1 (1265-bp fragment in pBluescript SK)
(33), 11ßHSD2 (1864-bp fragment in pCR3)
(34), MR (513-bp fragment in pGEM4) (35),
Na/K-ATPase 1 (332-bp fragment in Puc18) (36) and ß1
(271-bp fragment in pIBI30) (37), and 18S ribosomal DNA
(rDNA; 1070-bp fragment subcloned into pBluescript) (38)
were donated by Drs Keith Yamamoto, Perrin White, Elise Gomez-Sanchez,
Ronald Evans, Jerry Lingrel, and Ira Wool, respectively.
cDNA probes for 11ßHSD1, 11ßHSD2, Na/K-ATPase 1 and ß1, and
18S were synthesized and radiolabeled with
[32P]deoxy-CTP (3000 Ci/mmol) by
oligonucleotide random priming of the restriction endonuclease-excised
cDNA fragment (109 cpm/µg DNA) using
commercially available kits (Amersham Pharmacia Biotech)
as previously described (29). An antisense GR cRNA probe
was synthesized by T7 RNA polymerase-directed in vitro
transcription from the 1150-bp cDNA fragment after linearization of the
plasmid with XhoI as described previously (39).
An antisense MR cRNA probe was similarly synthesized by SP6 RNA
polymerase directed in vitro transcription from the 523-bp
cDNA fragment after linearization of the plasmid with
HindIII as previously described (39). The cRNA
probes were radiolabeled by incorporation of
[32P]UTP (3000 Ci/mmol). The in
vitro synthesis of cRNA probes employed transcription reagents and
enzymes from Promega Corp. UK (Southampton, UK) and
yielded probes comprising more than 90% full-length transcripts that
were analyzed by electrophoresis on a denaturing 6% polyacrylamide/7
M urea gel as previously described
(39).
SDS-PAGE and Western blot analysis
Stock protein samples were prepared for SDS-PAGE by the addition
of an equal volume of protein sample buffer [62.5 mM
Tris-HCl (pH 6.8), 4% SDS, 10% glycerol, 10% ß-mercaptoethanol,
and 0.2% bromophenol blue] and boiled for 10 min. For each tissue
equal quantities of total protein (20–100 µg) were loaded and
electrophoresed through an SDS-PAGE gel composed of a 4% stacking gel
(pH 6.8) and 10% separating gel (pH 8.8). Separated proteins were
transferred for 1 h at 0.8 mA/cm2 onto
nitrocellulose membrane (Amersham Pharmacia Biotech-Pharmacia) using a Trans-Blot semidry
blotting apparatus (Bio-Rad Laboratories, Inc.). Membranes
were blocked for nonspecific protein binding by incubation overnight at
4 C in Tris-buffered saline [TBS; 20 mM Tris-HCl
(pH 7.4) and 200 mM NaCl] containing 5% nonfat
milk. For detection of GR, membranes were incubated with polyclonal
rabbit antimouse GR antisera (Autogen Bioclear,
Santa Cruz Biotechnology, Inc., Salisbury, UK) diluted
1:200 with 1% milk/TBS, washed vigorously with TBS (three times, 20
min each time), and then incubated with a 1:2000 dilution of sheep
antirabbit IgG horseradish peroxidase conjugate (Sigma- Aldrich Corp.). After additional washing with TBS (three
times, 20 min each time), GR protein was visualized by autoradiography
using the ECL-Plus chemiluminescent system (Amersham Pharmacia Biotech-Pharmacia). In all analyses a single species,
approximately 97 kDa in size, was identified with negligible background
signal. The relative abundance of GR protein expression in each tissue
sample was quantified from the autoradiograph using Phoretix Gel
Analysis Software (NonLinear Dynamics, Newcastle upon
Tyne, UK) within the linear range of the autoradiographic film
(DuPont-Cronex) and expressed as OD units in relation to those of a
standard. In all protein analyses the standard refers to protein
harvested from pooled tissue samples.
Statistics
All data were found to be normally distributed by
Kolmogorov-Smirnov analysis and are presented as the mean ±
SEM. Birth weight, systolic blood pressure, and tissue
expression of GR, MR, 11ßHSD1, 11ßHSD2, and Na/K-ATPase 1 and
ß1 in each tissue were compared between the MLP and control groups by
Student’s t test, assuming unequal variance. One way ANOVAs
were also employed to determine whether there was a significant effect
of maternal undernutrition on offspring birth weight, blood pressure,
and expression of these genes in a range of offspring tissues from
before birth to well into adult life. Statistical analyses were
performed using SPSS software (SPSS, Inc., Chicago, IL).
P < 0.05 was considered significant. Analyses of
imprecision for Northern and Western blot analyses revealed percent
coefficient of variations within a gel of less than 10% and between
gels of less than 14%.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Effects of the MLP diet on offspring systolic blood pressure
Offspring that had been exposed to maternal protein restriction
in utero exhibited significantly higher systolic blood
pressures that were 25–35 mm Hg above those in control offspring from
when measurements commenced at 4 weeks of age (9). Thus,
systolic blood pressure (mm Hg, mean ± SEM)
for MLP vs. control offspring was 135 ± 5
vs. 111 ± 3 (P = <0.001) at 4 weeks,
140 ± 4 vs. 116 ± 3 (P =
<0.001) at 8 weeks, and 142 ± 4 vs. 115 ± 5
(P = <0.001) at 12 weeks.
The MLP diet resulted in increased blood pressure in the offspring that persisted in both males (n = 96) and females (n = 96) throughout adult life. Thus, this pattern was similarly evident at other age points in older offspring up to 20 weeks of age, when measurements halted. This confirms previous observations of life-long programmed hypertension in this rat model (9). In further accordance with previous studies (9), comparative studies of offspring up to 20 weeks of age revealed no effect of gender (two-way ANOVA) on the blood pressure programming effects of the MLP diet. However, similar analyses for the effect of postnatal age revealed an age-related increase in the blood pressure increment between the two groups (P < 0.05), i.e. the difference between blood pressures in the MLP vs. control offspring increased with age (P < 0.05).
Effects of the MLP diet on the expression of 11ßHSD isoforms in
the placenta and fetal and neonatal offspring tissues
On day 20 of gestation (term = 22 days), levels of 11ßHSD2
mRNA were markedly lower in the placentas from rats fed the MLP diet
throughout pregnancy compared with those in the control-fed animals
(Fig. 1
, A and B). The inhibitory effect
of the MLP diet on placental 11ßHSD2 expression was also evident on
day 14 of gestation, although the difference was slightly less marked
(Fig. 1, A and B). There was no effect of the MLP diet on placental
expression of 11ßHSD1 on day 20, and levels were undetectable on day
14 (data not shown). Reduced levels of 11ßHSD2, but unchanged levels
of 11ßHSD1, expression were also evident in late gestation (day 20)
fetal tissues from MLP offspring, in which expression of both
isoforms was detectable, e.g. in the lung (Fig. 2, A and B).
|
|
and 6). In keeping
with observations in the placenta, the MLP diet resulted in attenuated
levels of 11ßHSD2 expression in the offspring kidney from birth (Fig. 3). As described in detail below, this effect persisted through early
juvenile life and into adulthood (Fig. 6). In contrast, levels of
11ßHSD1 expression in neonatal offspring kidney were unaffected by
the MLP diet (Fig. 3).
|
|
and 3). The stimulatory effects of maternal
protein restriction during pregnancy on fetal and neonatal offspring
tissue expression of GR were also evident at the protein level. Figure 4 shows a 2-fold greater abundance of GR
protein expression in the kidney and lung (representative of classical
glucocorticoid and mineralocorticoid target tissues) from MLP
vs. control offspring at birth. Levels of MR expression were
unaffected by the maternal diet in the kidney of MLP offspring and were
undetectable in the lung.
|
and 6).
In keeping with the stimulatory effects of the MLP diet on renal GR
expression shortly after birth (Figs. 3 and 4), the abundance of GR
protein expression was also approximately 2-fold greater in MLP
vs. control offspring from preweaning (i.e. 2
weeks of age) through juvenile life and into adulthood (Fig. 5). Levels
of GR mRNA expression were also persistently 2-fold higher in the
kidney from MLP vs. control offspring. Northern blot
analyses of the effects of the MLP diet on renal GR expression in
12-week-old offspring, which are representative of similar patterns of
expression in the other age groups, are shown in Fig. 6.
|
), it resulted in a marked decline in 11ßHSD2
expression similar to that observed in the kidney at birth and in other
feto-placental tissues. The inhibitory effect of the MLP diet on renal
11ßHSD2 expression was also evident in offspring at other ages (data
not shown). These prenatally programmed increases in GR expression and
decreases in 11ßHSD2 expression in the kidney from MLP offspring were
accompanied by a concomitant 2- to 3-fold increase in the
expression of corticosteroid-responsive Na/K-ATPase 1- and
ß1- subunits (Fig. 6).
Effects of the MLP diet on the abundance of GR and 11ßHSD
isoforms and corticosteroid-responsive Na/ K-ATPase gene expression
in other offspring tissues during postnatal life
The effects of the MLP diet on levels of GR expression in the
offspring were tissue specific. The abundance of GR expression was
markedly greater in classical glucocorticoid target tissues (but not in
the heart or brain) from MLP vs. control offspring
throughout adult life. Levels of GR protein were approximately 2-fold
greater in the liver and 3-fold greater in the lung (Fig. 7) from MLP offspring compared with
controls at 2 weeks of age. The potent programming effect of the MLP
diet on GR expression in these tissues persisted from early life into
adulthood, as demonstrated by quantitatively similar elevated levels of
GR protein expression in both liver and lung from MLP vs.
control offspring at 12 weeks of age (Fig. 7).
|
). This pattern of
programmed gene expression was also evident in 4-, 8-, 16-, and
20-week-old offspring (data not shown).
|
1- and ß1-subunit mRNA
expression (Fig. 8). 11ßHSD2 expression was undetectable in liver and
lung. Levels of expression of the GR and corticosteroid-responsive
Na/K-ATPase 1- and ß1-subunits and
phenylethanolamine-N-methyl transferase (metabolizes
noradrenaline to adrenaline) were also persistently elevated in the
adrenal gland from MLP offspring compared with controls (Copin, N., and
C. B. Whorwood, unpublished observations).
In contrast with the MLP programmed up-regulation of GR expression in
classical glucocorticoid target tissues and the absence of any effect
in the heart, levels of GR mRNA and protein expression were markedly
decreased in the hypothalamus from MLP vs. control offspring
(Fig. 9). This was accompanied by a
concomitant 50% decline in corticosteroid-responsive Na/K-ATPase
1-subunit mRNA expression (Fig. 9). As in other tissues, levels of
11ßHSD1 expression in offspring hypothalami were unaffected by the
MLP diet during pregnancy. Expression of 11ßHSD2 and Na/K-ATPase
ß1-subunit was not detectable in this tissue.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
1-
and ß1-subunits in key peripheral target tissues, such as the kidney,
liver, and lung was observed. Furthermore, there were decreased levels
of GR and Na/K-ATPase gene expression in the hypothalamus, a central
tissue that mediates negative feedback regulation of
hypothalamic-pituitary-adrenal axis activity. An additional finding
from this study is that this relatively subtle isocaloric maternal
dietary manipulation also programs markedly attenuated levels of
11ßHSD2 expression in the kidney. Indeed, the prenatal programming
effect of the MLP diet appears to be not only tissue specific, but also
confined to the regulation of corticosteroid hormone action governed by
the GR and 11ßHSD2 rather than that dictated by the MR and
11ßHSD1. The MLP diet programming effects on GR, 11ßHSD2, and Na/K-ATPase gene expression persist from before birth, through juvenile life, and into adulthood. As such, they parallel the effects of the MLP diet on birth weight and the subsequent manifestation of hypertension seen in the MLP offspring generated in this study and accord with earlier observations of the effects of maternal undernutrition on offspring birth weight and later hypertension described previously (8, 9). The significance of our data is that they reveal potential molecular mechanisms underlying the key role of glucocorticoid hormone action in the link between maternal nutrition, fetal growth retardation, and programming of hypertension that has recently been identified in animal models (10, 14, 40) and human populations (22).
We have previously shown that the programming effects of the MLP diet are dependent on maternal circulating glucocorticoids (14), which are likely to pass into the fetal circulation at an inappropriately high level as a result of the inhibitory effects of maternal protein restriction on placental 11ßHSD2 activity (10). In normal physiology, this enzyme metabolizes active cortisol (or corticosterone in the rat) to inactive cortisone (11-dehydrocorticosterone in the rat) (24) and, as such, serves to protect the fetus from the deleterious effects of excess exposure to the higher levels of maternal glucocorticoid (13, 14, 26, 27). Indeed, two groups describing parallel increases in levels of 11ßHSD2 mRNA expression and 11-dehydrogenase activity in fetal sheep kidneys throughout the latter half of gestation to term (31, 41) suggest that placental 11ßHSD2 may also protect the renal MR from occupancy by the prepartum surge in fetal glucocorticoid levels that is necessary for parturition in this species. Data from the present study suggest that the inhibitory effect of the MLP diet on placental 11ßHSD2 activity is mediated by attenuating the levels of 11ßHSD2 gene transcription.
This nutrient/11ßHSD2 gene interaction was evident not only in the
placenta, but also in other fetal tissues in which this key enzyme is
expressed. The previously described close correlation between levels of
11ßHSD2 mRNA and activity (24, 31) suggests that the
maternal diet effects on fetal tissue 11ßHSD2 gene expression are
likely to result in attenuated fetal tissue inactivation of
glucocorticoid and hence greater levels of glucocorticoid hormone
action in these tissues. As there is tissue-specific ontogeny of both
GR and MR expression during late gestation in the rodent
(42), a decline in feto-placental 11ßHSD2 is likely to
potentiate increased glucocorticoid binding to both receptors in those
tissues in which they are coexpressed (e.g. kidney).
Increased levels of corticosteroid-responsive Na/K-ATPase 1- and
ß1-subunit (29) expression in fetal and neonatal tissues
from MLP offspring are consistent with this hypothesis.
The physiological importance of feto-placental 11ßHSD2 expression, with respect to the role of glucocorticoid hormone action in mediating the programming effects of the MLP diet (10, 14), is highlighted by studies employing a pharmacological inhibitor of 11ßHSD2, i.e. carbenoxolone (15). Maternal treatment with carbenoxolone results in increased fetal exposure to glucocorticoid-reduced fetal growth and programmed hypertension and insulin resistance in the offspring during later life (15). The adverse prenatal programming effects of maternal malnutrition during pregnancy on the blood pressure and insulin sensitivity of the offspring (8, 9, 10) can be mimicked in the rat and the sheep by maternal treatment with a synthetic glucocorticoid (i.e. dexamethasone) that is poorly metabolized by 11ßHSD2 and, as such, is able to bypass the placental barrier (16, 28, 43).
In addition to the inhibitory effects of the MLP diet on 11ßHSD2 gene expression, the present study suggests that maternal dietary composition during pregnancy may also regulate fetal and neonatal tissue expression of the GR. Indeed, the stimulatory effects of the MLP diet on GR expression in classical glucocorticoid and mineralocorticoid target tissues (i.e. kidney and lung) are likely to have further enhanced the levels of glucocorticoid hormone action in these tissues beyond those that would be predicted to have resulted solely from the attenuated levels of 11ßHSD2 expression. We report a clear interaction between maternal nutrient availability to the fetus and programmed expression of genes encoding key determinants of glucocorticoid hormone action in feto-placental tissue. The mechanisms underlying this interaction and the permanent establishment of these responses in the kidney and other tissues throughout adult life require further investigation. However, similar findings in an ovine model of maternal undernutrition programming of hypertension (40) suggest that this represents an important fundamental mechanism by which increased levels of glucocorticoid hormone action trigger programming events in utero that lead to disease in later life.
We have shown that the maternal dietary effects on the regulation of GR and 11ßHSD2 gene expression during fetal and neonatal life persist throughout adult life, i.e. levels of expression of these genes can be programmed in a tissue-specific manner by the nature of nutrient availability to the offspring before birth. Similarly, the permanent alterations in Na/K-ATPase subunit gene expression, which parallel those for the GR, strongly suggest that programmed increases in renal, hepatic, and pulmonary GR expression, as reported in this study, underpin concomitant increases in levels of glucocorticoid hormone action in these tissues. Conversely, programmed decreases in hypothalamic GR expression are likely to bring about persistently decreased sensitivity of this tissue to glucocorticoid regulation of trophic hormone release.
In support of our findings in the MLP offspring, treatment of pregnant rats with dexamethasone programs increased expression of GR and glucocorticoid-responsive phosphoenolpyruvate carboxykinase (the rate-limiting step in gluconeogenesis) in the liver (16) and also decreased expression of the GR in the hippocampal nuclei that mediate the central control of hypothalamic-pituitary-adrenal axis activity (28). The present study extends observations in this pharmacological model of programming to reveal that altered levels of GR expression and glucocorticoid hormone action are prenatally programmed by maternal nutritional status during pregnancy. This effect is evident in these and other important glucocorticoid target tissues in both the torso and central nervous system, including lung, adrenal medulla, and hypothalamus. It is unclear why altered levels of GR expression in the heart from these offspring were not similarly programmed by maternal protein restriction during pregnancy. Nevertheless, programmed increases in levels of GR expression in key peripheral glucocorticoid target tissues, in concert with diminished GR expression in central tissues are consistent with circulating ACTH and glucocorticoid hormone data, which suggest that the MLP diet programs hyperactivity of the hypothalamic-pituitary-adrenal axis.
The absence of a programming effect of the MLP diet on 11ßHSD1 gene expression in either central or peripheral tissues from the offspring is in keeping with similar findings in both pharmacological rat models of programmed hypertension (15, 16) and also our recent observations in an ovine maternal undernutrition model of programmed hypertension (44). Given the previously established close correlation between levels of 11ßHSD1 mRNA and 11-oxoreductase activity (24, 41, 45), this suggests that 11ßHSD1 does not contribute to programming of altered levels of glucocorticoid hormone action in these tissues. The lack of a programming effect of undernutrition or synthetic glucocorticoids on the expression of 11ßHSD1 in tissues with programmed elevation of GR expression is surprising given that 11ßHSD1 expression is up-regulated by glucocorticoid (24, 45, 46) and expression of other glucocorticoid target genes is increased (16, 44). However, as regulation of 11ßHSD1 by glucocorticoids occurs in a complex tissue- and temporal-specific manner (46) and is also initiated by other factors (24), these variables may contribute to maintaining unchanged levels of 11ßHSD1 expression despite the programming of increased levels of GR-mediated glucocorticoid hormone action in these tissues.
In contrast with the widespread tissue distribution of 11ßHSD1, abundant levels of 11ßHSD2 expression are confined to mineralocorticoid target tissues such as the kidney (24). This is in keeping with the well established role of 11ßHSD2 as a key protector of the renal MR from access by glucocorticoid, such that its specificity for aldosterone is maintained (24). The programmed decline in renal 11ßHSD2 expression in the MLP offspring from before birth and into adulthood is in keeping with observations in sheep, in which fetal exposure to nutrient restriction or chronic intermittent hypoxemia down-regulates renal 11ßHSD2 expression in the fetal lamb (44, 47), an effect that has been shown to persist during later life despite restoration of nutrient availability (44). A prenatally programmed decline in renal 11ßHSD2 expression may potentially enhance glucocorticoid availability to both the MR and the GR. Not only will this result in increased levels of glucocorticoid hormone action, but it is also likely to induce a state of apparent mineralocorticoid excess similar to that in patients with congenital or acquired deficiency of 11ßHSD2 activity (24).
The well documented hypertensive effects of congenital or acquired 11ßHSD2 deficiency in humans (24) suggest that reduced levels of 11ßHSD2 expression in MLP offspring and those described in the ovine maternal undernutrition model (44) may contribute to the manifestation of hypertension seen in the offspring from both species (9, 12). In support of this, it has been shown that in the kidney and adrenal gland that pharmacologically induced and nutritionally programmed decreases in 11ßHSD2 expression promote increased expression of glucocorticoid-responsive genes in the adrenal and both glucocorticoid- and mineralocorticoid-responsive genes in the kidney. These are known to have potent effects on sodium uptake, fluid-electrolyte homeostasis, and vascular tone (29, 44, 48). Programming of attenuated 11ßHSD2 expression may, therefore, represent a fundamental mechanism linking intrauterine life with the manifestation of hypertension in later life.
In addition to being a measure of the levels of corticosteroid hormone
action in the kidney and other tissues, Na/K-ATPase 1- and
ß1-subunit mRNA expression encodes an important transmembrane sodium
pump, which is the principal mechanism for ATP-dependent maintenance of
sodium and potassium electrochemical gradients across the cell membrane
(49). Na/K-ATPase is the main driving force for net sodium
reabsorption across renal tubular and distal colonic epithelia
(49), and in other tissues, such as skeletal and cardiac
muscle, it provides the electrolyte gradient required for mechanical
function (49). Corticosteroid-induced up- regulation
of Na/K-ATPase represents an important mechanism by which increased
levels of glucocorticoid hormone action promote sodium retention. This,
in turn, promotes fluid retention, resulting in raised blood pressure
as a consequence of maintaining fluid-electrolyte homeostasis
(49). The prenatal programming of increased levels of
Na/K- ATPase gene expression in the kidney from MLP offspring may,
therefore, contribute significantly to mechanisms mediating the
glucocorticoid-dependent programming of hypertension that has been
described in these animals (14).
The potent effects of glucocorticoids with respect not only to fetal growth and development (17), but also to induction of hypertension, insulin resistance, and glucose intolerance are well documented (18, 19, 20). Our findings, therefore, indicate a plausible molecular basis for the programming and manifestation of hypertension in the MLP rat model in which altered levels of glucocorticoid hormone action in the kidney appear to play a key role. These data also suggest molecular mechanisms by which the MLP diet programs the activity of the hypothalamic-pituitary-adrenal axis, impaired glucose tolerance, and insulin resistance in the offspring during adult life (11). This would occur by potentiation of both increased levels of glucocorticoid hormone action in peripheral tissues and decreased glucocorticoid hormone action in central tissues regulating negative feedback control of the adrenal.
In the rat MLP model and the ovine maternal undernutrition model of programmed hypertension (9, 12), the mild manipulations of dietary composition and calorie/protein restriction and of timing and duration through pregnancy (9, 12, 40, 44) represent close approximations to the suboptimal levels of nutrition during human pregnancy that similarly alter patterns of fetal growth and program raised blood pressure and increased CVD risk in populations worldwide (1, 2, 3, 50). Furthermore, these analyses were performed at postnatal ages similar to those employed in other animal model investigations (8, 9, 10, 11, 12, 13, 14, 15, 16, 17) and are broadly equivalent to those reported in more recent human epidemiological studies (1, 3). It remains to be seen whether the programming mechanisms identified in these animal models also contribute to the mechanisms linking events during intrauterine life and the programming of hypertension, insulin resistance, and other adverse risk factors for CVD in human populations, in which there is now compelling evidence that alterations of glucocorticoid hormone action play an important role.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Received October 26, 2000.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-subunit from rat
brain. Biochemistry 25:8125–8132[CrossRef][Medline]
This article has been cited by other articles:
 |
|
 |
|
  A. Dagan, H. M. Kwon, V. Dwarakanath, and M. Baum Effect of renal denervation on prenatal programming of hypertension and renal tubular transporter abundance Am J Physiol Renal Physiol, July 1, 2008; 295(1): F29 - F34. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 |
