Clinical Science (1996) 91,607-615 (Printed in Great Britain)
607
Weanling rats exposed to maternal low-protein diets during
discrete periods of gestation exhibit differing severity of
hypertension
Simon C. LANGLEY-EVANS, Simon J. M. WELHAM, Rachel C. SHERMAN and Alan A. JACKSON
Department of Human Nutrition, University of Southampton, Southampton, U.K.
(Received 16 May/l Juiy 1996; accepted 19 Juiy 1996)
1. In the rat, hypertension is induced by fetal
exposure to maternal low-protein diets. The effect
on blood pressure of undernutrition before conception and during discrete periods in early, mid or
late pregnancy was assessed using an 18% casein
(control) diet and a 9% casein diet to apply mild
protein restriction.
2. The offspring of rats fed % casein developed
raised blood pressure by weaning age. Feeding a
low-protein diet before conception was not a
prerequisite for programming of hypertension.
3. Hypertension was observed in rats exposed to low
protein during the following gestational periods:
days 0-7, days 8-14 and days 15-22. Blood
pressure increases elicited by these discrete periods
of undernutrition were lower than those induced by
feeding a low-protein diet throughout pregnancy.
The effect in early gestation was significant only in
male animals. Post-natal growth of male rats
exposed to low-protein diets was accelerated, but
kidneys were small in relation to body weight.
4. Biochemical indices of glucocorticoid action in
liver, hippocampus, hypothalamus and lung were
elevated in rats exposed to low-protein diets in
utero. The apparent hypersensitivity to glucocorticoids was primarily associated with undernutrition
in mid to late gestation.
5. Plasma renin activity was elevated in rats
exposed to Wo casein over days 15-22 of gestation.
Animals undernourished over days 0-7 and 8-14
produced pups with lower plasma angiotensin I1
concentrations at weaning.
6. Fetal exposure to maternal low-protein diets for
any period in gestation may programme hypertension in the rat. Alterations to renal structure,
renal hormone action or the hypothalamic-pituitary-adrenal axis may all play a role in the
programming phenomenon, either independently or
in concert.
~~
worQ: fetal pmgramming, glucocotticd, +don,
~~
INTRODUCTION
Hypertension and cardiovascular disease are in
part determined by intrauterine factors. Initial work
from the U.K. indicating relationships between
intrauterine growth restriction and later disease
[l-61 has received support from many parts of the
world [7, 81 and the ‘programming’ hypothesis
appears to be secure and reproducible. It is
suggested that, in addition to genetic factors and
maternal constraint, fetal growth is largely determined by maternal nutrition [9-121. Less than
optimum maternal nutrition may have two separate
effects upon the fetus, the first being the readily
measurable effect on growth and the second being a
more general resetting of metabolism and physiology that may manifest as disease in the offspring
many years later. Recent data have indicated that
primary determinants of impaired fetal and
placental growth may be high intakes of carbohydrate in early pregnancy and low intakes of
animal protein [13, 141.
Observations of the programming phenomenon in
the human population have been supported by a rat
model [15]. The feeding of low-protein diets during
rat pregnancy produces offspring that develop
higher blood pressures early in life [16] and remain
hypertensive into old age [17]. Exposure to low-protein diets is associated with disproportionate
patterns of fetal growth retardation [18], which in
the human population predict later cardiovascular
pathology [ll]. Investigations with the rat model
suggest possible roles for the renin-angiotensin
system [17] and glucocorticoids of maternal origin
[19-211 in the initiation of hypertension in utero.
The current paper presents studies that seek to
define whether hypertension may be initiated during
specific periods of rat pregnancy and to determine
whether these particular molecular mechanisms
~~~
~
~~
kidney, m
a
m
a
lnutrition, rats.
Abbreviations ACE, angi~tendn-cmrti& enzyme;ANG II, angiotensin II;GS, glutamine s y n w GPDH, g3.cwd-3-phmphate dehydrogenase; II BHSD. IIfl-hydroxystemid
&hydrogenare; PC phosphatidykhline; PRA, plasma twin activity.
Correspondence:Dr S. C. Langley-Evans, Deparenent of Human Nutrition, Uniwrdty of Southampton. Bassett Crescent East, Southampton SO1 6PX, U.K.
608
S.
C. Langley-Evans et al.
show evidence of programmed changes during the
same critical periods.
Plasma, lungs, livers, kidneys and hearts were
obtained from all animals and frozen at -80°C.
Hippocampus and hypothalamus were dissected
from male animals and frozen at -80°C.
METHODS
Chemicals
All chemicals and reagents used in the studies
were purchased from the Sigma (Poole, U.K.).
Animals
Animal experiments were performed under
licence from the British Home Office. All animals
used in the studies were bred in the Southampton
University Animal Unit and were housed individually, or in pairs, at a temperature of 24°C with a
12 h light cycle. A total of 37 adult female rats were
used to generate the 311 offspring used in the study.
Experimental protocols
Experiment 1. Previous studies with the low-protein diet model of hypertension have involved the
prehabituation of rat dams to the diets before
mating [15, 161. This study was directed to the determination of whether the period before conception
has a role in the programming of hypertension. Four
dams were fed an 18% casein (control) diet and
four dams a 9% casein (low-protein) diet for 14 days
before mating and throughout pregnancy, as previously described [15]. For the composition of the
diets used see Langley and Jackson [15]. A further
two groups of four dams were fed 18% or 9% casein
diets beginning on the day of conception (the day a
plug was detected on the floor of the mating cage).
On giving birth, litters were culled to a maximum of
eight pups and all dams were fed a standard laboratory chow diet; thus all offspring differed only in
prenatal dietary experience. At 7 weeks of age the
blood pressures of a random selection of these
offspring were determined.
Experiment 2. This study was directed towards the
determination of a critical period in pregnancy,
during which the fetus may be susceptible to
programming effects of the maternal diet. Four rats
were fed the 18% casein diet from conception until
full-term pregnancy (day 21-22). Four rats were fed
9% casein throughout pregnancy, while a further 13
rats were fed the low-protein diet for single-week
periods during their pregnancies (days 0-7, days
8-14 or days 15-22). During the periods of
pregnancy when the low-protein diet was not fed,
the rats were provided with the 18% casein control
diet. At birth, all litters were culled to eight pups
and transferred to a standard chow diet, as
described above.
The blood pressures of all the offspring were
determined at weaning (4 weeks). All male rats and
half the females were then killed by decapitation.
Determination of blood pressure
Systolic blood pressure was determined by tailcuff plethysmography [17]. All animals were acclimatized to a room temperature of 27°C for 2 h
before testing. Each rat was placed in a restraint
tube and allowed to rest for 3-5 min before inflation
of the tail cuff to 300 mmHg. The cuff was deflated
at 3 mmHg/s and the reappearance of the tail pulse
recorded. Three or four tests were performed over
5 min and the mean systolic pressure recorded. The
instrument used was the IITC model 29 blood
pressure monitor (Linton Instruments, Diss, U.K.),
which, unlike other tail-cuff systems, does not
require the animals to be heat shocked. As animals
used in the studies were of differing weights, care
was taken to use cuffs of appropriate size for each
animal to avoid a weight effect on the recorded
pressure. Other artifacts associated with tail-cuff
determinations were avoided by calculating
pressures using a preset algorithm, monitoring all
test runs using an IITC software system and through
the operator being unaware of the experimental
group to which each animal was assigned. Clearly,
with tail-cuff plethysmography influences of stress
are an important consideration, although it has been
reported by Bunag [22] and others [23] that tail-cuff
measurements performed with conscious rats are
appropriate and correlate closely with direct
measures from carotid artery. To reduce stress
effects, animals were conditioned to handling. In
validation experiments we found that, after a peak
in blood pressure immediately after placing in
restraint, systolic blood pressures were constant over
the period 2-7min (zero time, 103f6mmHg;
1min, 98 5 mmHg; 2 min, 93 10 mmHg; 4 min,
95 f10 mmHg; 5 min, 93 f7 mmHg; 7 min, 94 f6
mmHg; lOmin, 107f8mmHg). Thus, all our
measurements were routinely taken 3-7 min after
initial handling. Animals with a heart rate in excess
of 480 beats/min were excluded from the study as
this pulse rate is indicative of stress. In our hands
day-to-day variation and within-day variation of
blood pressure is approximately 5%.
Enzyme assays
Glucocorticoid-inducible enzyme activities were
determined in liver and brain regions of male
offspring as markers of glucocorticoid action.
Tyrosine aminotransferase in liver was determined
using the method of Shargill et al. [24]. Glutamine
synthetase (GS) and glycerol-3-phosphate dehydrogenase (GPDH) in hippocampus and hypothalamus
609
Intrauterineprogrammingof hypertension
were determined as previously described [25].
Pyruvate kinase activity was determined, as a glucocorticoid-insensitive control enzyme, in both hippocampus and hypothalamus, using the method of
Langley and York [25].
Determination of renin activity and angiotensin II
All plasma was treated with protease inhibitors
(0.5 g of o-phenanthroline, 4.64 g of disodium
EDTA, 0.2 g of neomycin sulphate in 100 ml water)
when blood was collected at the time of killing
(0.5 ml inhibitor/lO ml blood). Plasma renin activity
(PRA) and angiotensin I1 (ANG 11) concentrations
were determined as described previously [17].
Determination of lung phosphatidylcholine
Total lung phosphatidylcholine (PC) concentrations were determined as a marker of glucocorticoid
action within the tissue. Up until the age of
weaning, glucocorticoids may induce synthesis of
certain species of this phospholipid class, constituting approximately 40% of the total PC content of
the tissue [26].
Lungs were homogenized in 10 volumes of
30 mmolll potassium chloride, 10 mmol/l dipotassium hydrogen phosphate, pH 7.4. Homogenates
were sonicated at 20 mHz for three periods of 15 s
with 10s intervals. A 1OOpl volume of 10% (v/v)
Triton X-100 in buffer was added and the homogenate centrifuged at 1OOOg for 10 min. The supernatant was stored at -80°C until analysis. Samples
were diluted with 50 mmol/l Tris-HC1, pH 8.0,
10 mmol/l calcium chloride. A 10 pl aliquot of
sample was added to a microwell in an ELISA plate,
10 pl of phospholipase D (0.1 units/ml) was added to
the well and the plate was incubated at 42°C for 1h.
A 100pl volume of chromogen solution was added
and the plate was incubated for a further hour at
37°C. Chromogen solution contained 20 mg of
phenol, 12 mg of 4-aminoantipyreneY200 units of
choline oxidase and 158 units of horseradish peroxidase in 100ml of Tris-HC1 pH8.0, 10mmol/l
calcium chloride. Absorbance was read at 492nm.
PC in samples was determined from standard curves
of 0-1 pmol/ml choline chloride and 0-1 pmollml
PC.
Statistical analysis
All data are presented as means+SEM. Where
appropriate, a two-way analysis of variance was
performed, using a Tukey test as a secondary
analysis where statistical interactions were indicated.
Otherwise, data were analysed using Student’s t-test.
A probability of 5% or less was accepted as statistically significant.
RESULTS
Experiment I
Food intakes and weight gains of rats fed control
or low-protein diets and the outcome of pregnancy
in terms of litter sizes were similar (data not shown).
The blood pressures of male and female rats
exposed in utero to a 9% casein diet were significantly higher than those of offspring exposed to the
control diet (Table 1). Habituation of the dams to
the different diets before mating did not influence
the blood pressures of the offspring, but in the
litters of dams habituated to the low-protein diet the
increase in blood pressure elicited by the protocol
tended to be less than in litters of non-habituated
dams (habituated male, 15 mmHg; habituated
female, 10 mmHg; non-habituated male, 35 mmHg;
Table I. Systolic blood pressuresof 7-week-old rats exposed to dlierent maternal dietary repimens. Rat dams were fed I696
or 9% casein diets either for 14 days before mating and during pregnancy or during pregnancy only. The systolic blood pressures of
their offspring were determined at 7weeks of age. Data are shown as meansfSEM of n observations. Two-way analysis of Mnance
indicatedthat Mood pressure was influenced by maternal diet (P <0.OOOI, F = 17.72, I .68 degrees of freedom), but not by prefeeding
or sex. *Significantly different from offspring of the 18% casein-expoted, habituatedgroup of the same w. (P ~0.05).
Maternal
diet
18% casein
Habituation
time (days)
0
14
996 &n
0
14
Systolic blood pressure
sex
n
(mmMP)
M
F
8
6
116f7
130k4
M
8
132+3
F
8
132f5
M
F
7
II
151 f 12*
M
F
12
9
147f4*
142f6*
148+6*
S. C.Langley-Evans et at.
610
non-habituated female, 18 mmHg; P <0.05 for
non-habituated males versus habituated males).
Experiment 2
The feeding of a low-protein diet throughout
pregnancy did not significantly impair weight gain or
alter food intake relative to control animals (Table
2). Feeding of a low-protein diet for either the first
or second week of pregnancy only did tend to
produce a greater weight gain in pregnancy (not
significant). All dams produced litters of similar size,
and most of the recorded birth weights were
unchanged by the maternal dietary regimen. The
exception to this was the group exposed to 9%
casein only in the period days 8-14, which were of
significantly greater birth weight than control
animals (P<0.05).
At the age of weaning, male pups exposed to
low-protein diets throughout gestation or during
days 0-7 or 8-14 were significantly heavier than
control pups (Table 3). Animals in the group
exposed in the first week of pregnancy were particularly affected by maternal dietary restriction and
were heavier than animals in all other groups. Liver
and heart weights were similarly increased in the
heavier groups of animals, whereas kidney weight
was elevated, relative to control animals, in the rats
exposed to low-protein diets throughout gestation or
in the first week only. Determination of organ-body
weight ratios indicated that the growth of the kidney
had been disproportionately affected by maternal
protein restriction, whereas other organs were of a
size proportional to body weight in all groups.
Kidneys were small in relation to body weight in all
groups of rats that had experienced a period of
maternal protein restriction. This effect was greatest
in the rats exposed on days 0-7 and diminished with
ongoing pregnancy.
In females (Table 4), no significant effects of
maternal diet upon weight at weaning were
observed. In keeping with this, most organ weights
were also unchanged by the dietary manipulations
but, as with the males, kidney weight was influenced
by exposure to low-protein diets. Pups exposed to
Table 2. Maternal weight gain, food intake and reproductive outcome after protein restriction during discrete periods of
gestation. Rats were fed an 18% casein diet (control) from conception until birth or a 9% casein diet over the following periods of
pregnancy: days 0-22, 0-7, 8- 14 or 15-22. Body weight gain during pregnancy and the birth weights of all pups were determined.
*Statistically different from the 18% casein group (P<O.O5). ND, not determined.
Maternal weight
gain 0
Maternal food
intake (glday)
Litter size
(g)
n
Birth
weight
n
Days 0-22
I16f4
26.0 f I.O
11.3fl.O
4
4.99 f 0.08
45
Days 0-7
Days8-14
Days 15-22
Days 0-22
131 f 9
120f 12
109, I I
103f5
ND
ND
ND
I l.8f 1.0
I l.Ok 1.8
5
5
3
4
5. I6 k0.06
5.27+0.08*
5.1 I f0.15
5. I7 50.09
59
55
34
49
Maternal
diet
Timing
18% casein
9% casein
29.0 f I .O
11.3kO.9
l2.3f 1.0
Table 3. Body weight, organ weights and systolic blood pressures of male ratr exposed t o maternal low-protein diets
during d x r e t e periods of gestation. Rats exposed to different maternal dietary regimens in utero had systolic blwd pressure
determined at 4weeks and were then killed for analysis of organ weights. Data are means+SEM of n observations. *Statistically
significant difference relative to the 18% day 0-22 group (P<O.O5). tstatistically Significant difference relative to the 9% casein day
0-7 group (P <0.05). SEP, systolic blood pressure.
Maternal dietary regimen
18% casein
Body weight (g)
SEP (mmHg)
Liver weight (g)
Lung weight (g)
Kidney weight (g)
Heart weight (9)
Liver-body weight (%)
Lung-body weight (%)
Kidney-body weight (%)
Heart-body weight (%)
n
9% casein
Days 0-22
Days 0-7
Days 8-14
Days 15-22
Days 0-22
70f3
96+3
3. I5 f 0.12
0.72 f0.07
0.38 f0.02
0.40f0.0I
4.29 f0.07
0.97 f0.08
0.51 fO.01
0.55 f 0.02
16
97+5*
II1+4*
4.28 f 0.19:
0.89 f0.05
0.43 f 0.01 *
0.53 f0.02*
4.43 f0.07
0.93 k0.04
0.45 f 0.01 *
0.55 kO.01
20
88+5*t
105f3*
3.92 f0.22*t
0.75 f0.04
0.4I f0.02
0.50k0.02*
4.44 f0.06
0.86 f0.02
0.47 fO.Ol*
0.57 f0.02
20
79 f2t
I12+5*
3.61 f O . l l t
0.75 f0.02
0.38f0.01
0.47+0.02*t
4.55 fO.08
0.95 f0.03
0.48f0.01 *t
0.59 f0.02
82+4*t
I22 f 3*t
3.77+0.25*t
0.87 f 0. I0
0.43 f0.03'
0.47 f0.03*t
4.29 f 0. I2
0.98 0.08
0.48 fO.01 *t
0.53 f 0.01
16
II
Intrauterine programmingof hypertension
61 I
Table 4. Body weight, organ weights and systolic blood pressures of female rats exposed to maternal low-protein diets
during discrete periods of gestation. Rats exposed to different maternal dietary regimens in utero had systolic blood pressure
determined at 4 weeks and were then killed for analysis of organ weights. Data are means+_SEMof n observations. *Statistically
significant difference relative to the 18% day 0-22 group (P<0.05). tStatistically significant difference relative to the 9% casein day
0-22 group (P<O.O5). SBP, systolic blood pressure.
Maternal dietary regimen
9% w e i n
18% casein
Body weight (9)
SBP (mmHg)
Liver weight (9)
Lung weight (9)
Kidney weight (9)
Heart weight (9)
Liver-body weight (%)
Lung-body weight (%)
Kidney-body weight (%)
Heart-body weight (%)
n
Days 0-22
Days 0-7
83+5
91 f 4
3.42 f0.23
0.86 +O. 10
0.43 rl:0.03
0.43 rl:0.02
4. I5 f0.22
I.05 f0.I5
0.52f0.01
0.52 k0.03
16
90k4
102k3t
4.04 f0.I9
0.81 k0.04
0.40 k0.02
0.49 & 0.02
4.47 f0.07
0.90 k0.03
0.44 *O.OI*
0.55 k0.02
20
low protein on days 8-14 and 15-22 had significantly smaller kidneys in absolute terms, but when
expressed in terms of kidney-body weight ratio the
difference disappeared. However, female pups
exposed to low protein on days 0-7 did have disproportionately smaller kidneys than control animals.
Systolic blood pressures of all male offspring
exposed to low-protein diets were significantly
higher than in control animals (Table 3). The
magnitude of the hypertension observed varied with
the period of exposure to the diet, being greatest in
animals exposed throughout gestation (days 0-7,
15 mmHg; days 8-14, 9mmHg; days 15-22,
16 mmHg; days 0-22, 26 mmHg). In females (Table
4) blood pressure was significantly elevated in the
groups exposed to low protein on days 8-14
(15mmHg), 15-22
(20mmHg) and 0-22
(28 mmHg). No significant effect of maternal diet in
the first week of gestation was noted.
To determine the possible role of glucocorticoids
in programming hypertension, markers of glucocorticoid activity were determined in liver, brain and
lung. Hypertensive rats exposed to maternal
low-protein diets in utero have previously been
shown to have increased sensitivity to glucocorticoids, mediated by alterations to receptor populations [21]. Hepatic TAT activity was significantly
elevated by exposure to low protein throughout
gestation (Fig. 1). Exposure during specific periods
in pregnancy demonstrated that this effect of diet
upon glucocorticoid sensitivity in the liver was
programmed only during week 3 of gestation. TAT
activity in this group of animals was sevenfold higher
than in control animals.
Similarly, GS and GPDH activities (Fig. 2) in
hippocampus and hypothalamus were influenced by
maternal diet. These effects upon glucocorticoidinducible activities [27] were specific, and the
Days8-14
Days 15-22
81 +5
77k4
111f3*
3.50+0.22
0.68k0.03
0.37 +0.02*
0.46 k0.03
4.5 I k0.08
0.88 rl:0.04
0.47 f0.02
0.59 k0.03
13
106+4*t
3.52k0.19
0.76 k0.03
0.39 +0.02*
0.45 f0.02
4.38k0.10
0.95 kO.04
0.49 k0.0 I
0.56 k0.02
20
Days 0-22
92k5
I l9+3*
4.16k0.43
0.82 f0.04
0.48 f0.03
0.48 k0.03
4.49 k0.I9
0.91 kO.08
0.52 & 0.0 I
0.52 k0.01
16
activity of a non-inducible marker, pyruvate kinase,
was similar in all animals in each tissue. GS activity
was significantly elevated in the hypothalamus of all
animals exposed to low-protein diets during
pregnancy. As with hepatic TAT, the effects of
discrete periods of restriction were greater than
those of undernutrition throughout gestation. In the
hypothalamus, however, programming of sensitivity
to glucocorticoids appeared to occur during all
phases of gestation, the greatest effects being in the
periods days 8-14 and days 15-22. Similarly, GPDH
activity was elevated by all periods of low protein
exposure, although in the do-22 group the 25%
increase in activity was not statistically significant.
TAT (umoles/minlmg protein)
l4I
*
t
l2
I80- l
6 -
4
do-22
18%Casein
do-7
d8-14
d14-22
d01-22
9% Casein
Fig. 1. Hepatic tyrosine minotransferase activity in rats exposed to
the dfierent maternal dietary regimens. Data are means+SEM of 7-8
observations per group. *Statistically significant difference relative to the
18% casein day 0-22 group (P<O.O5).
S. C. Langley-Evanset al.
612
tions were determined in plasma of weanling male
rats (Table 5). PRA was similar in control animals
and rats exposed to low protein on days 0-22, 0-7
and 8-14. However, rats exposed during days 15-22
had significantly elevated (60%) activity. This group
of animals had similar ANG I1 concentrations to
control animals, but all other low protein-exposed
animals had significantly lower plasma ANG 11. This
effect was most notable in the day 0-7 group, in
which ANG I1 concentrations were 62% lower than
in control animals.
Sensitivity of this activity to maternal diet appeared
to be greatest in the period days 15-22.
The total PC content of the rat lung was
measured as a crude proxy of glucocorticoid stimulation of that tissue. All animals exposed to maternal
low-protein diets in utero exhibited significantly
higher tissue PC concentrations (Fig. 3).
In addition to stimulation of the fetus by glucocorticoids of maternal origin, a role for the reninangiotensin system in programming hypertension
has been postulated. PRA and ANG I1 concentra14)
GS (pmoledmidmg protein)
(B)
DISCUSSION
GPDH (nmoledmin/mg protein)
700
l,,ih\ 111,
500
400
300
200
A large body of epidemiological evidence from
countries in the developed and developing world
indicates that the risk of hypertension and coronary
heart disease is related to early life experience.
Exposure to an adverse environment in utero
appears to programme physiology and metabolism
permanently, with long-term consequences for the
health and well-being of the fetus. Body weight and
10
100
I
60-22
do-7 d8-14 d15-22 do-22
18% Casein
9% Casein
0
PK (nmoledmidmg protein)
d0-22
d0-7 d8-14 d15-22 d0-22
18% Casein
9% Casein
0)PK (nmoles/min/mg protein)
T
200
100
d0-22
do-7 68-14 d15-22 do-22
18% Casein
9% Casein
do-22
do-7 d8-14 dl5-22 do-22
18% Casein
9% Casein
i
Fig. 2. (A) GS activity in hypothalamus, (6) GPDH activity in hippocampus, (C) pyrwate kinase activity in hippocampus and (0)
p p v a t e kinase activity in hypothalamus of rats exposed in utero to
different maternal dietary regimens. All data are meansf SEM of 5-12
observations per group. *Statisticaliy significant difference relative to the
18% casein day 0-22 group (P <0.05).
do-22
40-7
d8-14
d16-22
do-22
9% Casein
18%Casein
Fig. 3. Lung PC concentrations in rats exposed to different maternal
dietary regimens. Data are meansfSEM of 6-8 observations. *Statistically significant difference relative to the 18% casein day 0-22 group
(P ~0.05).
Table 5. Plasma renin activiky (PRA) and angiotensin It (ANG II) concentrations in rats e x p a d to maternal law-protein
diets during discrete periods of gestation. Rats were exposed in utero to 18% (control) or 9% casein diets. At 4 weeks of age, the
animals were killed and PRA and ANG II concentrations determined (assays by Dr J. Morton, University of Glasgow). All data are
meansfSEM of n observations. *Statistically significant difference relative to the 18% casein day 0-22 group (P t0.05).
Maternal
diet
Timing
18%
casein
9% casein
ANG II
PRA
(pmol h-I ml-')
n
(Phi)
n
Days 0-22
4.94k0.79
7
22.0 k4.6
5
Days 0-7
Days8-14
Days 15-22
Days 0-22
4.68+0.80
4.64+0.59
7.93 I.58*
3.43 k0.74
8
8.5 k0.5*
13.3 +2.5*
15.1 k3.0
12.8 &2.l*
6
5
6
6
6
8
6
Intrauterine programmingof hypertension
proportions at birth appear to be predictive of
disease patterns in adulthood and are in turn early
manifestations of the effects of maternal undernutrition [ll].
The rat model reported in the present paper has
been of use in demonstrating, experimentally, that
modest manipulations of the maternal diet can
produce extreme metabolic changes in the rat fetus
that persist into adulthood. One of these
programmed changes is an increase in systolic blood
pressure. Blood pressure in weanling and young
adult rats is negatively correlated with maternal
protein intake, which explains 25% of the variation
in pressures in these animals [15, 161. Moreover,
changes in the nature of the fatty acid component of
the diet can also programme hypertension in the rat
[28]. The present work was intended to characterize
this rat model further, in terms of the effect of the
timing of a nutritional manipulation in pregnancy
and also in terms of the mechanisms that may
initiate hypertension in utero.
The first of the two studies investigated the effect
of the plane of nutrition before pregnancy on longterm blood pressure of the offspring. Studies in
Jamaica have indicated that in humans a poor
dietary status before conception may be a risk factor
for low birth weight and elevated blood pressure in
the offspring [29]. As in previous studies in our
laboratory, feeding a 9% casein diet before conception and during rat pregnancy resulted in the young
adult offspring with higher systolic blood pressures
than control animals [15]. The blood pressure difference at this age (10-15 mmHg) is fairly typical of
the model. Higher pressure differences are noted in
weanling animals [16] and in animals over 10 weeks
of age. The blood pressures of rats whose mothers
were not prehabituated to the low-protein diet were
also elevated relative to control animals, with a
pressure difference of 18-35 mmHg. Although it
was shown that preconception undernutrition is not
a prerequisite for programming of hypertension, it
did appear to exert some influence. With respect to
male offspring, the hypertension elicited by low-protein diets was more severe if the diet was adopted
only after conception.
The feeding of low-protein diets during specific
periods of pregnancy produced some major effects
upon fetal growth. No differences in weight at birth
were noted in rats exposed to 9% casein throughout
gestation. Our previous work has shown that an
effect of low-protein diets upon birth weight is
inconsistent and that 9% casein results in pups of
low to normal weight [18]. Feeding of the low-protein diet only in the second week of pregnancy
resulted in pups of greater birth weight. Interestingly, the rat fetus exposed to low-protein diets
shows accelerated body weight gain and an
increased placental weight at day 14 [18, 191, and
this increased weight gain persists up to day 20, at
which point the demands for growth can no longer
be met by the maternal substrate supply “1. In the
613
day 8-14 group it is possible that the enhancement
of fetal growth elicited by the low-protein diet was
maintained once an adequate maternal diet was
restored at day 15.
The effects of low-protein diets on body and
organ weights persisted into post-natal life, particularly among the male offspring. All male rats
exposed to low protein, except those exposed only in
the last week of gestation, were heavier than control
animals at weaning. Increased weight gain between
birth and weaning has been noted previously after
mild intrauterine protein restriction [15]. In the case
of the day 0-7 group, body weights were almost
40% higher than in the 18% casein group. The
apparently normal weight gain of the day 15-22
group may reflect the fact that protein restriction at
this time appears to have greatest effect upon linear
growth and growth of organs such as lung and liver
[18]. Organ weights in the male offspring of the
low-protein groups were in proportion to body
weight, with the notable exception of the kidney,
which showed some evidence of disproportionate
growth retardation in all the low-protein groups.
The growth of female rats was less sensitive to the
maternal diet but, again, kidneys appeared to be
small in proportion to body weight in the females
exposed to low-protein diets in the early stage of
pregnancy.
Exposure to low protein throughout gestation
increased blood pressure in both male and female
offspring by 26-28mmHg at the age of weaning.
This is consistent with our previous work with
weanling animals [16]. Again, in both males and
females, exposure in week 2 or 3 alone produced
significant elevations of systolic pressure. Only in
males, however, was the effect of very early
exposure (days 0-7) statistically significant. In all
cases exposure to maternal protein restriction for
discrete 7-day periods produced a less pronounced
hypertension than exposure throughout gestation. It
is, however, clear that no critical time window in
pregnancy is required for the programming of
hypertension. Maternal undernutrition at any time
appears able to elicit some programming response,
although the effect may be seen in males only, if
applied during the embryonic phase. This finding is
not inconsistent with the proposal made by Barker
Ill], in that, although specific periods of human
fetal undernutrition may produce different patterns
of growth and different patterns of disease, hypertension is associated with undernutrition in all three
trimesters of pregnancy.
Although all points in pregnancy may allow initiation of hypertension by maternal dietary factors,
different mechanisms may be implicated at different
points in development. With the rat model we are
currently investigating three candidate mechanisms:
(i) changes in renal structure, (ii) renin-angiotensin
disturbances and (iii) glucocorticoid hormone
action. Given the assertion that renal structure, and
in particular nephron number, may be programmed
614
S. C.Langley-Evans et al.
in utero [30], the changes in kidney size reported in
the present study are of great interest.
A role for the renin-angiotensin system in the
maintenance of hypertension induced by maternal
protein restriction has been previously asserted.
Hypertensive animals exposed to 9% casein diets
throughout fetal development are sensitive to angiotensin-converting enzyme inhibitor (ACE) therapy,
and there is evidence of elevated plasma and
pulmonary ACE activity [15, 171. In the present
study it has been clearly demonstrated that undernutrition in early to mid-pregnancy may modify
ANG I1 concentrations at weaning, whereas PRA is
programmed by protein restriction in the last week
of pregnancy. The finding of low ANG I1 concentrations in most of the groups exposed to low protein
in the presence of normal to high PRA is unusual.
Abolition of hypertension in these animals with
ACE inhibitors indicates a role for ANG I1 in
maintaining the higher blood pressure [17]. Low
concentrations of the peptide suggest that there may
be increased ANG I1 turnover in the low-protein
groups, or that these animals have increased
numbers of receptors for ANG 11, which remove
hormone from the circulation. The predominant
ANG I1 receptor in adulthood is the AT1 receptor,
which after hormone binding is internalized by cells
[311.
A role for glucocorticoids in programming hypertension was proposed on the basis of observations in
rats that activity of 11B-hydroxysteroid dehydrogenase (11BHSD) in placenta is related to birth
weight and inversely related to placental weight [32,
331. 11BHSD is proposed to protect the fetus from
glucocorticoids of maternal origin that may otherwise promote inappropriate gene expression. In the
low-protein diet model of hypertension, we have
demonstrated that protein restriction reduces
placental llpHSD activity [20], and our hypothesis
is that development of hypothalamic-pituitaryadrenal axis function is altered in the offspring
consequently overexposed to maternal steroid. To
support this argument, we have demonstrated that
pharmacological adrenalectomy of rats fed low-protein diets prevents initiation of hypertension in their
pups [19]. Rats exposed to low-protein diets in utero
have increased numbers of glucocorticoid receptors
in the hippocampus and blood vessels [21]. One
consequence of this may be a direct steroid effect on
blood pressure at the level of vascular smooth
muscle [34], or a central effect through the hippocampus or hypothalamus [35]. In keeping with
increased central sensitivity to glucocorticoids,
increased activities of the corticosterone-inducible
enzymes GS and GPDH have been noted in brains
of the hypertensive rats, despite low to normal
glucocorticoid levels [21].
In the present study we have again demonstrated
that in brain, lung and liver glucocorticoid-inducible
markers are elevated by maternal protein restriction.
These markers are unlikely to have any physiological
significance with respect to hypertension, but serve
as indicators of altered hormonal activity. In lung,
all periods of maternal dietary manipulation
produced an elevation of PC concentration. In brain
and liver, however, the effects were more specific in
timing and also in excess of the effect of protein
restriction throughout pregnancy. Glucocorticoid
effects in the liver appeared to be linked to the last
week of gestation and may result from stimulation
by maternal steroid that has crossed the placenta
without conversion by llpHSD, and also by fetal
steroids. The fetal adrenal becomes active around
day 16, and an inappropriate hormonal environment
around this time, resulting from stimulation by
maternal steroid, may have resulted in altered
patterns of fetal adrenal activity and regulation [36].
The central effects of glucocorticoids appeared to
be linked to mid or late gestation, although
increases in GS and GPDH activity also occurred in
the day 0-7 group. These effects may be the product
of the same stimulatory signals described above, but
the fact that such overstimulation of the developing
brain occurs in early and mid-gestation may be
suggestive of widespread alterations to brain structures and later function.
Considering the different mechanisms, it appears
that each may be initiated at different points in
gestation. Renal size and structure may be vulnerable in early gestation. Renal hormone systems may
be disturbed by early or mid-gestation undernutrition, and glucocorticoid-mediated effects, although
present at all points studied, were most pronounced
in late gestation. An interesting feature of the study
was that, although biochemical indices such as TAT,
GPDH or PRA showed clear critical periods of
sensitivity to protein restriction, blood pressure did
not, and the greatest increases in pressure were
elicited by undernutrition throughout the rat
pregnancy. It is possible that several independent
programming mechanisms may operate in this
model and that prolonged undernutrition produces
additive effects of different systems. As an example
of how this may occur, glucocorticoid receptor
activation in vascular smooth muscle has been
shown to increase pressor responsiveness to ANG I1
[37], and in the low protein-exposed rat glucocorticoid receptor binding is elevated in the thoracic
aorta [21].
ACKNOWLEDGMENTS
Assays of PRA and plasma ANG I1 were
performed by Dr J. J. Morton, University of
Glasgow. This work was funded by the Wellcome
Trust. R.C.S. is supported by an MRC Studentship.
REFERENCES
I. Barker DIP, Winter PD, Osrnond C, Margem 8, Sirnmd.5 SJ. Weight in infancy
and death fmm ischaemic heart disease. Lancet 1989; ii 577-82.
Intrauterine programming of hypertension
2. Barker DJ, Bull A, Osmond C, Simmonds SJ. Fetal and placental size and risk of
hypertensionin adult life. Br Med J 1990 301: 259-62.
3. Law CM, de Swiet M, Osmond C. et al. Initiation of hypertension in utero and its
amplificationthroughout life. Br Med J 1993; 3 0 6 24-7.
4. Barker DIP, Osmond C, GoldingJ, Kuh D, Wadsworth MEJ. Growth in utero,
blood pressure in childhood and adult life, and mortality from cardiovascular
disease. Br Med J 1989; 298 564-7.
5. Osmond C, Barker DIP, Winter PD, Fall CHD, Simmonds SJ. Early growth and
death from Cardiovascular disease in women. Br Med J 1993; 307: IS 19-24.
6. Barker DIP, Godfrey KM, Fall C, Osmond C. Winter PD, Shaheen SO. Relation
of birth weight and childhood respiratory infection to adult lung function and
death from chronic obstructive aitways disease. Br Med J 199 I; 3 0 3 671-5.
7. Forrester TE, Wilks RJ, Bennett FI, et al. Fetal growth and cardiovascular risk
factors in Jamaicanschoolchildren. Br Med J 1996; 312: 156-60.
8. Yajnik CS,Fall CHD, Vaidya U. et al. Fetal growth and glucose and insulin
metabolism in four-year-old Indian children. Diabet Med 1995; I2 330-6.
9. Barker DJP, Gluckman PD, Godfrey KM, HardingJE, Owens JA,Robinson IS.
Fetal nutrition and cardiovascular disease in adult life. Lancet 1993; 341: 938-41.
10. Prentice AM. Can maternal dietary supplements help in preventing infant
malnutrition?Acta Paediatr Scand 1991; Suppl. 374 67-77.
I I. Barker DIP. Mothers, babies and disease in later life. London: BMJ Publishing,
1994.
12. Godfrey KM, Barker DIP. Maternal nutrition in relation to fetal and placental
growth. Eur J Obstet Gynaecol 1995; 61: 15-22.
13. Godfrey KM, Robinson S, Barker DIP, Osmond C, Cox V. Maternal nutrition in
early and late pregnancy in relation to placental and fetal growth. Br Med J 1996;
312410-14.
14. Campbell DM, Hall MH, Barker DJP, Cross J, Shiell AW, Godfrey KM. Diet in
pregnancy and the offspring’s Mood pressure 40 years later. Br J Obstet
Gynaecol 1996; 103 273-80.
15. Langley SC, JacksonAA. Increased systolic blood pressure in adult rats induced
by fetal exposure to maternal low protein diet. Clin Sci 1994; 8 6 217-22.
16. Langley-Evans SC, Phillips GJ, Jackson MIn utero exposure to maternal low
protein diets induces hypertensionin weanling rats, independently of maternal
blood pressure changes. Clin Nutr 1994: 1 3 319-24.
17. Langley-Evans SC, Jackson A h Captopril normalises systolic blood pressure in
rats with hypertension induced by fetal exposure to maternal low protein diets.
Comparative Biochem Physiol 1995; I IOA 223-8.
18. Langley-Evans SC, Gardner DS, Jackson M Disproportionatefetal rat growth in
late gestation is associated with raised systolic blood pressure. J Reprod Fertil
1996; 106 307-12.
19. Langley-Evans SC, Phillips GJ, Gardner DS, JacksonM Me of glucocorticoids in
programmingof maternal diet-induced hypertensionin the rat. J Nutr Biochem
1996; 7: 173-8.
20. Langley-EvansSC. Phillips GJ, Benediktsson R, et al. Protein intake in pregnancy,
615
placental glucocorticoid metabolism and the programmingof hypertension.
Placenta 1996; 17: 169-72.
21. Langley-EvansSC, Gardner DS. Jackson M Evidence of programmingof the
hypothalamic-pituimy-adrenal axis by maternal protein restriction in pregnancy.
J Nutr 1996; 126 1578-8s.
22. Bunag, RD. Validation in wide awake rats of a tail-cuff method for measuring
systolic blood pressure. J Appl Physiol 1973; 34:279-82.
23. Pfeffer JM, Pfeffer MA, Frohlich ED. Validity of an indirect tail-cuff method for
determination of systolic arterial blood pressure in unanaesthetisednomtensive
and spontaneously hypertensiverats. J Lab Clin Med I97 I;7 8 957-62.
24. Shargill NS, York DA, Marchington DR Regulation of hepatic tyrosine amino
transferase in genetically obese rats. Biochim Biophys Acta 1983; 756 297-307.
25. Langley SC, York DA Increased type IIglucocorticoid-receptor numbers and
glucocorticoid-sensitive enzyme activities in the brain of the obese Zucker rat.
Brain Res 1990; 533: 268-74.
26. Langley SC, Rickett GWM, Hunt A, Postle AD,York DA, Kelly FJ. Maturational
effects of glucocorticoids on lung phosphatidylcholine and antioxidant enzyme
status in the obese Zucker rat. Biochem Pharmacol 1993; 4 5 543-5 I.
27. Schlatter LK, Ting S-m. Meserve, LA, D o h LA. Characterizationof a
glucocorticoid-sensitive hippocampal protein. Brain Res I990 522 215-23.
28. Langley-Evans SC. Intrauterine programmingof hypertension: nutrient
interactions. Comparative Biochem Physiol 1996; 14k 327-33.
29. Godfrey KM, Forrester T, Barker DIP, et al. The relation of maternal nutritional
status during pregnancy to blood pressure in childhood. Br J Obstet Gynaecol
1994; 101: 398-403.
30. Mackenzie HS, Brenner BM. Fewer nephrons at birth a missing link in the
etiology of essential hypertension?Am J Kidney Dis 1995; 2 6 91-8.
3 I. De Gasparo M, Levens NR Pharmacology of angiotensin II receptors in the
kidney. Kidney Int 1994; 4 6 1486-91.
32. Eenedikuson R, Lindsay RS. Noble J, Seckl JR, Edwards CRW. Glucocorticoid
exposure in utero: new model for adult hypertension. Lancet 1993: 341: 339-41.
33. Edwards CRW, Benediktsson R, Lindsay RS, Seckl JR Dysfunction of placental
glucocorticoid barrier: link between fetal environment and adult hypertension.
Lancet 1993; 341: 355-7.
34. Kornel L, KanamarlapudiN,Ramsay C, et al. Arterial steroid receptors and their
putative role in the mechanism of hypertension. J Steroid Biochem 1983; 1 9
333-44.
35. van den Berg DTWM, DeKJoet ER, van Dijken HH, de long W. Differential
central effects of mineralocorticoid and glucocorticoid agonists and antagonists
on blood pressure. Endocrinology I990 126: I 18-24,
36. Stewart PM, Whorwood CB, Mason JI.Type 2 I Ifl-hydroxysteroid
dehydrcgenase in fetal and adult life. J Steroid Biochem Mol Biol 1995; 5 5
465-7 I.
37. Provencher PH, Saltis J, Funder JW. Glucocorticoids but not mineralocorticoids
modulate endothelin-l and angiotensin I1 bindingto spontaneously hypertensive
~ x u l a smooth
r
muscle cells. J Steroid Biochem Mol Biol 1995; 5 2 219-25.