PERI-IMPLANTATION UNDERNUTRITION PROGRAMS BLUNTED ANGIOTENSIN II EVOKED BAROREFLEX RESPONSES IN YOUNG ADULT SHEEP

Abstract

An adverse environment restricted to the time of conception and implantation has been shown to influence later fetal growth and development to term in humans and sheep. In one rat study, protein undernutrition during the pre-implantation period only elevated the systolic blood pressure of the resulting adult offspring. No study has yet followed up adult cardiovascular function in a slower growing, non-litter bearing species after exposure to peri-implantation undernutrition. In the present study, eight ewes were fed to 50% equivalent food intake of twelve control ewes from 1-30 days (term ∼147 days) only. Following consumption of an adequate diet to term, natural lambing and then weaning, basal cardiovascular status and baroreflex function were examined in the resultant young adult offspring. We show that peri-implantation undernutrition has no effect on birth weight or postnatal growth rate but reduces brain weight in the offspring at one year of age i.e. as young adults. Early nutrient restricted sheep also show increased pulse pressure, a reduced rate pressure product and a leftward shift in their baroreflex function curve. However, baroreflex sensitivity in early nutrient restricted sheep is significant blunted during angiotensin II infusion, although not after phenylephrine and sodium nitroprusside treatment. In addition, peri-implantation undernutrition appears to potentiate the tachycardia produced after an experimentally controlled reduction in central blood pressure, relative to controls. The data suggest that peri-implantation undernutrition may program long-term cardiovascular dysfunction that ultimately increases the risk of hypertension in later life. An increase in regional angiotensin II activity during this critical early phase of development is a likely candidate mechanism for the effects observed. The data have broad implications for the health outcome of those offspring from mothers who were poorly nourished during early, often unknown pregnancy and for embryos artificially manipulated because of infertility treatment.

INTRODUCTION

Hypertension is a major risk factor associated with coronary heart disease (CHD) and represents a common cause of death in the population over the age of 50 ¹. Hypertension and coronary heart disease are both multifactorial in their etiology but epidemiological studies in a number of different human populations have shown that the prenatal environment has an important role in determining the incidence of these diseases ². Consequently, there has been renewed interest in the effect of poor gestational nutrition upon pre- and postnatal growth after the early studies of McCance & Widdowson ^3;4. These original hypotheses have been advanced to include changes in physiological function leading to adult pathology- the “developmental origins of adult disease hypothesis” ^5;6. The methodologies and inferences of the hypothesis have criticised ^7;8 but meta-analyses have illustrated the strength of the epidemiological findings ^1;9. Work from the many animal models that now exist strongly supports the hypothesis and suggests that programming of disease risk is both biologically plausible and of major importance in terms of public health ^5;10.

The majority of studies investigating the developmental origins of adult disease hypothesis have concentrated on exposing the developing fetus to a poor diet throughout gestation. However, transient undernutrition during defined critical periods of fetal development has been shown to have an equal ‘programming’ influence ^11-16. A major restriction of nutrient availability may be expected to influence the development of fetal form and function over late gestation, when the stoicheiometry of nutrient supply and demand are close ¹². However, while it is known that manipulation of embryos as a result of infertility treatments may alter fetal growth leading to ‘large offspring syndrome’ and perhaps influence later health outcomes ¹⁷, a number of studies have now shown that the periconceptual ^18-21 or even pre-implantation ^15;16;22 nutritional environment may also programme later physiological function - a time at which overall fetal demands for energy are remarkably low. In only two of these studies, both of which studied protein restriction in the rat ^21;22 - a species with a remarkable rate of protein accretion and thus sensitivity to protein restriction, were the exposed fetuses followed-up into adult life. In both, periconceptual ²¹ or pre-implantation ²² protein restriction elevated the systolic blood pressure of the young adult offspring. The mechanism underlying this effect remains elusive. While the follow-up evidence from the ‘Dutch Hunger Winter Famine’ suggests that undernutrition confined to early development has the greatest effects on the cardiovascular system ^23;24; no study to date has yet followed up adult cardiovascular function, after exposure to undernutrition during the peri-implantation period only, in a non-litter bearing species whose pre- and postnatal growth rate is more comparable to human infants. Hence, the present study investigated the hypothesis that global energy restriction during early gestation programmes alters cardiovascular control and increases blood pressure in the resultant offspring as young adults i.e. at one year of age. Resting cardiovascular status was assessed by measuring systolic, diastolic and mean arterial blood pressure and heart rate. Short-term cardiovascular control was assessed through an analysis of the range, sensitivity and set-point of the baroreceptor reflex using phenylephrine (PE) and sodium nitroprusside (SNP) infusion, to raise and lower arterial blood pressure, respectively. In addition, a cursory examination of baroreflex function was made through similar changes in pressure and heart rate during angiotensin II infusion - a protocol known to produce a different baroreflex response to the physiologically purer stimuli of PE and SNP ²⁵. Operation of the cardiovascular baroreflex is key to maintaining central pressure during ambulatory changes in blood pressure: if inadequate, then risk of later hypertension is increased ^26-28. We show that global undernutrition during this period does indeed programme cardiovascular dysfunction in the young adult offspring, and suggest that the primary triggering mechanism may be related to regional increases in angiotensin II activity, especially in centrally-located areas involved in baroreflex function.

MATERIALS AND METHODS

Animals

All procedures were performed under the UK Animals (Scientific Procedures) Act, 1986. Twenty Blue-faced Leicester x Swaledale ewes were used in the study. All ewes were of similar age (7-8 months), body weight and fat distribution as assessed from physical characteristics in the lumbar spine and loin. Ewes were mated with a single ram and within 12hrs of mating individually housed and randomly assigned to one of two groups; Group I, ewes were fed to adequately meet 100% metabolisable energy (ME) requirements as defined by the AFRC ²⁹ throughout gestation to term (Controls; term ∼ 147dGA); Group II, ewes were fed to 50% AFRC ME requirement from day 1-30 gestation and thereafter to 100% ME requirement to term (nutrient restricted; NR). All ewes were bedded on wood shavings and fed a diet consisting of chopped hay (ME content of 8.9 MJ/kg dry matter (DM) and a digestible crude protein (DCP) content of 45 g/kg DM) and a barley-based concentrate (ME, 8.9 MJ/kg DM; DCP, 218 g/kg DM) in a ratio of 3:1, respectively. All ewes were weighed and body condition scored at fortnightly internals and the nutritional regimen for each individual ewe was adjusted accordingly to allow for any weight gain and growth of the conceptus. At term, lambs were delivered naturally and birth weights recorded. The twin lambs that were delivered were reared as singletons with the ewe, the other twin being sacrificed and tissues collected. Of the control group n,6 were twins and n,6 were singletons; for the NR group n,3 were twins and n,5 were singletons. The offspring were ewe reared until weaning at 3 months of age and thereafter grass-fed until one year of age. During this time, the weights and growth of individual lambs were recorded monthly.

At one year of age, all sheep were group housed indoors for one week prior to surgery. For 24 h prior to surgery all food, but not water, was withdrawn from the animals. Anaesthesia was induced with sodium thiopentone (20 mg.kg^-1 I.V. Intraval Sodium; Rhone Mérieux, Dublin, Ireland) and maintained with 1-2% halothane in 50:50 O₂/N₂O. Carotid and jugular catheters were inserted into each sheep and the incision closed. All sheep received a dose of long-acting antibiotic (15 mg.kg^-1 I.M. amoxycillin, ‘Duphamox’; Fort Dodge Animal Health Ltd, Southampton, UK) and analgesia (1 mg.kg^-1 flunixin meglumine; ‘Finadyne’; Shering-Plough, Kenilworth, UK) post-operatively. Catheter patency was maintained by daily flushing with heparinized saline (50 I.U. heparin.ml^-1). All sheep had established normal feeding patterns within 1-2 hrs after surgery and showed no visible signs of discomfort for the duration of the experimental period.

Experimental protocols

A period of 2-4 days post-operative recovery was allowed with the investigator blinded to the dietary origin of the sheep prior to any experiment being performed. Over a 10-day period and on 5-6 separate occasions and days all sheep were placed in metabolism crates with ad libitum hay and water available. After at least an hour catheters were connected to pre-calibrated pressure transducers (SensorNor 840; S 4925) attached at heart level linked to a data acquisition system (Po-Ne-Mah; Version 3, Gould Instrument Systems Inc) and a baseline recording taken over a 30min period. Analogue signals for real-time systolic, diastolic, mean arterial pressure and heart rate were recorded at one-second intervals, digitized and stored on an Excel spreadsheet for further analysis. From these data, pulse pressure (systolic-diastolic) and the rate pressure product (mean arterial blood pressure (mmHg) x heart rate (beats min^-1); an index of myocardial work and thus oxygen consumption) were derived.

Angiotensin II

Step-wise I.V. increases in angiotensin II (0, 2.5,5,10,20,40 & 60 ng.kg^-1.min^-1) were administered every 10mins, followed by a 10min recovery period in which cardiovascular variables returned to baseline.

Phenylephrine

Sheep were administered a bolus dose (75 μg.kg^-1; I.V. given over 2 mins) of the sympathomimetic (α-adrenergic agonist), phenylephrine hydrochloride (PE; Sigma, Poole, UK) and cardiovascular variables followed for a 1-hour period.

Sodium nitroprusside

After return of cardiovascular variables to baseline the sheep were infused I.V. (2.5 μg.kg^-1.min^-1) with the endothelium-dependent vasodilator sodium nitroprusside (SNP; Abbott Laboratories, Maidenhead, UK) for 5 mins after a 5 min baseline recording period.

Statistical analyses

All data are expressed as Means ± S.E.M. unless otherwise stated. The data for maternal feed intake, birth weight, current weight, postnatal growth rates and cardiovascular variables (blood pressures, heart rate, rate pressure product) were continuous and analysed by one-way analysis of variance with repeated measures (SPSS Inc, Chicago, USA). Baroreceptor reflex curves describing the relationship between mean arterial blood pressure and heart rate were analysed by a four-parameter logistical sigmoid function (Sigmaplot; SPSS Inc, U.S.A.) using equation 1 ​^30;31:

$$\text{Baroreflex analysis}\mathit{MAP}=\left(\frac{a}{1+e^{b\cdot [HR-x_o]}}\right)+y_o$$ Equation 1

where a is the heart rate (HR) range (i.e. max HR − min HR; beats.min^-1), b is a measure of the slope over the linear portion of the curve (1/mmHg), x_o is the mid- or setpoint (mmHg) for equal pressor and depressor responses and y_o is the minimum HR. The sigmoidal parameters generated for each individual animal were compared by Students’ t-test. The maximal gain or baroreflex sensitivity was derived by calculus from the first derivative of equation 1.

$$\text{Calculation of maximal gain}\left(\frac{a}{4b}\right).$$ Equation 2

Areas under the curve (AUC) were calculated using a custom designed Excel spreadsheet according to Equation 3.

$$\text{Analysis of area under the curve}0.5•(a+z+\left(2•\sum _b^y\right).$$ Equation 3

Where a is the first data point, z is the last data point and b-y are the data points enclosed by the curve. Average values were then compared by one-way analysis of variance. For a comparison between the slopes of linear regression curves obtained during angiotensin II infusion the data were analysed by ANCOVA ³². For all comparisons, statistical significance was accepted when P<0.05.

RESULTS

Nutrition and Pregnancy

Ewe body weight and condition score at the time of conception was not different between dietary groups (Controls, 46.0±5.5 kg and 2.0 ± 0.5 units; NR, 45.9±5.1 kg and 2.5±0.5 units, means±1 S.D). Control ewes consumed between 7.68±0.24 to 17.7±0.59 MJ/day during weeks 1 and 21 of pregnancy, respectively and gained 11.6±1.5 kg over the duration of gestation. NR ewes consumed 3.61±0.11 MJ/day for the first 30 days of pregnancy i.e. ∼50% control intake and, after one week of realimentation, dietary intake was equivalent to control intake (Figure 1). Consequently, NR ewes lost more weight than controls over the first 4 weeks of pregnancy (4.1±0.4 vs. 1.5±0.4 kg, respectively; P<0.05) and their overall weight gain tended to be lower but not statistically different (8.2±1.2 kg).

Figure 1. Maternal metabolisable energy (ME) intake during gestation.

Values are means ± S.E.M. for control (-O-, n = 12) and nutrient restricted (NR; -●-, n = 8) ewes for weekly ME intake (MJ/day) from conception to term (∼147 days). Statistical differences are: *, P<0.05 control vs. NR intake. For dietary details see Materials and Methods.

Birth weights and postnatal growth rates

There were no differences between singles and twins of either group for any data and therefore subsequent values represent the combined data. Lamb weights were similar between maternal dietary groups (Control, 3.9±0.3; NR, 4.2±0.3 kg) and were appropriate given the size of the ewe and plane of nutrition during pregnancy, according to recommended guidelines for the nutrition of housed, pregnant, lowland ewes ²⁹. There was no difference in the growth rate (current weight-birth weight/time; kg/week) of control or NR lambs during the postnatal period (Controls: 1.93±0.15, 1.43±0.09, 0.92±0.07; NR: 1.80±0.14, 1.43±0.07, 0.92±0.06 kg/week for growth over 0-3, 4-6 and 7-12 months, respectively) but the rate of growth slowed over time in both groups (P<0.001). At one year of age there was no difference in body weight between the two groups of sheep (control, 51.5±3.5; NR, 51.6±3.0 kg).

Cardiovascular responses to angiotensin II infusion

Baseline cardiovascular status was assessed on different days between 9:00-10:00am on 5-6 occasions for each individual and data averaged to give a mean basal value for each sheep. Values for systolic (100±4 vs. 98±3 mmHg), diastolic (77±4 vs. 70±3 mmHg), mean arterial pressure (87±4 vs. 81±3 mmHg) and heart rate (107±5 vs. 100±4 beats.min^-1) were similar in control and NR sheep, respectively. However values for pulse pressure (28±1 vs. 23±1 mmHg, P=0.06) and the rate pressure product (7.91±0.46 vs. 9.36±0.72 (beats.min^-1).mmHg/10³; P=0.005) tended to be higher and lower in NR relative to control sheep, respectively. In control sheep, angiotensin II infusion resulted in dose-dependent increments in arterial blood pressure and decrements in heart rate (Figure 2A & 2C). In NR sheep, however, the increment in arterial blood pressure was similar but heart rate failed to decline significantly (Figure 2B & 2D). Consequently, the area under the curve for delta heart rate during the challenge was significantly (four-fold) lower in NR relative to control sheep (−125±120 vs. −841±107 units respectively; P=0.004). When plotted as a linear relationship between individual data points for mean arterial blood pressure and heart rate (baroreflex sensitivity) the slope was significantly blunted in NR (y=−0.15x + 109; r²=0.49, n=8) sheep relative to controls (y=−0.41x + 138; r²=0.92, t= 10.96; P<0.0001, n=12), as illustrated in Figure 3.

Figure 2. The change in mean arterial blood pressure and heart rate during angiotensin II infusion in control and periconceptually nutrient-restricted sheep.

Values are minute means ± S.E.M. for control (left panel, n = 12) and nutrient restricted (NR; right panel, n = 8) fetuses for a baseline period (10 min), 1 h of angiotensin II infusion (stepwise dose increments from 2.5-60 ng.kg^-1.min^-1 every 10 min) and 10 min of recovery. Statistical differences are: *, P<0.01 control vs. NR as analysed by area under the curve for the period between the start and end of angiotensin II infusion (time zero to 1 h: control, −125±120; NR, −841±107 units).

Figure 3. Regression of mean arterial blood pressure with heart rate during angiotensin II infusion.

Data points are the minute means for control (-O-, n = 12) and nutrient restricted (NR; -●-, n = 8) for paired mean arterial blood pressure and heart rate values obtained during the 1 h of angiotensin II infusion. For clarity, S.E.M. has been included with means every 9 min only. There was a significant change in the slope of the relationship between mean arterial blood pressure and heart rate in NR sheep (y=−0.15x + 109; r²=0.49, n=8) relative to controls (y=−0.41x + 138; r²=0.92, t= 10.96; P<0.0001, n=12).

Cardiovascular responses to phenylephrine (PE) and sodium nitroprusside (SNP) infusion

Resting mean arterial blood pressure and heart rate was similar prior to PE and SNP challenges in both groups of sheep (Control: PE, 92±3 mmHg and 115±4 beats.min^-1; SNP, 92±4 mmHg and 107±5 beats.min^-1; NR: PE, 89±5 mmHg and 99±5 beats.min^-1; SNP, 82±4 mmHg and 93±4 beats.min^-1). Injection of a bolus dose of PE significantly increased mean arterial blood pressure (by 120±11 and 100±13 mmHg) and decreased heart rate (by 88±4 and 80±4 beats.min^-1) in control and NR sheep, respectively (Figure 4). The slope of the linear relationship between individual data points for mean arterial blood pressure and heart rate during PE (i.e. parasympathetic arm of the baroreflex) was similar in control and NR sheep (Control, y=−1.00x + 208; r²=0.91; NR, y=−1.02x + 194; r²=0.88). Infusion of SNP significantly lowered arterial pressure; the reduction in diastolic pressure being significantly greater than the reduction in systolic pressure in both dietary groups (diastolic by 29±5 and 24±4; systolic by 19±3 and 15±4 mmHg, in control and NR sheep, respectively; P<0.05). Consequently, pulse pressure significantly increased in control and NR sheep (by 11±3 and 11±4 mmHg, respectively; P<0.05). The reduction in pressure was compensated for by significant increases in heart rate in control and NR sheep (Figure 5), however the mean increment from baseline was greater in NR sheep (by 12±6 and 30±6 beats.min^-1, respectively; P=0.04). Figure 6 shows the mean individual data points for the relationship between mean arterial pressure and heart rate, during the initial (2 min) phases of parasympathetic and sympathetic activation induced by PE and SNP, respectively. Statistical analysis of the curve indicated no significant difference in either the operating range of the baroreflex (a; 70±14 and 65±5 beats.min^-1), estimated slope (b; −6.5±2.1 and −6.6±1.5 mmHg) or derived maximal gain (b; −5.9±2.4 and −3.1±0.5 bpm/mmHg) and the minimum heart rate achieved (y_o; 58±4 and 62±5 beats.min^-1) in control and NR sheep, respectively. However, the set point (x_o) was significantly shifted to the left i.e. toward a lower mean pressure, in NR relative to control sheep (86±7 vs. 100±3 mmHg, respectively; P=0.05, Figure 6).

Figure 4. The change in mean arterial blood pressure and heart rate during phenyleprine (PE) infusion in control and periconceptually nutrient-restricted sheep.

Values are minute means ± S.E.M. for control (left panel, n = 11) and nutrient restricted (NR; right panel, n = 7) sheep for a baseline period (10 min) and for 30 min after a bolus dose of PE (75 μg.kg^-1).

Figure 5. The change in mean arterial blood pressure and heart rate during sodium nitroprusside (SNP) infusion in control and periconceptually nutrient-restricted sheep.

Values are minute means ± S.E.M. for control (left panel, n = 11) and nutrient restricted (NR; right panel, n = 7) sheep for a baseline period (5 min), 5 min of SNP infusion (2.5 μg.kg^-1.min^-1) and 5 min of recovery. Statistical differences are: *, P<0.05 control vs. NR for the mean increment in heart rate during SNP infusion.

Figure 6. The range and function of the baroreflex in control and nutrient restricted sheep.

Data points are (A) all individual points for a representative control (-O-) and nutrient restricted (NR; -●-) sheep and (B) for means ± S.E.M. (included every 20 sec for clarity) for control (-O-, n = 11) and NR (-●-, n = 7) for mean arterial blood pressure and heart rate values obtained during phenylephrine and sodium nitroprusside infusion. The slope, range, minimum heart rate value and sensitivity of the relationship between mean arterial blood pressure and heart rate were similar in control and NR sheep. However the set-point was significantly (P<0.05) shifted to the left in NR sheep relative to controls.

Sheep biometry at one year of age

There were no differences between the two dietary groups in the weights of any organ measured with the exception of the brain, which was significantly (P=0.03) smaller in NR relative to control sheep (Table 1). When organ weights were expressed relative to body weight there were, again, no significant differences between groups. The significance of the difference in the weight of the brain was, however, weakened when expressed relative to body weight (P=0.08; data not shown).

Table 1. Sheep biometrical measurements at post mortem.

Sheep were humanely put down using a lethal dose of sodium pentobarbitone (200mg.kg^-1) and body and all major organ weights measured.

Sheep biometry at one year of age

	CONTROLS	NR	P
Body weight (kg)	51.5 ± 3.5	51.6 ± 3.0	NS
Brain weight (g)	100 ± 2	94 ± 1	0.03
Heart weight (g)	214 ± 15	207 ± 8	NS
Left ventricular wall thickness (mm)	10.6 ± 0.6	9.8 ± 0.3	NS
Right ventricular wall thickness (mm)	5.6 ± 0.5	4.5 ± 0.2	NS
Septum thickness (mm)	12.9 ± 0.4	12.8 ± 0.6	NS
Liver weight (g)	634 ± 69	708 ± 58	NS
Kidney weight (g)	121 ± 5	121 ± 5	NS
Adrenal weight (g)	3.1 ± 0.2	3.3 ± 0.3	NS
Spleen weight (g)	125 ± 12	120 ± 19	NS
Lung weight (g)	516 ± 21	478 ± 29	NS
Pancreas weight (g)	47.6 ± 5.2	43.0 ± 5	NS
Perirenal adipose tissue (g)	154 ± 21	145 ± 20	NS
Cardiac adipose tissue (g)	43 ± 5	44 ± 4	NS
Omental adipose tissue (g)	278 ± 42	252 ± 31	NS

DISCUSSION

The present study provides the first documented evidence of programming influences operating prior to and during embryo attachment in a species that in terms of fetal number, pre- and postnatal growth is comparable to human infants. Furthermore, the peri-implantation nutritional environment of the embryo is clearly able to programme cardiovascular dysfunction in the young adult independently of reductions in birth weight or accelerated postnatal ‘catch-up’ growth - key processes thought to contribute to the developmental origins of adult cardiovascular disease ³³. It is known that baroreceptor sensitivity per se is blunted when analysed under conditions of high, but physiological infusion doses of angiotensin II, when compared to the purer stimulus produced with the synthetic α-adrenergic agonist, phenylephrine ²⁵ and this is reproduced in the current manuscript (witness the differing curve gradients produced by angiotensin II and PE). This may be due, in part, to the differing methodologies used to assess baroreflex function in the current study i.e. acute (PE & SNP) vs. a more prolonged stimulus with angiotensin II infusion, or a centrally-orientated mechanism e.g. angiotensin II-mediated inhibition of efferent parasympathetic outflow from the area postrema, rather than a direct effect on the heart or baroreceptors themselves (for review see Reid (1992) ²⁵. Nevertheless, in this respect, there is no difference between dietary groups. Importantly, while the full range and sensitivity of the baroreflex is similar between dietary groups when constructed with PE and SNP treatment, the baroreflex set-point is significantly shifted to the left in NR sheep. We speculate that this data, together with the observation of no greater pressor responses to peripherally infused angiotensin II infusion (Figure 2), but increased blunting of baroreflex sensitivity, suggests a central origin for the re-setting of baroreflex function in NR sheep. Possible mechanisms for these data are 1) increased angiotensin II action within the cardiovascular control centers of the brain i.e. the area postrema and/or nucleus of the solitary tract, or 2) a direct chronotrophic action on the heart. The present study cannot differentiate between these two possible mechanisms, but nutritional regulation of regional angiotensin II receptor populations (AT₁ & AT₂) has been demonstrated in term ovine neonates³⁴ and adult rat offspring ^35;36. Interestingly, increased densities of central AT₁ in mice has been shown to influence baroreflex setting and sensitivity ³⁷.

The current study has shown that peri-implantation undernutrition has no major effect on resting systolic or diastolic pressures ^25-27 in the resulting young adult sheep. However, at this age, we show evidence of altered baroreflex function and renin-angiotensin system activity. At three years of age, male sheep that have undergone a similar nutritional paradigm also have altered baroreflex function but, importantly, have elevated pre-feeding blood pressure ³⁸. We therefore speculate that the results of the current study may provide an early indication of compromised cardiovascular control that eventually leads to a pathophysiological outcome as the individual ages ³⁹.

Perspectives

Global undernutrition during peri-implantation, embryonic development only programmes cardiovascular dysfunction in the young adult offspring. The data suggest two, possibly interrelated, mechanisms for this phenomena: 1) altered baroreflex control and 2) regional increases in angiotensin II activity. The results of this study have wide-ranging implications with regard to the health outcomes of those offspring from mothers who were poorly nourished during early, often unknown pregnancy and from embryos artificially manipulated because of infertility treatment.