1024 Exp Physiol 94.9 pp 1024–1033

Experimental Physiology – Research Paper

Effects of pre- and periconceptional undernutritionon arterial function in adult female sheep are vascularbed dependent

Christopher Torrens1, Tristram H. Snelling1,2, Ryan Chau1,2, Meena Shanmuganathan1, Jane K. Cleal1,Kirsten R. Poore1, David E. Noakes3, Lucilla Poston4, Mark A. Hanson1 and Lucy R. Green1

1Institute of Developmental Sciences, School of Medicine, University of Southampton, Southampton SO16 6YD, UK2Wessex Cardiothoracic Unit, Southampton General Hospital, Southampton SO16 6YD, UK3Department of Veterinary Reproduction, Royal Veterinary College, London NE2 56H, UK4Division of Reproduction & Endocrinology, King’s College London, London SE1 7EH, UK

The nutritional environment during development and even prior to conception may contributeto cardiovascular risk. In mature adult female sheep, we investigated the effect of preconceptionaland periconceptional maternal nutritional restriction on the vascular reactivity of arteries fromfour vascular beds supplying the heart, thorax, kidney and hindlimb. Welsh Mountain ewesreceived 100% of nutrient requirements throughout gestation (control group, C, n = 18), or 50%of nutrient requirements for 30 days prior to conception (preconceptional group, PRE, n = 20) orfor 15 days either side of conception (periconceptional group, PERI, n = 31) and 100% thereafter.In 3.5-year-old female offspring, the left anterior descending coronary (LAD), left internalthoracic (LITA), right renal and second and third order femoral arteries were dissected and theirreactivity was assessed by organ bath or wire myography. Vasoconstrictor responses were greaterin both LAD and LITA from PERI offspring compared with C (P < 0.01), while vasoconstrictionwas unaffected by maternal diet in arteries from the renal and femoral circulations (P = n.s.).Endothelium-dependent and -independent vasodilatation was attenuated in third order femoralarteries of PRE and PERI groups compared with C (P < 0.05). Endothelium-independentvasodilatation was attenuated in both the LAD and renal arteries in the PERI group comparedwith C (P < 0.05). These data show that moderate maternal undernutrition either prior toor around conception affects vascular function of adult offspring. The effect depends on thetiming of the insult, but also on the vascular bed studied and vessel hierarchy in the vasculartree.

(Received 13 March 2009; accepted after revision 26 June 2009; first published online 26 June 2009)Corresponding author C. Torrens: Vascular Biology Group, Institute of Developmental Sciences (MP 887), Universityof Southampton, Tremona Road, Southampton S016 6YD, UK. Email: c.torrens@southampton.ac.uk

There is clear evidence that the early environment plays animportant role in the risk of cardiovascular and metabolicdisease in later life (Eriksson et al. 2001). In particular,the very early period of development appears to beimportant in determining cardiovascular risk. Evidencefrom the ‘Dutch Hunger Winter’ cohort shows thatindividuals exposed to the famine in early gestation hadan increased risk of coronary heart disease in later life,but this was not seen in individuals exposed in mid-or late gestation (Roseboom et al. 2000). In rodents,protein restriction in the first week of gestation and evenprior to conception led to raised blood pressure in adult

offspring (Kwong et al. 2000; Erhuma et al. 2007; Watkinset al. 2008), while in sheep, we have demonstrated thata 50% reduction in maternal food intake during the first31 days of pregnancy (term 147 days) led to cardiovascularand renal abnormalities in the adult offspring (Clealet al. 2007). Although nutrition during pregnancy isclearly important, there is growing evidence of theimportance of diet and/or body composition prior topregnancy and subsequent development of cardiovascularrisk. In sheep, periconceptional nutrient restriction (from!60 to 7 or 30 days gestational age) leads to raisedfetal blood pressure in late-gestation sheep (Harding

DOI: 10.1113/expphysiol.2009.047340 C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

Exp Physiol 94.9 pp 1024–1033 Early undernutrition alters vascular function 1025

& Johnston, 1995; Edwards & McMillen, 2002), whilematernal slimness was associated with reduced bloodpressure in the offspring (Gopalakrishnan et al. 2004). Inhumans too, prepregnancy maternal slimness is associatedwith altered liver blood flow in the fetus (Haugen et al.2005).

Peripheral vascular tone is important in regulatingblood pressure and the distribution of blood flow.Altered vascular reactivity is associated with metabolicand cardiovascular disease (Rizzoni et al. 2001; Taddeiet al. 2001; Landmesser et al. 2004). In rats, maternalundernutrition in pregnancy is associated with raisedblood pressure and endothelial dysfunction in theresistance arteries of the offspring (Brawley et al. 2003;Franco et al. 2004; Torrens et al. 2006), although this isnot a uniform response occurring in the whole vascularsystem, which suggests the involvement of multiplepathways dependent on the vascular bed in question(Torrens et al. 2003, 2009). In sheep, undernutrition inthe first half of gestation induces peripheral vasculardysfunction in femoral arteries of the mid- and late-gestation fetus (Ozaki et al. 2000; Nishina et al. 2003),and undernutrition in early gestation alters coronary andthoracic vascular function at 2.5 years of age (Khan et al.2005; Cleal et al. 2007).

To explore the importance of the very earlyenvironment, we investigated a 50% reduction in totalnutrient requirements for either 30 days before or 15 dayseither side of conception on vascular function in theoffspring at 3.5 years of age. Since the vascular tree isnot homogeneous, we have sought to investigate fourdifferent vascular beds, all of which are important inhuman cardiovascular disease or bypass surgery (Ganseraet al. 2007) and have been shown to be sensitive toperturbations in maternal nutrition (Ozaki et al. 2000;Nishina et al. 2003; Khan et al. 2005; Cleal et al. 2007).Furthermore, we have also assessed conduit and resistancearteries from the femoral bed, which are important inbrain-sparing redistribution of blood flow in response tofetal hypoxia (Giussani et al. 1993) and as a musculararterial bed are important in metabolic control.

Methods

Animal procedures were performed in accordance withthe UK Home Office Animals (Scientific Procedures) Act,1986.

Dietary protocol

Welsh Mountain ewes in their second or third parityreceived 100% (C, n = 18), or 50% of total nutrientrequirements from –30 to 0 days gestational age (PRE,n = 20) or from !15 to 15 days gestational age (PERI,n = 31) and then 100% thereafter (Agricultural & Food

Research Council, 1993). The pelleted diet consistedof barley, wheat, micronized full fat soya, grass meal,molasses, chopped straw, calcium carbonate, dicalciumphosphate salt and sheep vitamin/mineral supplement(GJW Titmus, St Albans, UK) and provided 9.6 MJ kg!1

(metabolizable energy) and 14.75% crude protein. At70 days gestational age, a subgroup of ewes from eachgroup (n = 5–11) were killed by an overdose of barbiturate(Pentobarbitone sodium, 160 mg kg!1 I.V.; AnimalcareLtd., York, UK) for fetal biometry of male and femalefetuses (C, male singleton n = 3, male twin n = 2, femalesingleton n = 3 and female twin n = 2; PRE, male singletonn = 3 and female singleton n = 2; PERI, male singletonn = 5, male twin n = 2 and female singleton, n = 4). Atterm, ewes were allowed to deliver and suckle their lambsnaturally, and the feed ration was adjusted to provide 100%nutrient requirements as previously described (Poore et al.2007).

Postweaning, all lambs received 100% of nutritionalrequirements with a diet consisting of free access to waterand hay, and creep pellets (Prestige Lamb Pellets + Decox;BOCM Pauls Ltd, Loughborough, UK) were provided eachmorning and afternoon. As fed, creep pellets provided10.51 MJ kg!1 (metabolizable energy) and 18% crudeprotein. At approximately 32 weeks of age, lambs weretransferred onto a standard ration of an adult completepelleted diet (Ewbol 18; BOCM Pauls Ltd) plus adlibitum hay. At 3.5 years of age, fat and muscle depthsin the third lumbar region were measured by ultrasound(Aloka SSD 210 DX11; BCF Technology, Livingstone,UK) with a 7.5 MHz linear array transducer. To ensuresimilar levels of circulating progesterone between thegroups, offspring were injected with progesterone for7 days (10 mg day!1 I.M., Gestone 50 mg ml!1; NordicPharmaceuticals Ltd, Reading, UK). Progesterone assay(Ridgeway Science, St Briavels, UK) showed equivalentplasma progesterone levels, although they were slightlyelevated in PRE compared with C (in ng ml!1: C,3.30 ± 0.25, n = 8; PRE, 4.59 ± 0.32, n = 12; and PERI,4.02 ± 0.32, n = 12; P < 0.05). After 7 days, offspringwere killed with an overdose of sodium pentobarbitone(0.8 ml kg!1 I.V., 200 mg ml!1; Animalcare Ltd, York, UK).The offspring used consisted of both singletons and twins,although no twin pairs were used. The breakdown offemale offspring in each group was as follows: PERI, n = 11(7 singleton and 4 twins); PRE, n = 12 (8 singleton and4 twins) and C, n = 8 (3 singleton and 5 twins).

At postmortem examination, the weight, crown–rumplength and abdominal circumference were recorded,and major organs dissected and weighed. For vascularreactivity studies, the left anterior descending coronaryartery (LAD), the left internal thoracic artery (LITA), theright renal artery (RA) and second order (conduit) andthird order femoral arteries (resistance, 342 ± 8.99 µmdiameter) were dissected.

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

1026 C. Torrens and others Exp Physiol 94.9 pp 1024–1033

Table 1. Placental weight and fetal body weight, biometry and organ weights of male and female fetusesat 70 days gestational age

Control (n = 10) PERI (n = 11) PRE (n = 5)

Fetal weight (g) 159.9 ± 5.4 180.7 ± 15.3 169.5 ± 6.8Total placental weight (g) 473.0 ± 48.8 507.3 ± 36.8 431.4 ± 46.9Fetal:placental weight ratio 2.9 ± 0.2 3.0 ± 0.2 2.6 ± 0.2Crown–rump length (mm) 165.2 ± 2.1 171.8 ± 4.8 159.6 ± 2.8Biparietal diameter (mm) 27.5 ± 0.5 28.9 ± 1.0 26.9 ± 1.0Femur length (mm) 34.3 ± 2.4 36.4 ± 1.6 29.1 ± 1.3Abdominal circumference (mm) 124.4 ± 2.9 129.8 ± 4.8 130.4 ± 2.0

Organ weight (% body weight)Brain 0.028 ± 0.001 0.029 ± 0.001 0.028 ± 0.001Heart 0.009 ± 0.001 0.010 ± 0.001 0.009 ± 0.001Lung 0.049 ± 0.002 0.051 ± 0.002 0.048 ± 0.002Liver 0.065 ± 0.002 0.065 ± 0.003 0.068 ± 0.002Total kidney 0.005 ± 0.001 0.005 ± 0.001 0.005 ± 0.001Total adrenal 0.0002 ± 0.00002 0.0002 ± 0.00004 0.0002 ± 0.00002

Vascular function

Dissected arteries were placed in ice-cold (#4$C)physiological salt solution (PSS, mM: 119 NaCl, 4.7 KCl,2.5 CaCl, 1.17 MgSO4, 25 NaHCO3, 1.18 KH2PO4,0.026 EDTA and 5.5 D-glucose), cleaned of connectivetissue and mounted on a wire myograph (Danish MyoTechnology A/S, Denmark), except the RA, which wasmounted in an organ bath (Linton Systems, Lutterworth,UK). The segments were bathed in PSS at 37$C andgassed continuously with 95% O2 and 5% CO2. All vesselswere normalized as previously described (Simonsen et al.1993; Docherty et al. 2001; Molnar et al. 2002), and thefunctional integrity of the smooth muscle was testedwith 125 mM KPSS (PSS solution with an equimolarsubstitution of KCl for NaCl). Arterial segments whichfailed to produce an active pressure of 13.3 kPa in responseto the KPSS solution were discarded. In viable arterialsegments, cumulative concentration–response curves(CRCs) were constructed for a range of agonists, whichare summarized in Table 3. All drugs and chemicals wereobtained from Sigma (Poole, UK), except endothelin-1(ET1; Bachem, St Helens, UK).

Calculations and statistical analysis

All data are expressed as means ± S.E.M. Growth wascalculated as the change in weight between birth and12 weeks of age. The weights of left and right kidneyand adrenal were not different and are expressed as amean adrenal or kidney weight. Constrictor responseswere calculated as the percentage of maximal constrictioninduced by 125 mM KPSS and relaxant responses as thepercentage inhibition of agonist-induced constriction.The resultant CRCs were fitted to a four-parameter logisticequation using non-linear regression, and pEC50 (theeffective concentration producing 50% of the maximalresponse) and maximal response values were derived.All data were analysed by one-way analysis of variance

(ANOVA) with Bonferroni post hoc correction (Prism 3.0,GraphPad Software Inc., San Diego, CA, USA). If the fit ofthe curve produced by non-linear regression was poor orcurves were not sigmoidal, calculation of the pEC50 wasnot made. Significance was accepted when P < 0.05.

Results

Offspring body and organ size

At 70 days gestational age, fetal weight, biometry andorgan weights were similar in all dietary groups (Table 1).Gestation length was unaffected by the dietary challenge(C, 145.0 ± 0.5 days, n = 8; PRE, 145.1 ± 0.9 days, n = 12;and PERI, 146.6 ± 0.3 days, n = 17; P = n.s.). At birth,weight (C, 3.78 ± 0.19 kg, n = 8; PRE, 3.33 ± 0.16 kg,n = 14; and PERI, 3.76 ± 0.16 kg, n = 21; P = n.s.) andcrown–rump length (C, 38.6 ± 0.8 mm, n = 8; PRE,40.3 ± 0.7 mm, n = 14; and PERI, 40.7 ± 0.6 mm, n = 21;P = n.s.) were similar in all dietary groups. Between birthand 12 weeks (weaning), the PRE group gained moreweight than both C and PERI groups (C, 17.6 ± 1.4 kg,n = 8; PRE, 21.1 ± 1.0 kg, n = 12; and PERI, 18.7 ± 0.7 kg,n = 11; P < 0.05). At 3.5 years, the body weight, biometryand organ weights were similar in all dietary groups(Table 2).

Vasoreactivity

A summary of all vascular responses is given inTable 3.

Left descending coronary artery

Vasoconstriction. Acetylcholine produced a concen-tration-dependent vasoconstriction that was significantlyenhanced in both the PRE and PERI groups comparedwith C (P < 0.01; Fig. 1A). Constriction in response to thethromboxane mimetic U46619 was significantly greaterin the PERI group compared with both PRE and C

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

Exp Physiol 94.9 pp 1024–1033 Early undernutrition alters vascular function 1027

Table 2. Biometry and organ weights of adult female offspring at 3.5 years of age

Control (n = 8) PERI (n = 12) PRE (n = 12)

Weight (kg) 65.25 ± 3.43 69.12 ± 1.79 65.97 ± 2.10Body condition score 3.19 ± 0.17 3.381 ± 0.1354 3.083 ± 0.1556Fat depth (mm) 6.13 ± 0.52 6.048 ± 0.3810 5.333 ± 0.4748Muscle depth (mm) 30.25 ± 1.00 31.10 ± 1.110 29.40 ± 0.8663Crown–rump length (cm) 103.9 ± 2.5 113.2 ± 1.9 110.1 ± 2.2Biparietal diameter (cm) 9.7 ± 0.7 11.1 ± 0.3 11.2 ± 0.4Shoulder height (cm) 78.1 ± 2.8 74.8 ± 0.8 77.5 ± 0.9Femur length (cm) 31.2 ± 1.0 30.9 ± 0.3 31.3 ± 0.7Abdominal circumference (cm) 117.5 ± 3.8 115.9 ± 1.4 118.0 ± 2.5

Organ weight (% body weight)Brain 0.115 ± 0.010 0.112 ± 0.004 0.119 ± 0.004Heart 0.367 ± 0.020 0.382 ± 0.013 0.394 ± 0.018Lung 0.817 ± 0.084 0.644 ± 0.037 0.643 ± 0.040Liver 1.046 ± 0.098 0.938 ± 0.021 0.970 ± 0.0328Total kidney 0.089 ± 0.003 0.087 ± 0.003 0.101 ± 0.007Total adrenal 0.0038 ± 0.0002 0.0034 ± 0.0002 0.0039 ± 0.0002

groups (P < 0.01; Fig. 1B). Constriction in response toangiotensin II (Ang II) was poor and similar in all thedietary groups (% max response: C, 14.5 ± 5.1%, n = 6;PRE, 13.7 ± 1.8%, n = 10; and PERI, 13.8 ± 2.7%, n = 11;P = n.s.).

Endothelium-dependent vasodilatation. The endothe-lium-dependent vasodilator bradykinin (BK) produceda concentration-dependent vasodilatation which wassimilar in sensitivity (pEC50: C, 9.28 ± 0.11, n = 7;PRE, 8.83 ± 0.15, n = 10; and PERI, 9.11 ± 0.03, n = 12;P = n.s.) and maximal response between all groups (%max response: C, 85.7 ± 3.5%, n = 7; PRE, 86.2 ± 3.8%,n = 10; and PERI, 74.1 ± 2.6%, n = 11; P = n.s.).

Endothelium-independent vasodilatation. Maximal va-sodilatation in response to isoprenaline (ISO) wassignificantly attenuated in both the PRE andPERI compared with C group (P < 0.05; Fig. 1C).Vasodilatation in response to the nitric oxide donorsodium nitroprusside (SNP) was similar in sensitivity(pEC50: C, 7.17 ± 0.04, n = 7; PRE, 6.99 ± 0.04, n = 10;and PERI, 7.25 ± 0.03, n = 12; P = n.s.) and maximalresponse (% max response: C, 97.0 ± 1.1%, n = 7; PRE,93.4 ± 1.4%, n = 10; and PERI, 96.8 ± 0.9%, n = 12;P = n.s.) between the groups.

Left internal thoracic artery

Vasoconstriction. Maximal constriction in response tophenylephrine (PE; P < 0.05; Fig. 2A) and sensitivity toET1 was enhanced in PERI compared with PRE, but notC, group (pEC50: C, 8.53 ± 0.07, n = 7; PRE, 8.33 ± 0.08,n = 10; and PERI, 8.74 ± 0.07, n = 11; P < 0.05 PERIversus PRE). Maximal constriction in response to Ang IIwas significantly enhanced in the PERI group compared

with PRE, and in both compared with C (P < 0.01;Fig. 2B).

Endothelium-dependent vasodilatation. Vasodilatationin response to BK was similar in sensitivity (pEC50: C,9.25 ± 0.05, n = 7; PRE, 9.18 ± 0.04, n = 10; and PERI,9.53 ± 0.13, n = 11; P = n.s.) and maximal response (%max response: C, 66.6 ± 4.8%, n = 7; PRE, 67.8 ± 4.5%,n = 10; and PERI, 62.3 ± 5.2%, n = 11; P = n.s.) in alldietary groups.

Endothelium-independent vasodilatation. Vasodilata-tion in response to isoprenaline was poor in comparisonwith BK, and was significantly attenuated in both the PERIand PRE groups compared with C (P < 0.01; Fig. 2C).Dilatation in response to SNP was similar in all dietarygroups in both sensitivity (pEC50: C, 7.17 ± 0.04, n = 7;PRE, 6.99 ± 0.04, n = 10; and PERI, 7.25 ± 0.03, n = 11;P = n.s.) and maximal response (% max response: C,86.7 ± 6.4%, n = 7; PRE, 92.0 ± 2.0%, n = 10; and PERI,88.4 ± 5.3%, n = 11; P = n.s.).

Second order femoral artery

Vasoconstriction. Sensitivity to both PE (P < 0.05;Fig. 3A) and ET1 (P < 0.05; Fig. 3B) was enhanced inPERI compared with PRE. Responses to U46619 were notsigmoidal, but maximal constriction was also enhancedin the PERI group compared with the PRE group (%max response to U46619: C, 42.9 ± 6.6%, n = 7; PRE,34.8 ± 8.0%, n = 12; and PERI, 67.0 ± 10.5%, n = 11;P < 0.05 PERI versus PRE).

Endothelium-dependent vasodilatation. Vasodilatationin response to ACh (pEC50: C, 7.57 ± 0.20, n = 8;PRE, 7.56 ± 0.14, n = 12; and PERI, 7.01 ± 0.19, n = 11;

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

1028C

.Torrensand

othersExp

Physiol94.9pp

1024–1033

Table 3. Summary of vascular responses in each vascular bed

PE U46619 Ang II Endothelin ACh BK Isoprenaline SNP

LAD Dose range — 0.1 nM–1 µM 10 pM–1 µM — 0.1 nM–10 µM 10 fM–10 µM 1 nM–10 µM 0.1 nM–10 µM

Preconstric — n.a. n.a. — n.a. U46619 U46619 U46619PRE — % % — &M % 'M %PERI — &M % — &M % 'M %

LITA Dose range 10 nM–100 µM — 10 pM–0.3 µM 1 pM–10 nM — 10 fM–10 µM 1 nM–10 µM 0.1 nM–10 µM

Preconstrict n.a. — n.a. n.a. — ET ET ETPRE '(M — &M '(E — % 'M %PERI &(M — &M &(E — % 'M %

2$ Femoral Dose range 1 nM–300 µM 0.1 nM–10 µM — 1 pM–10 nM 0.1 nM–10 µM 10 fM–10 µM 1 nM–10 µM 0.1 nM–10 µM

Preconstrict n.a. n.a. — n.a. PE PE PE PEPRE '(E '(M — '(E % % % %PERI &(E &(M — &(E % % 'M %

3$ Femoral Dose range 1 nM–300 µM 0.1 nM–10 µM — — 0.1 nM–10 µM 10 fM–10 µM 1 nM–10 µM 0.1 nM–10 µM

Preconstrict n.a. n.a. — — PE PE PE PEPRE % % — — 'E % 'M 'E

PERI % % — — 'E % % 'E

Renal Dose range 10 nM–100 µM 10 pM–1 µM — — 1 µM — 1 µM 0.1 nM–10 µM

Preconstrict n.a. n.a. — — PE — PE PEPRE '(E % — — % — % %PERI &(E % — — % — % 'E

% indicates no difference compared with control; & and ' indicate an enhanced and decreased response relative to control, respectively. Changes in pEC50 are signified byan E, changes in maximal response by an M. In all cases, the concentration for preconstriction was equivalent to pEC80, i.e. the concentration producing 80% of the maximalresponse. Arrows which are marked with an asterisk (e.g. '() indicate a change relative to the other experimental group. n.a., not applicable.

C "2009

TheA

uthors.Journalcompilation

C "2009

ThePhysiologicalSociety

Exp Physiol 94.9 pp 1024–1033 Early undernutrition alters vascular function 1029

Figure 1. Responses in left descending coronary arteries to ACh (A), U46619 (B) and isoprenaline (C) inC ($, n = 7), PRE (•, n = 11) or PERI group adult offspring (!, n = 12)(P < 0.05, ((P < 0.01 PERI and PRE versus C; ††P < 0.01 PERI versus PRE and C.

P = n.s.; and % max response: C, 32.7 ± 4.7%, n = 8;PRE, 31.8 ± 3.3%, n = 12; and PERI, 37.3 ± 4.8%, n = 11;P = n.s.) and BK (pEC50: C, 8.92 ± 0.33, n = 8; PRE,8.50 ± 0.24, n = 12; and PERI, 8.32 ± 0.49, n = 11;P = n.s.; and % max response: C, 40.0 ± 4.3%, n = 8;PRE, 40.7 ± 3.3%, n = 12; and PERI, 48.1 ± 8.0%, n = 11;P = n.s.) was similarly modest in all groups.

Figure 2. Responses in internal thoracic arteries to PE (A), Ang II (B) and isoprenaline (C) in C ($, n = 7),PRE (•, n = 11) or PERI group adult offspring (!, n = 12)(P < 0.05 percentage maximal response PERI versus PRE; ((P < 0.01 percentage maximal response PERI versusPRE and C; ††P < 0.01 percentage maximal response PRE versus C.

Figure 3. Responses in second order femoral arteries to PE (A), ET1 (B) and isoprenaline (C) in C($, n = 8), PRE (•, n = 12) or PERI group adult offspring (!, n = 12)†P < 0.05 PRE versus PERI; ((P < 0.01 PERI versus PRE and C.

Endothelium-independent vasodilatation. Isoprenaline-induced vasodilatation was poor and was significantlyattenuated in the PERI group compared with the PREand C groups (P < 0.01; Fig. 3C). Sodium nitroprusside-induced vasodilatation was similar in all dietary groupsboth in sensitivity (pEC50: C, 7.90 ± 0.20, n = 7; PRE,8.27 ± 0.14, n = 12; and PERI, 7.38 ± 0.24, n = 11;

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

1030 C. Torrens and others Exp Physiol 94.9 pp 1024–1033

Figure 4. Responses in third order femoral arteries to ACh (A), isoprenaline (B) and SNP (C) in C($, n = 6–7), PRE (•, n = 7–9) or PERI group adult offspring (!, n = 6–8)((P < 0.01, (((P < 0.001 pEC50 PERI versus C; †P < 0.05 pEC50 PRE versus C; ‡P < 0.05 percentage maximalresponse PRE versus C.

P = n.s.) and maximal response (% max response: C,41.1 ± 2.3%, n = 8; PRE, 38.2 ± 3.3%, n = 9; and PERI,48.0 ± 6.3%, n = 11; P = n.s.).

Third order femoral artery

Vasoconstriction. Vasoconstriction in response to PE wassimilar between the groups (pEC50: C, 4.89 ± 0.12, n = 8;PRE, 4.62 ± 0.13, n = 11; and PERI, 4.73 ± 0.11, n = 11;P = n.s.; and % max response: C, 55.4 ± 4.8%, n = 8;PRE, 60.3 ± 6.3%, n = 11; and PERI, 55.8 ± 5.0%, n = 11;P = n.s.). U46619 produced a non-sigmoidal responsethat was modest and not different between the groups (%max response: C, 31.0 ± 13.7%, n = 5; PRE, 33.3 ± 11.8%,n = 5; and PERI, 37.6 ± 12.3%, n = 6; P = n.s.).

Endothelium-dependent vasodilatation. Vasodilatationin response to ACh was significantly reduced in sensitivityin PRE (P < 0.05) and PERI compared with C (P < 0.001;Fig. 4A). Sensitivity to BK was similar between the groups(pEC50: C, 9.95 ± 0.40, n = 5; PRE, 8.83 ± 0.91, n = 9; andPERI, 8.50 ± 1.44, n = 6; P = n.s.).

Endothelium-independent vasodilatation. Isoprenaline-induced vasodilatation was significantly attenuatedin the PRE group compared with C (P < 0.05;Fig. 4B). Sensitivity to SNP-induced vasodilatation was

Figure 5. Responses in renal artery toPE (A) and SNP (B) in C ($, n = 7), PRE(•, n = 11) or PERI group adultoffspring (!, n = 12)(P < 0.05 pEC50 PERI versus PRE and C;†P < 0.05 pEC50 PRE versus PERI.

significantly reduced in the PERI group compared with C(P < 0.01; Fig. 4C).

Renal artery

Vasoconstriction. Sensitivity to PE-induced vaso-constriction was significantly enhanced in the PERIgroup compared with the PRE group (P < 0.05; Fig. 5A).Vasoconstriction in response to Ang II was poor andsimilar in all dietary groups (% max response: C,4.5 ± 2.5%, n = 8; PRE, 3.3 ± 3.2%, n = 11; and PERI,6.4 ± 3.5%, n = 11; P = n.s.).

Endothelium-dependent vasodilatation. A single 1 µM

dose of ACh failed to produce a relaxation in alldietary groups (% relaxation: C, 5.2 ± 2.7%, n = 8;PRE, 3.5 ± 0.9%, n = 12; and PERI, 5.4 ± 2.2%, n = 11;P = n.s.).

Endothelium-independent vasodilatation. A single 1 µM

dose of isoprenaline failed to produce a relaxation (%relaxation: C, 3.0 ± 1.3%, n = 8; PRE, 1.7 ± 1.1%, n = 12;and PERI, 3.7 ± 1.6%, n = 11; P = n.s.). Sensitivity toSNP-induced vasodilatation was significantly reduced inthe PERI group compared with the PRE and controlgroups (P < 0.05; Fig. 5B).

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

Exp Physiol 94.9 pp 1024–1033 Early undernutrition alters vascular function 1031

Discussion

The early life nutrient environment is implicated inthe aetiology of cardiovascular and metabolic diseasein later life (Eriksson et al. 2001). Peripheral vasculartone plays an important role in the maintenance ofblood pressure and the distribution of blood flow, whilevascular dysfunction, and endothelial dysfunction inparticular, are now recognized as common features ofcardiovascular disease as well as type II diabetes and themetabolic syndrome (Rizzoni et al. 2001; Taddei et al. 2001;Landmesser et al. 2004). This study shows for the firsttime that alterations in the vascular function of matureadult female offspring are induced by moderate, acutedietary restriction around or even prior to conception.Importantly, this study demonstrates that the effectobserved in the offspring is dependent on both the timeof the challenge and the vascular bed.

In the present study we report that bothpre- and periconceptional undernutrition enhancedvasoconstriction of the coronary arteries and reducedendothelium-dependent vasodilatation in the femoralresistance arteries in female adult offspring. This findingconfirms other previous studies which have shown alteredvascular function in response to unbalanced nutritionin pregnancy (Brawley et al. 2003; Franco et al. 2004;Torrens et al. 2003, 2006, 2009; Cleal et al. 2007; Watkinset al. 2008). What is distinct about the present studyis that we have targeted our challenge to very earlystages of development and even prior to conception.The periconceptiual challenge (from !15 to 15 daysgestational age) ran from before conception to the endof the pre-implantation period (#18 days gestationalage in the sheep), a period incorporating implantationand organogensis. This early period of implantation andorganogenesis has previously been demonstrated to besusceptible to changes in maternal food intake, thoughit produced changes in adult vascular function in theabsence of gross morphological changes in fetal biometry(70 days gestational age) or birth weight. In humans,exposure to famine in the first trimester increased therisk of coronary heart disease (Roseboom et al. 2000)and was associated with vascular remodelling in later life(Painter et al. 2007), a known risk factor for cardiovasculardisease (Mulvany, 1995). In rats, protein restriction duringthe first week of gestation or pre-implantation periodalone leads to raised blood pressure in the offspring(Kwong et al. 2000; Erhuma et al. 2007). In sheep, anutritional challenge of similar intensity to that reportedin the present study given between 1 and 31 days ofgestation led to enhanced vasoconstriction and attenuatedvasodilatation in the LAD, LITA and renal arteries (Khanet al. 2005; Cleal et al. 2007). A possible mechanism forthe vascular changes reported may involve glucocorticoids.In sheep, the administration of exogenous glucocorticoid

between 26 and 28 days gestational age leads to a decreasednephron number, raised blood pressure and alteredvascular function in the offspring, including similar effectson coronary arteries to those reported in the present study(Dodic et al. 1998; Wintour et al. 2003; Roghair et al.2005; Segar et al. 2006). This effect could be important,since maternal food restriction in pregnancy is knownto raise maternal plasma cortisol levels (Nishina et al.2003).

The same may also be true for the preconceptionalchallenge. Interestingly, despite nutritional restrictionbeing stopped before conception, the PRE group alsodemonstrated altered vascular responses compared withthe control group. The restriction of nutrition prior toconception alone in the present study led to alteredvascular function and increased postnatal growth in thefirst 12 postnatal weeks. Accelerated growth in infancyis linked to endothelial dysfunction in low-birthweightchildren (Singhal & Lucas, 2004). Maternal food intakeand body composition prior to pregnancy have beenshown to impact on fetal blood flow (Haugen et al.2005). In mice, we have demonstrated that maternalprotein restriction prior to conception alone leads toaltered vascular function in the offspring (Watkins et al.2008); an observation we extend to the sheep in thepresent study, where we show altered responses to bothACh and isoprenaline in a number of vascular bedsfollowing purely preconceptional undernutrition. Oneof the most striking observations from the presentstudy is the heterogeneity with which the vascularsystem responds to either challenge (summarized inTable 3).

In the rat, similar models of developmentalprogramming produce endothelial dysfunction in smallmesenteric arteries and have been linked to altered nitricoxide signalling and oxidative stress (Franco et al. 2004;Torrens et al. 2006, 2009). In the present study, endothelialdysfunction was also observed, but this was confinedto the small femoral arteries following either pre- orpericonceptional undernutrition and not conduit femoralarteries. In the sheep fetus, increased femoral resistanceis a brain-sparing adaptation to hypoxia (Giussani et al.1993), while attenuated dilatation of these arteries isseen in the midgestation fetus in response to maternalundernutrition (Ozaki et al. 2000; Nishina et al. 2003).Since the larger conduit arteries have a limited capacity toregulate blood flow distribution, this suggests a possiblevascular bed-specific adaptation aimed at redistributionaway from the periphery, and one which may havedetrimental consequences for glucose handling in theadult. Interestingly the femoral bed was the only one wherea decrease in endothelial function was reported. As this wasobserved only in the small resistance arteries and not theconduit vessels this may suggest that the small arteries aremore susceptible to endothelial damage than the larger

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

1032 C. Torrens and others Exp Physiol 94.9 pp 1024–1033

conduit arteries. Similar differences in responses betweenconduit and resistance arteries have been observedin the offspring of protein-restricted rat dams, wheredysfunction in the mesenteric arteries is not seen in thethoracic aorta (Torrens et al. 2003). These data suggest thateach artery responds to the challenge in a different way,rather than a global effect applied to the whole vascularsystem.

In three of the four beds studied, including the LADand LITA, altered vasoconstriction was observed. Wehave previously reported enhanced constriction in theLAD and LITA following a 50% nutrient restrictionfor first 31 days of gestation, which was associated withan increased myosin light chain kinase expression inthe vascular smooth muscle (Khan et al. 2005; Clealet al. 2007). The coronary arteries have some uniqueproperties in comparison with other beds. Two importantpoints are that they produce a vasoconstriction inresponse to ACh, rather than an endothelium-dependentvasodilatation (Simonsen et al. 1993), and they possesan 11!-hydroxysterone dehydrogenase isoenzyme, whichreactivates glucocorticoids (Hatakeyama et al. 2000).Indeed, this has been proposed as a mechanism forthe similarly enhanced coronary constriction in theoffspring of sheep following an early gestation exposureto dexamethasone (Roghair et al. 2005; Segar et al. 2006).

The most uniform response was the attenuation ofthe !-adrenoceptor-mediated vasodilatation, which wasnoted in all vascular beds except the RA, where it haspreviously been shown to be ineffective (Docherty et al.2001). Dilatation via !-adrenoceptors is an importantmechanism to increase perfusion in metabolically activetissues such as cardiac and skeletal muscle. Although itis mainly thought to act through activation of adenylatecyclase, responses to isoprenaline have been shown to bemodulated by levels of nitric oxide and cross-talk betweenother cyclic nucleotides, particularly cGMP (Delpy et al.1996). At present, the exact mechanism is unclear,although the attenuated responses to SNP in the renaland third order femoral arteries may suggest a defect inthe guanylate cyclase pathway.

In conclusion, this study supports the importanceof nutrition in very early development, even priorto conception, in determining vascular function inadulthood. Our data highlight the heterogeneity acrossdifferent vascular beds in response to early undernutrition,suggestive of a co-ordinated response at the vascular level.These data suggest that the use of several vascular beds isessential for gaining a complete picture of the effects ofsuch modest changes in nutrition. Overall, our findingsshow important implications for maternal prepregnancyand periconceptional nutritional status in determiningthe risk of cardiovascular disease, and may informfuture dietary advice given to women of child-bearingage.

References

Agricultural & Food Research Council (1993). Energy andProtein Requirements of Ruminants. An Advisory ManualPrepared by the AFRC Technical Committee on Responses toNutrients. CAB International, Wallingford, UK.

Brawley L, Itoh S, Torrens C, Barker A, Bertram C, Poston L &Hanson M (2003). Dietary protein restriction in pregnancyinduces hypertension and vascular defects in rat maleoffspring. Pediatr Res 54, 83–90.

Cleal JK, Poore KR, Boullin JP, Khan O, Chau R, Hambidge O,Torrens C, Newman JP, Poston L, Noakes DE, Hanson MA &Green LR (2007). Mismatched pre- and postnatal nutritionleads to cardiovascular dysfunction and altered renalfunction in adulthood. Proc Natl Acad Sci USA 104,9529–9533.

Delpy E, Coste H & Gouville AC (1996). Effects of cyclic GMPelevation on isoprenaline-induced increase in cyclic AMPand relaxation in rat aortic smooth muscle: role ofphosphodiesterase 3. Br J Pharmacol 119, 471–478.

Docherty CC, Kalmar-Nagy J, Engelen M & Nathanielsz PW(2001). Development of fetal vascular responses toendothelin-1 and acetylcholine in the sheep. Am J PhysiolRegul Integr Comp Physiol 280, R554–R562.

Dodic M, May CN, Wintour EM & Coghlan JP (1998). An earlyprenatal exposure to excess glucocorticoid leads tohypertensive offspring in sheep. Clin Sci (Lond) 94, 149–155.

Edwards LJ & McMillen IC (2002). Periconceptional nutritionprograms development of the cardiovascular system in thefetal sheep. Am J Physiol Regul Integr Comp Physiol 283,R669–R679.

Erhuma A, Salter AM, Sculley DV, Langley-Evans SC &Bennett AJ (2007). Prenatal exposure to a low-protein dietprograms disordered regulation of lipid metabolism in theaging rat. Am J Physiol Endocrinol Metab 292, E1702–E1714.

Eriksson JG, Forsen T, Tuomilehto J, Osmond C & Barker DJ(2001). Early growth and coronary heart disease in later life:longitudinal study. BMJ 322, 949–953.

Franco MC, Fortes ZB, Akamine EH, Kawamoto EM, ScavoneC, de Britto LR, Muscara MN, Teixeira SA, Tostes RC,Carvalho MH & Nigro D (2004). Tetrahydrobiopterinimproves endothelial dysfunction and vascular oxidativestress in microvessels of intrauterine undernourished rats. JPhysiol 558, 239–248.

Gansera B, Schmidtler F, Angelis I, Kiask T, Kemkes BM &Botzenhardt F (2007). Patency of internal thoracic arterycompared to vein grafts – postoperative angiographicfindings in 1189 symptomatic patients in 12 years. ThoracCardiovasc Surg 55, 412–417.

Giussani DA, Spencer JA, Moore PJ, Bennet L & Hanson MA(1993). Afferent and efferent components of thecardiovascular reflex responses to acute hypoxia in term fetalsheep. J Physiol 461, 431–449.

Gopalakrishnan GS, Gardner DS, Rhind SM, Rae MT, Kyle CE,Brooks AN, Walker RM, Ramsay MM, Keisler DH,Stephenson T & Symonds ME (2004). Programming of adultcardiovascular function after early maternal undernutritionin sheep. Am J Physiol Regul Integr Comp Physiol 287,R12–R20.

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society

Exp Physiol 94.9 pp 1024–1033 Early undernutrition alters vascular function 1033

Harding JE & Johnston BM (1995). Nutrition and fetal growth.Reprod Fertil Dev 7, 539–547.

Hatakeyama H, Inaba S, Takeda R & Miyamori I (2000).11!-Hydroxysteroid dehydrogenase in human vascular cells.Kidney Int 57, 1352–1357.

Haugen G, Hanson M, Kiserud T, Crozier S, Inskip H &Godfrey KM (2005). Fetal liver-sparing cardiovascularadaptations linked to mother’s slimness and diet. Circ Res 96,12–14.

Khan OA, Torrens C, Noakes DE, Poston L, Hanson MA, GreenLR & Ohri SK (2005). Effects of pre-natal and earlypost-natal undernutrition on adult internal thoracic arteryfunction. Eur J Cardiothorac Surg 28, 811–815.

Kwong WY, Wild AE, Roberts P, Willis AC & Fleming TP(2000). Maternal undernutrition during the preimplantationperiod of rat development causes blastocyst abnormalitiesand programming of postnatal hypertension. Development127, 4195–4202.

Landmesser U, Hornig B & Drexler H (2004). Endothelialfunction: a critical determinant in atherosclerosis?Circulation 109, II27–II33.

Molnar J, Nijland MJ, Howe DC & Nathanielsz PW (2002).Evidence for microvascular dysfunction after prenataldexamethasone at 0.7, 0.75, and 0.8 gestation in sheep. Am JPhysiol Regul Integr Comp Physiol 283, R561–R567.

Mulvany MJ (1995). Resistance vessel growth and remodelling:cause or consequence in cardiovascular disease. J HumHypertens 9, 479–485.

Nishina H, Green LR, McGarrigle HH, Noakes DE, Poston L &Hanson MA (2003). Effect of nutritional restriction in earlypregnancy on isolated femoral artery function inmid-gestation fetal sheep. J Physiol 553, 637–647.

Ozaki T, Hawkins P, Nishina H, Steyn C, Poston L & HansonMA (2000). Effects of undernutrition in early pregnancy onsystemic small artery function in late-gestation fetal sheep.Am J Obstet Gynecol 183, 1301–1307.

Painter RC, De Rooij SR, Hutten BA, Bossuyt PM, De Groot E,Osmond C, Barker DJ, Bleker OP & Roseboom TJ (2007).Reduced intima media thickness in adults after prenatalexposure to the Dutch famine. Atherosclerosis 193, 421–427.

Poore KR, Cleal JK, Newman JP, Boullin JP, Noakes DE,Hanson MA & Green LR (2007). Nutritional challengesduring development induce sex-specific changes in glucosehomeostasis in the adult sheep. Am J Physiol EndocrinolMetab 292, E32–E39.

Roseboom TJ, van der Meulen JH, Osmond C, Barker DJ,Ravelli AC, Schroeder-Tanka JM, van Montfrans GA, MichelsRP & Bleker OP (2000). Coronary heart disease after prenatalexposure to the Dutch famine, 1944-45. Heart 84, 595–598.

Rizzoni D, Porteri E, Guelfi D, Muiesan ML, Piccoli A,Valentini U, Cimino A, Girelli A, Salvetti M, De Ciuceis C,Tiberio GA, Giulini SM, Sleiman I, Monteduro C & Rosei EA(2001). Endothelial dysfunction in small resistance arteriesof patients with non-insulin-dependent diabetes mellitus. JHypertens 19, 913–919.

Roghair RD, Lamb FS, Miller FJ Jr, Scholz TD & Segar JL(2005). Early gestation dexamethasone programs enhancedpostnatal ovine coronary artery vascular reactivity. Am JPhysiol Regul Integr Comp Physiol 288, R46–R53.

Segar JL, Roghair RD, Segar EM, Bailey MC, Scholz TD &Lamb FS (2006). Early gestation dexamethasone altersbaroreflex and vascular responses in newborn lambs beforehypertension. Am J Physiol Regul Integr Comp Physiol 291,R481–R488.

Simonsen U, Prieto D, Rivera L, Hernandez M, Mulvany MJ &Garcia-Sacristan A (1993). Heterogeneity of muscarinicreceptors in lamb isolated coronary resistance arteries. Br JPharmacol 109, 998–1007.

Singhal A & Lucas A (2004). Early origins of cardiovasculardisease: is there a unifying hypothesis? Lancet 363,1642–1645.

Taddei S, Virdis A, Ghiadoni L, Sudano I & Salvetti A (2001).Endothelial dysfunction in hypertension. J CardiovascPharmacol 38(Suppl 2), S11–S14.

Torrens C, Brawley L, Anthony FW, Dance CS, Dunn R,Jackson AA, Poston L & Hanson MA (2006). Folatesupplementation during pregnancy improves offspringcardiovascular dysfunction induced by protein restriction.Hypertension 47, 982–987.

Torrens C, Brawley L, Barker AC, Itoh S, Poston L & HansonMA (2003). Maternal protein restriction in the rat impairsresistance artery but not conduit artery function in pregnantoffspring. J Physiol 547, 77–84.

Torrens C, Kelsall CJ, Hopkins LA, Anthony FW, Curzen NP &Hanson MA (2009). Atorvastatin restores endothelialfunction in offspring of protein restricted rats in acholesterol-independent manner. Hypertension 53, 661–667.

Watkins AJ, Wilkins A, Cunningham C, Perry VH, Seet MJ,Osmond C, Eckert JJ, Torrens C, Cagampang FR, Cleal J,Gray WP, Hanson MA & Fleming TP (2008). Low proteindiet fed exclusively during mouse oocyte maturation leads tobehavioural and cardiovascular abnormalities in offspring. JPhysiol 586, 2231–2244.

Wintour EM, Moritz KM, Johnson K, Ricardo S, Samuel CS &Dodic M (2003). Reduced nephron number in adult sheep,hypertensive as a result of prenatal glucocorticoid treatment.J Physiol 549, 929–935.

Acknowledgements

This work was supported by a British Heart Foundation (BHF)Programme Grant (RG-99011) and HOPE medical charity.M.A.H. is supported by the BHF. T.H.S. and R.C. were BUPAClinical Research Fellows, and we thank Mr S. Ohri FRCS for hissupport during their fellowships. L.P. is supported by Tommy’sthe Baby Charity.

C" 2009 The Authors. Journal compilation C" 2009 The Physiological Society