N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

Ava i l ab l e on l i ne a t www.sc i enced i r ec t . com

ScienceDirectwww.n r j ou rna l . com

Excess pregnancy weight gain leads to early indications ofmetabolic syndrome in a swine model of fetal programming☆

Emily J. Arentson-Lantza, Kimberly K. Buhmanb, Kolapo Ajuwonc, d,Shawn S. Donkin c, d,⁎a The University of Texas Medical Branch at Galveston, Galveston 77555-0144 TXb Department of Nutrition Science, Purdue University, Lafayette, INc Interdepartmental Nutrition Program, Purdue University, Lafayette, INd Department of Animal Sciences, Purdue University, Lafayette, IN

A R T I C L E I N F O

Abbreviations: MatNE, maternal nutritionHealth Institute of Medicine; BMI, body mapostweaning high energy; MatNE→PWnNE, mphosphate dehydrogenase; Pck1, phosphoenoγ coactivator 1a.☆ This work was supported by funds from t⁎ Corresponding author: Department of Anim

47907-2054. Tel.: +1 765 494 4847; fax: +1 765E-mail address: sdonkin@purdue.edu (S.S

0271-5317/$ – see front matter © 2014 Elsevihttp://dx.doi.org/10.1016/j.nutres.2014.01.001

A B S T R A C T

Article history:Received 15 March 2013Revised 24 December 2013Accepted 7 January 2014

Few data exist on the impact of maternal weight gain on offspring despite evidencedemonstrating that early-life environment precipitates risks for metabolic syndrome. Wehypothesized that excessive weight gain during pregnancy results in programming thatpredisposes offspring to obesity and metabolic syndrome. We further hypothesized thatearly postweaning nutrition alters the effects of maternal weight gain on indications ofmetabolic syndrome in offspring. Pregnant sows and their offspring were used for theseexperiments due to similarities with human digestive physiology, metabolism, andneonatal development. First parity sows fed a high-energy (maternal nutrition highenergy [MatHE]) diet gained 12.4 kg (42%) more weight during pregnancy than sows fed anormal energy (maternal nutrition normal energy) diet. Birth weight and littercharacteristics did not differ, but offspring MatHE gilts weighed more (P < .05) at age of3 weeks (4.35 vs 5.24 ± 0.35 kg). At age of 12 weeks, offspring from MatHE mothers thatwere weaned onto a high-energy diet had elevated (P < .05) blood glucose (102 vs 64 mg/dL, confidence interval [CI]: 67-91), insulin (0.21 vs 0.10 ng/mL, CI: 0.011-0.019), and lowernonesterified fatty acid (0.31 vs 0.62 mmol/L, CI: 0.34-0.56) than offspring from the sameMatHE sows weaned to the normal energy diet. These effects were not observed foroffspring from sows fed a normal energy diet during pregnancy. These data indicate thatexcessive gestational weight gain during pregnancy in a pig model promotes earlyindications of metabolic syndrome in offspring that are further promoted by a high-energy postweaning diet.

© 2014 Elsevier Inc. All rights reserved.

Keywords:Maternal weight gainMetabolic syndromeInsulin resistancePigProgramming

normal energy; MatHE, maternal nutrition high energy; GWG, gestation weight gain; IOM,ss index; NEFA, nonesterified fatty acid; PWnNE, postweaning normal energy; PWnHE,aternal nutrition normal energy to postweaning normal energy; GAPDH, glyceraldehyde 3-lpyruvate carboxykinase, cytosolic form; PGC1a, peroxisome proliferator-activated receptor

he Showalter Trust Foundation, Purdue University.al Sciences, Lilly Hall of Life Science, 915 W. State St, Purdue University, West Lafayette, IN494 9346.. Donkin).

er Inc. All rights reserved.

242 N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

1. Introduction

Weight gain during pregnancy, known as gestational weightgain (GWG), comprises growth of the developing conceptus andexpansion ofmaternal adipose tissue. The Institute ofMedicine(IOM) of the National Academy of Sciences advises an inverserelationship between prepregnancy bodymass index (BMI) andpregnancy weight gain [1]. Women considered to be “over-weight” or “obese” based on a BMI greater than 25 kg/m2 foroverweight and 30 kg/m2 for obese are encouraged to gain lessweight during pregnancy than those considered to be “normal”or “underweight” based on a BMI of 18.5 kg/m2 to 25 kg/m2andless than 18.5 kg/m2, respectively. These GWG guidelines wereestablishedby the IOM topromotehealthypregnancyoutcomesfor the mother and offspring. However, recent estimatesindicate that less than half of pregnancies in the United Statesmeet the IOM guidelines for weight gain, with 65% to 85% of thepregnancies that do not meet the recommendations gainingmore than the recommended amount of weight [1]. Further-more, over 40% of women with a normal prepregnancy BMIexceed IOM guidelines for recommended weight gain duringpregnancy [2]. Excess weight gain during pregnancy is associ-ated with increased neonatal adiposity [3]. Dynamic mathe-matical modeling of energy balance during pregnancy linksexcessive gestational weight gain with increased energy intake[4], but the relation of energy-generating nutrients with weightgain and offspring health has not been adequately investigated.Consequently, gestational weight gain may represent a modi-fiable lifestyle factor to reduce incidence of childhood over-weight and obesity.

The programming effects of maternal gestational weightgain on offspring are not well understood. Pregnancy weightgain exceeding IOM recommendations is associated withchildhood adiposity and a risk for the offspring being over-weight at ages 3 years [5], 7 years [6], and 9 years [7]. Fatmass, amore direct measure of childhood adiposity, is associated withexcessiveGWG [7,8] at age of 9 years [7] but not age of 6 years [8].In addition, women who exceed the IOM GWG recommenda-tions, especially those who had a high rate of weight gain inearly pregnancy, are more likely to have children that, at age 9years, have elevated leptin and systolic blood pressure [7].Although several animal studies have explored the negativeimpacts of maternal obesity before conception [9-11], excessmaternal weight gain during pregnancy, with normal prepreg-nancy BMI, has not been extensively characterized. Wehypothesized that excessive weight gain during pregnancywould predispose offspring to early-life obesity and aspects ofthe metabolic syndrome and that early postweaning nutritionwould modulate the effects of maternal diet.

The objectives of this study were to determine effects ofexcess maternal gestational weight gain on offspring duringearly life at the key developmental milestones of birth,weaning, and early adolescence because it is now widelyaccepted that developmental plasticity extends past gestationto include neonatal life [12,13]. Primigravid commercial swinewere used as a model of first pregnancy weight gain due toavailability, the ability to control pregnancy weight gain,abundant overlap in physiology and metabolism with humanbeings, and the close developmental similarities of the piglet

and human neonate [14]. Because dysregulation of glucosemetabolism in offspring is a hallmark of a challenging in uteroenvironment and a well-characterized outcome for models ofboth maternal energy restriction [15,16] and obesity duringpregnancy [17], we focused our analysis in the piglets onindicators of metabolic syndrome in blood and transcripts inliver and intestine that are sensitive to insulin and in uteroprogramming, specifically phosphoenolpyruvate carboxyki-nase, cytosolic form (Pck1) and peroxisome proliferator-activated receptor γ coactivator 1a (PGC1a). Because liverand intestine account for approximately 45% of whole bodyoxygen consumption in pigs [18], changes in their activity inresponse to maternal diet are likely to have importantconsequences with regard to energy metabolism and predis-position to adiposity.

To evaluate the offspring response tomaternalweight gain,we measured birth weight, weight gain, adiposity, glucose,insulin, nonesterified fatty acids (NEFAs), and abundance ofkey insulin and developmentally responsive transcripts. Weevaluated the interaction of the maternal and postweaningdiet energy density on offspring growth, glucose metabolism,and intestinal and hepatic gene expression. We hypothesizedthat material weight gain during pregnancy leads to alter-ations in metabolism for offspring partly through changes inexpression of insulin-responsive genes in liver and intestine.

2. Methods and materials

2.1. Diets and animals

Crossbred gilts from the Purdue University Animal SciencesResearch and Education Center were artificially inseminatedand individually housed in 0.6 × 2.1 m gestation stalls. Uponconfirmation of pregnancy, gilts were blocked at 3 weeks ofgestation by weight and boar used for insemination andassigned to either a normal energy maternal diet (maternalnutrition normal energy [MatNE], n = 9) or a maternal high-energy diet (maternal nutrition high energy [MatHE], n = 5)during gestation (Table 1). All diets were formulated to meetNational Research Council Requirements for Swine (1998)using commercially available ingredients. Gilts fed the normalenergy diet received 2.05 kg of feed per day, and those on thehigh-energy diet received 3.0 kg; both groups were allowed adlibitum access to water. The gilts were fed once per day at0630. The MatHE diet was designed to provide a 50% increasein metabolizable energy intake and pregnancy weight gain.Weight of gilts was determined at 21, 48, 63, and 77 of daysgestation and at day 21 of lactation. Back fat at the 12th rib wasmeasured in pregnant gilts using B-mode ultrasound (AlokaAmerican Ltd, Wallington, CT) at 21, 48, 63, and 77 days ofgestation and at the end of lactation on day 21. Blood samplesfor serum and plasma were collected, approximately 4 hoursafter feeding, by venipuncture from the jugular vein at 21, 48,and 77 days of gestation. Serum was allowed to clot at roomtemperature for 10 minutes. All plasma and serum sampleswere stored on ice until they were centrifuged at 2000g for 20minutes. Serum and plasma were stored at −20°C pendinganalysis for glucose, insulin, and NEFA.

Table 1 – Ingredient and nutrient composition of maternal gestation and lactation diets

Ingredient Maternal diets

Normal energy e High energy f Lactation

Corn, g/kg 52.80 74.00 60.90Soybean meal, g/kg 12.20 6.50 24.10Dried distillers grains, g/kg 30.0 15.0 7.5Choice white grease, g/kg 1.0 1.65 3.0Limestone, g/kg 1.55 1.06 1.44Monocal, g/kg 0.75 0.50 1.42Vitamin premix a, g/kg 0.25 0.17 0.25Vitamin premix b, g/kg 0.25 0.17 0.25Selenium 600 premix c, g/kg 0.05 0.035 0.05Total mineral premixd, g/kg 0.125 0.85 0.125Phytase, g/kg 0.10 0.10 0.10Salt, g/kg 0.50 0.35 0.50Rabon larvicide, g/kg 0.13 0.13 0.13Diffusion plus, g/kg 0.25 0.25 0.25Feed intake, kg/day 2.05 3.00 ad libProtein intake, g/day 370 395 NDg

Lysine intake, g/day 16.03 15.8 NDg

Carbohydrate intake, g/day 741 1457 NDg

Fat intake, g/day 119 178 NDg

Metabolizable energy intake, kJ/day 28,306 42,470 NDg

Abbreviations: g/kg, gram per kilogram; kg/day, kilogram per day; g/day, gram per day; kcal/day, kilocalorie per day; ad lib, ad libitum.a Purdue swine vitamin premix: vitamin A, 544,680 IU/kg; vitamin D3, 54,448 IU/kg; vitamin E, 3631 IU/kg; mendione (vitamin K), 182 mg/kg;vitamin B12, 3.2 mg/kg; riboflavin, 726 mg/kg; D-pantothenic acid, 1816 mg/kg; niacin, 2723 mg/kg.b Sow vitamin premix: Kansas State University (KSU) sow vitamin add pack with carnichrome from Archer Daniels Midland Company (ADM)(biotin, 18.1 mg/kg; folic acid, 136 mg/kg; choline, 45,390 mg/kg; pyridoxine, 409 mg/kg; vitamin E, 1816 IU/kg; chromium, 16.3 mg/kg; carnitine,4805 mg/kg).c Selenium 600 premix: calcium, 28% to 31%; selenium, 0.06% equivalent to 123.6 mg/kg.d Purdue nonsulfur trace mineral premix: iron, 51.05%; zinc, 20.73%; manganese, 2.86%; copper, 1.56%; iodine, 0.046%.e Diet used for MatNE treatment during gestation.f Diet used for MatHE treatment during gestation.g Not determined, lactation diet was fed for ad libitum intake.

243N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

One week before expected farrowing, gilts were moved to0.6 × 2.1 m farrowing pens equipped with a 46-cm opening oneach side to allow sows to nurse and piglets to lie safely. Allgilts gave birth spontaneously. During lactation, all sows werefed a standard lactation diet for ad libitum intake (Table 1).Milk samples were collected from sows at day 13, 14, or 15 oflactation. Piglets were removed from the pen, and sows wereinjected with 10 IU of oxytocin intramuscularly to facilitatemilk let-down. Milk was manually expressed, and milksamples were analyzed for fat, lactose, ash, and total solidcontent at a commercial laboratory (Dairy One, Ithaca, NY).

2.2. Diets and piglets

Within 3 days of birth, piglets were tail docked and received anintramuscular injection of iron as per normal commercialproduction practices. Male offspring were not castrated tomodel early-life development of human males. Offspringremained with their mothers until they were weaned at ageof 18 to 21 days. At weaning, bothmale and female littermateswere blocked by weight and maternal diet and assigned,within litter, to either a postweaning normal energy (PWnNE)or postweaning high-energy (PWnHE) treatment. Male andfemale offspring were housed together in pens. Feed intakewas not measured in the offspring. Piglets were moved tonursery pens at weaning and were fed a common starter diet

ad libitum (Table 2). On day 7, piglets were transitioned to agrower diet and assigned to postweaning treatments (PWnNE,PWnHE). The high-energy phase 2 diet (PWnHE) was made byadding 68 kg of a milk replacer energy additive, Milk Energizer7:60 (Milk Specialties, Dundee, IL), which was 60% fat and 7%protein to 909 kg of the phase 2 mash (Table 2). At age of 6weeks, pigs were moved to a grow-finish unit and fed afinisher diet ad libitum to age of 84 days as per their previoustreatment groups (PWnNE or PWnHE, Table 2). This resulted ina 2 × 2 factorial arrangement of maternal diet (MatNE andMatHE) by 2 offspring diets. The 4 treatment groups aredesignated by maternal background followed by postwean-ing diet: maternal nutrition normal energy to postweaningnormal energy (MatNE→PWnNE) (n = 10), maternal nutritionnormal energy to postweaning high energy (n = 8), maternalnutrition high energy to postweaning normal energy (n = 6),maternal nutrition high energy to postweaning high energy(n = 6). All pigs were weighed weekly during the growerphase and biweekly during the finisher phase.

2.3. Tissue and blood collections and handling

At age of 48 hours, age of 3 weeks, and age of 12 weeks, maleand female offspring were euthanized, and samples werecollected for messenger RNA (mRNA) analysis. The 48-hourand 3-week-old offspring were removed from their sows and

Table 2 – Ingredient and nutrient composition of offspring diets

Ingredient Weaning to 4 weeks 4-6 weeks 6-12 weeks

All pigs PWnNE PWnHE PWnNE PWnHE

Corn, g/kg 32.25 32.25 37.35 60.98 52.98Soybean meal, g/kg 13.72 13.72 19.00 24.64 24.64Dried distillers grains, g/kg – – – 7.50 7.50Soybean oil, g/kg 5.00 5.00 5.00 – 5.00Limestone, g/kg 0.72 0.72 0.61 1.35 1.35Monocal phosphate, g/kg 0.53 0.53 0.75 0.74 0.74Vitamin premix a, g/kg 0.25 0.25 0.25 0.25 0.25Total mineral premix b, g/kg 0.12 0.12 0.12 0.12 0.12Selenium 600 premix c, g/kg 0.05 0.05 0.05 0.05 0.05Dried whey, g/kg 25.0 25.0 25.0 – –Lactose, g/kg 5.00 5.00 – – –Fish meal, g/kg 4.00 4.00 4.00 – –Phytase (600 PU/g), g/kg 0.10 0.10 0.10 0.10 0.10Salt, g/kg 0.25 0.25 0.25 0.35 0.35Blood meal, g/kg 6.50 6.50 3.75 – –Zinc oxide, g/kg 0.37 0.37 0.37 – –Soy concentrate, g/kg 4.00 4.00 2.50 – –Carbadox (10g/lb), g/kg 0.25 0.25 0.25 1.00 1.00L-Lysine hydrochloride, g/kg 0.11 0.11 0.25 0.40 0.40DL-methionine, g/kg 0.20 0.20 0.22 0.12 0.12L-threonine, g/kg 0.04 0.04 0.12 0.16 0.16L-tryptophan, g/kg – – 0.02 0.30 0.30Rabon larvicide, g/kg 0.025 0.025 0.25 0.25 0.25Bansmith dewormer, g/kg – – – 0.10 0.10Copper sulfate, g/kg – – – 0.07 0.07Crude protein, kcal/kg 23.96 23.96 22.8 19.40 17.9Total lysine, % 1.73 1.73 1.65 1.29 1.21Metabolizable energy, kJ/kg 14,842 14,842 14,745 14,214 15,788

a Purdue swine vitamin premix: vitamin A, 544,680 IU/kg; vitamin D3, 54,448 IU/kg; vitamin E, 3631 IU/kg; mendione (vitamin K), 182 mg/kg;vitamin B12, 3.2 mg/kg; riboflavin, 726 mg/kg; d-pantothenic acid, 1816 mg/kg; niacin, 2723 mg/kg.b Purdue nonsulfur trace mineral premix: iron, 51.05%; zinc, 20.73%; manganese, 2.86%; copper, 1.56%; iodine, 0.046%.c Selenium 600 premix: calcium, 28% to 31%; selenium, 0.06% equivalent to 123.6 mg/kg.

244 N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

euthanized immediately tominimize the impact of separationstress on the data. The 12-week-old offspring were fastedovernight before euthanasia to minimize the impact of timeafter food consumption on blood metabolites. Body lengthwas determined using the distance from the crown to the baseof the tail for 48-hour and 3-week-old offspring. Heart girthcircumference was measured around the abdomen justbehind the fore legs. All offspring were euthanized by CO2

overexposure followed by severance of the jugular vein andexsanguination. Blood was collected into glass tubes forserum or into heparinized tubes and processed as describedabove. Liver was excised and weighed. A 0.5 g sample of liverwas placed in 5 mL of TRIzol Reagent (15596-018; InvitrogenLife Technologies, Carlsbad, CA, USA), frozen in liquidnitrogen, and stored at −80°C until analysis for mRNA. Smallintestine, from the pylorus sphincter juncture with thestomach to the ileocecal valve at the cecum, was excised,flushed with saline, and weighed. The first 10 cm of intestineimmediately proximal to the stomach corresponding toduodenum was cut longitudinally and the lumen carefullyscraped with a glass slide to yield mucosa. Mucosal sampleswere placed in TRIzol Reagent (Invitrogen Life Technologies),frozen in liquid nitrogen, and stored at −80°C until analysis formRNA. In addition, in offspring euthanized at age of 12 weeks,the depth of intrascapular back fat wasmeasured using a rulerplaced at scapulae along the midline of the split carcass.

2.4. Biochemical analyses

Plasma glucose was determined using the glucose oxidasemethod [19] and reagents supplied as Autokit Glucose (439-90901; Wako Diagnostics, Mountain View, CA, USA). SerumNEFA were measured using acylation of coenzyme A by NEFAin the presence of acyl-CoA synthetase and reagents suppliedas HR Series NEFA-HR (2) (999-34691; Wako Diagnostics) in acolorimetric enzymatic assay. Serum insulin concentrationswere evaluated using immunoassay method and reagentssupplied as Porcine Insulin ELISA Kit (10-1200-01; Mercodia,Uppsala, Sweden). Metabolite and insulin assays were vali-dated using a test for parallelism of diluted serum or plasmato the standard curve and for recovery using samples spikedwith a known quantity of the metabolite of interest. The intraand interassay coefficients of variation for insulin were 5.82%and 3.04%, respectively. The homeostasis model assessmentof insulin resistance was calculated as fasting serum glucosemultiplied by fasting serum insulin divided by 2430 [20].

2.5. RNA isolation, cleanup, and reverse transcription

Approximately 0.5 g of mucosa or liver was homogenized inTRIzol Reagent (Invitrogen Life Technologies) using a Polytronhomogenizer, and total RNAwasextractedandquantifiedusingaNanoDrop1000 spectrometer (ThermoScientific,Wilmington,

245N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

DE,USA). A 50-μg sample of total RNAwas further purifiedusingthe RNeasy Mini Kit (74104; Qiagen, Germantown, MD, USA)with the addition of RNase-Free DNase (79254; Qiagen) to digestgenomicDNA. A 2-μg aliquot of DNase-treated RNAwas reversetranscribed to complementary DNA (cDNA) using OmniscriptRT Kit (205113; Qiagen) with 2 μL of 10-μM Oligo-dT (79237;Qiagen) and 0.4 μL of 10-μM Random Decamer (0702002;Ambion, Austin, TX, USA) per reaction.

2.6. Real-time polymerase chain reaction analysisof transcripts

Primers for quantitative real-time polymerase chain reaction(PCR) were designed using PrimerQuest software (IntegratedDNA Technologies, Iowa City, IA, USA) for glyceraldehyde 3-phosphate dehydrogenase (GAPDH), Pck1, PGC1a. The forwardand reverse primers, respectively, for each transcript wereGAPDH, ATGCCAGTGAGCTTCCCTTT and CCCAGAACAT-CATCCCTGCTTCTA; Pck1, GCCAGAAGAAATACTTTGCCGCAGand GGTCGAATTTCATCCAGGCGATGT; and PGC1a, AGCCATG-GATGGCCTGTTTGATGA and AGCCATGGATGGCCTGTTTGATGA.

Real-time PCR was performed with 2 μL of RT productdiluted 1:40, 100 μMof both forward and reverse primers, 0.375μL of 1 mM ROX diluted 1:500 (Stratagene, La Jolla, CA, USA),12.5 μL of 2x master mix (Brilliant III UltraFast SyberGreenQPCR Stratagene, 600882; Stratagene), and adjusted to a totalvolume of 25 μL per reaction with nuclease-free water.Quantitative real-time PCR amplification and detection wasperformed in 96 well plates using Stratagene Mx3005Pmachine (Stratagene). The reactions were initialized at 95°Cfor 3 minutes, denatured at 95°C for 5 seconds, and annealedand elongated at 60°C for 20 seconds for 40 cycles. Dissociationcurve was achieved by melting the DNA at 95°C for 1 minute,incubating the DNA at 55°C for 30 seconds, and followed by aramp up to 95°C for 30 seconds. A no-template control(nuclease-free water), standard curve of pooled cDNA, andpooled cDNA sample diluted 1:40 were included on each plateto check for any contamination or interplate variation(samples were reanalyze if between plate variation exceeded5%). Samples and controls were analyzed in triplicate.Expression of mRNA was calculated using the ΔΔCt method[21]. Both hepatic and intestine GAPDH abundance did notdiffer between treatment groups and was used as thehousekeeping gene to normalize abundance of genes ofinterest. Hepatic transcripts of offspring ages of 48 hoursand 3 weeks were arbitrated to the 48-hour offspring from thecontrol dams (MatNE), and hepatic transcripts of offspring ageof 12 weeks were arbitrated to the offspring from the controldams fed the control postweaning diet (MatNE→PWnNE).Similarly, intestinal transcripts from the 48-hour and 3-week-old offspring were normalized to the duodenum of the 48-hour offspring from the control dams (MatNE). For 12-week-old offspring, intestinal transcripts were normalized to theduodenum of the offspring from the control dams that wereweaned on the control postweaning diet (MatNE→PWnNE).

2.7. Statistical analyses

Data were tested for normality of error variances using thePROC UNIVARIATE procedure of SAS version 7.2 (SAS Institute

Inc, Cary, NC, USA) and transformed as appropriate if theShapiro-Wilks value was P < .05. Transformed values wereback-transformed to physiologically meaningful values andreported with a 95% confidence interval (CI). Data from giltsand offspring were analyzed using PROC MIXED procedure ofSAS with a model that accounted for the main effect ofmaternal diet, random effect of gilt nested within treatment,and age of offspring. Post hoc analysis was performed using 1-way 1 degree of freedomTukey lysergic acid diethylamide testor predicted difference test and preplanned single degree offreedom comparisons. Means were considered different whenP < .05 and tended to differ when .05 ≤ P ≤ .1.

3. Results

3.1. Maternal outcomes

Gilts fed the high-energy diet gained significantly (P < .05)more weight than sows fed the normal energy diet (Table 3).This increase was, at least in part, due to increased adiposeaccumulation because back fat thickness was also increasedfor MatHE gilts. In addition, there was a tendency (P = .05) forgreater blood glucose concentrations in sows fed the high-energy diet, but serum insulin and NEFA did not differbetween the maternal treatment groups. Likewise, postpar-tum milk composition did not differ between the treatmentgroups, whereas the percent fat, solid, and energy contenttended to be greater (P = .08) in milk from sows fed the high-energy diet during gestation.

3.2. Effect of maternal diet on 48-hour and 3-week-oldoffspring

There was no difference in body weight for MatNE and MatHEoffspring at 48-hour, but at age of 3 weeks, MatHE offspringweighed more (P < .05) than MatNE offspring (Table 4). TheMatHE piglets had greater heart girth circumference at age of 3weeks, but there were no differences in crown-to-rump lengthbetween MatNE and MatHE offspring. The lack of difference inweights at age of 48 hours but greater body weight for MatHEoffspring at 3 weeks represents more rapid growth during thesuckling phase for MatHE offspring. Furthermore, the greatergirth circumference for MatHE piglets at age of 3weeks and lackof difference in crown-to-rump length suggest a difference incompositionofgainbetweenMatNEandMatHEoffspring. Therewas no interaction ofmaternal diet and age on offspring length,liver weight, liver relative to body weight, and intestine weightas a percentage of body weight, insulin, or NEFAs.

3.3. Effect of maternal diet at 48-hour and 3-week-oldon Pck1 and PGC1a mRNA transcript profiles in liverand intestine

Liver and duodenal Pck1 and liver PGC1a were more abundantin 48-hour-old pigs, whereas duodenal PGC1a was moreabundant at age of 3 weeks (Table 4). In general, offspringfrom sows fed the MatHE diet had lower liver Pck1 mRNAexpression. At age of 3 weeks, MatHE offspring exhibitedsignificantly lower (P < .05) hepatic PGC1a mRNA.

Table 3 – Effect of maternal diet on maternal weight change, adipose tissue depth, bloodmetabolites, andmilk compositionin gilts a

Maternal diet SE P b

MatNE MatHE

Weight and adipose tissueGestational weight gain, kg 29.6 41.9 c 3.1 <.05Gestational back fat accumulation, mm 4.36 7.60 c 1.2 <.05Lactation, d 21, weight, kg 173.5 192 a 7.1 .07Lactation, d 21, back fat, mm 15.9 19.9 1.9 .21

Blood metabolite serum insulinGlucose (mg/dL) 78.6 88.8 6.4 .05Insulin (ng/mL) 0.12 0.12 0.025 .75NEFA (mmol/L) 0.10 0.08 002 .54

Milk compositionFat, % 7.20 8.79 a 0.54 .08Protein, % 5.27 5.45 0.10 .18Lactose, % 4.89 4.91 0.11 .90Solids, % 18.2 19.9 a 0.56 .08Ash, % 0.81 0.79 0.04 .74Energy, kJ/g 443 a 502 a 2.1 .08

Data are least-squared means and standard errors (SE).a Tended (.05 ≤ P ≤ .10) to differ from MatNE group.b P value-associated main factor level.c Differs (P < .05) from gilts in MatNE group.

246 N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

3.4. Effect of maternal and postweaning diet on12-week-old offspring

Postweaning diet was without effect on weight gain (Table 5),whereas there was a tendency (P = .08) for the PWnHE to

Table 4 –Main effects of maternal diet and age and the interactimetabolites at 48 hours postparturition and age of 3 weeks a

48 hours

Maternal diet

MatNE (n = 10) MatHE (n = 6) MatNE (n

Body weight (BW), kg 1.63 1.40 4.3Length, cm 25.6 25.2 35.6Girth, cm 24.9 25.7 36.0Liver, g 46.1 42.9 125Liver, % BW 3.11 2.93 2.9Intestine, g 42.1 34.2 124.6Intestine, % BW 2.85 2.33 3.0Insulin, ng/mL 0.063 0.041 0.0NEFAc, mmol/L 0.13 0.13 0.2Glucose, mg/dL 196 176 159DuodenumPGC1a 1.16 1.32 1.9Pck1 1.00 1.81 0.1

LiverPGC1a 1.00 1.25 0.1Pck1 0.90 1.38 0.0

Abbreviation: Mat, maternal nutrition.a Data are least-squared means and SE or 95% CI.b P value-associated main factor level and interaction.c Serum nonesterified fatty acids.d Differs (P < .05) from MatNE at same age.e Differs (P < .05) from MatNE at 48 hours.

increase back fat thickness at age of 12 weeks (23 vs 19 ± 1.6mm for PWnHE and PWnNE). Pigs fed the PWnHE diet hadgreater (P < .05) fasting plasma glucose than PWnNE (91 vs 67mg/dL, CI: 71.2-86.5; PWnHE vs PWnNE) at age of 12 weeks.There was no main effect of postweaning diet on insulin,

on of maternal diet and age on offspring weight and serum

3 weeks SE P b

Maternal diet Mat age Mat × Age

= 10) MatHE (n = 6)

5 5.24d 0.35 .32 <.05 .0636.2 1.30 .89 <.05 .5839.7 d 1.01 <.05 <.05 .12

153 11.7 .20 <.05 .115 2.92 0.18 .49 .57 .62

151.3 13.9 .42 <.05 .140 2.94 0.30 .25 .14 .3543 0.050 0.016 .59 .72 .301 0.24 0.04 .69 <.05 .57

201 21.6 .55 .75 .09

7 2.09 0.37 .65 <.05 .698 0.16 (0.16, 0.54) .19 <.05 .27

8 0.063 e (0.27, 0.66) .21 <.05 .0835 0.020 (0.13, 0.23) .78 <.05 <.05

247N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

NEFA, or homeostasis model assessment of insulin resistanceindex,whereas offspring frommothers in theMatHE group fedPWnNE diets had elevated insulin concentrations relative topigs from MatNE dams fed the PWnHE diet.

3.5. Effect of maternal and postweaning diet on Pck1and PGC1a mRNA transcript profiles in liver and intestinein 12-week-old offspring

Hepatic Pck1 mRNA abundance was elevated for pigs fromMatHE sows and subsequently fed the PWnHE diet com-pared with offspring from MatNE fed the PWnHE diet. At ageof 3 months, duodenal PGC1a and Pck1 mRNA were notaffected by the postweaning diets, maternal background, ortheir interactions.

4. Discussion

Excess gestational weight gain is associated with adverseoutcomes for both the mother and offspring, includingmaternal gestational diabetes and postpartum weight reten-tion [1] and an increase in offspring birth weight [22] and fatmass in early life [5,7]. The data reported here support ourhypothesis that excess gestational weight gain acts toprogram the pattern of growth in offspring, serum metabo-lites, and abundance of mRNA in liver and intestine. Theprogramming effect of gestational weight gain is furthermodulated by a high-energy postweaning diet. Offspring ofsows that experienced excess weight during pregnancy andwere weaned onto a high-energy postweaning diet displayedincreased fasting glucose and insulin, early indicators ofmetabolic syndrome. In addition, offspring from MatHE sows

Table 5 – Interactions of the main effects of maternal and postwage of 12 weeks a.

Maternal diet

MatNE Ma

PWnNE PWnHE PWnNE

Body weight gain, kg 30.1 30.9 31.3Intrascapular back fat, mm 18.3 21.2 19.0Insulin (ng/mL) 0.016 0.013 0.010NEFA (mmol/L) c 0.43 0.43 0.62Glucose, mg/dL 70.0 80.0 64HOMAd 0.43 0.42 0.48DuodenumPGC1a 0.89 0.88 0.43Pck1 0.82 1.21 1.40

LiverPGC1a 1.00 0.78 1.07Pck1 1.14 0.33 1.88

Abbreviations: HOMA, homeostasis model assessment; PWn, postweanina Data are least-squared means and SE or 95% CI.b P value-associated main factor level.c Serum nonesterified fatty acids.d Homeostatic model assessment.e Differs (P < .05) from PWnNE within MatHE groups.f Differs (P < .05) from MatNE offspring fed PWnHE.

displayed altered intestinal and hepatic transcript abundanceof PGC1a and Pck1 at weaning and age of 12 weeks.

Gilts in the MatHE group were challenged with moreenergy, which resulted in approximately 30% greatermaternalweight gain compared with gilts fed the normal energy diet.The maternal diets were matched for daily protein and lysineintake to minimize the impact of change in lean gain betweenthe MatNE and MatHE groups. Consequently, the increase inenergy intake is primarily from carbohydrate and resulted in a45% greater accretion of back fat for gilts fed the high-energydiet. The increase in back fat deposition is consistent withhuman data that link excess gestational weight gain with fatmass accretion [23]. As a result, therewas an altered pattern ofgrowth in piglets in response to maternal diet that resulted inaccelerated weight gain that was not matched by increasedstature. One of the underlying assumptions of fetal program-ming is that programming events occurring in utero andduring the early postnatal phase serve to prepare the fetus toanticipate the nutritional and metabolic environment duringadult life [24]. If the environment in adult life does not matchthe environment during early life, offspring are at anincreased risk for developing adult onset diseases [25]. Wefound that offspring of sows that gained excessive weightduring pregnancy were susceptible to early weight gain.Exposure to an energy-rich postweaning diet further promot-ed aspects of the metabolic syndrome in offspring, and theseeffects are accompanied by changes in PGC1a and Pck1expression, key insulin sensitive metabolic regulators ofglucose, and energy metabolism.

Gestational weight gain in women in excess of IOMrecommendations is linked with an increased risk of largefor gestational age defined as infant birth weight more than4000 g [26-28]. In the present study, sows fed the high-energy

eaning diet on offspring weight and serum metabolites at

SE or CI P b

tHE Mat PWn Mat × PWn

PWnHE

32.0 3.0 .67 .78 .9923.8 2.5 .49 .12 .69

0.021 e (0.011, 0.019) .92 .24 .080.31 f (0.34,0.56) .87 .12 .11102 f e (67.3, 90.9) .47 <.05 .210.36 0.11 .93 .44 .57

0.62 (0.46, 0.99) .12 .62 .590.58 (0.61, 1.46) .80 .54 .12

0.57 (0.58,1.18) .71 .18 .550.62 f 0.38 .17 <.05 .54

g.

248 N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

diet gained more weight and had elevated blood glucose, butoffspring birth weight and body size did not differ betweenmaternal diet groups. These data are inconsistent withprevious observations in swine, where a high-energy dietduring fed to gilts during gestation increased weight gain andback fat accumulation but did not alter piglet birth weight orpostnatal weight [29]. However, our observations are consis-tent with other data demonstrating that increased energyintake during gestation was without effect on birth weight orpiglet growth rate [30]. In either case, the current data extendprevious observations to indicate a change in the pattern ofearly postnatal growth, and composition of growth is aconsequence of excess weight gain in pregnant sows.

Nutrition during suckling may program metabolic out-comes in offspring [31]. The present study indicates thatoffspring of MatHE sows exhibited increased body weight andheart girth circumference at the conclusion of the sucklingperiod despite a lack of differences at birth. This increase inbody weight and heart girth circumference for MatHEoffspring was not matched by an increase in length, whichsuggests increased adiposity. Similarly, 3-week-old offspringfrom obese rat dams exhibit more subcutaneous fat tissuecompared with offspring of lean dams [17]. Increased adipos-ity observed in the MatHE offspring is similar to the catch-upgrowth observed for programming in other species [32,33] andis consistent with production and consumption of milk withgreater lipid and energy content by MatHE sows.

Insulin resistance in offspring is a well-documentedconsequence of maternal dietary insults during pregnancythat is exacerbated by an imbalanced postnatal diet[16,34,35]. Similar responses are seen in the MatHE offspringweaned on to the high-fat postweaning diet. Offspring fromthe MatHE group given the PWnHE diet exhibited signifi-cantly increased fasting glucose and insulin, indicative ofearly insulin resistance. The increase in Pck1 in liver ofoffspring from gilts fed a high-energy diet during gestationand receiving a high-energy diet at weaning along withelevated blood glucose and insulin is consistent with insulinresistance in this offspring group. Coordinated changes inPGC1a and Pck1 and impaired insulin action have beenobserved in liver of intrauterine growth retarded offspring[36]; data from MatHE offspring also support a programmingeffect on glucose metabolism.

Unfortunately, the present study is limited by the lack ofintake measures in the offspring due to the housing facilitiesused during the postweaning period. Differences in voluntaryintake are linked to fetal programming [37], and a pro-grammed difference in food intake between the MatNEoffspring and MatHE offspring may account for a portion ofincreased weight gain for the MatHE offspring.

In the current study, we found that increasing maternalenergy intake during pregnancy altered glucose homeostasisin sows and increased overall gestation weight gain andadiposity. Excess maternal gestational weight gain had asignificant impact on pattern of intestine and liver transcriptabundance at weaning. These changes may be the result ofincreased energy supply from milk of MatHE sows; however,the lack of information onmilk intake precludes amore robustinterpretation of altered growth to weaning. Controlledexperiments using formula feeding are needed to more fully

appreciate the impact of maternal weight gain to potentiateoffspring growth to weaning. This may be especially truewhen extrapolating the current findings to weight gain duringpregnancy in human beings and postnatal nutrition andgrowth. Our data would suggest that excess weight gainpotentiates the effects of increased postweaning energyintake on indicators of metabolic syndrome. Additional workis needed to further understand the underlying causes ofdisrupted metabolism in offspring and the impact of neonatalnutrition on these outcomes.

R E F E R E N C E S

[1] Guidelines CtRIPW, Council IOMNR. Weight gain duringpregnancy. In: Rasmussen KM, Yaktine AL, editors.Reexamining the guidelines. The National Academies Press:Washington DC; 2009. p. 324.

[2] Chu SY, Callaghan WM, Bish CL, D’Angelo D. Gestationalweight gain by bodymass index among US women deliveringlive births, 2004-2005: fueling future obesity. Am J ObstetGynecol 2009;200:271.e1-271.

[3] Josefson JL, Hoffmann JA, Metzger BE. Excessive weight gainin women with a normal pre-pregnancy BMI is associatedwith increased neonatal adiposity. Pediatr Obes 2013;8:e33–6.

[4] Thomas DM, Navarro-Barrientos JE, Rivera DE, Heymsfield SB,Bredlau C, Redman LM, et al. Dynamic energy-balance modelpredicting gestational weight gain. Am J Clin Nutr Jan.2012;95:115–22.

[5] Oken E, Taveras EM, Kleinman KP, Rich-Edwards JW, GillmanMW. Gestational weight gain and child adiposity at age 3years. Am J Obstet Gynecol Apr. 2007;196:322.e1–8.

[6] Wrotniak BH, Shults J, Butts S, Stettler N. Gestational weightgain and risk of overweight in the offspring at age 7 years in amulticenter, multiethnic cohort study. Am J Clin Nutr Jun.2008;87:1818–24.

[7] Fraser A, Tilling K, Macdonald-Wallis C, Sattar N, Brion MJ,Benfield L, et al. Association of maternal weight gain inpregnancy with offspring obesity and metabolic and vasculartraits in childhood. Circulation Jun. 2010;121:2557–64.

[8] Crozier SR, Inskip HM, Godfrey KM, Cooper C, Harvey NC, ColeZA, et al. Weight gain in pregnancy and childhood bodycomposition: findings from the Southampton Women’sSurvey. Am J Clin Nutr Jun. 2010;91:1745–51.

[9] Zambrano E, Martínez-Samayoa PM, Rodríguez-González GL,Nathanielsz PW. Dietary intervention prior to pregnancyreverses metabolic programming in male offspring of obeserats. J Physiol 2010;588:1791–9.

[10] Niculescu MD, Lupu DS. High fat diet-induced maternalobesity alters fetal hippocampal development. Int J DevNeurosci 2009;27:627–33.

[11] Kirk SL, Samuelsson AM, Argenton M, Dhonye H,Kalamatianos T, Poston L, et al. Maternal obesity induced bydiet in rats permanently influences central processesregulating food intake in offspring. PLoS One 2009;4:e5870.

[12] Hanson MA, Gluckman PD. Developmental origins ofhealth and disease: moving from biological concepts tointerventions and policy. Int J Gynaecol Obstet2011;115(Suppl 1):S3–5.

[13] Taylor PD, Poston L. Developmental programming of obesityin mammals. Exp Physiol 2007;92:287–98.

[14] Puimana P, Stolla B. Animal models to study neonatalnutrition in humans. Curr Opin Clin Nutr Metab Care2008;11:601–6 [2008].

[15] Gluckman PD, Lillycrop KA, Vickers MH, Pleasants AB, PhillipsES, Beedle AS, et al. Metabolic plasticity during mammalian

249N U T R I T I O N R E S E A R C H 3 4 ( 2 0 1 4 ) 2 4 1 – 2 4 9

development is directionally dependent on early nutritionalstatus. Proc Natl Acad Sci U S A Jul. 2007;104:12796–800.

[16] Armitage JA, Khan IY, Taylor PD, Nathanielsz PW, Poston L.Developmental programming of the metabolic syndrome bymaternal nutritional imbalance: how strong is the evidencefrom experimental models in mammals? J Physiol Dec.2004;561:355–77.

[17] Zambrano E, Martínez-Samayoa PM, Rodríguez-González GL,Nathanielsz PW. Dietary intervention prior to pregnancyreverses metabolic programming in male offspring of obeserats. J Physiol May 2010;588:1791–9.

[18] Yen JT, Neinaber JA, Hill DA, Pond WG. Oxygen consumptionby portal vein-drained organs and by whole animal inconscious growing swine. Proc Soc Exp Biol Med1989;190:393–8.

[19] Miwa I, Okudo J, Maeda K, Okuda G. Mutarotase effect oncolorimetric determination of blood glucose with D-glucoseoxidase. Clin Chim Acta Mar. 1972;37:538–40.

[20] Cacho J, Sevillano J, de Castro J, Herrera E, Ramos MP.Validation of simple indexes to assess insulin sensitivityduring pregnancy in Wistar and Sprague-Dawley rats. Am JPhysiol Endocrinol Metab Nov. 2008;295:E1269–76.

[21] Livak KJ, Schmittgen TD. Analysis of relative gene expressiondata using real-time quantitative PCR and the 2(-delta deltaC(T)) method. Methods Dec. 2001;25:402–8.

[22] Ludwig DS, Currie J. The association between pregnancyweight gain and birth weight: a within-family comparison.Lancet Sept. 2010;376:984–90.

[23] Lederman SA, Paxton A, Heymsfield SB, Wang J, Thornton J,Pierson RN. Body fat and water changes during pregnancy inwomen with different body weight and weight gain. ObstetGynecol Oct. 1997;90:483–8.

[24] Langley-Evans SC. Nutritional programming of disease:unravelling the mechanism. J Anat Jul. 2009;215:36–51.

[25] Bruce KD, Hanson MA. The developmental origins, mecha-nisms, and implications of metabolic syndrome. J Nutr Mar.2010;140:648–52.

[26] Crane JM, White J, Murphy P, Burrage L, Hutchens D. Theeffect of gestational weight gain by body mass index onmaternal and neonatal outcomes. J Obstet Gynaecol Can Jan.2009;31:28–35.

[27] Viswanathan M, Siega-Riz AM, Moos MK, Deierlein A,Mumford S, Knaack J, et al. Outcomes of maternal weightgain. Evid Rep Technol Assess (Full Rep); May 2008 1–223.

[28] Butte NF, Ellis KJ, Wong WW, Hopkinson JM, Smith EO.Composition of gestational weight gain impacts maternal fatretention and infant birth weight. Am J Obstet Gynecol Nov.2003;189:1423–32.

[29] Xue JL, Koketsu Y, Dial GD, Pettigrew J, Sower A. Glucosetolerance, luteinizing hormone release, and reproductiveperformance of first-litter sows fed two levels of energyduring gestation. J Anim Sci Jul. 1997;75:1845–52.

[30] Prunier ACM, Guadarrama J Mourot, Quesnel H. Influence offeed intake during pregnancy and lactation on fat bodyreserve mobilisation, plasma leptin and reproductive func-tion of primiparous lactating sows. Reprod Nutr Dev2001;41:333–47.

[31] Khan IY, Dekou V, Douglas G, Jensen R, Hanson MA, Poston L,et al. A high-fat diet during rat pregnancy or suckling inducescardiovascular dysfunction in adult offspring. Am J PhysiolRegul Integr Comp Physiol Jan. 2005;288:R127–33.

[32] Atinmo T, Pond WG, Barnes RH. Effect of maternal energy vs.protein restriction on growth and development of progeny inswine. J Anim Sci Oct. 1974;39:703–11.

[33] Beltrand J, Nicolescu R, Kaguelidou F, Verkauskiene R, SibonyO, Chevenne D, et al. Catch-up growth following fetal growthrestriction promotes rapid restoration of fatmass but withoutmetabolic consequences at one year of age. PLoS One 2009;4:e5343.

[34] Jones RH, Ozanne SE. Fetal programming of glucose-insulinmetabolism. Mol Cell Endocrinol Jan. 2009;297:4–9.

[35] Barker DJ, Clark PM. Fetal undernutrition and disease in laterlife. Rev Reprod May 1997;2:105–12.

[36] Lane RH, MacLennan NK, Hsu JL, Janke SM, Pham TD.Increased hepatic peroxisome proliferator-activated recep-tor-gamma coactivator-1 gene expression in a rat model ofintrauterine growth retardation and subsequent insulinresistance. Endocrinology 2002;143:2486–90.

[37] Vickers MH, Breier BH, McCarthy D, Gluckman PD. Sedentarybehavior during postnatal life is determined by the prenatalenvironment and exacerbated by postnatal hypercaloric nu-trition. Am J Physiol Regul Integr Comp Physiol 2003;285:R271.