Search
Author
Title
Vol.
Issue
Year
1st Page

Journal of Animal Science - Animal Growth, Physiology, and Reproduction

Effects of low and high protein:carbohydrate ratios in the diet of pregnant gilts on maternal cortisol concentrations and the adrenocortical and sympathoadrenal reactivity in their offspring12

 

This article in JAS

  1. Vol. 91 No. 6, p. 2680-2692
     
    Received: Nov 8, 2012
    Accepted: Feb 25, 2013
    Published: November 25, 2014


    3 Corresponding author(s): otten@fbn-dummerstorf.de
 View
 Download
 Share

doi:10.2527/jas.2012-6080
  1. W. Otten 3,
  2. E. Kanitz*,
  3. M. Tuchscherer*,
  4. M. Gräbner*,
  5. G. Nürnberg,
  6. O. Bellmann,
  7. U. Hennig§,
  8. C. Rehfeldt# and
  9. C. C. Metges§
  1. Behavioral Physiology
    Genetics and Biometry
    Research Units
    Nutritional Physiology “Oskar Kellner”
    Muscle Biology and Growth. Leibniz Institute for Farm Animal Biology (FBN Dummerstorf), Wilhelm-Stahl-Allee 2, D-18196 Dummerstorf, Germany

Abstract

Inadequate maternal nutrition during gestation may cause an adverse environment for the fetus leading to alterations of the hypothalamic-pituitary-adrenal (HPA) and sympatho-adrenomedullary (SAM) systems later in life. In the present study, we investigated the effects of diets with low and high protein:carbohydrate ratios on cortisol concentrations of pregnant gilts as well as the long-term effects on the function of the HPA and SAM axes in their offspring. Throughout gestation, 33 German Landrace gilts were fed high (HP, 30%), low (LP, 6.5%), or adequate (AP, 12.1%) protein diets, which were made isocaloric by adjusting the carbohydrate content. The salivary cortisol concentrations of the sows were measured in the course of the gestation period. The offspring were cross-fostered, and the plasma cortisol and catecholamine concentrations of the offspring were determined on postnatal d (PND) 1 and 27 and under specific challenging conditions: after weaning (PND 29) and after ACTH and insulin challenges (PND 68 and 70, respectively). Glucocorticoid receptor (GR) binding and neurotransmitter concentrations were measured in stress-related brain regions, and histological analyses of the adrenal were performed. Maternal salivary cortisol concentrations increased throughout gestation (P < 0.001) and the LP gilts had greater salivary cortisol compared with the AP and HP gilts (P < 0.05). No differences between diets were found for cortisol, corticosteroid-binding globulin, and catecholamine concentrations in plasma and for GR binding in hippocampus and hypothalamus in piglets at PND 1 and 27. However, the cortisol response to weaning was increased in LP piglets (P < 0.05), and in HP offspring the basal plasma noradrenaline concentrations were increased (P < 0.05). The cortisol response to the ACTH and the insulin challenge did not differ between diets. On PND 81, an increased adrenal medulla area was observed in LP offspring compared with the AP offspring (P < 0.05). Our results show that maternal diets with aberrant protein:carbohydrate ratios during gestation have moderate long-term effects on the function of the HPA and SAM system in the offspring, which indicates that pigs show a considerable plasticity to cope with maternal malnutrition.



INTRODUCTION

Studies in rodents and humans have shown that inadequate maternal nutrition or the action of maternal glucocorticoids during gestation may cause an adverse environment for the fetus (reviewed by McMillen and Robinson (2005) and Langley-Evans (2009)) leading to intrauterine growth retardation (IUGR) and pathological consequences later in life (Huizink et al., 2004; McMillen and Robinson, 2005).

We recently reported that high and low protein:carbohydrate diets fed to adolescent pregnant sows lead to IUGR (Rehfeldt et al., 2011) and an altered “setting” of fetal cortisol regulation and gene expression involved in central glucocorticoid action (Kanitz et al., 2012). However, it is not clear whether these imbalanced diets affect stress-sensitive systems in the offspring of pigs in the long term. The consequences of adverse fetal environmental conditions are of particular interest because of their central role in body functions and the application of this knowledge to human health studies and sustainable farm animal production.

In the present study, we examined the hypotheses that low and high protein:carbohydrate ratios in the diet of pregnant adolescent sows disrupt maternal cortisol regulation and have long-term consequences on the functions of the hypothalamic-pituitary-adrenal (HPA) and sympatho-adrenomedullary (SAM) axes in the offspring. Maternal cortisol regulation was measured by salivary cortisol concentrations during the gestation period. In the offspring, basal and stress-related function of the HPA and SAM system were determined by plasma cortisol and catecholamine concentrations on postnatal d (PND) 1 and 27 and under specific challenging conditions: after weaning (PND 29) and after ACTH and insulin challenges (PND 68 and 70, respectively). Central and peripheral alterations of these systems were measured by glucocorticoid receptor (GR) binding and neurotransmitter concentrations in stress-related brain regions and by histological analyses of the adrenal cortex and medulla at different ages.


MATERIAL AND METHODS

All procedures were in accordance with German animal protection law and were approved by the local ethics committee (Landesamt für Landwirtschaft, Lebensmittelsicherheit und Fischerei Mecklenburg-Vorpommern, Rostock, Germany; LVL M-V/TSD/7221.3–1.1–006/04; LALLF M-V/TSD/7221.3–1.2–05/06; LALLF M-V/TSD/7221.3–1.2–013/06).

Animals

A total of 36 primiparous German Landrace sows and their offspring (Sus scrofa domestica) were examined in this study.

Prenatal Treatments

Gilts were estrus synchronized and inseminated twice on 2 consecutive days with semen from pure German Landrace boars. One day before the first insemination, gilts were randomly allocated to the 3 dietary groups with 12 gilts each. Starting on the day of the first insemination, gilts were fed corn-barley and soybean meal diets (∼13.7 MJ ME/kg), which were made isocaloric by adjusting the carbohydrate content with crude protein concentrations of 65 g/kg (low protein; LP), 121 g/kg (adequate protein; AP), or 300 g/kg (high protein; HP) throughout the gestation period. The protein:carbohydrate ratios were 1:10.4, 1:5, and 1:1.3 for the LP, AP, and HP diets, respectively. From early to late gestation, the amount of feed provided was 2.3 to 2.9 kg/d to achieve an average target energy intake of ∼34 MJ ME/d according to the recommendations for primiparous sows (GfE, 2006). Pregnancy was confirmed by ultrasound detection on gestational d (GD) 28 and 50, and pregnant gilts were moved to group pens (2.4 × 6.9 m) with concrete floor and a maximum of 4 animals per group. The gilts were fed discretely twice daily at 0700 and 1500 h with 50% of the daily ration each time, and water was provided ad libitum. At GD 109, gilts were moved to individual farrowing stalls with a slatted floor (2.0 × 3.0 m). The experiment was conducted over 6 temporally successive replicates with the comparison of all 3 diets in each replicate (usually 2 gilts per diet and replicate). Three sows were removed from the experiment due to infertility or miscarriage so that 33 sows remained for investigations. Finally, 11, 12, and 10 gilts fed the LP, AP, and HP diets, respectively, produced a litter. Additional experimental details on these items were previously reported: the diet, housing, and treatment of gilts; induction of farrowing on GD 114; the dietary effects on the BW gain of gilts and offspring; and reproductive data (Rehfeldt et al., 2011; Metges et al., 2012).

Before and during gestation, saliva samples were collected in the morning between 0800 and 1000 h on GD −5 to −3, GD 22 to 24, GD 64 to 66, and GD 106 to 108 for cortisol analyses. To reduce the variability of cortisol from a single sample, saliva samples taken on 3 consecutive days were analyzed for each gilt and a mean value was calculated for further statistical analysis. Gilts were allowed to chew on veterinary cotton buds (Weisweiler GmbH and Co. KG, Münster, Germany), and the thoroughly moistened buds were placed in tubes and centrifuged at 2,500 × g for 10 min at 4°C. Saliva samples were then stored at −20°C for subsequent cortisol analysis.

Offspring

At birth, litter size and individual birth mass were recorded. Additional details on gestation outcome and postnatal growth of offspring were reported previously (Rehfeldt et al., 2011, 2012). Poor viable piglets weighing less than 800 g were excluded from subsequent experiments. Remaining piglets were ear-tagged. From each litter, 8 piglets were randomly selected and assigned to 1 of 3 subsequent times of sampling. In this way, 4 piglets per litter were assigned to plasma and tissue sampling on PND 1 and another 2 piglets per litter to plasma and tissue sampling on PND 27. The remaining 2 piglets per litter were sampled between PND 67 and 81. Thus, 2 to 4 piglets from each sow were intended for each time of sampling. Piglets assigned to the sampling times later than PND 1 were cross-fostered as littermates within 48 h of birth to multiparous foster sows. The fostered litters were standardized to 11 piglets with piglets from the foster sows. These foster sows were fed standard diets during gestation and lactation. Male piglets were castrated at 4 d of age.

Blood and Tissue Sampling on PND 1 and 27

For the piglets allocated for sampling on PND 1 and 27, blood samples (3 mL) were taken by a fast venipuncture of the anterior vena cava with the piglets restrained in a supine position. Samples were collected in ice-cooled polypropylene tubes containing EDTA solution, placed on ice and subsequently centrifuged at 2,000 × g for 15 min at 4°C for plasma extraction. The collected plasma was stored at −80°C until subsequent analysis of cortisol, corticosteroid-binding globulin (CBG), and catecholamine concentrations. Immediately after blood sampling, piglets were euthanized by an intravenous (i.v.) injection of 1 mL (PND 1) or 3 mL (PND 27) T61 (embutramide 200 mg/mL, mebezonium iodide 50 mg/mL, tetracaine hydrochloride 5 mg/mL; Intervet, Unterschleißheim, Germany). The brains were quickly removed and the hippocampus, the hypothalamus, and the region of the locus coeruleus for each was dissected, frozen in liquid nitrogen, and stored at −80°C until further analysis. The stereotaxic atlas of the pig brain (Félix et al., 1999) served as a reference for dissection. The adrenal glands were also removed, weighed, frozen in liquid nitrogen, and stored at −80°C until further analysis.

Weaning (PND 28), Challenge Tests (PND 68 and 70), and Tissue Sampling (PND 81)

Blood samples (2 mL) were also taken from the 2 remaining experimental piglets of each litter on PND 27 by venipuncture of the anterior vena cava. Piglets were weaned on PND 28 by moving them as a litter to a postweaning room with group pens (2.5 m × 1.8 m). On PND 29, an additional blood sample was taken, and the samples were treated as described above and stored at −20°C until analysis of cortisol and CBG. After weaning, animals had free access to standard diets formulated for the postweaning period.

Starting on PND 32, piglets were kept in single pens (2.0 × 1.0 m). Between PND 60 and 62, animals were surgically fitted with a venous and an arterial catheter as previously described (Metges et al., 2005). On PND 68 ± 1 d, an ACTH challenge test was performed followed by an insulin challenge test 2 d later (PND 70 ± 1 d). For the ACTH challenge, 0.5 IU ACTH (Synacthene, Novartis Pharma AG, Basel, Switzerland) per kilogram BW was administered i.v. in the venous catheter. Blood samples (1 mL) were taken from the arterial catheter 20 min before ACTH application and every 20 min thereafter until 180 min after the application. Samples were treated as described above and stored at −20°C until analysis of cortisol and CBG. For the insulin challenge, 0.5 IU porcine insulin (27 IU/mg; Sigma-Aldrich, Taufkirchen, Germany) per kilogram BW was administered i.v. in the venous catheter. Blood samples (2 mL) were collected from the arterial catheter 15 min and immediately before insulin application as well as 5, 10, 15, 30, 45, 60, 75, 90, 120, 150, 180, and 240 min after application. After plasma extraction, samples were stored at −80°C for further analysis of catecholamines and cortisol.

At PND 81, the animals were slaughtered, and the adrenal glands were removed and treated as described above.

Cortisol Analyses

The analysis of saliva cortisol concentrations of gilts was performed in duplicate using 25-μL samples and a commercial enzyme immunoassay kit (DSL Inc., Sinsheim, Germany) according to the manufacturer’s instructions. The assay was validated for use with porcine saliva. The sensitivity was 0.96 nmol/L, and the intra- and interassay CV were 2.0% and 8.7%, respectively. Plasma cortisol concentrations in samples from offspring were analyzed in duplicate using a commercially available 125I-RIA kit (DSL Inc., Sinsheim, Germany) according to the manufacturer’s guidelines. The test sensitivity was 8.1 nmol/L, and the intra- and interassay CV were 8.2% and 9.8%, respectively.

Corticosteroid-Binding Globulin (CBG)

Plasma samples from offspring were examined for CBG using a modified binding assay previously described by Kanitz et al. (2002). Briefly, after removing endogenous steroids from plasma by dextran-coated charcoal treatment, 25 μL plasma was incubated with 0.78 nM unlabeled cortisol (Hydrocortisone; Merck, Darmstadt, Germany) and 25 pM 3H-cortisol (specific radioactivity 68 Ci/mmol; Amersham Pharmacia Biotech, Freiburg, Germany). Nonspecific binding was determined in parallel using a 100-fold excess of cold cortisol. The separation of bound and free 3H-cortisol was performed by precipitation with dextran-coated charcoal at 4°C and subsequent centrifugation at 1,000 × g for 10 min at 4°C. The intra- and interassay CV were 7.8% and 9.1%, respectively.

Glucocorticosteroid Receptor Binding Assay

Glucocorticosteroid receptor binding in pig brain was performed as previously described by Kanitz et al. (1998, 2003). Briefly, tissues of the hypothalamus and right hippocampus were homogenized in a buffer solution (10 mM Tris-HCL, 12.5 mM EDTA, 10 mM sodium molybdate, 0.25 mM sucrose, 1 mM dithiothreitol) and centrifuged at 4°C at 120,000 × g for 60 min to obtain cytosol (i.e., the supernatant fraction). Hippocampal GR binding was evaluated in saturation experiments using 3H-dexamethasone (specific activity 43 Ci/mmol; Amersham Pharmacia Biotech, Freiburg, Germany) over a concentration range of 0.2 to 24 nM. Glucocorticosteroid receptor binding in the hypothalamus was determined by incubation of cytosol in the presence of a saturating concentration of 10 nM 3H-dexamethasone in a single-point assay. In both experiments, bound 3H-dexamethasone was separated from unbound steroid by precipitation with dextran-coated charcoal, and the receptor-3H-steroid complexes were counted in a spectral liquid scintillation counter (LKB Wallac, Turku, Finland). Available GR was determined from the amount of total 3H-dexamethasone binding that was displaced by the selective GR agonist RU 28362 (kindly donated by Roussel Uclaf, Romainville, France). Protein concentrations for each sample were determined by the method of Lowry et al. (1951), with bovine serum albumin as the standard. Data were expressed as fmol/mg protein.

Plasma Catecholamines and Monoamines in the Locus Coeruleus

Plasma concentrations of adrenaline (ADR) and noradrenaline (NA) were analyzed in duplicate using HPLC with electrochemical detection. Briefly, 1 mL plasma and 50 μL internal standard (500 pg dihydroxybenzylamine; DHBA) were added to extraction tubes with 20 mg aluminum oxide previously activated with 600 μL 2 M Tris/EDTA buffer. Samples were stirred for 10 min and centrifuged at 800 × g for 1 min at 4°C. The aluminum oxide with bound catecholamines was then washed 3 times with 1 mL of 16.5 mM Tris/EDTA solution followed by centrifugation. The elution of catecholamines was achieved by the addition of 120 μL of 200 mM perchloric acid and subsequent centrifugation at 800 × g for 1 min at 4°C. Aliquots of 40 μL were injected into the HPLC system (SHIMADZU Deutschland, Duisburg, Germany), which was equipped with a 250 × 4 mm column packed with Prontosil C18 AQ (Bischoff Analysentechnik, Leonberg, Germany). The mobile phase consisted of 58 mM sodium dihydrogen phosphate buffer containing 1.2 mM octanesulfonic acid, 0.3 mM EDTA, 0.2 mM potassium chloride, and 8% methanol (vol/vol) at pH 5.6. Electrochemical detection was performed using an ISAAC cell with a glassy carbon working electrode set at a potential of 600 mV (Axel Semrau, Sprockhövel, Germany). Single-point calibration curves were calculated using peak area versus analyte concentration, and response factors were evaluated for each compound in relation to DHBA. The intra- and interassay CV were 1.6% and 5.9%, respectively, for ADR and 1.0% and 2.4%, respectively, for NA.

HPLC with electrochemical detection was used to determine the concentrations of the following compounds in the locus coeruleus region: NA and its metabolite 3-methoxy-4-hydroxyphenylglycol (MHPG); dopamine (DA) and its metabolites 3,4-dihydroxyphenylacetic acid (DOPAC) and homovanillic acid (HVA); and 5-hydroxytryptamine (5-HT) and its metabolite 5-hydroxyindole-3-acetic acid (5-HIAA). The methods used were previously described in detail by Otten et al. (2010), and the concentrations of neurotransmitters and metabolites were expressed as pmol/mg tissue weight.

Adrenal Glands

The relative adrenal weight was calculated as the ratio of the weight of both adrenals to the BW of the animals. For histological analyses, the right adrenal gland was sectioned (7 µm) with a cryostat microtome (Reichert-Jung; Leica, Nussloch, Germany) transversely to the long axis in the midglandular region. Sections were stained using hematoxylin and eosin dye for visualization of the cortex and the medulla. Three sections per adrenal were analyzed using a video camera (CF15/2RGB; KAPPA, Gleichen, Germany) and an image analysis system (SIS, Münster, Germany). The adrenal capsule, the line of demarcation between the adrenal cortex and the medulla and any open spaces where a blood vessel resided were traced to calculate the areas of the entire gland, the cortex, and the medulla. The CV for the area of the 3 sections within the same gland was 1.6% for the adrenal cortex and 1.9% for the adrenal medulla. For each selected section, the number of labeled cells was counted, and the data were expressed as densities (number of cells/mm2).

Statistical Analysis

Data were evaluated by ANOVA using the MIXED procedure (SAS Inst. Inc., Cary, NC). The model for salivary cortisol of gilts comprised the fixed effects diet (LP, AP, HP) and replicate (1 to 6), the repeated factor GD (GD −4, GD 23, GD 65, GD 107), and all 2-way interactions with diet. Litter size was analyzed with a model comprising the fixed effect diet, replicate, and the 2-way interaction. The model for birth weight of offspring included the fixed effects of diet, sex (female, male), and replicate; the 2-way interactions with diet; and the random effect of sow (nested in diet and replicate) to model the correlation between offspring of the same sow. The model for basal blood and tissue parameters in offspring comprised the fixed effects diet, age (PND 1, PND 27, and/or PND 81), sex, replicate, birth weight class (light, medium, heavy), all 2-way interactions with diet, and the random sow effect. Age was used as a fixed effect because different animals were probed at the different ages. The birth weight class of piglets was included to account for possible weight effects on endocrine and tissue data and was determined according to the frequency distribution of birth weight within litters (25% light, 50% medium, 25% heavy), with each class represented within each dietary group. The model for blood parameters in response to weaning, ACTH, and insulin challenge included the fixed effects diet, sex, replicate, birth weight class, the repeated factor sampling time, all 2-way interactions with diet, and again the random sow effect. In case of NA during the insulin challenge test, significant differences between diet groups were already found for basal concentrations before insulin administration. Therefore, NA response to insulin was evaluated using Area under the Curve (AUC) adjusted for baseline. For AUC, the same model was used as for blood parameters in response to challenge but without the repeated factor sampling time. All pairwise differences among factors were tested using the Tukey-Kramer correction for multiple testing. The significance level for all statistical tests was set at P ≤ 0.05, and a tendency was considered when 0.05 < P < 0.1. The results are reported as least squares means ± SE.


RESULTS

General Observations

Litter size was not affected by the different maternal diets (LP = 12.6 ± 0.5, AP = 11.8 ± 0.5, HP = 11.9 ± 0.6; P = 0.521). However, significant effects on individual birth weight of piglets were found for maternal diet (LP = 1.21 ± 0.05 kg, AP = 1.41 ± 0.05 kg, HP = 1.24 ± 0.06 kg; P < 0.05) with decreased birth weights of LP compared with AP piglets (P < 0.05). In addition, there was also an effect of sex on birth weight (males = 1.31 ± 0.03 kg, females = 1.26 ± 0.03 kg; P < 0.05).

Salivary Cortisol Concentrations in Gilts

The maternal diet had an effect on salivary cortisol concentrations of gilts throughout gestation (P < 0.01). Tukey-Kramer test revealed increased values in LP gilts compared with AP and HP gilts (both P < 0.05; Fig. 1). In addition, salivary cortisol concentrations were affected by gestational day (P < 0.001) and increased from GD −5 until GD 108. There were significant differences between gestational days (P < 0.05) for all pairwise comparisons. No influences were found for the interaction diet × age.

Figure 1.
Figure 1.

Salivary cortisol concentrations in gilts 4 d before insemination and on gestational d 23, 65, and 107. Throughout the gestation period, gilts were fed low (6.5%), adequate (12.1%), or high (30%) protein diets, which were made isocaloric by adjusting the carbohydrate content. Values are least squares means ± SE and calculated from 10 to 12 sows per diet group. *P < 0.05 and #P < 0.1 according to the Tukey-Kramer test.

 

Plasma Catecholamine, Cortisol, and CBG Concentrations in Offspring

No effects of maternal diet were found for the plasma concentrations of ADR, NA, cortisol, and CBG (Table 1) in blood samples taken by venipuncture on PND 1 and 27. The age factor was significant for all blood variables (P < 0.01). According to the Tukey-Kramer test, plasma concentrations of all variables were increased on PND 1 compared with PND 27 (P < 0.01). No influences were found for diet × age, sex, and diet × sex.


View Full Table | Close Full ViewTable 1.

Plasma adrenaline, noradrenaline, cortisol, and corticosteroid-binding globulin (CBG) concentrations in piglets on postnatal d 1 and 27. Piglets were born to gilts fed diets with low, adequate, or high protein levels throughout gestation1

 
Diet2
P-value
Item Age, d n/diet LP AP HP Diet Age Sex Diet × Age Diet × Sex
Plasma adrenaline, nmol/L 1 35 to 42 4.66 ± 0.48 5.23 ± 0.51 4.56 ± 0.43 0.729 < 0.001 0.509 0.203 0.704
27 16 to 18 2.61 ± 0.72 2.66 ± 0.73 3.71 ± 0.61
Plasma noradrenaline, nmol/L 1 34 to 42 10.92 ± 1.06 11.90 ± 1.11 10.76 ± 0.91 0.465 < 0.001 0.270 0.368 0.929
27 16 to 18 4.42 ± 1.57 6.90 ± 1.58 7.43 ± 1.31
Plasma cortisol, nmol/L 1 17 to 23 349 ± 52 393 ± 74 337 ± 48 0.757 < 0.001 0.363 0.617 0.870
27 10 to 11 113 ± 76 67 ± 86 28 ± 62
Plasma CBG, pmol/L 1 21 to 26 40.5 ± 4.1 38.2 ± 4.4 44.7 ± 3.6 0.490 0.004 0.111 0.985 0.947
27 10 to 11 28.2 ± 7.2 27.6 ± 6.3 33.7 ± 5.4
1Values are least squares means ± SE.
2Diets: LP = low protein (6.5%); AP = adequate protein (12.1%); HP = high protein (30%).

Brain Measures

No effects of diet or diet × age were found for GR binding in the hypothalamus (PND 1: LP = 23.2 ± 4.9 fmol/mg protein, AP = 26.3 ± 4.6 fmol/mg protein, HP = 18.6 ± 3.8 fmol/mg protein; PND 27: LP = 32.5 ± 3.5 fmol/mg protein, AP = 25.5 ± 5.2 fmol/mg protein, HP = 28.8 ± 4.6 fmol/mg protein; diet: P = 0.693, diet × age: P = 0.214). In the hippocampus, GR binding was also not affected by diet or diet × age (PND 1: LP = 36.3 ± 1.9 fmol/mg protein, AP = 32.4 ± 2.3 fmol/mg protein, HP = 37.0 ± 2.6 fmol/mg protein; PND 27: LP = 50.8 ± 2.4 fmol/mg protein, AP = 44.4 ± 2.9 fmol/mg protein, HP = 47.5 ± 2.9 fmol/mg protein; diet: P = 0.179, diet × age: P = 0.637). The age factor was significant for GR binding in the hypothalamus (P < 0.05) and hippocampus (P < 0.001), with increased values on PND 27 compared with PND 1. No effects of sex or diet × sex were found for GR binding in these brain regions.

There were no effects of diet on the monoamine concentrations in the locus coeruleus region. However, the significant interaction of diet × age for DA (P < 0.05) revealed that DA concentrations decreased from PND 1 to 27 only in HP piglets (P < 0.05), whereas no alterations were found in the other diet groups (Table 2). In addition, NA, MHPG, DOPAC, HVA, 5-HT, and 5-HIAA were significantly influenced by age. On PND 27, NA concentrations were increased (P < 0.001), whereas MHPG, DOPAC, HVA, 5-HT, and 5-HIAA were reduced compared with the values on PND 1 (all: P < 0.01) independent of diet (Table 2). The significant effect of sex on DOPAC and HVA revealed lower concentrations of DOPAC (castrated males = 1.67 ± 0.11 pmol/mg, females = 1.32 ± 0.11 pmol/mg, P < 0.01) and HVA (castrated males = 2.56 ± 0.14 pmol/mg, females = 2.18 ± 0.14 pmol/mg, P < 0.05) in female piglets. The F-test for the diet × sex interaction was significant for NA, but the Tukey-Kramer test indicated no significant pairwise differences.


View Full Table | Close Full ViewTable 2.

Concentrations of monoamines and metabolites in the locus coeruleus of piglets at 1 and 27 d of age, born to sows fed diets with low, adequate or high protein levels throughout gestation1

 
Diet2
P-value
Item3 Age, d n/diet LP AP HP Diet Age Sex Diet × Age Diet × Sex
NA, pmol/mg 1 40 to 48 6.02 ± 0.43 6.27 ± 0.47 5.89 ± 0.55 0.846 < 0.001 0.752 0.952 0.026
27 20 to 24 7.36 ± 0.52 7.44 ± 0.58 7.03 ± 0.60
MHPG, pmol/mg 1 13 to 22 0.46 ± 0.10 0.74 ± 0.08 0.53 ± 0.09 0.187 < 0.001 0.228 0.743 0.083
27 7 to 8 0.30 ± 0.12 0.51 ± 0.10 0.40 ± 0.12
DA, pmol/mg 1 40 to 48 1.01 ± 0.08 1.06 ± 0.10 1.08 ± 0.10A 0.286 0.123 0.530 0.028 0.619
27 20 to 24 0.95 ± 0.11 1.17 ± 0.12 0.72 ± 0.12B
DOPAC, pmol/mg 1 32 to 35 1.63 ± 0.17 1.65 ± 0.17 1.68 ± 0.19 0.803 0.009 0.004 0.429 0.527
27 17 to 18 1.37 ± 0.21 1.48 ± 0.23 1.14 ± 0.20
HVA, pmol/mg 1 40 to 48 2.63 ± 0.19 2.58 ± 0.22 2.90 ± 0.24 0.741 < 0.001 0.014 0.270 0.374
27 20 to 24 1.87 ± 0.25 2.28 ± 0.28 1.97 ± 0.27
5-HT, pmol/mg 1 40 to 48 4.99 ± 0.29 5.18 ± 0.31 5.58 ± 0.37 0.565 0.002 0.360 0.598 0.144
27 20 to 24 4.68 ± 0.33 4.43 ± 0.37 4.92 ± 0.40
5-HIAA, pmol/mg 1 40 to 48 7.22 ± 0.41 7.21 ± 0.45 7.39 ± 0.52 0.832 < 0.001 0.326 0.826 0.541
27 20 to 24 3.18 ± 0.49 2.90 ± 0.55 3.52 ± 0.57
A,BWithin diet HP: only HP offspring showed age-dependent differences (P < 0.05) according to Tukey-Kramer test.
1Values are least squares means ± SE.
2Diets: LP = low protein (6.5%); AP = adequate protein (12.1%); HP = high protein (30%).
3NA = noradrenaline; MHPG = 3-methoxy-4-hydroxyphenylglycol; DA = dopamine; DOPAC = 3,4-dihydroxyphenylacetic acid; HVA = homovanillic acid; 5-HT = 5-hydroxytryptamine; 5-HIAA = 5-hydroxyindole-3-acetic acid.

Adrenal Gland

The relative adrenal weight was not affected by diet or diet × age. The diet × age interaction had an effect on the total adrenal area (P = 0.050) and the adrenal medulla area (P < 0.01; Table 3). LP offspring exhibited a greater total adrenal area compared with AP offspring and a greater adrenal medulla area compared with HP offspring on PND 81 (both P < 0.05). Cell density in the adrenal medulla tended to be affected by the maternal diet (P = 0.090), with a lower cell density observed in LP offspring compared with HP offspring (P = 0.093). Cell density in the adrenal medulla was influenced by diet × sex (P < 0.05), with castrated male LP offspring having a lower density compared with castrated male AP offspring (LP males = 5388 ± 130 cells/mm2, AP males = 5901 ± 141 cells/mm2; P < 0.01). The age factor was significant for all adrenal measures (P < 0.001); as age increased, the values for relative adrenal weight and cell density in the medulla and cortex decreased, and the values for all area measures increased (Table 3). Differences between sexes were found for the total adrenal area (castrated males = 20.2 ± 0.5 mm2, females = 18.9 ± 0.5 mm2; P < 0.05) and the cortex area (castrated males = 15.2 ± 0.4 mm2, females = 14.2 ± 0.4 mm2; P < 0.05).


View Full Table | Close Full ViewTable 3.

Relative adrenal weight, areas, and cell density of adrenal cortex and medulla of piglets and juvenile pigs at 1, 27, and 81 d of age, born to sows fed diets with low, adequate, or high protein levels throughout gestation1

 
Diet2
P-value
Item Age, d n/diet LP AP HP Diet Age Sex Diet × Age Diet × Sex
Relative adrenal weight, mg/kg BW 1 35 to 43 227 ± 6 216 ± 7 217 ± 6 0.954 < 0.001 0.137 0.660 0.746
27 18 to 21 126 ± 7 125 ± 8 128 ± 8
81 16 to 19 78 ± 8 83 ± 9 81 ± 8
Total adrenal area, mm2 1 29 to 35 9.0 ± 0.9 8.9 ± 1.0 8.0 ± 0.9 0.082 < 0.001 0.030 0.053 0.106
27 18 to 21 19.3 ± 1.0 18.2 ± 1.1 17.7 ± 1.1
81 18 to 21 34.6 ± 1.0a 29.8 ± 1.2b 30.3 ± 1.1a,b
Cortex area, mm2 1 29 to 35 6.4 ± 0.7 6.3 ± 0.7 5.8 ± 0.7 0.144 < 0.001 0.037 0.275 0.064
27 18 to 21 14.7 ± 0.8 13.3 ± 0.8 13.7 ± 0.8
81 18 to 21 25.7 ± 0.8 22.8 ± 0.9 23.5 ± 0.8
Medulla area, mm2 1 29 to 35 2.57 ± 0.38 2.58 ± 0.42 2.18 ± 0.40 0.152 < 0.001 0.131 0.007 0.476
27 18 to 21 4.61 ± 0.43 4.84 ± 0.45 4.10 ± 0.45
81 18 to 21 8.88 ± 0.43a 6.96 ± 0.48a,b 6.89 ± 0.45b
Cell density cortex, number of cells/mm2 1 29 to 35 3874 ± 102 3988 ± 114 3851 ± 109 0.219 < 0.001 0.806 0.925 0.290
27 18 to 21 3740 ± 117 3976 ± 126 3662 ± 121
81 18 to 21 3387 ± 119 3571 ± 134 3370 ± 123
Cell density medulla, number of cells/mm2 1 29 to 35 6291 ± 129 6734 ± 142 6736 ± 138 0.090 < 0.001 0.440 0.470 0.029
27 18 to 21 5423 ± 146 5501 ± 154 5762 ± 152
81 18 to 21 4657 ± 148 4975 ± 162 4945 ± 154
a,bWithin a row: values not sharing a common superscript differ significantly (P < 0.05) according to Tukey-Kramer test.
1Values are least squares means ± SE.
2Diets: LP = low protein (6.5%); AP = adequate protein (12.1%); HP = high protein (30%).

Response to Weaning, ACTH, and Insulin Challenge

The cortisol and CBG concentrations in plasma samples taken before (PND 27) and 1 d after weaning (PND 29) revealed an effect of time (all: P < 0.01), but no main effect of diet. For the plasma cortisol concentrations, the Tukey-Kramer test indicated an increase only in LP piglets (P < 0.05; Fig. 2A). Plasma concentrations of CBG decreased due to weaning stress (effect of time: P < 0.001; Fig. 2B). No influence was found for sex or diet × sex.

Figure 2.
Figure 2.

(A) Plasma cortisol and (B) corticosteroid-binding globulin (CBG) concentrations in piglets on postnatal d 27 (before weaning) and 29 (1 d after weaning). Piglets were born to gilts fed diets with low (6.5%), adequate (12.1%), or high (30%) protein levels throughout gestation, and the diets were made isocaloric by adjusting the carbohydrate content. Values are least squares means ± SE and calculated from 16 to 18 piglets per diet group. *P < 0.05 according to the Tukey-Kramer test.

 

After i.v. ACTH injection on PND 68, plasma cortisol concentrations increased, reached peak concentrations 60 min after administration and returned to baseline after 120 min (effect of sampling time: P < 0.001; Fig. 3A). In contrast, ACTH induced a decrease in plasma CBG concentrations with a nadir after 40 min and a return to baseline concentrations after 120 min (effect of sampling time: P < 0.001; Fig. 3B). No effects of maternal diet, sex, or interactions were found with regard to the HPA response.

Figure 3.
Figure 3.

(A) Plasma cortisol and (B) corticosteroid-binding globulin (CBG) concentrations of juvenile pigs before and after intravenous ACTH challenge (0.5 IU per kg BW) on postnatal d 68. Pigs were born to gilts fed diets with low (6.5%), adequate (12.1%), or high (30%) protein levels throughout gestation, and the diets were made isocaloric by adjusting the carbohydrate content. Values are least squares means ± SE calculated from 13 to 23 animals per diet group.

 

In basal samples taken immediately before insulin administration on PND 70, ANOVA revealed an effect of maternal diet on NA plasma concentrations (P < 0.05) but not on ADR, cortisol, or CBG concentrations. The basal plasma NA concentrations of HP offspring were increased compared with AP offspring (P < 0.05) and tended to be increased compared with LP offspring (P = 0.068; LP = 1.06 ± 0.12 nmol/L, AP = 1.06 ± 0.12 nmol/L, HP = 1.49 ± 0.11 nmol/L). The i.v. injection of insulin-induced hypoglycemia, with a minimum plasma glucose concentration 15 min after administration and a return to the normal range after 3 h; no differences were observed between dietary groups (data not shown). The factor sampling time was significant for all variables examined (P < 0.001). ADR, NA, and cortisol increased due to hypoglycemia, reached peak concentrations after 45 min (cortisol after 60 min) and returned to basal concentrations after 120 min (Fig. 4A to 4C). The maximum concentrations reached were roughly 23-fold (ADR), 2.8-fold (NA), and 3.4-fold (cortisol) greater than the respective basal concentrations. Concentrations of CBG decreased after insulin administration with a nadir after 45 min and a return to basal concentrations after 90 min (Fig. 4D). The maternal diet had no effects on plasma ADR, cortisol, and CBG concentrations after insulin administration. Evaluation of AUC corrected for baseline also revealed no effect of diet on NA response to insulin (LP = 32,042 ± 5871, AP = 19,471 ± 4590, HP = 24,526 ± 4721; P = 0.266). Plasma CBG was affected by diet × sex with female LP pigs having greater concentrations than castrated male LP pigs (castrated LP males = 32.8 ± 5.9 pmol/L, LP females = 53.6 ± 5.9 pmol/L; P < 0.05).

Figure 4.
Figure 4.

(A) Plasma adrenaline (ADR), (B) noradrenaline (NA), (C) cortisol, and (D) corticosteroid-binding globulin (CBG) concentrations of juvenile pigs before and after intravenous insulin challenge (0.5 IU per kg BW) on postnatal d 70. Pigs were born to gilts fed diets with low (6.5%), adequate (12.1%), or high (30%) protein levels throughout gestation, and the diets were made isocaloric by adjusting the carbohydrate content. Values are least squares means ± SE calculated from 13 to 23 animals per diet group.

 


DISCUSSION

We recently reported that diets with low and high protein:carbohydrate ratios fed to gilts during gestation alter maternal macronutrient metabolism and cause IUGR (Rehfeldt et al., 2011; Metges et al., 2012). To elucidate the long-term physiological consequences of this maternal low and high protein:carbohydrate model in the progeny, we investigated the function of the HPA and SAM axes in the offspring of LP, AP, and HP gilts under basal and specific challenge conditions as well as the dietary effect on maternal salivary cortisol concentrations.

Salivary Cortisol in Sows

Our results showed a significant effect of gestation stage on maternal salivary cortisol with increasing concentrations in all 3 diet groups throughout gestation. Similar increases of cortisol during gestation were previously reported for pigs (Kattesh et al., 1997; Hay et al., 2000) and humans (Allolio et al., 1990). It is hypothesized that increased cortisol concentrations may be the result from antiglucocorticoid effects of elevated progesterone concentrations over the course of pregnancy (Allolio et al., 1990), or that placental corticotropin-releasing hormone may influence maternal adrenocortical function. Furthermore, placenta-derived ACTH represents an autonomous continuous source that may overlay the normal pituitary ACTH production (Lindsay and Nieman, 2005). In the present study, the LP diet caused greater salivary cortisol concentrations compared with AP and HP sows, an effect which was more pronounced during late gestation. The increased salivary cortisol reflects a greater concentration of unbound biologically active cortisol in LP sows. In an additional study using the same dietary model, we also found increased concentrations of salivary cortisol in LP sows during gestation and a disruption of the normal daily rhythm in the circulating plasma cortisol concentration at the end of gestation (Kanitz et al., 2012). Our results are consistent with previous findings pertaining to undernutrition during gestation, including elevated maternal cortisol concentrations in sheep (Edwards and McMillen, 2001; Chadio et al., 2007) and increased maternal plasma corticosterone concentrations in rats (Lesage et al., 2001). Similarly, increased glucocorticoid concentrations were observed in a rat model for isoenergetic protein restriction (Herbert and Carrillo, 1982). Protein-caloric malnutrition disturbs the energy balance of the organism proopiomelanocortin (POMC) of the endogenous responses to preserve this balance is the activation of the HPA axis. Activation of the HPA axis and increased glucocorticoid concentrations due to protein malnutrition may be partially explained by an increased expression of pituitary POMC mRNA, which elevates ACTH plasma concentrations. This stimulatory effect on the HPA axis was previously found to be specific for protein restriction (Jacobson et al., 1997).

HPA Axis in Offspring

The maternal diet was not found to influence HPA axis regulation in the offspring on the basis of basal plasma cortisol and CBG or GR binding in the hippocampus and hypothalamus. Activity of the HPA axis is controlled via several negative feedback loops mediated by central corticosteroid receptors. Although hippocampal mineralocorticoid receptors play a major role in the control of proactive feedback of glucocorticoids involved in the maintenance of basal HPA activity, the progressive occupancy of GR after the increase of glucocorticoid concentrations seems to mediate, in coordinated manner with mineralocorticoid receptors, the reactive feedback aimed at controlling the HPA response to stress (de Kloet et al., 1998). With respect to these receptor regulations, it was found in rats that moderate maternal protein deficiency (50%) during gestation can modulate central glucocorticoid action through decreased abundance of GR mRNA and protein expression in the hypothalamus of offspring (Bertram et al., 2001). Another study in rats showed increased GR binding in the hippocampus but not in the hypothalamus in offspring exposed to an LP diet in utero (Langley-Evans et al., 1996). Our study is the first investigating the effects of different maternal dietary protein:carbohydrate ratios on GR binding in the brains of pig offspring. Previous studies of our group on pigs revealed that maternal stress or glucocorticoid exposure during gestation decrease GR binding in the hypothalamus of neonates, which may attenuate HPA axis feedback (Kanitz et al., 2003, 2006). The relative adrenal weight, the area of the adrenal cortex, and the cortical cell density did also not exhibit diet-dependent differences but were influenced by age. The lack of differences between the effects of the treatments in the present study indicates that HPA regulation in offspring is not affected by the maternal diets under basal conditions.

Weaning stress caused an increase of plasma cortisol in LP piglets 1 d after separation from the foster mother and transferal to a new housing environment. This indicates that a maternal dietary protein deficiency causes a moderate hyperactivity of the HPA axis in the offspring under stress situations such as weaning. At PND 68 and 70, stimulation of adrenal cortisol secretion by exogenous ACTH or HPA stimulation by insulin-induced hypoglycemia revealed no differences between the diet groups as measured by plasma cortisol and CBG. The divergent effect of the maternal diets on the cortisol response to weaning compared with stimulation with ACTH and hypoglycemia may be due to the different age of the animals at the time of the tests. Supporting this notion, studies in rats show that maternal food restriction differently affects the HPA axis activity in the offspring when measured at weaning, in young, and in older adults (Lesage et al., 2006). On the other hand, the different effect of the maternal diet on the HPA response of the offspring in our study may be also due to the different challenges used. Whereas insulin-induced hypoglycemia induces a counter-regulatory response and ACTH acts directly at the adrenal, weaning is a complex stress situation and includes the perception of the stressor and central activation of the HPA axis. Thus, possible alterations in stress perception and regulation of the offspring due to a protein-deficient maternal diet may be seen in their cortisol response to stress but not in their response to an ACTH or insulin challenge.

Neurotransmitters and SAM Axis in Offspring

The locus coeruleus is a key brainstem region of the noradrenergic system involved in sympathetic activation, stress responses, arousal, and anxiety (Berridge and Waterhouse, 2003; Berridge 2008; Valentino and Van Bockstaele, 2008). The maternal diets had no effects on concentrations of NA and its metabolite MHPG in this region. However, only HP piglets showed a decrease in DA concentrations from PND 1 to 27, as indicated by the significant diet × age interaction. This finding may indicate rather a decreased DA synthesis than an enhanced DA turnover in HP offspring at PND 27 as there were no differences in the concentrations of DA metabolites. Dopamine in this region may derive from the high density of noradrenergic neurons and may serve as the precursor for NA. The maternal diets in our study differed in their tyrosine content, with the HP diet having 19% and 46% more tyrosine compared with the AP and LP diets, respectively. The HP and AP sows also exhibited greater plasma tyrosine concentrations compared with LP sows (Metzler-Zebeli et al., 2012). Thus, postnatal catecholamine synthesis and degradation may be programmed by a greater prenatal availability of tyrosine. Interestingly, LP maternal diets in rats led to decreased basal DA concentrations in the medial prefrontal cortex and a reduced stress-induced DA response in this region (Mokler et al., 2007). Furthermore, prenatally stressed rats also exhibited reduced DA concentrations and increased DA turnover in the locus coeruleus (Takahashi et al., 1992), indicating that these alterations of neurotransmitter systems may contribute to the changes observed in the reactivity of stress-sensitive systems in prenatally malnourished or stressed offspring.

A high or low protein intake may also affect the availability of essential AA such as tryptophan. Previous studies in pigs and other species have shown that deficiency or supplementation of dietary tryptophan is able to impact the central serotonergic system as indicated by altered 5-HT and 5-HIAA concentrations within different brain regions (Henry et al., 1996; Pastuszewska et al., 2007), and may affect the HPA and SAM responses to stress (Koopmans et al., 2005; Guzik et al., 2006). In a parallel experiment using the same dietary model as in our study, LP sows showed decreased plasma tryptophan concentrations at the end of gestation, whereas no differences between diets were found in fetal plasma tryptophan (Metzler-Zebeli et al., 2012). This may explain the lack of differences between diets for 5-HT and 5-HIAA levels in the brain of offspring.

Hypoglycemia due to the insulin challenge at PND 70 induced a counter-regulatory endocrine response with a strong increase of ADR concentrations and increases of NA and cortisol with lesser magnitudes. No effect of the maternal diets was found for the ADR, NA, and cortisol response to insulin-induced hypoglycemia. However, HP offspring exhibited increased plasma NA concentrations compared with AP offspring at basal concentrations before insulin administration. The increase of basal NA concentrations may be of both central and adrenal origin. Plasma NA concentration is normally seen as an index of sympathetic neural activity under basal conditions, thus indicating an increased activation in HP offspring, whereas the NA response to hypoglycemia is derived largely from the adrenal medulla (DeRosa and Cryer, 2004). On the other hand, perinatal food restriction induced morphological alterations of noradrenergic but not adrenergic clusters in the adrenal medulla of rat offspring, suggesting that NA chromaffin cells are more sensitive to maternal malnutrition than ADR chromaffin cells (Laborie et al., 2011).

In LP offspring, an increased total adrenal area was observed on PND 81, which was largely attributable to an enhanced adrenal medulla area. Cell densities in the adrenal medulla tended to be less only in LP pigs, indicating that the enlargement was primarily due to hyperplasia rather than cell hypertrophy. However, castrated male LP pigs also exhibited a decreased cell density compared with castrated male AP pigs, which suggests that the proliferative process may differ between castrated males and females. In rats, maternal pre- and postnatal dietary protein restriction results in an increased adrenal-to-body weight ratio, an increased medullary area, decreased glucose concentrations (Fernandez-Twinn et al., 2006), and enhanced blood catecholamine concentrations in offspring after feeding (Petry et al., 2000). In the rat model of perinatal undernutrition, maternal food restriction caused enhanced plasma catecholamine concentrations and altered noradrenergic chromaffin cell aggregation and nerve fiber fasciculation in the adrenal medulla of the offspring (Molendi-Coste et al., 2006, 2009). Using our pig model with low and high protein:carbohydrate ratios, we previously found that the HP diet is accompanied by glucose and energy deficits during gestation (Metges et al., 2012). Furthermore, we found that offspring of both LP and HP sows were less sensitive to insulin and required more insulin to metabolize an i.v. glucose load at 10 wk of age (Metges et al., 2009). The restriction of nutrients by inadequate maternal diets may therefore result in adaptive changes in the developing adrenals that help maintain the metabolic and glycemic balance in the offspring through the counter-regulatory action of catecholamines.

In summary, our results indicate that maternal diets during gestation with aberrant low protein:high carbohydrate or high protein:low carbohydrate ratios have moderate but long-lasting effects on the HPA and SAM system in the offspring. In the protein-deficient group, the action of glucocorticoids may mediate these processes as indicated by increased maternal cortisol concentrations and an increased cortisol response of piglets to weaning. In HP animals, the different bioavailability of amino acids during gestation (Metzler-Zebeli et al., 2012), serving as precursors for neurotransmitter and catecholamine synthesis, may also contribute to the alterations of central DA and plasma NA concentrations in the offspring. In conclusion, our data suggest that imbalanced maternal protein:carbohydrate nutrition may moderately alter structural and functional maturation of the HPA and SAM axes in the offspring of pigs, which show a considerable plasticity to cope with maternal malnutrition.

 

References

Footnotes


Comments
Be the first to comment.



Please log in to post a comment.
*Society members, certified professionals, and authors are permitted to comment.