Pre-conceptional diet manipulation and fetus number can influence placenta endocrine function in sheep

Abstract

Changes in maternal nutrition during pregnancy can result in profound effects on placental function and fetal development. Although the pre-conceptional period holds the potential to reprogram embryonic and placental development, little is known regarding the effects of pre-mating nutritional manipulation on placental function and fetal and postnatal offspring growth. To test this, Polypay-Dorset sheep (n=99) were assigned to one of three nutritional treatments (n=33/treatment) receiving 50% (UN: undernutrition), 100% (C: control), or 200% (ON: overnutrition) of maintenance energy requirements for 21 days before mating during April-May (increasing photoperiod). Thereafter, diets were the same across groups. We evaluated maternal reproductive variables and maternal and offspring weight and body mass index through weaning. Maternal plasma was collected through pregnancy until postnatal day 1 to assay pregnancy associated glycoproteins (PAG) and progesterone. Fertility rate was similar among treatments, but ON females had a higher reproductive rate (UN: 82%; C:100% ON:145%). When correcting by total birth weight, twin pregnancies had lower PAG and progesterone vs. singleton pregnancies (P<0.001). At birth, UN lambs were heavier than C lambs regardless of birth type (P<0.01). Growth velocity, daily gain, and weaning weight were similar, but UN and ON females grew faster and were heavier at weaning vs. C females. We demonstrated that a 3-week pre-conceptional maternal under- or overnutrition, when correcting by total birth weight, results in lower endocrine capacity in twin pregnancies. Pre-conceptional maternal under- and overnutrition increased postnatal female lamb growth, suggestive of reprogramming of pathways regulating growth before conception. This highlights how pre-conceptional nutrition can result in marked sex-specific differences.

1. Introduction

The relationship between gestational malnutrition and adult metabolic diseases was first described by Barker in 1989 [1] and later conceptualized in the developmental origins of health and disease (DOHaD) theory. Gestational undernutrition results in reduced birth weight and can ultimately lead to chronic diseases in later life, including insulin resistance, type 2 diabetes, and metabolic syndrome in humans [2,3] and animals [4]. Although the evidence supportive of overnutrition programming in humans is scarce [5], animal studies have demonstrated that gestational overnutrition can result in increased risk of insulin resistance and obesity in the offspring [6,7]. Additionally, gestational undernutrition can lead to a puberty delay and reduced pregnancy rate [8,9], while gestational overnutrition can lead to early puberty onset and decreased pregnancy rate [10,11]. However, to our knowledge, the impact of pre-conceptional overnutrition on placenta function in large animals has not been evaluated.

The fetus is sensitive to maternal nutrition and adapts to nutritional challenges by modulating its growth [12]. Studies have demonstrated the importance of the window of exposure on offspring birth weight, with gestational undernutrition in mid- or late gestation resulting in reduced birth weight; but not when undernutrition occurred during early pregnancy [13,14]. Because a relative body fatness is required for pubertal onset in humans [15] and sheep [16,17], offspring that are born small and display slow postnatal growth can have delayed puberty onset and decreased reproductive success. However, smaller offspring can also grow faster – referred to as catch-up growth - and be at risk of metabolic diseases in later life, such as insulin resistance, obesity, and metabolic syndrome [18,19,20]. On the other side of the spectrum, gestational overnutrition increases birth weight and postnatal offspring growth [21,22]. Faster postnatal growth tends to increase adiposity and the development of insulin resistance [5,21,23]. Although extensive efforts have focused on the effects of gestational maternal nutrition on offspring growth and health [20,24], whether these effects extend to the pre-conceptional period remain unknown.

Placental size and nutrient transfer capacity determine fetal growth and birth weight [25,26]. Gestational under- or overnutrition can change the structure and function of the placenta, thus altering the transport efficiency of essential nutrients to the fetus [27] at the trophoblast layer within the placentomes, where nutrient and gas exchange occurs [28,29]. Moreover, trophoblast binucleate cells secrete pregnancy-associated glycoproteins [PAGs; 30,31]. PAGs have been proposed to play a key role in pregnancy maintenance [32,33] and are considered as good indicators of feto-placental well-being [34,35]. Importantly, circulating PAGs can be increased or reduced by gestational undernutrition [36] or overnutrition [37], respectively.

In the U.S., at least 30 million people suffer from eating disorders that are associated with poor health [38,39]. Among them, a high proportion of pregnant females do not know they are pregnant up to ~6 weeks after conception; thus, it is highly likely that they kept their eating habits before and during the first weeks of pregnancy [40]. Similarly, in animal production systems, producers increase feed intake of the reproductive flock before conception and during breeding to increase ovulation rate. Depending on the duration of the period of increased diet, females can gain weight before conception [41]. The period around conception is potentially a vulnerable period where prenatal programming might initiate and lead to chronic diseases later in life [42]. In this study, we aimed to determine if a short (21 days) pre-conceptional nutritional manipulation (overnutrition and undernutrition) alters placental endocrine function, fetal growth, and postnatal growth trajectory in the progeny. Therefore, we hypothesized that pre-conceptional undernutrition will decrease the placental endocrine function resulting in reduced fetal growth and retarded postnatal growth trajectory in the progeny; whereas pre-conceptional overnutrition will increase the placental endocrine function resulting in increased fetal growth and accelerated postnatal growth trajectory in the progeny.

2. Material and Methods

All procedures in this study were approved by the Institutional Animal Care and Use Committee of Michigan State University (MSU), are consistent with National Research Council’s Guide for the Care and Use of Laboratory Animals, and meet the ARRIVE guidelines for reporting animal research [43]. The study was conducted at the MSU Sheep Teaching and Research Center (East Lansing, MI; 42.7°N, 84.4°W).

2.1. Experimental design

To investigate the effect of pre-conceptional nutrition on placental function, pregnancy outcomes and progeny growth, 99 crossbreed Polypay x Dorset mature sheep (average age of 2.7 ± 1.1 years old) were included in the study from the MSU in-house flock managed on an 8-month accelerated production system. Within this system, mating periods take place three times per year (August/September, December/January, or April/May), allowing for a consistent supply of lambs throughout the year and improved flock production efficiency. Three pre-conceptional nutrition groups were established randomly while ensuring that groups were blocked at enrolment by animal body weight, body score condition and subcutaneous adipose tissue depth (Figure 1). The pre-conceptional nutrition treatments were designed to achieve 50% (n = 33; undernutrition group: UN), 100% (n = 33; control group: C) or 200% (n = 33; overnutrition group: ON) of the energy requirements for maintenance [44] 21 days before breeding. Details of diets are described below. During the pre-conceptional period, animals were housed in separate indoor pens by group, and a vasectomized teaser ram introduced into each pen to stimulate, induce and detect the onset of estrus activity. Teaser rams were rotated among groups daily (pre-conception period; Figure 1).

Figure 1.

Schematic of experimental design with four periods; pre-conception (April), breeding (April - May), pregnancy (June - September), and postnatal offspring growth (October - November). Shaded area during the pregnancy period represents time that animals were housed outdoors (see text for details). Day 0 represents the start of the nutritional treatment. Closed arrows represent the three timepoints (before and after the pre-conceptional period and after the breeding period) at which body condition score (BCS) and subcutaneous adipose tissue depth were performed. Nutritional treatment occurred during the pre-conceptional period (noted as double dashed lines).

On Day 21, animals were housed in three breeding indoor pens where they all received the same diet (see Diet section). Breeding soundness examinations were performed to evaluate fertility in three Polled Dorset rams used for breeding. A ram bearing a marking harness (BreedingMark; Hamilton, New Zealand) was introduced into each breeding pen. Rams were rotated every day among breeding pens and removed after 34 days (two full reproductive cycles). The males marking harnesses with crayons were checked daily and changed as needed. Crayon marks on the females’ rump were checked and recorded daily to establish day of conception.

The breeding period occurred during April and May, as part of the 8-month accelerated system. In this latitude, the increase in day length during Spring results in sub-optimal reproductive performance, when both, conception rate and litter size are lower as compared to mating during a time of declining photoperiod (September to January). After the end of the breeding period, all ewes were combined into a single group and moved into a rotational grazing in a 25-ha paddock with access to clean water (shaded pregnancy period, Figure 1). Thirty days prior to lambing, pregnant ewes were moved back indoors and stalled (20 x 10 m pens) according to the predicted number of fetuses by ultrasonography (see details below).

2.2. Maternal diets

Indoor diets, provided daily between 3 to 4 pm, consisted of a total mixed ration (TMR) according to the nutritional requirements during pre-conception, breeding, and early pregnancy [44]. As in previous studies [45, 46, 47], treatments were applied to in a group setting. In this study, a total mixed ration was provided in a feeder with sufficient space (36 cm of linear space per ewe) to minimize competition allowing each animal free access to consume its feed allocation. During the pre-conception period, diets were based on corn silage and corn grain which were added as needed to modify the energy content. Thereafter, all groups had the same diet (Table 1). Nutritional composition of diets included dry matter, crude protein and additional fractions required to derive metabolizable energy concentration (non-fiber carbohydrates, ether extract, neutral detergent fiber) and were assessed by wet chemistry methodologies (Dairy One, Ithaca, New York, USA) and are detailed in Table 1.

Table 1.

Nutrient composition (DM basis) of the total mixed ration diet offering during pre-conception, breeding and early pregnancy.

Diet treatment groups % Nutrient Requirements	Pre-conception Period			Breeding	Early Pregnancy
	UN 50%	C 100%	ON 200%
Dry Matter (DM)	0.43	0.43	0.44	0.43	0.39
Metabolizable Energy (ME; Mcal/kg)	2.48	2.48	2.66	2.48	2.48
Crude Protein (%)	12.30	12.30	12.00	12.30	16.80
ME intake (Mcal/ewe/d)	1.53	3.06	6.10	3.56	3.82

Metabolizable energy content in diet differ only during the pre-conception period. ME content was calculated from digestible energy as defined in NRC [48]. C: control group, UN: underfed group, ON: overfed group.

2.3. Pregnancy evaluation

All animals in the study were scanned on four dates (June 6, June 15, June 27, and July 11) to confirm pregnancy status and reproductive rate (number of fetuses) by transabdominal ultrasonography. At those dates, pregnant females were between gestational days 30 and 80 based on day of observed mating. An ultrasound (GE LOGIQ Book XP Vet; Boston, Massachusetts, USA) fitted to a 4 MHz transabdominal convex probe was used. Pregnancy was confirmed by detection of uterine fluid, placentomes, fetal membranes, and/or fetus (es). Gestational age was confirmed by assessing uterine depth (in early pregnancy), fetus length from crown to rump, biparietal diameter, and calcification of the fetal ribs and/or skull. Fetal demise was also noted on one ewe between ultrasonographic scans. Fetal demise was determined by fetal membrane detachment in combination with lack of fetal heartbeat.

2.4. Maternal body condition and fatness assessment

Throughout the experiment, weekly maternal body weights were recorded and used to determine the body weight change during the diet and breeding periods. Maternal body condition score (BCS) and subcutaneous adipose tissue depth were recorded at the start (day 0) and end (day 21) of the pre-conceptional diet and end (day 56) of the breeding period (see Figure 1). BCS was assessed by the same operator using a 5-point scale, with 1 being emaciated and 5 being obese [49]. Subcutaneous adipose tissue depth was assessed at a point 45 mm from the midline over the twelfth rib using ultrasound. Transdermal ultrasonographies were performed with an ultrasound machine (GE; Boston, Massachusetts, USA) equipped with a transabdominal multifrequency probe (2.2-10 MHz). The lumbar region was first clipped and then, with the animal standing, ultrasound gel was applied to the transducer, as a coupling medium for probe contact. The transducer was then placed directly in contact with the skin and the subcutaneous adipose tissue recorded. The probe generated a cross-sectional image of the subcutaneous adipose tissue depth.

2.5. Circulating hormonal and protein concentrations

Maternal plasma samples were collected weekly at 8:00 AM and in the same day of the week throughout the experiment. Only samples on gestational days (GD ± SEM) 15 ± 1.1, 30 ± 0.8, 60 ± 1.2, 90 ± 0.9, 120 ± 0.7, 140 ± 0.9 and the day after lambing were assayed for PAG and progesterone. The Bovine Pregnancy Test Kit (IDEXX Laboratories; Michigan, USA) was used to semi-quantitatively determine PAGs and has been validated for ovine species as previously described [50]. This kit detects the variants PAG-4, PAG-6, PAG-9, PAG-16, and PAG-19. In brief, maternal plasma were pipetted into 96-well anti-PAG antibody coated plates along with sample diluent, sealed, and incubated for 60 min at 37 °C in a forced air incubator. Between assay steps, plates were washed 4 times with wash solution (405 LS Washer, BioTek, Winooski, VT, USA). Samples were then incubated with anti-PAG antibody, anti-IgG-horseradish peroxidase, and tetramethylbenzidine. Absorbance (450 nm) was read on a microtiter plate spectrophotometer and values reported as plasma sample minus negative controls after subtracting the mean absorbance value of the negative controls from the absorbance of each sample value. Values are expressed as optical density (OD). The intra-plate and inter-plate CVs for the positive controls were 2.7 and 6.3%, respectively. In singleton and twin pregnancies, the PAGs concentration was corrected by total offspring birth weight and the following equation was used: (Maternal PAG concentration per date) / (birth weight [singleton] or sum of both birth weights [twins]).

Plasma progesterone concentrations were determined using a commercially available direct, competitive ELISA assay (Ridgeway Science, Gloucestershire, United Kingdom) following manufacturer instructions as previously described [50]. Briefly, standards (0.6 to 15.0 ng/ml), serum samples and a plate control were diluted 1:20 with progesterone-enzyme conjugate and added into anti-progesterone-coated 96-well plates in duplicate. Plates were incubated at 37 °C for 2 hours on a rotating shaker at 250 rpm/min. After incubation, plates were washed and then incubated with the substrate until absorbance (550 nm) reading. Intra- and inter-plate CV for the plate control were 3.6 and 8.1%, respectively.

2.6. Newborn outcomes and offspring growth

On the day of lambing, the date, sex, and birth weight were recorded to calculate fertility rate (pregnant ewes per 100 ewes exposed to rams) and reproductive rate (number of fetuses in utero per 100 ewes exposed to rams). To assess offspring growth, body weights were recorded weekly from birth until weaning when lambs averaged 65 ± 5.7 days old (range 49 to 77 days old). Growth velocity was calculated every two weeks from birth until post-natal day (PND) 46 and from PND46 until weaning as follows: (new weight*100)/old weight.

In addition, anthropomorphic measures (body length from rump to shoulder; body height from the ground up to the withers; head length from the nose to the back head [occipital bone] and head width from eye to eye) of the progeny were recorded at one week of age and weaning. Anthropomorphic measures were used to determine body mass index (BMI) as an indicator of body growth and development. BMI was calculated using equations as previously described [51,52]; BMI = (body weight (kg) / withers height (m) / body length (m)) x 10.

2.7. Statistical analysis

Data analyses used the SAS statistical package SAS version 9.3 [53]. Maternal body weight, body weight change, BCS and subcutaneous adipose tissue depth, separated for the diet and breeding periods, were analyzed using the linear mixed model procedures (PROC-MIXED). Fixed effect in the model was treatment. Maternal body weight, BCS and subcutaneous adipose tissue depth were included as covariates as appropriate. PAGs and progesterone separated by gestational age were analyzed using linear mixed model procedures (PROC-MIXED). Fixed effects were treatment, birth type and sex of the progeny. Maternal body weight at sampling and progeny birth weight (BWT) were included as covariates. The relationship between PAGs or progesterone across pregnancy and BWT was computed using mixed models (PROC-MIXED) allowing for repeated measures.

Birth weight, body weight gain, weaning weight, growth velocity and BMI were analyzed using linear mixed model procedures (PROC-MIXED). Fixed effects in the model were treatment, birth type and sex of the progeny. Birth weight, body weight gain, weaning data, growth velocity and BMI were included as covariates as appropriate. Ewe body weight change and lamb body weight gain were fitted in a linear regression model of weight on time for each individual and estimated the regression coefficient as a measurement of change on weight by unit of time. Fertility rate (percentage of ewes pregnant per 100 ewes mated) was analyzed using the generalized linear mixed model procedures with a binomial distribution and logit link function (PROC-GLIMMIX). Fixed effect was treatment. Body weight at start of breeding, body weight change during breeding, BCS and subcutaneous adipose tissue depth were included independently as covariates. Data for reproductive rate (number of fetuses per 100 ewes mated) were analyzed using the generalized linear mixed model procedures with a multinomial distribution and logit link function (PROC-GLIMMIX). The same fixed effect and covariates were used as for the analysis of fertility.

The area under the curve (AUC) of PAG or progesterone throughout the sampling period (GD15 to birth) was calculated using the trapezoidal formula, as previously reported [54]. Differences in AUC were compared using ANOVA and fixed effects in the model were treatment and birth type.

The correlations among weight at birth and weaning, body weight gain, growth velocity, perinatal and weaning BMI of the progeny were computed using PROC GLM with MANOVA option, which allows removal of major fixed effects. Fixed effects included in the model were maternal treatment, birth type and sex of the lambs. All 2-way interactions among the fixed effect and covariates were included in each model and non-significant (P > 0.05) interactions were removed from the model. Significant differences among mean differences of treatments within variables were analyzed using LSD of PROC GLM. The data for fertility and reproductive rate are presented as logit values and back-transformed percentages. All data are presented as mean ± SEM.

3. Results

3.1. Effect of pre-conceptional diet on maternal body condition

Body weight, BCS and subcutaneous adipose tissue depth at the start of the pre-conceptional period did not differ among treatments (Table 2). Body weight change during the pre-conceptional period differed among treatments (UN: − 462 ± 30, C: − 104 ± 30, ON: + 230 ± 26 g/day; P < 0.001) resulting in different body weights (P < 0.001), subcutaneous adipose tissue depth (P < 0.001) and BCS (P < 0.001) at the start of the breeding period (P < 0.001; Table 2). Body weight change during the breeding period differed among treatments with the UN group gaining weight compared to a weight loss in C and ON groups (UN: 49 ± 31, C: − 60 ± 11, ON: − 54 ± 12 g/day; P < 0.01; Table 2).

Table 2.

Effect of pre-conceptional dietary manipulation in mature sheep fed either 50% (UN), 100% (C) or 200% (ON) of maintenance energy requirements during the 21-day period before breeding on maternal body condition at the start and end of the diet and breeding periods and body weight change (BWC) during the diet and breeding periods.

Variable	treatment groups			P
	UN (n = 33)	C (n = 33)	ON (n = 33)
BW start diet (kg)	75.7 ± 2.4	76.2 ± 2.2	76.3 ± 2.5	0.98
BW start breeding (kg)	69.3 ± 2.1b	75.2 ± 2.2ab	79.5 ± 2.5a	0.007
BWC during diet (g / day)	− 462 ± 30c	− 104 ± 30b	230 ± 26a	< 0.001
BWC during breeding (g / day)	49 ± 31a	− 60 ± 11b	− 54 ± 12b	0.003
BCS start diet (units)	3.0 ± 0.07	3.0 ± 0.06	3.0 ± 0.07	0.97
BCS start breeding (units)	2.6 ± 0.05c	3.0 ± 0.06b	3.3 ± 0.08a	< 0.001
SC fat start diet (cm)	0.26 ± 0.02	0.28 ± 0.02	0.27 ± 0.02	0.88
SC fat start breeding (cm)	0.18 ± 0.01c	0.21 ± 0.01b	0.38 ± 0.04a	< 0.001

Values are expressed as mean ± SEM. Superscripts represent differences among treatment groups within variables. BW: body weight (kg), BCS: body condition score (units), SC fat: subcutaneous adipose tissue depth (mm).

3.2. Pre-conceptional diet on fertility and reproductive rate

Fertility rate was similar among treatment groups (UN: 82% [27/33], C: 82% [27/33], ON: 94% [31/33]; P > 0.05). Body condition score at the start of breeding positively influenced fertility rate (P < 0.05); but not body weight and subcutaneous adipose tissue depth at the start breeding or daily body weight change during breeding (P > 0.05). The ratio of singleton and twin pregnancies was influenced by nutritional treatment group (UN: 20:7, C: 21:6, and ON: 14:17; P < 0.05). The reproductive rate was influenced by nutritional treatment group (UN: 103%, C: 100%, and ON: 145%; P < 0.05). Body weight (P < 0.05), body condition score (P < 0.001) and subcutaneous adipose tissue depth (P < 0.01) at breeding positively influenced reproductive rate; but not the daily body weight change during breeding (P > 0.05).

3.3. Pre-conceptional diet effect on pregnancy proteins and hormones

In all groups, circulating PAGs at GD15 were similar, and then changed over time (P < 0.001) by increasing until GD120, and then declining thereafter until the day after birth. The interaction between sampling date and nutritional treatment was not significant (P > 0.05). Pre-conceptional ON resulted in higher total PAG concentrations on GD60 (0.19 ± 0.01 OD) and GD120 (0.33 ± 0.02 OD), compared to C (0.17 ± 0.01 and 0.25 ± 0.02 OD, respectively) and UN treatments (0.17 ± 0.01 OD and 0.27 ± 0.01 OD, respectively; P < 0.05), but not the rest of the sampling dates (P > 0.05). After correcting the relative PAG contribution by total birth weight (see Methods section for formula), circulating PAGs differed by birth type (P < 0.001).

Circulating PAGs relative to birth weight were ~36% lower in twin vs. singleton pregnancies (P < 0.001; Figure 2). Furthermore, when we separate the effect by birth type and treatment, we observed that the difference in total PAG concentration between singletons and twins remained within treatments. The total PAG concentration across pregnancy among treatments for singletons did not differ statistically (P > 0.05); meanwhile, the total PAG concentration across pregnancy among treatments for twins favored the overnutrition treatment only on sampling GD15 and GD120 (P < 0.05; Figure 2). PAG AUC was higher in singletons compared to twins (singleton: 29.2 ± 1.3 vs twin: 16.5 ± 1.7 OD; P < 0.001), but not significantly different among treatment groups (UN: 21.8 ± 1.8, C: 20.8 ± 1.8, ON: 25.9 ± 1.6 OD; P > 0.05). The interaction between treatment and birth type was not significant (P > 0.05).

Figure 2.

Maternal circulating pregnancy associated glycoprotein (PAGs; O.D.) (mean ± SEM) in pregnant females with singletons (solid lines) or twins (dashed lines) that received 50% (light gray), 100% (dark gray), or 200% (black) of maintenance energy requirements during the 21-day pre-conceptional period (see diet details in text). UN: 19 singletons and 14 twins; C: 21 singletons and 13 twins; ON: 14 singletons and 34 twins. GD: gestational day; O.D.: optical density; PDN1: postnatal day 1; ** P< 0.01.

Like for PAGs, in all groups, maternal progesterone was similar among treatments on GD15, followed by an increase until GD120, and decline until the day after birth to concentrations below those of GD15. Pre-conceptional ON resulted in higher progesterone compared to the C group on GD30 (0.28 ± 0.02 ng/ml; P < 0.05), GD60 (0.43 ± 0.03 ng/ml; P = 0.08) and GD120 (1.1 ± 0.07 ng/ml; P < 0.05), compared to C and UN treatment (UN: 0.21 ± 0.02, 0.32 ± 0.04, 0.94 ± 0.07 ng/ml, respectively; C: 0.24 ± 0.02, 0.31 ± 0.04, 0.81 ± 0.08 ng/ml, respectively). However, on the sample at birth, the control group had higher progesterone concentration (0.11 ± 0.019 ng/ml) compared to UN (0.05 ± 0.019 ng/ml) or ON (0.05 ± 0.016 ng/ml; P > 0.05) groups. After correcting the relative progesterone contribution by total birth weight, circulating progesterone differed by birth type and was greater in singletons than twins at GD15 (P < 0.05), GD30 (P < 0.001), GD60 (P < 0.01) and GD140 (P < 0.05). Progesterone concentration corrected for fetal mass across pregnancy and was 12% greater in singleton compared to twin pregnancies. When accounting for the effects of birth type and treatment, we observed the same trend between singletons and twins within treatments (P > 0.05). In singletons, progesterone concentration across pregnancy did not differ among treatments (P > 0.05) except for PND1 when the control treatment had higher progesterone concentrations (P < 0.05; Figure 3). In twins, progesterone concentration across pregnancy among treatments was higher in the overnutrition treatment at GD30 (P < 0.05), GD90 (P < 0.01) and GD120 (P < 0.001; Figure 3), but was higher in the control treatment at PND1 (P < 0.05; Figure 3). Progesterone AUC was higher in singletons compared to twins (singletons: 76.7 ± 2.7 ng/ml vs twins: 60.3 ± 3.8 ng/ml; P < 0.001). Progesterone AUC was also higher in ON (77.5 ± 3.6) vs. UN or C (UN: 66.5 ± 4.0 OD, C: 61.4 ± 4.0 OD; P < 0.05). The interaction between treatment and birth type was not significant (P > 0.05).

Figure 3.

Dynamic maternal profiles of circulating (mean ± SEM) progesterone (ng/ml) in in pregnant females with singletons (top; solid lines) or twins (bottom; dashed lines) that received 50% (lightgray), 100% (dark gray), or 200% (black) of maintenance energy requirements during the 21-day pre-conceptional period (see diet details in text). UN: 19 singletons and 14 twins; C: 21 singletons and 13 twins; ON: 14 singletons and 34 twins GD: gestational day; PDN1: postnatal day 1; * P<0.05; ** P<0.01, *** P<0.001.

3.4. Pre-conceptional diet effect on lamb outcomes

When all the animals from all the treatments were considered for the analyses, on average; single-born lambs were heavier at birth than twin-born lambs (P < 0.001; Table 3). Single-born lambs grew 25% faster (P < 0.001) and were 17% heavier at weaning than twin-born lambs (P < 0.001; Table 3). BMI was greater at birth and weaning for single-born lambs (P < 0.001; Table 3). Growth velocity did not differ between single- and twin-born lambs (P > 0.05; data not shown). When all the animals from all the treatments were considered for the analyses, on average, male lambs were heavier at birth than female lambs (P < 0.001; Figure 4). Male lambs grew 12% faster and were 11% heavier at weaning than female lambs (P < 0.05; Figure 4). Perinatal BMI did not differ between female and male lambs (P > 0.05); however, weaning BMI was greater for male lambs (P < 0.05; Table 3). Growth velocity did not differ between female and male lambs (P > 0.05; data not shown). At birth, UN lambs were heavier than C or ON lambs regardless of birth type (Table 3). Daily body weight gain and weaning weight did not differ among treatments (P > 0.05; Table 3). However, UN lambs grew faster and were heavier at weaning (322 g day⁻¹ and 26.5 kg, respectively) than C lambs (317 g day⁻¹ and 25.3 kg, respectively) or ON lambs (314 g day⁻¹ and 25.5 kg, respectively). Perinatal BMI was greater (P < 0.01; Table 3) at birth for UN lambs (0.66±0.024) than C lambs (0.58±0.023) and ON lambs (0.56±0.019); whereas wean BMI did not differ among treatments (UN: 2.4±0.11 vs C: 2.3±0.10 vs ON: 2.3±0.09; P > 0.05; Table 3). Growth velocity did not differ among treatments (P > 0.05; data not shown). Supplemental Table 1 presents the weight and growth variables of progeny considering maternal nutritional treatment, birth type and lamb sex.

Table 3.

Effect of pre-conceptional dietary manipulation on progeny weight and growth variables at birth and weaning of progeny from sheep that received 50% (UN) 100% (C) or 200% (ON) of maintenance energy requirements during the 21-day pre-conceptional period.

		Birth weight (kg)	LW gain (g / day)	Weaning weight (kg)	Newborn BMI	Weaning BMI
Treatment P > f	n	**	NS	NS	**	NS
UN	31	5.4 ± 0.15a	322 ± 14a	26.6 ± 1.0a	0.63 ± 0.02a	2.4 ± 0.10a
C	34	5.0 ± 0.20ab	318 ± 17a	25.3 ± 1.2a	0.59 ± 0.02b	2.3 ± 0.11a
ON	48	4.8 ± 0.10b	314 ± 13a	25.5 ± 0.9a	0.55 ± 0.02b	2.3 ± 0.09a
Birth-rear type
P > f		***	***	***	***	***
Single (1-1)	54	5.3 ± 0.1a	354 ± 12a	27.9 ± 0.8a	0.63 ± 0.02a	2.6 ± 0.08a
Twin (2-2)	59	4.7 ± 0.1b	283 ± 10b	23.7 ± 0.7b	0.54 ± 0.02b	2.1 ± 0.07b
Sex type
P > f		***	*	*	NS	*
Female	57	4.8 ± 0.1b	300 ± 11 b	24.5 ± 0.8b	0.56 ± 0.02a	2.2 ± 0.08b
Male	56	5.3 ± 0.1a	335 ± 13a	27.1 ± 0.8a	0.60 ± 0.02a	2.4 ± 0.08a

P-values:

^*P ≤ 0.05;

^**P ≤ 0.01;

^***P ≤ 0.001; NS: not significant.

1-1: Born and raised as singleton

2-2: Born and raised as twin

Figure 4.

Birth weight (left panel) and postnatal growth (right panel; mean ± SEM) from week1 to weaning of male (top) and female progeny (bottom) from ewes that received 50% (UN; light gray), 100% (C; dark gray), or 200% (ON; black) of the nutritional requirements 21 d before breeding. The data combined the birth type of the lambs (singleton and twins). See diet details in text. In males, weight profile from birth to weaning was similar among treatments (P > 0.05). In females, asterisks denote significant differences between the UN or ON vs. the C group: * P<0.05, ** P< 0.01. On weeks 1, 3 and weaning, female progeny from 50% and 200% (50%: 7.4±0.3, 12.6±0.4 and 29.1±1.1 kg; 200%: 7.4±0.6, 13.2±0.4, and 28.8±1.3 kg) were similar but heavier compared to the 100% group (6.5±0.5, 8.9±0.5 and 24.7±1.6 kg; P < 0.05. WK: week. UN: 17 females and 14 males; C: 16 females and 18 males; ON: 24 females and 24 males.

When all the animals from all the treatments were considered for the analyses, body weight gain tended to be positively related to birth weight (P = 0.07) and for every kg increases in birthweight there was an 18 g day⁻¹ increase in body weight gain. Weaning weight was positively related to birth weight (P < 0.05) and body weight gain (P < 0.001). For every 1 kg increase in birthweight there was a 1.6 kg increase in weaning weight. Moreover, for every 50 g day⁻¹ increase in body weight gain there was a 3.3 kg increase in weaning weight. When all treatment groups were combined, birth weight was positively correlated with both, body weight gain and weaning weight. The correlation between body weight gain and weaning weight was positive and strong (r = 0.95; P < 0.001). Perinatal BMI was positive and very low related to birth weight (r = 0.02; P > 0.05); but the correlation between perinatal BMI and body weight gain (r = 0.47; P < 0.001) and weaning weight (r = 0.48; P < 0.001) was positive and moderate. Weaning BMI was positive and low correlated to birth weight (r = 0.05; P > 0.05) and perinatal BMI (r = 0.47; P < 0.001); however, it was positive and strong correlated to body weight gain (r = 0.95; P < 0.001) and weaning weight (r = 0.99; P < 0.001). Across pregnancy, the total progesterone concentration influenced the birth weight (P < 0.05); but not the total PAGs concentration (P > 0.05). The relationship between PAGs and progesterone was significant and positive (P < 0.001).

4. Discussion

The effects of maternal nutrition during pregnancy on fetal and offspring growth and health have been well established [55,56]. However, less understood are the effects of pre-conceptional maternal nutrition, which was the aim of this study. We have demonstrated that maternal pre-conceptional nutrition results in moderate changes in placental endocrine secretion with overnutrition resulting in higher maternal PAG and in higher maternal progesterone concentrations during mid-gestation (GD60 to GD120) suggestive of an enhanced placental endocrine function upon pre-gestational overnutrition. Similar to our previous observations [50], maternal circulating PAGs and progesterone were lower in twin pregnancies when corrected by birth weight. Furthermore, pre-conceptional undernutrition resulted in lambs that were heavier and with higher BMI at birth independent of sex.

4.1. Effect of pre-conceptional diet on maternal body outcomes

The 21-days of pre-conceptional diet at 50% of the nutritional requirements resulted in a decrease in body weight (8.5%) that is consistent with previous studies in Rambouillet/Columbia females with a similar, but longer, nutritional restriction [11%; 57]. Body weight loss in our study was also accompanied by a subcutaneous adipose tissue reduction (31%), which is associated with energy store mobilization in response to the nutrient deficit [41]. As expected, pre-conceptional overnutrition increased maternal body weight by 4.2% and maternal subcutaneous adipose tissue depth by 41%. These findings are similar to the 31.5% increase in body weight observed in female sheep fed with ad libitum diet (170-190% ME of the nutritional requirements) from before conception to 6 days after mating [58] and are the reflection of an increase in excess energy storage as glycogen and triglycerides [59,60,61].

4.2. Pre-conceptional diet effect on placental function

Similar to our previous studies [50,52], we observed that PAGs increased until GD120 and declined thereafter until the day after birth, independent of the pre-conceptional diet. Without correcting for total newborn weight, circulating PAGs were higher in pre-conceptional overnutrition across pregnancy, but only significant between GD60 and GD120. Because twin pregnancies had higher circulating PAGs, higher PAGs in the ON group was likely driven by a greater proportion of females bearing twins (52%) compared to that of the control (21%). This is similar to previous work in Texel, Poll-Dorset and Suffolk ewes where females bearing twins had higher PAGs concentration in the first trimester of pregnancy to those bearing singletons [50,62].

Because placental size is positively associated with fetal and newborn size [63,64] and endocrine output [25,57,65], accounting for newborn weight in maternal PAG concentrations provides an index of placental secretory capacity. Using this correction, we have observed that despite twin pregnancies having the highest PAG concentrations, PAG per newborn weight is ~36% lower in twin vs. singleton pregnancies. Using this correction, we have observed that singleton pregnancies among treatments were not affected by maternal pre-conceptional nutrition. Similar to previous studies, we have observed that females bearing twins had higher PAG concentrations compared to females bearing singletons. The difference in maternal PAGs between singleton and twin pregnancies may relate to the fact that the number of placentomes is lower in twin pregnancies per conceptus and the birth weight of the progeny is lighter per placental unit [66].

Maternal progesterone also increased across pregnancy only to decline after GD120 in all groups. These findings align with previous work regarding a steady progesterone increase across pregnancy to decline before birth and a positive relationship between PAGs and progesterone [47,52,67]. Once fetal maturation has been reached, fetal glucocorticoid concentration increases rapidly resulting in progesterone reduction which triggers labor [68]. Similar to our PAGs results, and before correcting by total newborn weight, progesterone concentration was on average ~18% higher in twin pregnancies to that of singleton pregnancies (ranging from 8% to 27%). This is similar to previous studies where serum progesterone concentration increased as the number of fetuses increased [69,70] and supports the concept that progesterone enhances fetal weight [71] and that both, progesterone concentration and placental efficiency increase with litter size [69,72]. After correcting by total birth weight, and similar to that observed in PAGs, progesterone concentration was higher in singletons than twin pregnancies across pregnancy. This difference remained between singletons vs. twins within treatments, demonstrating that pre-conceptional nutrition did not influence progesterone concentration across pregnancy.

Overall, the three-week maternal dietary intervention did not result in overt alterations in placental secretory function (PAGs and progesterone) regardless of the number of fetuses in contrast with previous work when maternal dietary restriction occurring during mid-pregnancy resulted in a reduction in mass of the fetal component of the placenta [73]. Profound placental effects have also been reported upon maternal gestational overnutrition, including reduction in circulating progesterone, placental growth, placentome weight and fetal growth [28,74]. In multiparous sheep, maternal overnutrition resulted in increased insulin resistance, circulating concentration of free fatty acids, cholesterol and triglycerides leading to an inflammatory response in the placenta [56,75]. In sheep, placentation begins on day 15 to 16 and is thought to be complete around day 50 to 60 of pregnancy and placental mass is known to reach its peak by gestational day 80 to 90 [76,77,78]. Altogether, this suggests that the pre-conceptional period may not be as vulnerable of a window for placental steroid and protein biosynthesis function as during placental development per se. Additional studies are needed to determine whether longer pre-conceptional windows or stronger dietary interventions can compromise placental function.

4.3. Pre-conceptional diet effect on lamb outcomes

Similar to previous reports [79,80], and after combining the lambs from all the treatment groups, male lambs were heavier and had larger BMI at birth, grew faster and were heavier at weaning compared to female lambs. The sexual dimorphism observed during gestation and postnatal growth might be due to the differences in the secretory capacity of the somatotropic axis between males and females [81,82]. Maternal diet manipulation during gestation either below or above nutritional requirements may impact fetal neuroendocrine maturation, fetal growth and birth weight [83,84]. Even though the pre-conceptional period encompasses the time when the ovulatory follicle grows to its preovulatory size and the oocyte matures to metaphase II, limited studies are available on the effect of maternal diet during the periconceptional period on placental function or progeny outcomes. In this study, pre-conceptional diet manipulation influenced the size and weight at birth with UN lambs being the heaviest. The fact that these differences were independent of birth type and despite an uneven distribution of singles and twins in the different diet groups, highlights the effect of pre-conceptional nutrition in programming birth weight.

In our study, the breeding period lasted 34 days (2 reproductive cycles) with 82 to 94% of females across all treatment groups conceiving in their first reproductive cycle and 9% in their second reproductive cycle (data not shown). It is likely that the body weight, body condition score and fat depth loss observed in the undernourished group was not enough to reduce fertility rate. Although periconceptional undernutrition has been shown to reduce prolificacy and fecundity (mean number of live and dead lambs born per ewe lambing and per ewe presented to the ram, respectively) in sheep [85] and periconceptional maternal high protein diet can modify the uterine environment and increase embryo loss [86], our data indicate that the fertility rate was not affected by the maternal diet during follicular growth (from primary to preovulatory in sheep or oocyte maturation). In sheep, periconceptional undernutrition has been shown to reduce oocyte maturation and embryo survival; resulting in lower prolificacy and fecundity [85,86]. On the other hand, in Merino sheep, periconceptional overnutrition increased fertility rate and body size of their progeny; although survival was similar [87,88]. In multiovulatory species such as pigs, periconceptional overnutrition also leads to increased blastocyst cell number and embryo survival [89,90].

Growth velocity between male and female lambs among treatments was similar. However, female offspring derived from maternal pre-conceptional under- and overnutrition were heavier from postnatal week 1 until weaning compared to control females. Previous reports indicate that the diet manipulation during the oocyte maturation and early embryonic development can modify the metabolism and affect postnatal growth of the offspring and result in long term, permanent effects in the offspring [41,91]. Evidence in rats indicates that periconceptional maternal diet manipulation (eg. low protein diet) resulted in reduced expression of the growth-regulating imprinted genes which reduced postnatal development [92]. Our findings are similar to those reported in offspring of nutrient-restricted sheep during early gestation resulting in increased body weight and backfat thickness at PND140 [18]. In our study, all female lambs were born of similar weight, but ON and UN female offspring reached weaning weights similar to that of the control males, known to grow faster and be heavier at weaning than females [79,80]. The enhanced growth in UN and ON offspring females suggests that peri-conceptional maternal diet either below or above requirements leads to altered pathways of metabolism leading to an increase in postnatal growth and adipose tissue accumulation as previously suggested [41,91]. Notably, these postnatal changes in body weight were only evident in females. This sex-specific effect has also been observed in other studies, in females [93,94] and in males [18,95], highlighting sex-specific responses to in utero pre-conceptional and gestational insults. Further research should focus on understanding the underlying mechanisms by which postnatal body weight changes occur and how sex-specific differences are determined in utero and are manifested postnatally.

In conclusion, we demonstrated that a 3-week periconceptional maternal diet manipulation either under or over the nutritional requirements did not modify the synthetic/secretory capacity of PAGs and progesterone in single or twin pregnancies. However, correcting by total birth weight, lower placental PAG and progesterone synthetic/secretory capacity was observed in twin pregnancies. Pre-conceptional maternal under- and overnutrition increased postnatal female lamb growth through weaning, suggestive of reprogramming of pathways regulating growth before conception. Importantly, this effect was only observed in female offspring highlighting how pre-conceptional nutrition can result in marked sex-specific differences. Since this study was conducted during the suboptimal breeding season (long day season) whether our findings were influenced by season remain to be determined.

Table 4.

Effect of pre-conceptional dietary manipulation on progeny weight and growth variables at birth and weaning according to birth type (singleton [1] or twin [2]) of progeny from sheep that received 50% (UN) 100% (C) or 200% (ON) of maintenance energy requirements during the 21-day pre-conceptional period.

Nutritional Treatment	Lamb Sex	Birth Type	n	BWT	LWG	WWT
UN	Female	1	10	5.3±0.1	360±15	29.2±1.2
		2	7	4.7±0.3	272±32	22.6±2.3
	Male	1	9	5.7±0.3	337±30	28.3±2.0
		2	5	5.8±0.4	278±25	23±1.4
C	Female	1	9	4.9±0.3	292±35	23±2.3
		2	7	4.2±0.2	276±20	23±2.0
	Male	1	12	5.8±0.2	376±34	28.7±2.2
		2	6	4.7±0.2	293±16	25.2±1.2
ON	Female	1	8	4.9±0.2	371±14	28.9±1.4
		2	16	4.6±0.2	259±20	21.9±1.5
	Male	1	6	5.3±0.4	412±23	30.7±2
		2	18	4.8±0.2	314±21	26±1.5
P value/TRT				0.01	0.9	0.6
P value/BT				0.008	0.001	0.003
P value/Sex				0.003	0.3	0.03
Interaction TRT*BT				NS	NS	NS
Interaction TRT*Sex				NS	NS	NS
Interaction BT*Sex				NS	NS	NS

Highlights.

• Preconceptional maternal undernutrition led to heavier and larger newborns

• Preconceptional maternal unde- and overnutrition female lambs grew faster

• Preconceptional nutrition can modify the placental endocrine capacity