1
Impact of donor and recipient adiposity on feto-placental growth in adolescent sheep
Jacqueline M Wallace, John S Milne, Clare L Adam and Raymond P Aitken
The Rowett Institute, University of Aberdeen, Aberdeen AB25 2ZD, UK.
Corresponding author:
Short title: Peri-conception adiposity and fetal growth
Abstract
The influence of maternal obesity during oocyte development and its putative interaction with nutrient reserves at conception on pregnancy outcome were examined in an adolescent sheep model. Donor ewes were nutritionally managed to achieve contrasting adiposity (control [CD]/obese [ObD]) for 6 weeks prior to superovulation, and inseminated by a non-obese sire. Morulaefrom 6 CD and 7 ObD were transferred in singleton into adolescent recipients of identical age but differing adiposity, classified as relatively fat or thin, respectively. Thereafter all were overnourished to promote rapid growth/adiposity (2x2 design, 13/14 pregnancies/group). A fifthrecipient group of intermediate adiposity received embryos from another 5 CD, were offered a moderate intake to maintain adiposity throughout gestation and acted as controls for normal pregnancy outcome (optimally-treated-control [OTC], 19 pregnancies). Donor obesity did not influence ovulation, fertilization or recovery rates or impact embryo morphology. Gestation length and colostrum yield were unaffected by donor or recipient adiposityand were reduced relative to OTC. Total fetal cotyledon and lamb birthweights were independent of initial donor adiposity but reduced in relatively thinversus relatively fat recipients, and lower than in the OTC group. In spite of high placental efficiency, the incidence of fetal growth-restriction was greatest in the thin recipients. Thus maternal adiposityat conception, but not pre-conception maternal obesity, modestly influences the feto-placental growth trajectory,whilecomparison with the OTC indicates that high gestational-intakes to promote rapid maternal growth remain the dominant negative influence on pregnancy outcome in young adolescents. These findings inform dietary advice for pregnant adolescent girls.
Introduction
Birth weight is a valuable aggregate of fetal nutrient supply and a robust prognosticator of health and wellbeing immediately after delivery and throughout the life-course. Our focus is the low end of the birth weight spectrum as these babies are most likely to die in infancy or experience a range of physical and development issues that can limit their life chances (Sharma et al.2016). Further, low birth weight is a risk factor for the later development of a number of life-limiting diseases that drain health service resources including diabetes, stroke and cardiovascular disease: as such decreasing the proportion of babies born too early and/or too small is a pressing public health objective (WHO 2012; Hanson and Gluckman, 2015).
The most consistent risk of poor outcome is when pregnancy coincides with adolescence. Accordinglyboth population-wide and single-center studies in low, middle and high income countries reliably report a higher risk of spontaneous miscarriage, premature delivery, low birth weight, and neonatal mortality in adolescent compared with adult pregnancies (Shrim et al. 2011; Malabarey et al. 2012; Ganchimeg et al. 2013; Kozuki et al. 2013; Weng et al. 2014; Torvie et al. 2015).These hazards are most pronounced in very young girls when pregnancy potentially overlaps with continued or incomplete growth of the mother, setting up a competition for nutrients between the maternal body and the gravid uterus which results in attenuated fetal growth (Frisancho, 1997; Frisancho et al. 1985; Scholl et al. 1994, 1997).A similar maternal-fetal growth competition for nutrients has been partly reproduced in a highly controlled sheep model whereby overfeeding young singleton-bearing adolescents throughout pregnancy promotes high gestational weight gains andsupports continued maternal growth and increasedadiposity. These rapid maternal growthrates are associated with a greater incidence of miscarriage and stillbirth and robustly result in the premature delivery of low birth weight lambs compared to control-fed (slow-growing) adolescents of the same age (Wallace et al. 2004).In thisovernourished model, defects in early placental development including lowercellular proliferation,reduced blood vessel development and impaired secretory functionprogressively compromise the growth trajectory and haemodynamic function of the placenta. Consequently by the final third of gestation and regardless of nutrient excess in the mother, the small size of the placenta and associated reduction in uteroplacental blood flows and nutrient uptakes limits fetal nutrient supply (Wallace et al. 2006a). This leads to a slowing of fetal growth (Carr et al. 2012) and at parturition about50% of these lambs are classified as prenatally growth restricted relative to the normal birth weight offspring of the optimally nourished controls (Wallace et al. 2004, 2006a).
In the foregoinginitial sheep studies weconcentrated on varying dietary intake and hence growth status immediately after pregnancy had been established and thus the adolescentswere of similar age, weight and adiposity at conception. Acknowledging that adolescent girls enter pregnancy from diverse nutritional backgrounds and with different nutrient reserves at conception we have additionally shown that,irrespective of gestational intake and growth status, adolescents who were relatively light and thin at conception gave birth to lighter lambs than those who were heavier and fatter(Wallace et al. 2010). This reduction in fetal growth was again mediated by the placenta,implying a direct effect of nutrient reserves at conception on the metabolism of the dam and her early uterine environment,leading to a compromised placental growth trajectory. However,variable nutrient intakes and/or nutritional statusin the peri-conception period may additionallycontribute to pregnancy outcome via effects on follicle recruitment, oocyte maturation, fertilization and early embryo development (Sinclair and Watkins, 2014). As pregnancy is normally a continuum from oocyte to fetus to delivery, it is difficult to segregate the impact of nutrition at any precise stage due to potential carry over effects between sequential stages. Howeverin our animal model,by using assisted conception procedures we are able touncouplepre-, peri- and post-conception nutritional exposures in order to assess their separate or interdependent influences.
Herein we evaluate whether the nutritional status of the embryo donor ewe (relatively obese versus normal) interacts with that of the embryo recipient (relatively fat versus thin) to influence conception rate, fetal growth, pregnancy outcome and early offspring growthin adolescent animals that were overnourished throughout pregnancy. Obese versus control (normal adiposity) donor ewes were used because obesityimpacts fertility with negative effects on follicle recruitment, oocyte viability, fertilization and early embryo development reported in animal models and in humans seeking assistance to conceive via assisted reproduction technologies (ART, Gonzalez-Anover et al. 2011; Purcell & Moley, 2011;Kumbak et al. 2012;Sinclair and Watkins, 2014; Velazquez, 2015). Moreover in adult humans conceiving naturally, peri-conception obesity is widely associated with a plethora of pregnancy complications including hypertensive disorders, gestational diabetes, stillbirth, fetal malformations, premature delivery and derangements in fetal growth resulting in either high birthweight or relative fetal growth restriction (McDonald et al.2010; Anderson et al. 2013; Aune et al. 2014; Lutsiv et al. 2015; Marchi et al. 2015). Therefore, this study tests the hypothesis that pregnancy rate and conceptus development would be mostnegatively disturbed in pregnancies generated from embryos of relatively obese compared with control donors and this would be exacerbated in adolescentrecipients who were relatively thincompared with those who were fatter at conception.
Materials and Methods
Experimental design
All procedures were licensed under the UK Animals (Scientific Procedures) Act of 1986 and approved by the Rowett Institute’s Ethical Review Committee. The main design is a 2 x 2 factorial to examine the impact of embryo donor adiposityduring oocyte development versus embryo recipient nutritional status at conception on pregnancy outcome at term. These adolescent ewes were subsequently overnourished throughout gestation to promote rapid maternal body growth. A fifth contemporaneous group of optimally treatedcontrol adolescents acted as a reference point for normal fetal growth (Figure 1).
Animals: Donors
Adult ewes (Border Leicester × Scottish Blackface) of equivalent age and parity (~2.5 years old and 1 previous pregnancy/lactation) destined to become potential embryo donors were selected from the Institute’s flock, group housed and nutritionally manipulated by varying quantity of diet offered over a three month period to achieve different adiposity levels (control versus relatively obese). For the four weeks prior to starting superovulation protocols (6 weeks prior to embryo recovery) animals were individually housed under natural lighting conditions and offered maintenance rations of a complete diet (see below) to maintain weight and adiposity score. The latter was evaluated,based on manual palpation of lumbar spine, ribs and tail-head, by one experienced operator on a scale of 0 to 5, where 0 = extremely emaciated and 5 = extremely obese, according to the criteria of Russel et al. (1969). The accuracy of this scoring system has been validated against whole carcass chemical analyses and is sensitive to within 0.25 score units (Wallace et al. 1999). On the day prior to insemination, the adiposity (mean±sem) of potential control compared with relatively obese donors was2.2±0 units versus 3.4±0.11 units, equivalent to approximately 22 and 33% body fat, respectively (Russel et al. 1969). Donor ewes were adult rather than adolescent since their embryos are inherently more viable (Quirke and Hanrahan, 1977; McMillan and McDonald, 1985).
Recipients
Meanwhile two groups of adolescent ewe lambs of identical age (~7.5 months) but markedly different adiposity scores were selected from another flock (Dorset Horn × Greyface) four weeks prior to assisted conception procedures, destined to become embryo recipients. During this four week period animals were individually offered maintenance rations of a complete diet to maintain their initial weight and thereby adiposity score. For ease of presentation, adolescents who were light and had a low adiposity score were classified as relatively thin, while those that were heavier and had a higher adiposity score were classified as relatively fat. Immediately prior to embryo transfer the adiposity score of these“thin”versus “fat” groups was 2.0±0 and 2.7±0.02 units equivalent to approximately20 and 26%body fat, respectively.
Embryo transfer
Using techniques described previously (Wallace et al. 1997), donor ewes were intrauterine inseminated on Day 0by a single sire with optimal adiposity for breeding (Dorset Horn, body score 3) and embryos recovered at laparotomy on Day 4 after oestrus. Corpora lutea were counted to establish ovulation rate and embryos morphologically graded under a stereomicroscope. Embryos classified as good early morula (grade 1) and from either control or obese donors were synchronously transferred singly into the uterus of either fator thin recipients (main study, 2 x 2).A contemporaneous third group of adolescents of intermediate orcontroladiposity (2.3±0.01 units, 23% body fat) received embryos from control donors (optimally treated control (OTC) reference group).
Embryo transfers were carried out on six separate days during the mid-breeding season.Equivalent numbers of fat and thinadolescents were set-up and synchronized as potential recipients for each day (7 or 8 animals per adiposityclassification per day). On the first and third day potential embryo donors were controls and on the second and fourth days donors were obese. On each of these days the aim was to transfer equal numbers of embryos from an individual donorinto the two groups of recipients while taking care not to over-represent a particular donor’s genetics. Thus a maximum of eight embryos per donor were used in the study (average 5.5, range 2-8).The fifth and sixth days were reserved for the transfer of control donor embryos into the control (intermediate) adiposity recipients destined to receive a control dietary intake(OTCgroup). Three potential embryo donors (2 x control, 1 x obese) had either regressing corpora lutea or 100% unfertilized oocytes and these animals took no further part in the study. Similarly a small number of potential recipients either failed to ovulate or had regressing corpora lutea when examined by laparoscope prior to embryo transfer and were discharged from further study. For the main study 37 and 34embryos representative of 6 control and 7 obese adult donor genetics were transferred into 71 adolescent recipients (36x thin, and 35 x fat). This yielded 4 main study groups, namely control donor-thin recipient (CD-TR, n=19), control donor-fat recipient (CD-FR, n=18), obese donor-thin recipient (ObD-TR, n=17), obese donor-fat recipient (ObD-FR, n=17). A further 26 embryos representative of a further 5 control donor ewes’ genetics were transferred into the control adiposity reference groupdestined to become the OTC group (Figure 1).
Nutritional management
The complete diet used throughout supplied 12 MJ metabolizable energy (ME) and 140 g crude protein per kg and was offered in two equal portions at 08:00 and 16:00 h daily (see Wallace et al. 2006b for full details of the diet composition and analyses). This diet was used to achieve the contrasting nutritional states in donors and recipients pre-conception, and in recipients during pregnancy and lactation. Following embryo transfer all recipients in the main four donor x recipient adipositygroups were offered high dietary intakes to promote rapid gestational weight gain and increasing adiposity throughout gestation (overnourished; equivalent to ~ 2 x estimated ME requirements for optimum conceptus growth in ewes of this age and genotype). Following a three day post-surgery re-alimentation period, high intakes were achieved by increasing the level of complete diet offered gradually over a two week period until the level of the daily food refusal was ~15% of the total offered (equivalent to ad libitum intakes).The dietary level offered in the OTC group was calculated to maintain normal maternal adiposity throughout gestation (i.e. no change from initial starting adiposity) and hence meet the estimated ME requirements for optimum conceptus growth (AFRC, 1993). To achieve this objective, the OTC group was fed to promote a modest maternal weight gain of ~75 g per day during the first two thirds of gestation, followed by step-wise increases in maternal intake during the final third of gestation, calculated to meet the increasing demands of the developing fetus. During pregnancy the level of food offered was reviewed three times weekly and adjusted, on an individual basis as and when appropriate, on the basis of daily food refusal rates (overnourished groups) and adiposity score (OTC group). Following parturition all ewes were offered the complete diet to appetite (i.e. ad libitum) to maximize milk availability. For the OTC dams this was achieved step- wise over a period of approximately 10 days. Recipient ewes were weighed fortnightly and external adiposity score assessed approximatelymonthly through-out pregnancy. Ewes were also weighed ~ 24h after parturition and at the end of lactation (77 days).
Conception rate, parturition management and neonatal care
Conception rate was determined by transabdominal ultrasonography at day 50 of gestation. Pregnancy outcome was determined after spontaneous delivery and ewes were supervised throughout the expected delivery period from day 135 onwards (the earliest point commensurate with live birth in overnourished adolescents of this genotype). A standardized proactive regimen of neonatal care was used to prevent high neonatal mortality due to prematurity and impaired passive immunity and/or nutrient intake secondary to inadequate colostrum supply (Wallace et al. 1996).Lambs were dried, weighed and girth at the umbilicus measured after delivery. The height measurement was delayed until ~12h after birth. Ewe colostrum yield was measured before lamb suckling and within 30 min of parturition. After intravenous injection of oxytocin (Oxytocin-S® 10 i.u. per ewe; Intervet Ltd, Cambridge, UK), ewes were milked by hand until all the colostrum had been removed from the udder. The colostrum was weighed, sampled for IgG analysis, and then fed to the ewe’s own lamb by bottle or feeding tube at a rate of 50 ml/kg body weight. In cases where the dam had insufficient colostrum, frozen pooled ewe colostrum collected previously immediately after birth (day 145-147 of gestation) from optimally nourished adult twin-bearing ewes was substituted to ensure lamb survival. All lambs were weighed at 4 hourly intervals throughout the first 72h of life and at 8 hour intervals from 72 to 168h. Any lamb which failed to suckle or gain in weight over an 8 hour period was offered supplementary colostrum (first 24h) or ewe milk until the ewe-lamb bond and appropriate lactation were established (by 120h after birth in all cases). The frequency of supplementary feeds per lamb was recorded. To further reduce the risk of infection, each lamb’s navel was dipped in iodine at birth and at 12 hours after delivery, and all lambs received intramuscular vitamin E / selenium supplementation at birth and prophylactic antibiotics for five days. After the placenta (fetal cotyledons and membranes) was delivered, its weight was recorded, and then the cotyledons were dissected, counted and weighed. Lamb weight and height were measured weekly throughout the 11 week lactation. For both parameters absolute growth rate (AGR) was linear and was recorded as the slope of the line of best fit, determined by linear regression analysis. Current FGR (CFGR) at weekly intervals was calculated as the AGR for 0 to 77 days divided by the value of a parameter (weight or height) at the start of each 7 day period.