Introduction Effects of nutrition on the ovarian activity

Ž . Animal Reproduction Science 60–61 2000 185–197 www.elsevier.comrlocateranireprosci Influence of the nutritional status on ovarian development in female pigs A. Prunier , H. Quesnel Unite Mixte de Recherches sur le Veau et le Porc, INRA, 35590 Saint-Gilles, France ´ Abstract In female pigs, undernutrition may influence growth of antral follicles from various size classes, decrease ovulation rate, delay puberty and return to oestrus after weaning. It could also affect the oocyte maturation and hence the number of viable embryos per litter. Inhibition of the gonadotrophin release due to undernutrition is presumably involved in these phenomena. Presence of receptors, as well as in vitro and in vivo studies suggest that insulin and hormones from the somatotrophic axis are able to alter folliculogenesis directly at the ovarian level. They should act as hormones controlling nutrition, proliferation, growth and differentiation of the cells andror as amplifiers of the action of gonadotrophins. Information are needed to determine whether their availability at the ovarian level may become insufficient or excessive in case of nutritional deficit. Increase in plasma concentrations of progesterone due to lower hepatic metabolic rate in underfed females probably contributes to inhibit folliculogenesis. q 2000 Elsevier Science B.V. All rights reserved. Keywords: Gilt; Sow; Ovarian follicle; Nutrition; Physiology

1. Introduction

Numerous data from various mammalian species have shown that nutrition may influence the reproductive function. Most experimental studies involving farm animals have focussed on the negative consequences of undernutrition on reproductive perfor- mance. This is particularly true in the pig species and is related to the fact that nutritional deficit may occur spontaneously in gilts and adult sows under farm condi- Ž . tions Prunier and Quesnel, 2000 . Physiological mechanisms explaining the links Corresponding author. Tel.: q33-2-23-48-50-56; fax: q33-2-23-48-50-80. Ž . E-mail address: prunierst-gilles.rennes.inra.fr A. Prunier . 0378-4320r00r - see front matterq 2000 Elsevier Science B.V. All rights reserved. Ž . PII: S 0 3 7 8 - 4 3 2 0 0 0 0 0 0 9 3 - 2 between nutrition and reproduction are partly known. It is clear that the GnRH pulse Ž generator system and the gonadotrophin release are inhibited by undernutrition for . review, see Booth, 1990; Cameron, 1996; Wade et al., 1996 . As a consequence, the gonadotrophin support to folliculogenesis may become insufficient. In addition, metabolic Ž . mediators hormones or metabolites may act directly at the ovarian level and amplify Ž the gonadotrophin-mediated effects of nutrition for review, see Downing and Scara- . muzzi, 1991; Cosgrove and Foxcroft, 1996; Prunier and Quesnel, 2000 . However, doubts concerning the nature and the relative importance of the physiological mecha- nisms explaining the effects of nutrition directly at the ovarian level are still existing. Therefore, the present paper will focus on these potential mechanisms, after a review of the consequences of undernutrition on folliculogenesis and ovulation. We will concen- trate on the pig species since inadequate nutrition of gilts and sows is thought to play a Ž . major role in reproductive disturbances Dourmad et al., 1994 .

2. Effects of nutrition on the ovarian activity

2.1. Folliculogenesis To our knowledge, there are only two reports relating the effects of nutrition on the overall population of antral follicles: one in prepubertal gilts from two different Ž . genotypes Dufour et al., 1985 and the second in primiparous sows before and shortly Ž . after weaning Quesnel et al., 1998a . Results failed to show any influence of feed restriction on the repartitioning of healthy follicles between size classes in prepubertal gilts from one genotype and in luteal-phase gilts. Contrarily, in prepubertal gilts from Ž . the second genotype and in adult sows, the proportion of small 0.4 to 1.0 mm healthy follicles to the total number of antral follicles was increased in feed-restricted females at Ž . the expense of the proportion of bigger healthy follicles Fig. 1 . Regardless of its influence on the repartitioning of the follicles between size classes and whatever the Fig. 1. Average proportion of healthy antral follicles per size class to the total number of antral follicles in Ž . lactating sows submitted to high or low levels of feeding adapted from Quesnel et al., 1998a . P -0.05 between groups. Ž physiological stage, nutrition had no effect on the rate of atresia prepubertal gilts: Dufour et al., 1985; cyclic gilts: Quesnel et al., unpublished data; lactating and weaned . sows: Quesnel et al., 1998a . The effects of feed restriction or of realimentation after feed restriction on character- istics of the follicles present at the ovarian surface have been studied in prepubertal gilts Ž . Cosgrove et al., 1992; Charlton et al., 1993; Booth et al., 1994, 1996 and in sows Ž . around weaning Quesnel et al., 1998a . These follicles measured G 2 mm in diameter Ž . and were healthy or atretic. In one study Cosgrove et al., 1992 , gilts received a progestagen that blocks LH pulsatility but may also alter follicular growth and matura- Ž . tion directly at the ovarian level Guthrie and Bolt, 1990 . All these studies have shown that growth of the surface follicles was influenced by the feeding regimen. For instance, the mean diameter and volume of the 10 largest follicles were decreased in underfed Ž females, regardless of the physiological stage and the use of a progestagen Cosgrove et . al., 1992; Charlton et al., 1993; Quesnel et al., 1998a . However, the biochemical characteristics of these follicles, such as the follicular fluid concentrations of oestradiol Ž and testosterone, the in vitro synthesis of oestradiol, the thecal binding of hCG human . Ž chorionic gonadotrophin were not clearly modified by the pattern of feeding Cosgrove . et al., 1992; Charlton et al., 1993 . Feed intake may influence the ability of the oocytes to be fertilized andror to Ž . develop into viable embryos as suggested by Zak et al. 1997a . These authors have shown that the percentage of corpora lutea represented by an embryo at day 28 of Ž . gestation s embryo survival was low in sows submitted to feed restriction during their last week of lactation and inseminated at their first oestrus after weaning. This effect of undernutrition may be due to an impaired quality of the oocytes which may, itself, be Ž . related to a poor ability of the follicles to support oocyte maturation Zak et al., 1997b . These hypotheses were tested in vitro by comparing the rate of maturation of oocytes harvested in sows submitted or not to feed restriction during their last week of lactation, as well as the ability of the follicular fluid, recovered from the same sows, to support Ž . oocyte maturation Zak et al., 1997b . Data showed advantages for both oocytes and follicular fluid recovered from preovulatory follicles of non-restricted sows. However, it should be noted that the duration between weaning and ovarian collection was different in the two groups of sows, assuming that oestrus and preovulatory surge of LH may Ž . occur 12 h later in non-restricted than in restricted sows Zak et al., 1997a . Moreover, the reduction in embryo survival associated with feed restriction during the last week of lactation was no more observed in a more recent experiment from the same laboratory Ž . Mao et al., 1999 . Therefore, it seems that nutrition may influence growth of antral follicles from various sizes but consequences on their ability to further mature or to become atretic are still not clear. Similarly, experiments are needed to definitively conclude on the consequences of undernutrition on the oocyte maturation. 2.2. OÕulation occurrence Ovulation is accurately detected with measurement of plasma progesterone or by observation of the ovaries by ultrasonography, laparoscopy or after removal. However, it is generally inferred from oestrus detection with a mature boar, the occurrence of ‘‘silent’’ ovulation and of oestrus without ovulation being observed in only a minority of Ž . gilts Eliasson, 1991 . In prepubertal gilts, feed restriction delays the occurrence of first ovulation in most Ž . Ž studies Table 1 . In cyclic gilts, very low level of feeding daily maintenance levels of essential amino acids and of protein but less than 25 of the daily energy requirements . Ž for maintenance was able to cause cessation of the ovulatory oestrous cycles Armstrong . and Britt, 1987; Rozeboom et al., 1993 . However, gilts experienced three cycles in average before becoming anoestrus. Duration of these cycles was normal and proges- terone levels during the luteal phase were within a common range. A less severe level of Ž . energy restriction about 75 of the energy requirements for maintenance applied Ž during six to seven oestrous cycles was not sufficient to induce anoestrus Rozeboom et . al., 1993 . In most studies, feed restriction during lactation delays the occurrence of oestrus in Ž . sows after weaning Table 2 . Similarly, protein restriction during lactation prolongs the Ž weaning-to-oestrus interval in primiparous sows King and Williams, 1984a; Jones and . Ž Stahly, 1999 . In contrast, feed restriction after weaning does not delay oestrus Table . 2 . However, it should be noted that, in this latter situation, feed supply covers more than 100 of the requirements for maintenance in both control and restricted sows. Table 1 Influence of feed restriction during growth on the age at puberty of gilts † ‡ Ž . Reference Feed supply 300 to 250 250 to 200 200 to 150 150 to 100 Age at first oestrus or at first oÕulation days Friend et al., 1981 a b Trial 1 159 – 170 – Trial 2 165 166 – – ns Etienne et al., 1983 – 232 225 – ns a a b Den Hartog and Noordewier, 1984 234 237 244 – a b Ogle and Dalin, 1989 – 234 – 261 Dyck, 1989 193 201 – ns a b Dourmad et al., 1990 – 187 198 – a b Beltranena et al., 1991 170 187 – – Percentage of cyclic gilts at a fixed age Prunier et al., 1987 – 30 20 – ns a a b Newton and Mahan, 1992 91 89 77 – a b Prunier et al., 1993 40 – – † Ž Data published before 1980 were discarded since characteristics of female pigs reproductive and growth . performance, metabolic body reserves have greatly changed from that period. ‡ Ž . Feed supply is the estimated ratio between metabolic energy intake and requirements for maintenance Ž . MAINT . When this ratio was not indicated in the publication, it was necessary to calculate it. The following Ž . equations derived from Noblet et al. 1999 were used: metabolic energy intakes 0.95=digestible energy Ž . Ž . 0.60 intake; MAINT MJ s1.0=liveweight kg . P -0.05, ns P 0.05 within a row, between feeding levels. a,b Within a row, means with a different superscripts differ at P -0.05. Table 2 Ž . Influence of feed restriction in cyclic, lactating or weaned female pigs on the ovulation rate OR and the Ž . a weaning-to-oestrus interval WOI in days b Ž . Period of feed Reference Feed supply OR WOI restriction High Low High Low High Low Luteal phase Quesnel et al., 1999 240 80 20.3 20.3 – – Luteal and Cox et al., 1987 270 135 15.8 13.4 – – follicular phases Flowers et al., 1989 270 135 16.0 9.4 – – Beltranena et al., 1991 – – 13.5 11.1 – – Ashworth, 1991 380 90 19.2 14.2 – – c Luteolysis and Quesnel et al., 1999 240 80 20.6 16.9 – – follicular phases Lactation King and Williams, 1984b 85 40 14.4 13.4 7.6 19.9 Kirkwood et al.,1987 85 45 18.1 18.6 4.3 5.8 Kirkwood et al.,1990 80 40 17.6 17.7 6.0 8.9 Baidoo et al., 1992 80 45 16.2 16.7 5.9 7.5 d Zak et al., 1997a 80 45 19.9 15.4 3.7 5.6 Quesnel and Prunier, 1998 90 60 19.2 20.7 5.7 5.9 Zak et al., 1998 85 50 14.4 15.6 4.2 6.3 d Mao et al., 1999 87 48 17.7 16.7 4.6 5.1 e van den Brandt et al., 2000 79 67 18.2 16.9 5.1 5.4 After weaning Den Hartog and – – 15.2 14.8 9.1 8.2 van der Steen, 1981 King and Williams, 1984b 285 115 14.6 13.2 13.4 14.1 Baidoo et al., 1992 245 155 16.6 16.2 6.0 5.9 P -0.05 between control and feed-restricted groups, within criterium. a Ž Data published before 1980 were discarded since characteristics of female pigs reproductive and growth . performance, metabolic body reserves have greatly changed from that period. b Ž . Feed supply is the estimated ratio between metabolic energy intake and requirements for maintenance Ž . MAINT in prepubertal, mature and pregnant gilts, and in sows restricted after weaning or for maintenanceq Ž . milk production MAINTqMILK, in lactating sows . When this ratio was not indicated in the publication, it Ž . was necessary to calculate it. The following equations derived from Noblet et al. 1990, 1999 were used: Ž . Ž . 0.60 metabolic energy intakes 0.95=digestible energy intake; MAINT MJ s1.0=liveweight kg for prepu- Ž . Ž . 0.75 bertal and mature gilts; MAINT MJ s 0.44=liveweight kg for pregnant gilts and weaned sows; Ž . Ž . 0.75 Ž . MAINTqMILK MJ s 0.46=liveweight kg q28.59=daily gain of the litter kg y0.52 for lactating sows. c Number of follicles larger than 5 mm, 5 days after induced luteolysis. d Feed restriction was limited to the last week of lactation. e Ž . Sows with delayed return to oestrus after weaning 10 days were discarded. In summary, oestrous cycles can be disrupted in cyclic gilts only in case of very severe and prolonged feed restriction. Delayed ovulation at puberty and after weaning may occur in less drastic conditions. Therefore, under farm conditions, inadequate appetite is suspected to have negative effects on the age at puberty and on the weaning-to-oestrus interval. 2.3. OÕulation rate Feed restriction of growing gilts before puberty has no marked effect on the ovulation Ž . Ž . rate at the first oestrus: Etienne et al. 1983 and Beltranena et al. 1991 did not show Ž . any significant difference between feeding levels, whereas King 1989 observed a lower ovulation rate in underfed females. In cyclic gilts, imposing feed restriction only during the luteal phase seems to have no effect on the number of corpora lutea at the Ž . subsequent ovulation Table 2 . Contrarily, feed restriction starting during the luteal phase or at luteolysis and maintained during the follicular phase reduces the ovulation rate at next oestrus. In this case, feed restriction occurs, at least, during recruitment and selection of the preovulatory follicles. In reproductive sows, most researchers have determined the effects of feeding level during lactation on ovulation rate after weaning. Therefore, they have imposed the nutritional deficit before recruitment of the preovulatory follicles. This might explain Ž . why, in most studies, feed restriction has no clear effect on ovulation rate Table 2 . Ž . Likewise, King and Williams 1984a showed similar ovulation rates in primiparous sows submitted or not to protein restriction during lactation. Few studies have evaluated Ž the effects of feed intake between weaning and oestrus s during and after recruitment . and selection . Out of the three studies reported, only one showed an altered ovulation Ž . rate in the restricted group Table 2 . Therefore, it seems that, in cyclic gilts, the critical period for the ovulation rate is at luteolysis and during the follicular phase, i.e., during recruitment and selection of the preovulatory follicles. In reproductive sows, more studies are necessary to elucidate whether nutrition may also influence the ovulation rate by focussing experiments around Ž . weaning day of weaning and the following days .

3. Physiological mechanisms mediating the effects of nutrition directly at the ovarian level