Search
Author
Title
Vol.
Issue
Year
1st Page

# Journal of Animal Science - Article

1. Vol. 84 No. 9, p. 2316-2337

Published: December 8, 2014

2 Corresponding author(s):
View

doi:10.2527/jas.2006-156

# BOARD-INVITED REVIEW: Intrauterine growth retardation: Implications for the animal sciences1

1. G. Wu*2,
2. F. W. Bazer*,
3. J. M. Wallace and
4. T. E. Spencer*
1. Department of Animal Science, Texas A&M University, College Station, TX 77843; and
Rowett Research Institute, Bucksburn, Aberdeen, AB21 9SB, UK

## Abstract

Intrauterine growth retardation (IUGR), defined as impaired growth and development of the mammalian embryo/fetus or its organs during pregnancy, is a major concern in domestic animal production. Fetal growth restriction reduces neonatal survival, has a permanent stunting effect on postnatal growth and the efficiency of feed/forage utilization in offspring, negatively affects whole body composition and meat quality, and impairs long-term health and athletic performance. Knowledge of the underlying mechanisms has important implications for the prevention of IUGR and is crucial for enhancing the efficiency of livestock production and animal health. Fetal growth within the uterus is a complex biological event influenced by genetic, epigenetic, and environmental factors, as well as maternal maturity. These factors impact on the size and functional capacity of the placenta, uteroplacental blood flows, transfer of nutrients and oxygen from mother to fetus, conceptus nutrient availability, the endocrine milieu, and metabolic pathways. Alterations in fetal nutrition and endocrine status may result in developmental adaptations that permanently change the structure, physiology, metabolism, and postnatal growth of the offspring. Impaired placental syntheses of nitric oxide (a major vasodilator and angiogenic factor) and polyamines (key regulators of DNA and protein synthesis) may provide a unified explanation for the etiology of IUGR in response to maternal undernutrition and overnutrition. There is growing evidence that maternal nutritional status can alter the epigenetic state (stable alterations of gene expression through DNA methylation and histone modifications) of the fetal genome. This may provide a molecular mechanism for the role of maternal nutrition on fetal programming and genomic imprinting. Innovative interdisciplinary research in the areas of nutrition, reproductive physiology, and vascular biology will play an important role in designing the next generation of nutrient-balanced gestation diets and developing new tools for livestock management that will enhance the efficiency of animal production and improve animal well being.

### INTRODUCTION

Growth (an increase in the number and size of cells or in the mass of tissues) and development (changes in the structure and function of cells or tissues) of the fetus are complex biological events influenced by genetic, epigenetic, maternal maturity, as well as environmental and other factors (Redmer et al., 2004; Gootwine, 2005). These factors affect the size and functional capacity of the placenta, uteroplacental transfer of nutrients and oxygen from mother to fetus, conceptus nutrient availability, the fetal endocrine milieu, and metabolic pathways (Bell and Ehrhardt, 2002; Fowden et al. 2005; Reynolds et al., 2005).

The effects of uterine capacity, which can be defined as the physiological and biochemical limitations imposed on conceptus growth and development by the uterus (Bazer et al., 1969a,b) and maternal nutrition on fetal growth have clearly been demonstrated by studies involving embryo transfer (Dickinson et al., 1962; Ferrell, 1991; Allen et al., 2002) and altered maternal nutrient intake (Redmer et al., 2004), respectively. Further, uterine environment can affect the size of the fetus, as demonstrated in different breeds of pigs (Wilson et al., 1998). There are a plethora of studies aimed at identifying nutritionally sensitive periods of conceptus (i.e., embryo/fetus, associated placental membranes, and fetal fluids) development. Available evidence suggests that the prenatal growth trajectory of all eutherians (placental mammals) is sensitive to the direct and indirect effects of maternal nutrition at all stages between oocyte maturation and birth (Robinson et al., 1999; Rehfeldt et al., 2004; Ferguson, 2005).

Intrauterine growth retardation (IUGR) can be defined as impaired growth and development of the mammalian embryo/fetus or its organs during pregnancy. Because it is easy to measure practically on farms and in clinics, fetal weight or birth weight relative to gestational age is often used as a criterion to detect IUGR. Naturally occurring and environmentally (e.g., over-and underfeeding, heat stress, disease, and toxins) induced IUGR are well documented for livestock (including cattle, goat, horse, pig, and sheep; Pond et al., 1969; Baker et al., 1969; Wallace et al., 2005b) and litter-bearing small mammals (e.g., dog, mouse, and rat; Wootton et al., 1983).

Despite improvement of management techniques and intensive research on mammalian nutrient requirements over the past half-century, IUGR remains a significant problem in animal agriculture because of our incomplete knowledge concerning the impact of nutrition on the mechanisms regulating fetal growth. The major objective of this article is to critically review the literature on IUGR in domestic animals, its implications for the animal sciences, its putative biological mechanisms, and its potential solutions. Readers are referred to recent reviews for discussion of IUGR in rodents and humans (Wu et al., 2004a; McMillen and Robinson, 2005; Murphy et al., 2006).

## INTRAUTERINE GROWTH RETARDATION IN LIVESTOCK

Significant losses of embryos/fetuses occur during early, mid, and late gestation (Geisert and Schmitt, 2002; van der Lende and van Rens, 2003; Jonker, 2004). In addition to the genetic contribution from both parents, fetal growth and development are affected by a variety of environmental and other factors. These include maternal nutrition (low or high feed intake, and nutrient imbalance), maternal intestinal malabsorption, inadequate provision of amniotic and allantoic fluid nutrients, the ingestion of toxic substances, environmental temperature and stress, disturbances in maternal or fetal metabolic and homeostatic mechanisms, insufficiency or dysfunction of the uterus, endometrium, or placenta, and poor management (Mellor, 1983; McEvoy et al., 2001; Redmer et al., 2004; Wu et al., 2004a).

The outcome of stressful conditions in utero depends on their nature, severity, stage of gestation, and duration. Thus, multiple factors regulate conceptus growth and contribute to IUGR (Figure 1). Insufficient uterine capacity and inadequate maternal nutrition are 2 major factors that impair fetal growth.

## Uterine Capacity and IUGR

### Experimentally Induced Uterine Insufficiency.

Although the fetal genome plays an important role in growth potential in utero, increasing evidence suggests that the intrauterine environment is an important determinant of fetal growth (Wilson, 2002). Further, the intrauterine environment of the individual fetus may be of greater importance in the etiology of chronic diseases in adults than the genetics of the fetus (Wu et al., 2004a). Additionally, in domestic animals (including sheep, cattle, and horses), when the embryo from a genetically larger mother was transferred to a recipient dam with a lower uterine capacity, IUGR was a pregnancy outcome (Dickinson et al., 1962; Ferrell, 1991; Allen et al., 2002). Conversely, when the embryo from a genetically smaller mother was transferred to a recipient dam with a greater uterine capacity, fetal growth was enhanced (Dickinson et al., 1962; Ferrell, 1991; Allen et al., 2002) but cardiovascular dysfunction occurred (Giussani et al., 2003). Notably, genetic selection of gilts for high uterine capacity led to increased litter size and total litter weight without a change in average piglet weight (Vallet et al., 2002).

### Production-Imposed Uterine Insufficiency.

Through improving reproductive technologies (e.g., embryo transfer and hormonal induction of ovulation), twinning in cattle offers an important means to increase the efficiency of beef production, because the overhead costs for maintaining single-calving cows account for more than 50% of the total costs of beef production (Guerra-Martinez et al., 1990). Herd input costs per unit of beef output value can be reduced by 24% in twin compared with single births (Guerra-Martinez et al., 1990). However, twin-bearing heifers and cows can lose 10 to 12% of empty BW during the last one-third of pregnancy. Also, twinning reduces fetal growth and calf birth weight. In sheep, high prolificacy is a desirable trait under intensive management systems, and increasing the prolificacy of ewes through genetic selection is an effective means to increase the profitability of lamb production (Gootwine et al., 2001). However, an increased number of fetuses within the uterus results in relative placental insufficiency and low birth weights (Gootwine et al., 2006). These observations indicate a challenge for reducing the risk of IUGR associated with current reproductive technologies that increase ovulation rates.

Maternal or gynecological immaturity is a prototype for production-imposed uterine insufficiency in livestock. Domestic animals are often bred at immature BW to maximize their production performance. For example, it is commonly recommended and practiced that ewe lambs be bred with fertile rams at two-thirds of their mature BW with the goal of first lambing at 12 to 13 mo of age (Chappell, 1993). Similarly, heifers and gilts enter the first pregnancy at 70 to 80% of their mature BW. The birth weights of the first-parity progeny (e.g., in lambs, calves, piglets, and foals) from immature dams are generally 10 to 15% lower compared with the offspring born from dams of mature adult BW (Bellows and Short, 1978; Quiniou et al., 2002; Wilsher and Allen, 2003). This can be explained by the fact that mother and fetus grow substantially and compete for nutrients during pregnancy (Redmer et al., 2004; Wu et al., 2004a). Interestingly, when heifers began the first pregnancy at a more mature BW achieved at 21 rather than 15 mo of age, the parity of the dam had no effect on fetal growth rate (Tudor, 1972). Thus, it is the maturity of the dam rather than parity that affects intrauterine growth.

### Natural Uterine Insufficiency.

Natural IUGR is frequently observed in dams with multifetal pregnancies (Wootton et al., 1983). Although total placental weight is increased in these animals, placental mass per fetus is reduced, resulting in relative placental insufficiency (Redmer et al., 2004). Thus, in ewes, the individual birth weight of a lamb in triplet and twin pregnancies is only 62 and 78%, respectively, of a singleton pregnancy (Gootwine, 2005). Twins account for 38 to 52% of all pregnancies in sheep (USDA, 2003). Even in well-fed ewes, multifetal pregnancy impairs fetal growth, including reduction in the skeletal muscle mass and myofiber number of neonates (Greenwood et al., 2000). Similar findings have been observed for heifers and cows (Guerra-Martinez et al., 1990) as well as horses (Rossdale and Ousey, 2002) when natural multifetal pregnancy occurs.

Among domestic animals, pigs exhibit the most severe, naturally occurring IUGR. Before d 35 of gestation, porcine embryos are uniformly distributed within the uterine horn (Anderson and Parker, 1976). After this date, uterine capacity becomes a limiting factor for fetal growth, even though the fetuses are distributed relatively uniformly (Knight et al., 1977). Consequently, porcine fetal development may depend on the position and number of fetuses in the uterus, in that fetuses near both ends of the uterus (i.e., the uterotubal junction and the cervix) are generally larger than those in the middle of the horn (Perry and Rowell, 1969). This differential growth of porcine fetuses in relation to their position within the uterus is evident, particularly during late gestation, in pregnancies in which the number of fetuses exceeds 5 per horn (Perry and Rowell, 1969).

However, runts can be present at any position within the uterus and are more related to the size of the placenta. At birth, runt piglets may weigh only one-half or even one-third as much as the largest littermates (Widdowson, 1971). The small intestine, liver, and skeletal muscle of the runt pig are disproportionately smaller than those of the largest littermates at birth (Widdowson, 1971). Whereas IUGR could be considered as a natural mechanism to protect the dam in cases of undernutrition, it may not be beneficial for the survival and growth performance of the progeny or the efficiency of livestock production in animal agriculture.

## Undernutrition and IUGR

### Undernutrition Under Practical Production Conditions.

Animals of agricultural importance are raised under various production conditions (e.g., intensive and extensive systems), depending on the species, region, and season. Pigs are commonly housed in pens and fed primarily plant-based formulated diets (an intensive system). Ruminants (e.g., cattle, sheep, goats, and deer) and horses are allowed to graze pasture in rangeland and consume forages (an extensive system); receive feedlot diets containing various supplemental levels of energy, protein, vitamins, and minerals; or a combination of these. Pasture grazing is the most common practice for managing dairy cows worldwide, and it is gaining renewed interest in the United States (Boken et al., 2005; Fontaneli et al., 2005). Also, beef herds in the United States and worldwide are managed under conditions varying from confinement cow-calf production units to the more common grazing systems. However, the quality of forages and roughages is often poor, particularly in dry and winter seasons, and is inadequate for optimal nutrition of growing, gestating, and lactating herbivores (including ruminants and horses) without high-quality protein and energy supplements (Lippke, 1980; Hoaglund et al., 1992; Huston et al., 1993; Fontaneli et al., 2005). In extensive production systems worldwide, there is little or no supplement provided for grazing ruminants (Fontaneli et al., 2005).

Thus, fetal undernutrition frequently occurs in animal agriculture, leading to reduced fetal growth. For example, Thomas and Kott (1995) reported that, without any supplement, the nutrient uptake of grazing ewes in the western United States is often less than 50% of the National Research Council (NRC) recommendations (NRC, 1985). Unsupplemented, grazing ewes lose a significant amount of BW during pregnancy, and their health, fetal growth, and lactation performance are seriously compromised (Thomas and Kott, 1995). Also, the content of MP in the grazed forage, particularly during winter, is low (often <8% on a DM basis) and is inadequate for supporting optimal reproductive performance of beef heifers or cows (Patterson et al., 2003; Ferguson, 2005). Additionally, the sheep is a seasonal breeder. In the United States, ewes usually enter pregnancy in late fall or early winter seasons, and therefore, most of the gestational period coincides with winter, when the grazed forage is of low quality (Hoaglund et al., 1992). Also, gestating heifers in feedlot situations often have inadequate intakes of nutrients and poor pregnancy outcomes (including reduced fetal growth; Kreikemeier and Unruh, 1993). Another example of production-imposed fetal undernutrition is shortening of the postpartum-to-breeding or interpregnancy intervals. Although this practice is desirable for increasing the potential economic return from livestock production (Ferguson, 2005), it results in maternal nutritional depletion at the outset of pregnancy (Wu et al., 2004a).

Finally, in tropical or subtropical regions, high environmental temperatures reduce feed intake by pregnant dams that graze pasture in open rangelands or by pigs housed without air conditioning. The thermal stress will cause IUGR in animals (Reynolds et al., 1985; 2005; Wallace et al., 2005c). Conversely, exposure to a cold climate can increase the utilization of dietary energy for maintaining maternal and fetal body temperatures, thereby reducing the availability of nutrients for fetal growth (Ferguson, 2005).

### Undernutrition Due to Maternal Physiological Extremes.

Various physiological extremes of gestating or lactating dams often result in fetal undernutrition. Ewes commonly exhibit ketosis during late gestation due to an energy deficit (Wastney et al., 1982), and acidotic conditions are associated with increased catabolism of branched-chain AA and glutamine by skeletal muscle and kidneys, respectively (Wu and Marliss, 1992). Indeed, ditocous ewes fed a 12%-CP diet (current NRC requirement) exhibited negative protein balance in maternal tissues between d 110 and 140 of gestation, indicating significant mobilization of protein reserves (McNeill et al., 1997). Similarly, bovine fetal undernutrition often occurs during late pregnancy, particularly in multiparous cows. In heifers and mature cows, voluntary feed intake usually decreases by 30 to 35% during the last 3 wk before calving (Grummer, 1995), when the absolute rate of fetal growth is most rapid (Ferrell, 1991). The reduced feed intake during this transition period is further decreased by conditions such as twin pregnancies, increase in body condition, primiparous pregnancy, and thermal stress (Grummer, 1995). Complicating the metabolic challenge during late pregnancy, much of gestation is concurrent with lactation in multiparous cows, where additional amounts of nutrients are required for conceptus growth (Knight, 2001). Low protein intake (e.g., 80% of NRC requirement) prepartum further reduces DMI in pregnant cows (Chew et al., 1984). These metabolic interplays cause negative energy and protein balances prepartum in heifers and cows (Grummer, 1995; Bell et al., 2000).

Low precalving BW of the cow is associated with low birth weight of the calf (Bellows et al., 1971). There is also a nutritional inadequacy before mating in lactating heifers and cows beginning at the second or greater parity. In these animals, milk output generally peaks at about 2 mo postpartum, but feed intake usually takes at least 2 mo (in some cases up to 4 to 5 mo despite provision of a high-quality diet) to reach its maximum, therefore resulting in negative nutrient balance (particularly energy and protein deficits) during early or mid-lactation (Bauman and Currie, 1980; Bar-Peled et al., 1998). This period usually coincides with the early stage of pregnancy in multiparous cows, and undernutrition affects embryonic and fetal development (Redmer et al., 2004). The findings that calf birth weight was increased in heifers and cows in response to dietary supplementation with protein and energy concentrates during late gestation (Clanton and Zimmerman, 1970; Bellows and Short, 1978) suggest that undernutrition caused by the maternal physiological extremes impairs fetal growth in unsupplemented dams.

Besides the ruminant, low feed intake remains a significant problem for lactating sows before breeding, when the mobilization of nutrient reserves for milk production results in a severe catabolic state and a prolonged interval from farrowing to estrus (Cole, 1990). A 3-yr study of 10,200 lactating sows on 120 farms in the United States showed that feed consumption could be as low as 70% of the NRC requirements (Johnson, 1993). Inadequate nutrition increased losses of BW and backfat in lactating sows and also prolonged weaning-to-estrus intervals (Johnson, 1993). When sows enter pregnancy, the suboptimal nutritional status (namely premating maternal undernutrition), coupled with restricted feed intake (Ji et al., 2005), may negatively affect the growth and development of early embryos and fetuses (Vinsky et al., 2006).

Maternal insulin resistance gradually develops in cows (Bell et al., 2000), ewes (Wastney et al., 1982), horses (Hoffman et al., 2003), and sows (Kemp et al., 1996) during late pregnancy (Bell et al., 2000), likely because of the inability of the liver and skeletal muscle to oxidize the fatty acids released from adipose tissue in response to a negative energy balance (Ferguson, 2005). An increase in plasma and tissue levels of free fatty acids is a major factor contributing to the occurrence of insulin resistance (Jobgen et al., 2006). There is evidence that low glucose tolerance of pregnant sows is associated with high postnatal mortality of piglets (Kemp et al., 1996).

Whereas insulin resistance in the dam may have the potential to increase the availability of glucose and AA for the fetus, the transfer of nutrients from mother to fetus may be impaired under this condition. Because insulin stimulates muscle protein synthesis and inhibits muscle protein degradation, insulin resistance increases the net rate of whole-body proteolysis and thus plasma levels of methylarginines (protein-derived inhibitors of endothelial nitric oxide [NO] synthesis; Marliss et al., 2006). Because NO is a major regulator of uteroplacental blood flows (Bird et al., 2003), severe insulin resistance likely compromises the placental delivery of nutrients and oxygen during late gestation. In support of this view, IUGR is associated with elevated concentrations of plasma asymmetric dimethylarginine in obese subjects (Savvidou et al., 2003).

### Experimentally Induced Undernutrition.

In addition to the above practical production and physiological conditions of undernutrition that can result in IUGR, well-controlled experimental studies have demonstrated that maternal undernutrition during the periconceptual or gestational periods reduces fetal growth in sheep (Mellor, 1983; Osgerby et al., 2002; Vonnahme et al., 2003), cows (Tudor, 1972), pigs (Pond et al., 1969), and horses (Pugh, 1993). In adult sheep, severe under-nutrition during the periconceptual period accelerates maturation of the fetal hypothalamic-pituitary-adrenal axis and causes preterm delivery (Fowden et al., 1994). Low prepregnancy weights, followed by undernutrition during midpregnancy, result in reduced placental growth and lower birth weights at term (Redmer et al., 2004). Studies involving the restricted intake of nutrients solely during midgestation reveal variable effects on the placental and fetal growth trajectory; however, if undernutrition is prolonged during late pregnancy, fetal growth is compromised, particularly in twin pregnancies (Redmer et al., 2004; Luther et al., 2005a). Reduced provision of all nutrients to the ovine fetus through a combination of reduced maternal feed intake and carunclectomy also resulted in IUGR and particularly impaired growth of the fetal gastrointestinal tract (Trahair et al., 1997). Intrauterine growth retardation in undernourished sheep is often associated with fetal hypoglycemia and hypoxemia as well as with increased risks of fetal death and premature birth (Mellor, 1983).

In Hereford cows, submaintenance levels of nutrition during the last trimester reduced calf birth weight (Tudor, 1972). In heifers bred at 15 mo of age, reducing nutrient intake from high to maintenance to low levels via decreasing the amounts of feedlot rations or pasture availability during the last 3 mo of pregnancy also caused a progressive decrease in birth weights of calves (Kroker and Cummins, 1979). Likewise, in beef heifers fed a low-level protein diet, a BW loss of 0.5 kg/d during the last trimester was associated with weak labor, increased incidence of dystocia, increased perinatal mortality, reduced postnatal growth of calves, and prolonged postpartum anestrus (Kroker and Cummins, 1979). Also, in heifers, decreasing daily TDN from 6.4 to 3.4 kg for 90 d before calving led to a substantial loss of maternal tissues during pregnancy, reduced calf birth weight, and prolonged postpartum-to-estrus intervals (Bellows and Short, 1978). In mares whose fetus has a limited ability to synthesize glucose during the entire gestation, maternal fasting caused an increased uteroplacental production of PGF and uterine contractility, impaired fetal growth, premature delivery of nonviable foals in most animals (>80% of pregnancies) during late gestation, and low birth weight (Fowden et al., 1994).

In contrast to the ruminant and horse, the pig generally has a remarkable ability to mobilize maternal nutrient reserves to support placental and fetal development during prolonged inanition in the presence of adequate progesterone and estrogen (Anderson, 1975). Thus, a modest reduction in the dietary intake of energy alone is not sufficient to cause IUGR in pigs. For example, in gilts fed adequate amounts of protein, vitamins and minerals, restriction of dietary energy intake (50% of controls) did not affect birth weight of piglets (Atinmo et al., 1974). However, with a more severe reduction in energy intake by gilts during the entire gestation from 8.0 to 2.2 Mcal of DE/d caused a reduction in birth weights, the number of gastrocnemius muscle fibers, muscle weight, liver weight, liver glycogen content, and serum protein concentrations of newborn piglets (Buitrago et al., 1974).

Energy deficiency likely reduces protein synthesis in the liver and skeletal muscle. Results of the following extensive studies indicate that maternal underfeeding of energy and protein impairs embryonic/fetal growth in pigs. First, reducing the intake of complete rations by 50% for 2 estrous cycles before mating decreased fetal weight at d 30 of pregnancy in gilts (Ashworth, 1991). Similarly, in primiparous sows, restricting feed intake by 50% during lactation (a reduction from 5.0 to 2.5 kg/d between d 14 and 21 of lactation) before mating reduced the weight of both male and female fetuses as well as the survival of female embryos at d 30 of gestation (Vinsky et al., 2006). Second, birth weight of piglets decreased in response to restriction of feed intake (e.g., 0.9 vs. 1.9 kg/d) or increased litter size (Baker et al., 1969). Third, decreasing feed intake after d 80 of gestation reduced fetal growth in gilts (Noblet et al., 1985). Finally, birth weights as well as brain and liver weights were reduced in the progeny of gilts fed a protein-deficient diet throughout gestation (Pond et al., 1969; Atinmo et al., 1974). These findings suggest that porcine fetal growth can be influenced by a severe maternal protein-energy imbalance during pregnancy.

### Overnutrition and IUGR

Increasing energy intake increases the rate of ovulation in farm animals (including cattle, sheep, pigs, and horses). Thus, the practice of increasing feed intake during a short period of time (termed flushing) around the time of conception has been employed by producers in an attempt to increase the number of embryos/fetuses (Cole, 1990). Overnutrition can result from increased intake of energy, protein, or both. Thus, overfeeding of livestock and companion animals occurs when excess amounts of diets (particularly concentrates) are provided to dams before breeding or during pregnancy (Han et al., 2000; Luther et al., 2005b). Indeed, overconditioning of cows during the dry period still occurs on many farms, particularly among high-producing herds (Ferguson, 2005).

Maternal overnutrition (high energy, high protein feeding, or both) during the premating period or early pregnancy often results in increased porcine embryo and fetal mortality (Ashworth, 1991; Einarsson and Rojkittikhun, 1993). Interestingly, like underfeeding, overfeeding once pregnancy is established retards fetal growth in pigs (Cole, 1990) and adolescent sheep (Wallace et al., 2004). Strikingly, feeding mares to obesity before or after mating can also reduce fetal growth and cause fetal death (Pugh, 1993). Overfeeding of dairy cows during late pregnancy is associated with an increased risk of metritis, ketosis, milk fever, cystic ovaries, and subsequent infertility. Further, overconditioned cows are more susceptible to a prepartum decrease in voluntary feed intake, thereby compromising nutritional status in the mother and fetus (Ferguson, 2005).

Increased feed intake by sows during all or part of gestation has a negative effect on feed intake during lactation (Han et al., 2000). In multiparous sows, increasing dietary intakes of both protein and energy by 43% during the first 50 d of gestation, relative to a standard gestational diet (10.7 MJ of DE/kg and 12.0% CP), decreased the birth weights of the 2 lightest and 2 heaviest piglets in litters (Bee, 2004). Likewise, overfeeding both energy and protein between d 25 and 50 of gestation had no beneficial effect on muscle fiber number or area in the offspring but instead reduced skeletal muscle weight of newborn piglets due to smaller fiber size (Nissen et al., 2003). Furthermore, overfeeding gilts by 40% of the NRC requirements (NRC, 1998) during the entire gestation impaired fetal development and postnatal survival (Han et al., 2000). These results indicate that overfeeding during all or part of the gestation has a detrimental effect on pregnancy outcomes in domestic animals.

## IMPLICATIONS OF IUGR FOR THE ANIMAL SCIENCES

The major goals of animal production are to enhance the efficiency of feed/forage utilization, produce abundant, healthy meats, eggs, milk, and wools for increasingly health-conscious consumers, and improve the quality of human life. Fetal growth restriction is a significant obstacle to achieving these goals. The available evidence suggests that IUGR has permanent negative impacts on neonatal adjustment, preweaning survival, postnatal growth, feed utilization efficiency, lifetime health, body composition, and meat quality, as well as reproductive and athletic performance (Table 1). Thus, IUGR has important implications for the animal sciences.

Low birth weight is associated with high neonatal morbidity and mortality rates in domestic animals, particularly under adverse climatic conditions (Mellor, 1983; Azzam et al., 1993; Van Rens et al., 2005). Recent data show that preweaning deaths in neonates of domestic animals in the United States remain high, with most mortality occurring within the first days of postnatal life (Table 2). These high preweaning mortality rates result in economic loss and emotional stress for animal owners, particularly on farms raising heifers/cows and horses with long gestational periods (280 and 335 d, respectively).

Intestinal and respiratory dysfunctions, which occur in IUGR neonates (Thornbury et al., 1993; Trahair et al., 1997; Rossdale and Ousey, 2002), are major factors contributing to preweaning mortality in livestock (Table 2). Neonates with IUGR that survive the first days of life are often at increased risk for subsequent neurological, respiratory, intestinal, and circulatory disorders during the neonatal period (Wu et al., 2004a). Neonates whose intrauterine growth is retarded due to small placentae or severe malnutrition often become hypoglycemic and hypoxemic, and are susceptible to fatal hypothermia in response to cold stress due to impaired thermogenic mechanisms (both shivering and non-shivering) and low energy reserves (Mellor, 1983). In addition, the reduced quality (nutrient composition and immunoglobulin content) and quantity of colostrum produced at parturition by dams underfed or overfed during pregnancy also negatively affects neonatal survival (Mellor, 1983; Wallace et al., 2001).

Of note, the physical appearance or behavior may appear to be normal in some neonates (e.g., foals), but their various organs may not be functionally mature (Ginther and Douglas, 1982). Thus, special care is required for managing young animals that experience IUGR, which adds additional costs to animal production. This is compounded when fetal growth restriction is accompanied by a major reduction in gestation length (premature delivery), which frequently occurs in over-nourished adolescent sheep (Wallace et al., 2001) and in underfed mares (Rossdale and Ousey, 2002). Thus, future research is warranted to identify the proportion of neonatal death caused by IUGR in livestock.

Compared with high birth weight offspring, IUGR newborn lambs (Greenwood et al., 1998), calves (Bellows et al., 1971), piglets (Milligan et al., 2002; Quiniou et al., 2002), and foals (Ginther and Douglas, 1982) suffered from greater rates of neonatal mortality and took a longer period of time to adapt to postnatal life. Heavier offspring (including calves, lambs, and piglets) at birth are more viable and more rapidly adjust to the extrauterine environment (Cundiff et al., 1986). Below 0.8 kg of birth weight, 35% of piglets are stillborn, in comparison with 4% for birth weights ranging from 1.2 to 1.4 kg (Quiniou et al., 2002). Preweaning survival rates decrease progressively from 95 to 15% as piglet birth weights decrease from 1.80 to 0.61 kg (Quiniou et al., 2002). Approximately 15 to 20% of piglets are born with a birth weight less than 1.1 kg, and their survival and postnatal growth rates are severely reduced (Wu et al., 2004a).

Newborn piglets with IUGR suffer from necrotizing enterocolitis (a serious disorder of the small intestine; Thornbury et al., 1993), which impairs intestinal function, including the synthesis of arginine, an essential AA for neonatal pigs, but remarkably deficient in sow’s milk (Wu et al., 2004d). Necrotizing enterocolitis is a major cause of death in neonates, including piglets (Thornbury et al., 1993), and can be ameliorated by dietary arginine supplementation (Wu et al., 2004c). Foals with IUGR exhibit organ dysfunction (e.g., skeletal and respiratory problems, and reduced immune function). Twin foals have poor prospects for postnatal survival of one or both foals (Ginther and Douglas, 1982). The second-day syndrome in neonatal horses, defined as a condition in which foals markedly deteriorate on the second day postpartum, may result from IUGR due to impaired pulmonary and metabolic functions (Rossdale and Ousey, 2002).

### Postnatal Growth and Efficiency of Feed/Forage Utilization

The gut and muscle coordinate nutrient metabolism in animals. The small intestine plays an important role in terminal digestion and absorption of nutrients and, therefore, in postnatal growth of animals (Wu, 1998). In growing animals, protein deposition in skeletal muscle is a high priority, and accounts for approximately 15% of total energy expenditure (Wu and Self, 2005). In contrast to fat (a hydrophobic substance), protein deposition is associated with retention of a large amount of water, with a ratio of approximately 1 to 3 on a gram basis (Wu and Marliss, 1992). Thus, with water content of 75 to 80% in the body, muscle protein balance is the major determinant of postnatal growth rate in young livestock. In other words, skeletal muscle growth is energetically more efficient than fat synthesis and accretion. Interestingly, natural or experimentally induced IUGR is associated with abnormal gastrointestinal morphologies and gastrointestinal dysfunction (Thornbury et al., 1993; Trahair et al., 1997; Wang et al., 2005) as well as the impaired development of skeletal muscle (Hegarty and Allen, 1978; Greenwood et al., 2000). These conditions will contribute to a reduced efficiency of nutrient utilization in the IUGR progeny.

Studies have shown that IUGR has a permanent stunting effect on postnatal growth and reduces the efficiency of feed/forage utilization. Some researchers have reported reduced postnatal growth of IUGR lambs under artificial rearing (Schinckel and Short, 1961; Villette and Theriez, 1981) or practical production conditions (Gootwine et al., 2006). Compared with high birth weight lambs, the IUGR newborn lambs grew slower within the first 2 wk, exhibited lower rates of efficiency of energy utilization for protein and fat deposition (Greenwood et al., 1998), and had lower intramuscular concentrations of DNA and lower rates of postnatal skeletal muscle growth (Greenwood et al., 2000). Reduced myofiber number in IUGR lambs limits the capacity for postnatal compensatory growth of skeletal muscle. Under feedlot and forage grazing conditions, the efficiency of feed utilization is lower for twins than singletons (Guerra-Martinez et al., 1990), and small birth weight calves grew more slowly before weaning than high birth weight calves (Cundiff et al., 1986). Recent studies involving embryo transfer have shown that IUGR led to a permanent stunting effect on postnatal growth of horses throughout life (Allen et al., 2004). These results indicate a negative effect of IUGR on postnatal nutrient utilization and growth performance in animals.

In addition to herbivores, maternal undernutrition during gestation stunted the postnatal growth and development of swine (Schoknecht et al., 1993). Low birth weight pigs also fail to increase their muscle fiber number or muscle growth during the postnatal period even when fed adequately (Hegarty and Allen, 1978). Thus, the small intestine, liver, and skeletal muscle of the runt pig continued to be disproportionately smaller than the largest littermates at 3 yr of age (Widdowson, 1971). Runt or fostered runt pigs exhibited lower rates of skeletal muscle and whole-body growth between birth and slaughter, and utilized feeds less efficiently for growth, compared with high birth weight littermates (Hegarty and Allen, 1978; Powell and Aberle, 1980). The lower piglet birth weight is associated with lower ADG during the suckling, nursing, and growing-finishing periods (Quiniou et al., 2002). Compared with progeny of gilts fed a diet containing adequate protein, postnatal growth rates between birth and weaning (5 wk of age) and between weaning and slaughter (90 kg) were markedly reduced in all progeny of gilts fed a protein-deficient, isocaloric diet during all or part of gestation, regardless of birth weights (Pond et al., 1969; Atinmo et al., 1974).

As with IUGR piglets born from underfed dams, progeny from sows overfed between d 0 and 50 of gestation exhibited slower growth rates during both lactation and growing-finishing periods, as well as lower efficiency of feed utilization for gain (G:F) in comparison with pigs born from underfed sows (Bee, 2004). Similarly, overfeeding both energy and protein between d 25 and 50 of gestation reduced the postnatal ADG, muscle deposition rate, and carcass weight at slaughter weight (104 kg; Nissen et al., 2003) of the offspring. During the newborn period, the fractional rate of protein synthesis (%/d) did not differ in tissues (skeletal muscle, heart, liver, pancreas, and jejunum) between normal and IUGR piglets under fasting or fed conditions (Davis et al., 1997). The inability of the IUGR piglets to increase tissue protein synthesis beyond that of the normal littermates explains their incomplete compensatory growth after birth. The longer the period of intrauterine nutrient deprivation, the lesser the ability of IUGR pigs to recover from the insult (Pond et al., 1969).

### Body Composition and Meat Quality

The function of an animal critically depends on its body composition of protein, fat, carbohydrates, minerals, vitamins, and water, which in turn influences the rate of postnatal growth. In addition, the contents of skeletal muscle, fat, and connective tissue as well as muscle fiber number and area are major factors that affect the postmortem quality of meat. Further, the intramuscular concentration of glycogen and the glycolysis rate postslaughter affect the production of lactic acid and the pH of meat, as well as its water-holding capacity. Also, an increase in the amount of intramuscular fat promotes lipid peroxidation postslaughter (Fang et al., 2002). This results in the oxidation of muscle proteins, including oxymyoglobin (the main pigment responsible for the bright red color of fresh meat) and, therefore, changes in the color and taste of meat (Gorelik and Kanner, 2001). A greater amount of connective tissue results in tougher meat. Fast-growing animals containing a high number of muscle fibers with a small cross-sectional area generally yield a greater quality meat (Gondret et al., 2005).

There is evidence showing that IUGR is associated with altered composition of the whole body and muscle, as well as the distribution of muscle fiber type, of the offspring (Wigmore and Stickland, 1983). During late gestation, growth-retarded fetuses from overfed adolescent mothers have greater relative fetal carcass fat content and perirenal fat mass than normally growing control fetuses (Matsuzaki et al., 2006). In contrast, the more modest fetal growth restriction in undernourished adolescent pregnancies is associated with preservation of fetal skeletal growth and depletion of fetal fat stores (Luther et al., 2005a). In lambs, low birth weight is associated with lower percentages of bone and muscle and a greater percentage of fat in the slaughter-weight (46 kg) carcass (Makarechian et al., 1978). Compared with high birth weight lambs, low birth weight lambs had more fat and less minerals in the whole body regardless of whether they exhibited slow or fast postnatal growth rates (achieved by feeding different levels of high-quality liquid diet; Greenwood et al., 1998).

### Long-Term Consequences for Health as Well as Reproductive and Athletic Performance

Most domestic animals are raised for producing meat at relatively young ages when muscle protein deposition approaches a plateau. For those animals selected for breeding, lactation, or both, IUGR may influence their subsequent reproductive performance. For example, under some production conditions (a combination of grazing and feed supplement), ewe lambs born as singletons with a greater birth weight reach puberty at both a younger age and a heavier weight than twin-born lambs (Southam et al., 1971). Additionally, female fetuses from overfed adolescent sheep pregnancies have fewer ovarian follicles than normally growing fetuses at mid- and late gestation (Da Silva et al., 2002; 2003) and hence a limited pool for follicular recruitment in adult life. Similarly, follicular development is delayed in IUGR piglets at birth (Da Silva-Buttkus et al., 2003). Furthermore, low birth weight lambs produced by an in utero crowding model have fewer uterine caruncles than normal birth weight lambs, and this may affect subsequent placental growth and uterine capacity (Aitken et al., 2003). For male lambs, low birth weight is associated with a delay in the onset of endocrine puberty and attenuated testis growth (Da Silva et al., 2001).

All of the above findings suggest that selection of IUGR offspring for breeding purposes is best avoided in animal production. Additionally, the lifespan of animals (including horses, cats, and dogs) that are raised for racing or for human companionship is increasing due to improved medical and nutritional care. Because IUGR results in smaller skeletal muscle and liver (Widdowson, 1971), and these organs play a crucial role in the metabolism of energy substrates (Jobgen et al., 2006), their reduced functional capacity may help explain impaired glucose utilization and dyslipidemia in the adult life of IUGR offspring. Likewise, abnormal composition and the reduced size of muscle fibers in the animal that experiences IUGR (Hegarty and Allen, 1978; Powell and Aberle, 1980) may impair energy metabolism, protein turnover, force generation, locomotion, strength, endurance, and coordination of skeletal muscle. Thus, IUGR likely has an adverse impact on lifetime health, reproductive performance, and athletic performance under practical livestock production and management conditions. In addition, results from well-controlled experiments with sheep show that IUGR progeny develop metabolic abnormalities, including reduced insulin secretion, insulin resistance, dyslipidemia, and cardiovascular dysfunction in adult life (Table 1). There is also evidence indicating that IUGR has a negative impact on athletic performance in horses (Rossdale and Ousey, 2002).

### Fetal Programming

The compelling evidence summarized in the preceding sections suggests that the intrauterine environment of the conceptus may alter expression of the fetal genome and have lifelong consequences. This phenomenon is termed fetal programming, which has led to the recent theory of the fetal, or developmental, origins of adult disease (Barker and Clark, 1997). Namely, alterations in fetal nutritional and endocrine status may result in developmental adaptations that permanently change the structure, physiology, and metabolism of the offspring, thereby predisposing individuals to metabolic, endocrine, and cardiovascular diseases in adult lives of animals and humans. Because growth performance, which depends on both the rates and the efficiency of metabolic transformations of nutrients, is also a major concern in animal agriculture, the theory of fetal programming can be extended to include fetal origins of postnatal growth retardation, reduced feed efficiency, and reduced meat quality. This concept of fetal programming has far-reaching implications for the animal sciences.

## MECHANISMS OF IUGR

The lack of knowledge about the mechanisms for IUGR has prevented the development of effective means to enhance fetal growth in animals. Due to the lack of intervention options, the current management of IUGR livestock fetuses is only empirical and primarily aimed at improving neonatal care and adopting artificial rearing with liquid replacer milk. Artificial rearing is costly at present because it requires expensive facilities and diets (Wu et al., 2004d). Only by elucidating the mechanisms of IUGR, can we design effective means to prevent fetal growth restriction. Available evidence suggests that impaired placental growth (including vascular growth) or function, possibly owing to reduced placental synthesis of vasodilators and metabolic regulators, may contribute primarily to IUGR in response to undernutrition and overnutrition.

## Impaired Placental Growth and IUGR

### Crucial Role for Placental Growth and Uteroplacental Blood Flows in Fetal Growth.

The placenta is the organ that transports nutrients, respiratory gases, and the products of their metabolism between the maternal and fetal circulation. Placental growth (including vascular growth) is crucial for fetal growth and development (Gootwine, 2004; Reynolds et al., 2005). Thus, an increase in placental growth through elevated expression of placental anabolic proteins (e.g., prolactin and placental lactogen) is associated with enhanced fetal growth in sheep (Gootwine, 2004). During normal pregnancy, uterine and placental blood flows increase throughout gestation to meet the metabolic needs of the growing conceptus (Reynolds et al., 2005). Umbilical blood flow also increases markedly during late gestation in livestock (including sows, ewes, and cows) to satisfy the metabolic needs of the rapidly growing fetus (Ford, 1995; Père and Etienne, 2000). Thus, uteroplacental blood flow is a major factor that influences the availability of nutrients for fetal growth and development. Available evidence from well-controlled studies shows that impaired placental growth is associated with IUGR (Mellor, 1983; Schoknecht et al., 1994; Wallace et al., 1996, 2003a).

Rates of uteroplacental blood flows depend in large part on placental vascular growth, which results from angiogenesis (the growth of new vessels from existing ones) and placental vascularization (Vonnahme and Ford, 2004; Reynolds et al., 2005). Consistent with increased uterine and placental blood flows (Ford, 1995), placental angiogenesis increases markedly from the first to the second third of gestation and continues to increase further during late gestation (Reynolds and Redmer, 2001). Both nutrient restriction of adult ewes (Redmer et al., 2004) and overnourishment of adolescent ewes (Redmer et al., 2005) during pregnancy reduced placental proliferation in the fetal trophectoderm and placental expression of angiogenic factors. In overfed adolescent ewes, these changes at midgestation may underlie the attenuated uteroplacental blood flows and IUGR that characterize late pregnancy (approximately d 130) in these rapidly growing animals (Wallace et al., 2002).

Indeed, we recently found that uterine blood flow was reduced by 56% as early as on d 90 of gestation, which occurred before any reduction in fetal or placental weight was observed (J. M. Wallace, unpublished data). In other ovine models of IUGR induced by heat stress or multiple fetuses, decreases in placental angiogenesis and vascularity are also associated with reduced uteroplacental blood flows as well as reduced placental and fetal growth (Reynolds et al., 2005). Thus, placental efficiency is not reflected just by placental weight or size but also depends on other factors, such as placental microvascular density, interdigitation of the placenta with the maternal endometrium to increase surface, and placental blood flow.

Nutrient uptake by the uterus or the fetus can be determined experimentally on the basis of the Fick principle: Uptake = Blood Flow Rate × (A-V), where (AV) represents the difference in arteriovenous concentration across the uterus or the fetus (Bell and Ehrhardt, 2002). Thus, the transuterine or transplacental exchange of a substance is determined by both blood flow rate and its concentrations in the arterial and venous blood (Reynolds et al., 2006). Blood concentrations of metabolites in the uterine artery and vein as well as the umbilical vein and artery are regulated by 1) the activities and amounts of nutrient transporters on the plasma membranes of cells of the uteroplacental unit, 2) the amounts of the substances entering the circulation from dietary and endogenous sources, and 3) rates of oxidation of the substances. There is evidence that reductions in placental growth, angiogenesis, and presumably placental vascularization are associated with decreased placental transport of O2 and nutrients from mother to fetus in compromised ovine pregnancies (Wallace et al., 2002, 2005c).

### Insufficient Uteroplacental Blood Flows and Reduced Transport Activity in Natural Uterine Insufficiency.

Fetal growth restriction in ruminants carrying multiple fetuses is associated with reduced uteroplacental blood flows and placental function (Ferrell and Reynolds, 1992). In sows, at d 77 to 110 of gestation, there are significant correlations between placental weight and placental blood flow, between placental weight and fetal weight, and between placental blood flow and fetal weight (Wootton et al., 1977). Between d 44 and 111 of gestation, total blood flow to the porcine uterus does not increase linearly with an increase in the number of fetuses, and uterine blood flow per fetus decreases with increasing litter size (Père and Etienne, 2000). In comparison with its littermate, the runt fetal pig is associated with a small placenta and a low rate of placental blood flow (Wootton et al., 1977). In addition to the compromised placental blood flow, placental transport of leucine was reduced in the small porcine fetus compared with the average-size fetus at d 45, 60, and 100 of gestation because of the impaired development of transport systems and their reduced capacity (Finch et al., 2004).

## NO and Polyamines and IUGR

### Crucial Roles of NO and Polyamines in Placental and Fetal Growth.

Arginine is a common substrate for NO and polyamine syntheses via NO synthase (NOS) and ornithine decarboxylase, respectively (Wu and Morris, 1998). Nitric oxide is a major endothelium-derived vasorelaxing factor, and plays an important role in regulating placental-fetal blood flows and, thus, the transfer of nutrients and O2 from mother to fetus (Bird et al., 2003). Likewise, polyamines regulate DNA and protein synthesis and, therefore, cell proliferation and differentiation (Flynn et al., 2002). Growing evidence shows that NO and polyamines are key regulators of angiogenesis and embryogenesis as well as placental and fetal growth (Reynolds and Redmer, 2001; Zheng et al., 2006).

Excitingly, we recently discovered that arginine is particularly abundant in porcine allantoic fluid (4.1 to 6 mM) at d 40 of gestation (term = 114 d; Wu et al., 1996, 1998a). Remarkably, concentrations of arginine and its precursor ornithine in porcine allantoic fluid increased by 23- and 18-fold, respectively, between d 30 and 40 of gestation (Figure 2), with their N accounting for approximately 50% of the total free α-amino acid N in allantoic fluid (Wu et al., 1996). The absence of arginase activity from the porcine placenta ensures maximum transfer of arginine from mother to fetus (Wu et al., 2005). Most recently, we found that citrulline (an immediate precursor of arginine) is unusually rich (10 mM) in ovine allantoic fluid at d 60 of gestation (term = 147 d; Kwon et al., 2003). Concentrations of citrulline and its precursor glutamine in ovine allantoic fluid increase by 34- and 18-fold, respectively, between d 30 and 60 of gestation (Figure 3), with their N representing 60% of total α-amino acid N in ovine allantoic fluid (Kwon et al., 2003a). Citrulline derived from the uterus and/or placenta is effectively converted into arginine via argininosuccinate synthase and lyase in fetal tissues (Wu and Morris, 1998). Because the ovine placenta contains a high arginase activity (Kwon et al., 2004b) that would catabolize arginine, the placental transfer of citrulline and its storage in allantoic fluid provide an effective strategy to conserve arginine in the ovine conceptus. The unusual abundance of the arginine-family AA in fetal fluids is associated with the greatest rates of NO and polyamine syntheses in the ovine and porcine placentae during the first half of pregnancy, when its growth is most rapid (Kwon et al., 2003b, 2004b; Self et al., 2004; Wu et al., 2005). These novel findings from the 2 diverse animal models are consistent with the proposed crucial roles of the arginine-dependent metabolic pathways in conceptus growth and development (Figure 4).

Evidence from studies with pregnant pigs and sheep indicates impaired syntheses of NO and polyamines in the conceptuses of underfed and overfed dams (Wu et al., 2004a). Dietary protein deficiency or hypercholesterolemia reduces the availability of arginine and ornithine in maternal plasma, fetal plasma, amniotic fluid, and allantoic fluid, as well as the placental synthesis of NO and polyamines in pregnant pigs (Wu et al., 1998a; 1998b). In addition, we found that maternal undernutrition in sheep (50% of NRC requirements; NRC, 1985) between d 28 and 78 of gestation decreased concentrations of arginine, citrulline, and polyamines in maternal plasma, fetal plasma, amniotic fluid, and allantoic fluid at d 78 of gestation (Kwon et al., 2004a). Notably, concentrations of biopterin (an indicator of de novo synthesis of tetrahydrobiopterin, BH4, an essential cofactor for NOS; Shi et al., 2004) in amniotic and allantoic fluids are reduced by 32 to 36% in underfed ewes, compared with control ewes (G. Wu, unpublished data). These changes would impair placental and fetal NO synthesis, thereby theoretically resulting in reduced placental-fetal blood flows in underfed ewes (Redmer et al., 2004).

Consistent with these findings, maternal undernutrition impairs NO-dependent vasodilation and increases arterial blood pressure in the ovine fetus (Ozaki et al., 2000). Similarly, uterine and umbilical blood flows are reduced in overnourished adolescent sheep (Wallace et al., 2002, 2003a), suggesting a reduction in NO generation by vascular endothelial cells of the uterus and placentae. The underlying mechanisms may include 1) reduced availability of BH4 due to oxidative stress, 2) reduced expression of NOS, 3) inactivation of NOS due to its close association with caveolin-1, and 4) increased plasma concentrations of homocysteine, saturated lipids, and asymmetric dimethylarginine, which are all inhibitors of endothelial NO generation (Wu and Meininger, 2002; Fu et al., 2005). Collectively, these results have generated the novel hypothesis that impaired placental syntheses of NO and polyamines may provide a unified explanation for the same pregnancy outcome (namely, IUGR) in response to the 2 extremes of nutritional problems (both maternal undernutrition and overnutrition) during gestation (Wu et al., 2004a).

### Possible Role of NO and Polyamines in the Growth and Development of Fetal Muscle Cells and Adipocytes.

Whereas the endocrine system plays an important role in fetal growth (Fowden et al., 2005), the changes in fetal muscle growth rate in response to maternal undernutrition and overnutrition is not always correlated with changes in growth regulatory hormones (Davis et al., 1997; Rehfeldt et al., 2004). Thus, we have been prompted to investigate the role of specific nutrients in the growth and development of fetal muscle cells. Because AA are the only source of both C and N for energy metabolism and protein synthesis, we focused on these macronutrients using our porcine and ovine models of IUGR (Wu et al., 1998a).

Myocytes and adipocytes are derived from a common mesenchymal precursor (Sordella et al., 2003). Thus, excessive amounts of adipose tissue are developed at the expense of skeletal muscle when embryonic myogenesis is impaired (Kablar et al., 2003). There are 2 developing types of muscle fibers in fetal pigs: primary fibers (formed by the rapid fusion of primary myoblasts between d 25 and 50 of gestation) and secondary fibers (formed on the surface of primary fibers between approximately d 50 and 90 of gestation). In late gestation, the number of secondary fibers is much greater than that of primary fibers. During prenatal development, muscle fibers undergo contractile differentiation, resulting in the formation of slow-twitch oxidative, fast-twitch oxidative-glycolytic, and fast-twitch glycolytic fibers. At birth, most muscle fibers are oxidative (Bee, 2004).

The numbers of secondary muscle fibers, but not primary muscle fibers, are affected by conditions in utero (Dwyer et al., 1994). The total number of muscle fibers is fixed at birth and is a major factor affecting the growth of skeletal muscle and thus the postnatal growth of the animal. Postnatally, skeletal muscle development continues such that there are more glycolytic fibers with increasing age (Bee, 2004). Comparison between small and large littermates in pigs indicates that the differences in their prenatal and postnatal growth rates are related to a lower ratio of secondary to primary muscle fibers as well as a smaller size of the fibers in the former (Handel and Stickland, 1987).

Polyamines are necessary for the proliferation and differentiation of cells (Flynn et al., 2002) and likely mediate the growth and development of fetal muscle fibers and adipocytes. In support of this suggestion, we recently noted that concentrations of arginine, ornithine, and polyamines were reduced in skeletal muscle of IUGR fetal pigs compared with average-weight littermates (G. Wu, unpublished data). Similarly, concentrations of arginine, putrescine, and spermidine were lower in the gastrocnemius muscle of IUGR fetal lambs in response to maternal undernutrition (G. Wu, unpublished data). Further, elevated levels of NO inhibit the growth of adipocytes (Fu et al., 2005; Jobgen et al., 2006). Because adipose tissue of fetal lambs in underfed ewes have reduced levels of endothelial NOS (G. Wu, unpublished data), reduced NO availability is expected to facilitate the growth of preadipocytes in IUGR lambs.

### Possible Role of AA in mTOR Signaling and IUGR.

Fetal muscle growth may be regulated by the signaling mechanism of the mammalian target of rapamycin (mTOR; Du et al., 2005). This serine/threonine protein kinase is an evolutionarily conserved member of the phosphoinositol kinase-related kinase family of proteins. The phosphorylation of mTOR in response to nutrients (e.g., amino acids and glucose) results in the phosphorylation of p70 S6 kinase and eukaryotic initiation factor 4E-binding protein-1, which promotes the formation of the active initiation complex for polypeptide synthesis (Meijer and Dubbelhuis, 2004).

In mature cows, maternal undernutrition (50% of feed intake for the control group) during early gestation (d 30 to 125) had no effect on the content of calpains I and II (calcium-dependent cysteine proteases) in maternal and fetal skeletal muscles (Du et al., 2004). However, underfeeding increased and reduced the concentration of calpastatin (a specific inhibitor of calpains) in maternal and fetal muscles, respectively (Du et al., 2004). Malnutrition also reduced the concentrations of phosphorylated mTOR in both maternal and fetal skeletal muscle and increased concentrations of ubiquitinated proteins in maternal muscle (Du et al., 2005).

Importantly, the activation of the mTOR signaling pathway in skeletal muscle is under the control of the arginine-family amino acids (e.g., arginine and glutamine) and leucine (Meijer and Dubbelhuis, 2004), whose concentrations are reduced in the IUGR fetus (Wu et al., 1998a; Kwon et al., 2004b). Glutamine and leucine also inhibit protein degradation in skeletal muscle (Wu and Thompson, 1990; Meijer and Dubbelhuis, 2004). Thus, during pregnancy, reduced concentrations of AA in the conceptus in response to maternal undernutrition may contribute to an increase in protein degradation and a decrease in protein synthesis in maternal skeletal muscle. Such a mechanism may also reduce both protein degradation and protein synthesis in fetal muscle. These differential effects of maternal underfeeding may help to mobilize the maternal protein reserve to supply AA for metabolic utilization by the mother and the fetus, thereby providing a protective mechanism for ensuring fetal growth.

### MECHANISMS OF FETAL PROGRAMMING

As noted above, nutritional insults during a critical period of gestation may have a permanent effect on the progeny throughout postnatal life. There is also evidence that fetal undernutrition due to placental insufficiency impaired vascular function in two generations of rats (Anderson et al., 2006), indicating an inter-generational effect. Interestingly, some of the effects, such as embryo survival (Vinsky et al., 2006) and endothelium-dependent relaxation (Anderson et al., 2006), appear to be sex-specific. These effects likely result from genomic imprinting, which is defined as the parent-of-origin-dependent expression of a single allele of a gene in the embryo/fetus, namely a parental influence on the gene expression of the progeny. There is growing evidence that maternal or fetal nutritional status can alter the epigenetic state of the fetal genome and gene expression of imprinted genes (e.g, Igf2 and H19), where the methylation of DNA and proteins plays a crucial role. Interestingly, Igf2 is paternally expressed and maternally silent, whereas H19 is paternally silent and preferentially expressed from the maternal allele (Doherty et al., 2000).

Epigenetic alterations (i.e., stable alterations of gene expression through covalent modifications of DNA and core histones) in early embryos may be carried forward to subsequent developmental stages (Waterland and Jirtle, 2004). Two mechanisms mediating epigenetic effects are DNA methylation (occurring in the 5′-positions of cytosine residues within CpG dinucleotides throughout the mammalian genome) and histone modification (acetylation, methylation, etc.; Jaenisch and Bird, 2003; Oommen et al., 2005). The CpG methylation can regulate gene expression by modulating the binding of methyl-sensitive DNA-binding proteins, thereby affecting regional chromatin conformation. Histone modifications can alter the positioning of histone-DNA interactions and the affinity of histone binding to DNA, thereby affecting gene expression (Jaenisch and Bird, 2003).

The DNA and protein methylation are catalyzed by specific DNA and protein methyltransferases, with S-adenosylmethionine (SAM) being the methyl donor in these reactions (Jaenisch and Bird, 2003). S-adenosylmethionine is synthesized from methionine and ATP by methionine adenosyltransferase, and its placental concentration is greatest when placental growth is most rapid (Wu et al., 2005). Besides methylation reactions, SAM is utilized for the synthesis of spermidine and spermine from putrescine through the generation of decarboxylated SAM (Figure 5).

The synthesis of creatine is quantitatively the most important pathway for SAM utilization and thus is a major regulator of methyl donor availability in the body (Stead et al., 2001). When the diet is deficient in cysteine/taurine or contains excess methionine, an increase in cysteine/taurine synthesis from methionine consumes a large amount of SAM. One-carbon-unit metabolism, which depends on serine, glycine, histidine, choline, and B vitamins (including folate, vitamin B12, and vitamin B6), in addition to methionine, plays an important role in regulating the availability of SAM (Figure 5). Thus, DNA methylation and histone modifications may be altered by the overall availability of AA and micronutrients (Oommen et al., 2005). Epigenetics may provide a molecular mechanism for the impact of maternal nutrition on the fetal programming of postnatal growth performance and disease susceptibility.

## POTENTIAL SOLUTIONS TO PREVENT IUGR

Because of the incomplete knowledge about the mechanisms of IUGR, attempts to alleviate the detrimental effects of IUGR on postnatal growth performance in livestock have so far achieved only limited success. The recognition of fetal programming suggests that strategies to promote postnatal growth and health of livestock should be initiated at the key stages of prenatal development (Finch et al., 2004). Thus, targeting an effective window of opportunity during a specific period of pregnancy would be most beneficial for preventing IUGR. Despite much failure, the largely trial-and-error approaches to treating pregnant dams have generated some promising results. These approaches include hormonal therapy; dietary supplementation of energy, protein concentrates, or both; provision of antioxidant nutrients; and manipulations of the arginine-NO pathway. Although these methods are diverse in nature, they appear to directly or indirectly promote placental growth and uteroplacental blood flow in pregnant dams via increasing the availabilities of arginine, NO, or both.

### Hormonal Therapy

Changes in endocrine systems during prenatal or postnatal life in response to altered maternal nutrition may contribute to the programming of metabolism and physiology in later life (Fowden et al., 2005). Specifically, maternal concentrations of progesterone are usually low in severely underfed and overfed pregnant dams, likely because of a decrease and an increase in its synthesis and catabolism, respectively (Dziuk, 1992). Thus, administration of progesterone is required to maintain pregnancy and fetal growth in pigs during prolonged inanition (Anderson, 1975). Although progesterone treatment may not affect litter size or fetal growth in gilts fed a restricted diet (2 kg/d) during pregnancy (Yu et al., 1997), some evidence indicates that, beginning 24 h after the onset of estrus, administration of progesterone to gilts fed a diet providing 2-times maintenance requirements enhanced embryonic survival (Jindal et al., 1997). Progesterone also may be required for maintaining adequate placental growth (including vascular growth) via NO- and polyamine-dependent mechanisms (Chwalisz and Garfield, 1997; Kwon et al., 2003b). In this regard, it is noteworthy that treatment with progesterone partially ameliorated fetal growth restriction in overfed adolescent ewes (Wallace et al., 2003b).

Alternatively, maternal growth hormone treatment from d 35 to 80 of gestation alters maternal nutrient partitioning in favor of uteroplacental growth in the overnourished adolescent sheep (Wallace et al., 2004). When overnourished adolescent dams were treated with growth hormone in late pregnancy once placental growth was complete, a modest increase in fetal weight was observed but was associated with a major increase in fetal adiposity, which may have negative implications for long-term health (Wallace et al., 2005a). Similarly, daily administration of growth hormone to pregnant sows during late or a large part of gestation increased fetal weight (Rehfeldt et al., 2004). Further, intramuscular administration of growth hormone to ewes at breeding led to a more efficient placenta, larger birth weight lambs, and more rapid postnatal growth (Costine et al., 2005). In addition to its effects on maternal carbohydrate metabolism, growth hormone may increase the availability of arginine in the conceptus by stimulating maternal and fetal intestinal synthesis of citrulline (the precursor of arginine) and inhibiting hepatic degradation of amino acids in pregnant dams, as previously reported for growing pigs (Bush et al., 2002). Significantly, a single injection of human chorionic gonadotropin on d 12 postmating increased placental and fetal growth in sheep carrying two fetuses (Cam and Kuran, 2004), further supporting the view that placental growth positively influences fetal growth.

### Dietary Supplementation with Energy, Protein Concentrates, or Both

Maintenance represents 75 to 85% of the total requirement in pregnant dams because of substantial increases in maternal body and tissue weights as well as metabolic rates (Cole, 1990; Grummer, 1995). Nutrients provided beyond the maintenance requirement are used for maternal tissue accretion and fetal growth. As noted above, forages are often deficient in both protein and energy and, therefore, they are inadequate for providing both macro- and micronutrients to support the maximal growth performance of unsupplemented ruminants and horses during pregnancy (Pugh, 1993; Bell et al., 2000). In addition, dams are usually in a severe catabolic state during late gestation because of either inadequate voluntary feed intake for ruminants or restricted feed provision for sows. Thus, dietary supplementation with energy, protein concentrates, or both may provide a means to enhance fetal growth.

### CONCLUSION AND PERSPECTIVES

Fetal growth is controlled by complex interactions among genetic, epigenetic, and environmental factors, as well as maternal maturity. These factors regulate placental growth (including placental angiogenesis and vascular growth) and, therefore, uteroplacental blood flows and the transfer of nutrients from mother to fetus. The IUGR results from disturbances of these maternal and fetal homeostatic mechanisms and occurs under various practical conditions of animal production. The IUGR reduces neonatal survival, has a permanent stunting effect on postnatal growth performance and the efficiency of feed/forage utilization in offspring, negatively affects whole body composition and meat quality, and impairs lifetime fertility, health, and athletic performance. Available evidence suggests that the placental or fetal growth trajectory is vulnerable to maternal undernutrition or overnutrition throughout gestation but that the most profound effects arguably occur when nutritional insults are applied during the period of rapid placental development. Additionally, arginine-derived signaling and regulatory molecules (NO and polyamines) are crucial for placental and fetal growth. New knowledge on the mechanisms regulating fetal growth and development will be beneficial for designing new, rational, and effective strategies to prevent and treat IUGR in livestock. Further, understanding the multiple roles of nutrients in DNA methylation (which can influence genome stability, viability, expression, and imprinting) will have a broad impact on reproductive health and disease prevention. We expect that studies utilizing domestic animal models of IUGR will provide the necessary scientific basis for the development of management practices that will improve pregnancy outcome in domestic animals. In view of the crucial roles of the arginine-dependent metabolic pathways, intravenous or oral administration of arginine may provide a potentially novel solution to enhance uteroplacental blood flows (and therefore transfer of nutrients from mother to fetus), thereby ameliorating or preventing IUGR. Promoting an optimal intrauterine environment will not only ensure optimal fetal development but also will enhance growth performance postnatally and reduce the risk of chronic diseases in adults. Because IUGR remains a major problem in mammalian pregnancies, innovative interdisciplinary research in the areas of nutrition, reproductive physiology, and vascular biology are critical to design the next generation of nutrient-balanced gestational diets and develop new tools for livestock management, which will enhance the efficiency of animal production and improve the well being of animals.