Developmental programming of the metabolic syndrome

Evidence from numerous epidemiological studies has suggested that there is a relationship between intrauterine growth restriction (IUGR) and the subsequent development of the metabolic syndrome (Hansen, Reference Hansen1999; Horvath and Bruning, Reference Horvath and Bruning2006; Palinski, Reference Palinski2007; Taylor and Poston, Reference Taylor and Poston2007). Sheep, rats, guinea pigs and more recently pigs have extensively been used to investigate the mechanisms which link prenatal programming to adult diseases (Hoet and Hanson, Reference Hoet and Hanson1999; Clarke et al., Reference Clarke, Kirth, Heasman, Juniper, Budge, Stephenson and Symonds2000; Poore and Fowden, Reference Poore and Fowden2004a).

Placental and Foetal growth and development

It is well established that in mammals the principal determinant of IUGR is placental nutrient supply, which in turn depends on the size, morphology, blood supply and transporter abundance of the placenta, and on synthesis and metabolism of nutrients and hormones by the uteroplacental tissues (Fowden et al., Reference Fowden, Ward, Wooding, Forhead and Constancia2006). Both human and rat placentas belong to the general category of haemochorial placentas; however, there are fundamental differences (Table 3). For example, adiponectin and adiponectin receptor 2 (Adipo-R2) are present in cytotrophoblast and syncytiotrophoblast cells in humans, whereas adiponectin expression occurs in rat trophoblast cells and Adipo-R2 in giant cells and viteline membranes (Caminos et al., Reference Caminos, Nogueiras, Gallego, Bravo, Tovar, García-Caballero, Casanueva and Diéguez2005). Pigs, on the other hand, have epithelial chorial (or diffuse) placentas making them a good model for studies during pregnancy as the maternal and foetal components of the placenta can be cleanly dissected without cross contamination (Tayade et al., Reference Tayade, Black, Fang and Croy2005).

Table 3

Placental and metabolic comparisons between foetal and neonatal humans, swine and rats

NEFA = non-esterified fatty acids.

*Values in bracket are for premature human infants.

Foetal growth is a combination of numerous interrelated mechanisms that include cell replication and tissue differentiation, matrix formation and cell death, all of which need to be regulated (Gluckman, Reference Gluckman1986). There is a well-defined, programmed, sequence of developmental changes that occurs at specific times in the growth of the embryo and foetus (Figure 1). The underlying principles of foetal maturation appear to be comparable between mammals, but it should be noted that quantitative differences occur between species. For example, stage 3 foetal development occurs in utero in humans and swine but postnatally in rats (Figure 1). In addition, rat pups are much smaller than either babies or piglets (Figure 1), which means that it is often necessary to pool tissue samples, thereby reducing indications of natural variation. When compared with the young of other commonly used research animals such as the dog, cat, goat and sheep, the piglet appears to show the greatest similarity to the human infant in terms of anatomy, physiology, nutrition and metabolism (Book and Bustad, Reference Book and Bustad1974; Miller and Ullrey, Reference Miller and Ullrey1987; Mei and Xu, Reference Mei and Xu2003).

Figure 1

Comparison of prenatal development and crown-to-rump length of human (—), pig (– -) and rat (- - - -) foetuses throughout gestation. (a) Adapted from Book and Bustad (Reference Book and Bustad1974) and Hill (Reference Hill2008). (b) Adapted from Book and Bustad (Reference Book and Bustad1974), and Evans and Sack (Reference Evans and Sack1973).

Measurements of embryos and foetuses are generally cited as crown-to-rump length (Figure 1), but while this parameter is quite reliable for man, it must be used with caution for domestic mammalian foetuses because of the longer neck and the increased chance in error due to changes in posture (Evans and Sack, Reference Evans and Sack1973). In humans, the rate of foetal growth is relatively slow up to the 20th week of pregnancy and, as with the pig and rat; the majority of growth occurs during the second half of pregnancy (Manners and McCrea, Reference Manners and McCrea1963; Ashworth, Reference Ashworth2006). Human foetal growth accelerates to reach a maximum around 30 to 36 weeks, declining thereafter until birth as the mother fails to meet the increasing energy demand of her growing foetus. On the contrary, the pig and rat have not reached their peak growth rate at birth and so the decline in growth rate observed during late gestation in humans does not occur (Figure 1). In litter-bearing species the number of developing foetuses is inversely related to the ultimate size of each individual at full term (Pond and Houpt, Reference Pond and Houpt1978). The pattern of foetal growth in pigs is determined by the genome of the foetus, maternal nutrition and health, uterine capacity and foetal position in the uterine horn (McCance and Widdowson, Reference McCance and Widdowson1974; Antipatis et al., Reference Antipatis, Finch and Ashworth2008).

Protein accumulation tends to occur early in human foetal development to reach its maximum (∼300 g) around week 35, and precedes fat accumulation, most of which is subcutaneous, and only exceeds the weight of protein deposition by week 38 (Prentice et al., Reference Prentice, Spaaij, Goldberg, Poppitt, van Raaij, Totton, Swann and Black1996). By term, there is approximately three-times more energy stored as fat than as protein in human infants. Unlike humans and sheep, both pigs and rats are born with little adipose tissue, but in this sense their early post partum development can be viewed as equivalent to the late development of human babies in utero (Table 3). It should be noted that the amino acid compositions differ appreciably among species that have relatively short periods of gestation such as the rat and species that have relatively long periods of gestation such as the pig and human for the amino acids: histidine, glycine, lysine, proline and hydroxyproline (Wu et al., Reference Wu, Ott, Knabe and Bazer1999).

The newborn pig’s body fat is minimal and usually constitutes no more than 2% of total body weight (BW). However, this situation is reversed postnatally; piglets can double their birth weight due to high rates of protein deposition and lean tissue growth (Wood and Groves, Reference Wood and Groves1965; Whittemore and Kyriazakis, Reference Whittemore and Kyriazakis2006), and there can be a 10- to 20-fold increase in body fat in 7 to 10 days, whereas the human infant doubles its birth weight by about 5 months of age (Shulman, Reference Shulman1993; Pond and Mersmann, Reference Pond and Mersmann2001). This means that piglets can provide an excellent model of accelerated growth and development (Odle, Reference Odle1997), which is advantageous when studying the effects of postnatal nutrition on health in later life (Taylor and Poston, Reference Taylor and Poston2007). Moreover, the newborn piglet is more akin to the premature than to the full-term human infant in several aspects, such as body composition (i.e. low fat and hepatic glycogen content) and a reduced thermoregulatory ability (Widdowson, Reference Widdowson1971; Book and Bustad, Reference Book and Bustad1974; Herpin et al., Reference Herpin, Damon and Le Dividich2002; Mostyn et al., Reference Mostyn, Litten, Perkins, Euden, Corson, Symonds and Clarke2005).

Low birth weight (LBW) modern domestic pigs exhibit impaired mitochondrial metabolism in adipose tissue, which may in part explain their reduced body temperature (Mostyn et al., Reference Mostyn, Litten, Perkins, Euden, Corson, Symonds and Clarke2005). However, physiological studies have shown that porcine genotype must be taken into consideration when selecting an appropriate model because despite the minipig being smaller, it possesses more fat (0.117 to 0.158 kg fat for a 1.25 kg 5-day old neonatal piglet, (Sheng et al., Reference Sheng, Adolph, O’Brian Smith and Garza1988)) compared with modern domestic breeds (0.03 kg fat for a 1.5 kg 1 day old neonatal piglet, Shields et al., Reference Shields, Mahan and Graham1983)) and exhibits altered uncoupling protein (UCP) expression and endocrine profile at birth compared with modern domestic breeds of pigs (Knol et al., Reference Knol, Leenhouwers and van der Lende2002; Mostyn et al., Reference Mostyn, Litten, Perkins, Alves-Guerra, Pecqueur, Miroux, Symonds and Clarke2004, Reference Mostyn, Sebert, Litten, Perkins, Laws, Symonds and Clarke2006). Compromised UCP abundance in adipose tissue may contribute to excess fat deposition in later life resulting in obesity, hence providing a natural model for obesity (Dyson et al., Reference Dyson, Alloosh, Vuchetich, Mokelke and Sturek2006).

Size and shape at birth

More recently, human studies (Jarvelin et al., Reference Jarvelin, Elliott, Kleinschmidt, Martuzzi, Grundy, Hartikainen and Rantakallio1997; Laitinen et al., Reference Laitinen, Taponen, Martikainen, Pouta, Millwood, Hartikainen, Ruokonen, Sovio, McCarthy, Franks and Jarvelin2003 and Reference Laitinen, Pietilainen, Wadsorth, Sovio and Jarvelin2004) have demonstrated that body shape at birth, rather than weight per se, may be a better indicator of future health and development of the human infant. The modern domestic pig, rather than the ancient domestic pig or minipig, is an excellent model for developmental programming due to the 2- to 3-fold variation in BW among littermates, which provides a naturally occurring form of foetal growth restriction with less genetic variation than seen in man or other monotocous species (Bauer et al., Reference Bauer, Walter, Hoppe, Gaser, Lampe, Kauf and Zwiener1998; Corson et al., Reference Corson, Laws, Laws, Litten, Lean and Clarke2008a). In addition, it is usually possible to identify asymmetrical and symmetrical body shapes within a domestic pig litter. Piglets that are long and thin at birth grow more slowly compared with those that are short for weight (i.e. fatter) over the first month of life and have a lower developmental score (Litten et al., Reference Litten, Drury, Corson, Lean and Clarke2003). Moreover, LBW piglets that exhibited ‘catch-up’ growth had a higher neonatal developmental score than those animals that remained on their prenatal growth trajectory.

There is a growing body of evidence to suggest that postnatal ‘catch-up’ or ‘catch-down’ growth may also be a contributory factor to predisposition to adulthood diseases (Ong et al., Reference Ong, Ahmed, Emmett, Preece and Dunger2000 and Reference Ong, Kratzsch, Kiess and Dunger2002 and Ong and Dunger, Reference Ong and Dunger2004). Practically it is difficult to follow the growth of individual rat pups from birth and the importance of size at birth remains controversial. In addition, the programming of higher blood pressure appears to be greatly amplified in rats compared with human epidemiological and large animal studies (Langley-Evans, Reference Langley-Evans2001). The pig provides an excellent model for longitudinal studies as individuals are easily identifiable, large volumes of blood can be taken and exogenous substances can be infused in both chronic and acute studies (Poore et al., Reference Poore, Forhead, Gardner, Giussani and Fowden2002; Poore and Fowden, Reference Poore and Fowden2003, Reference Poore and Fowden2004a and Reference Poore and Fowden2004b; Litten et al., Reference Litten, Mostyn, Perkins, Corson, Symonds and Clarke2005 and Reference Litten, Mostyn, Laws, Corson, Symonds and Clarke2008; Corson et al., Reference Corson, Laws, Laws, Litten, Lean and Clarke2008a and Reference Corson, Laws, Litten, Lean and Clarke2009). Furthermore, the pig can be trained to remain conscious but relatively unstressed while procedures are carried out (Larsen and Rolin, Reference Larsen and Rolin2004). Postnatal dietary studies can continue as individual feed intake can be recorded using feed intake recording equipment feeders without removing the pig from its social group (Litten et al., Reference Litten, Corson, Hall and Clarke2004). The similarity of pathological response to high caloric intakes with humans supports the use of pig models for identifying genes and their variants associated with energy storage defects through the activation of both hormonal and biochemical pathways (Brambilla and Cantafora, Reference Brambilla and Cantafora2004).

Centile charts for BW and height are routinely used to track the normal growth and development of children alongside their birth percentile, allowing their health and progress to be monitored. Many children deviate from their early percentile, but it is still not clear as to when the point at which they deviate becomes of pathological importance (Wright et al., Reference Wright, Matthews, Waterston and Aynsley-Green1994). Percentile curves for pigs, similar to those used for human studies, have recently been developed, and allow an efficient method of following an individual pig’s growth, particularly during experimental manipulations (Corson et al., Reference Corson, Laws, Laws, Litten, Lean and Clarke2008a). Such curves could be used to determine if a pig’s growth deviates from the expected normal growth curve, and identify when, or if, the point of deviation is of pathological importance. However, an air of caution must be taken if using these porcine growth curves as data were collected only for a specific domestic breed combination (i.e. Large White X Landrace), reared under standard commercial conditions, not experimental conditions; it is known that growth and development varies between porcine genotypes (Mostyn et al., Reference Mostyn, Litten, Perkins, Alves-Guerra, Pecqueur, Miroux, Symonds and Clarke2004) and also under different management regimes (Kusec et al., Reference Kusec, Baulain, Kallweit and Glodek2007).

Studies in older pigs have demonstrated that LBW is associated with glucose intolerance (Poore and Fowden, Reference Poore and Fowden2004a; Corson et al., Reference Corson, Laws, Litten, Lean and Clarke2009), reduced insulin sensitivity and increased mean arterial blood pressure (Poore et al., Reference Poore, Forhead, Gardner, Giussani and Fowden2002) and altered Hypothalamo–pituitary–adrenal axis function (Poore and Fowden, Reference Poore and Fowden2003) in conjunction with changes in body composition and endocrine profile (e.g. plasma leptin) (Poore and Fowden, Reference Poore and Fowden2004b; Morise et al., Reference Morise, Louveau and Le Huërou-Luron2008). LBW pigs exhibit glucose intolerance (Figure 2), irrespective of whether they are reared under standard commercial (Corson et al., Reference Corson, Laws, Litten, Lean and Clarke2009) or experimental conditions (Poore and Fowden, Reference Poore and Fowden2004b), and the methodology used for a glucose tolerance test (GTT). This suggests that their postnatal environment had little effect on their risk of developing adulthood diseases, and that the LBW pig provides a robust model for studying the metabolic syndrome.

Figure 2

The effect of low (L) or normal (N) birth weight on the area under the curve following a glucose tolerance test (GTT) at 3, 6 and 12 months of age and during pregnancy. Values are presented as Mean ± s.e.m., L pigs are represented by black bars, N pigs by white bars. GTT: intravenous administration of 0.5 g/kg BW dextrose 40%. Data taken from Poore and Fowden, Reference Poore and Fowden2004a (3 and 12 months) and Corson et al., Reference Corson, Laws, Litten, Dodds, Lean and Clarke2008b (6 months and pregnancy).

Studies have shown that the birth weight of a human baby is related to both their mother’s and maternal grandmother’s birth weight (Hackman et al., Reference Hackman, Emanuel, Belle and Daling1983; Klebanoff et al., Reference Klebanoff, Graubard, Kessel and Berendes1984). Women who were themselves small for gestational age (SGA) are twice at risk of having a SGA baby (Klebanoff et al., Reference Klebanoff, Meirik and Berendes1989; Skjærven et al., Reference Skjærven, Wilcox, Øyen and Magnus1997) as well as being at a greater risk of developing adulthood diseases. Moreover, women who had parents with NIDDM had children with higher birth weights than women with non-diabetic parents. Grandchildren of grandparents with NIDDM are usually in the top percentiles in growth charts (i.e. 90th centile and above) compared with non-diabetic grandparents (McCarron et al., Reference McCarron, Smith and Hattersley2004). In the past intergeneration studies have been somewhat limited to rodent models (Harrison and Langley-Evans, Reference Harrison and Langley-Evans2009), but more recently it has been demonstrated that an intergeneration relationship between maternal size at birth and altered glucose tolerance in later life exists in the domestic pig (Corson et al., Reference Corson, Laws, Litten, Lean and Clarke2009). The added advantage of using a porcine model for intergeneration investigations is that it allows both LBW and high birth weight piglets within a litter to be studied, further reducing the number of animals required as well as decreasing genetic variation.

Cardiovascular disease and the metabolic syndrome

Pigs are frequently used as models for the study of cardiovascular disease and atherosclerosis as the morphology and physical function of the cardiovascular system in swine are similar to humans’ (Smith and Swindle, Reference Smith and Swindle2006; Tables 1 and 2). Exceptions are that the swine heart has a left azygous vein that contains systemic blood from the inter-costal vessels; the conduction system is more neurogenic than myogenic; and the nerve cells within the atrioventricular node also differ. Having said that, the blood supply to the myocardium and the conduction system of the swine is similar to 90% of the human population (Smith and Swindle, Reference Smith and Swindle2006). Atherosclerosis in the pig develops both spontaneously and when induced by an experimental atherogenic, high-cholesterol diet in pigs (Gerrity et al., Reference Gerrity, Natarajan, Nadler and Kimsey2001). Plaque histology and pathogenesis in the atherogenic swine appears to be similar to that of humans (Gal and Isner, Reference Gal and Isner1992; Rand, Reference Rand2009). Atherogenic diets offered to rodents result in some of characteristics of atherosclerosis and the metabolic syndrome observed in humans (Buettner et al., Reference Buettner, Schölmerich and Cornelius Bollheimer2007; Oron-Herman et al., Reference Oron-Herman, Kamari, Grossman, Yeger, Peleg, Shabtay, Shamiss and Sharabi2008), however they generally provide poor models of the changes in plasma lipids associated with coronary heart disease.

Some metabolic and cardiovascular disorders occur naturally in domestic pigs as a consequence of intense selective breeding and selection for increased growth rate and leanness (Brambilla and Cantafora, Reference Brambilla and Cantafora2004). Feeding the modern domestic pig excess amounts of lard and cholesterol can induce atherosclerosis (Gerrity et al., Reference Gerrity, Natarajan, Nadler and Kimsey2001; McDonald et al., Reference McDonald, Gerrity, Jen, Chen, Wark, Wight, Chait and O’Brien2007; Table 4) without some of the other clinical signs of the metabolic syndrome. The minipig, particularly the ossabaw swine, has proved to be an excellent model because an atherogenic diet induces three or more of the clinical signs of the metabolic syndrome (Johansen et al., Reference Johansen, Hansen, Richelsen and Malmlöf2001; Spurlock and Gabler, Reference Spurlock and Gabler2008).

Table 4

A comparison of diet-induced metabolic syndrome in different breeds of pig

Yes indicates whether the animal expresses diseases.

The natural history of restenosis (i.e. the reoccurrence of stenosis) in pigs, both with respect to neointimal formation and remodelling, resemble the human disease closely. A combination of denudation and atherogenic diet can be used to induce atherosclerosis in pigs. Although the Yucatan breed of minipig is not a successful model for metabolic-syndrome-induced atherosclerosis, as it fails to exhibit insulin resistance and obesity, it does provide a model for studying the arterial response after balloon angioplasty and natural remodelling (De Smet et al., Reference De Smet, Van Der Zande, Van Der Helm, Kuntz, Borst and Post1998).

Over the past decade, pigs have been increasingly used to study changes in the collagen of the extra cellular matrix of the heart during hypertrophic cardiomyopathy (Liu et al., Reference Liu, Chiu, Shyu, Factor, Chu, Lin, Hsou, Fox and Yang1994; Chiu et al., Reference Chiu, Liu, Liu, Chen, Lin, Mao and Chu1999); abdominal aortic aneurysm (Ruiz et al., Reference Ruiz, Zhang, Butt and Whittaker1997) and chronic ischaemia because their coronary anatomy, with minimal preexisting coronary collateral vessels (Maxwell et al., Reference Maxwell, Hearse and Yellon1987), cardiac physiology, and cardiac conduction systems are very similar to humans (Swindle et al., Reference Swindle, Horneffer, Gardner, Gott, Hall, Stuart, Baumgartner, Borkon, Galloway and Reitz1986). Likewise, the heart weight to BW ratio for the typical 30 kg minipig used in most laboratory studies is identical to that of humans at around 0.005:1 or 5 g heart tissue per kilo of BW (Joseph, Reference Joseph1908; Hughes, Reference Hughes1986; Kist et al., Reference Kist, Thomas, Horner and Laughlin1999) and from a metabolic standpoint non-esterified fatty acids (NEFA) are the preferred substrate of myocardial energy production in both species (Abdel-Aleem et al., Reference Abdel-Aleem, St Louis, Hughes and Lowe1999). The development of a porcine model of myocardial ischaemia to evaluate the efficacy of therapeutic angiogenesis in the preclinical setting is ongoing (Sodha et al., Reference Sodha, Boodhwani, Ramlawi, Clements, Mieno, Feng, Xu, Bianchi and Sellke2008). Despite the success in some of these models (Hughes, 2003), randomized placebo controlled clinical trials failed to demonstrate a similar degree of success (Kastrup et al., Reference Kastrup, Jørgensen, Rück, Tägil, Glogar, Ruzyllo, Bøtker, Dudek, Drvota and Hesse2005). There are multiple reasons for the disparate results seen between animal models and human patients, including the presence of comorbidities such as hypercholesterolemia and diabetes, and medication usage, which may alter the milieu of signalling molecules involved in the angiogenic cascade (Simons et al., Reference Simons, Bonow, Chronos, Cohen, Giordao, Hammond, Laham, Li, Pike, Sellke, Stegmann, Udelson and Rosengart2000).

Diabetes and the metabolic syndrome

NIDDM

Atherosclerosis is accelerated in diabetic patients and contributes to the majority of deaths of diabetic patients (Ruderman et al., Reference Ruderman, Williamson and Brownlee1992). Pigs represent a valuable model for diabetes-induced accelerated atherosclerosis. Administration of streptozotocin (STZ) induces DNA strands to break within β-cells. The activation of repair mechanisms leads to a reduction in cellular nicotinamide adenine dinucleotide and adenosine triphosphate levels to below physiological levels, resulting in cell death and ultimately diabetes (Yamamoto et al., Reference Yamamoto, Uchigata and Okamoto1981). Various studies have investigated the effectiveness of different dose rates of STZ to induce diabetes (Table 5) and the differences in response have been attributed to age and gender variations (Larsen and Rolin, Reference Larsen and Rolin2004).

Table 5

Response to increasing dose rates of streptozotocin in juvenile pigs

Total parenteral nutrition

When an infant is born prematurely, although its gastrointestinal (GI) tract is fully formed at about 20 weeks of pregnancy, it is not fully functional at that time; for example peristalsis does not commence until around 29 weeks of gestation and it may not be able to synthesize the enzymes required to digest milk and formula (Sangild, Reference Sangild2006). As a consequence, it can often not tolerate enteral feeding due to its immature GI tract and so must be fed intravenously (known as total parenteral nutrition (TPN)). As nutrition provided by TPN bypasses the gut and absorption at the intestinal mucosa, it results in varying degrees of gut atrophy, depending on the period of treatment and whether the individual is receiving any nutrient enterally. Most infants on TPN are subject to numerous clinical problems, which complicate the nutritional and metabolic requirements and make interpretation of any experimental results difficult (Jadhav et al., Reference Jadhav, Parimi and Kalhan2007). The presence of central venous catheters and mechanical ventilation associated with TPN may increase the incidence of late-onset sepsis in these very LBW (<1500 g) babies (Kansagra et al., Reference Kansagra, Stoll, Rognerud, Niinikoski, Ou, Harvey and Burrin2003). TPN results in deprivation of luminal nutrition that also adversely affects the mucosal integrity and may compromise the barrier function of the neonatal gut, resulting in ‘gut-derived’ sepsis, and the subsequent translocation of luminal bacteria and toxins into the blood (Berg, Reference Berg1995).

Rat pups are not a suitable model for studying human premature infants receiving TPN due to their immaturity. When rats have reached a BW of around 180 g (∼60% mature BW), they can be used as a model to study gut function and epithelial regeneration following TPN after GI trauma, surgery or cancer (Jordinson et al., Reference Jordinson, Goodlad, Brynes, Bliss, Ghatei, Bloom, Fitzgerald, Grant, Bardocz, Pusztai, Pignatelli and Calam1999). As mentioned above, piglets possess poor thermoregulatory mechanisms, have high metabolic rates and are prone to hypoglycemia, thus providing an excellent model for premature human neonates receiving TPN (Mei and Xu, Reference Mei and Xu2003). Although immature at parturition, piglets have the ability to survive when delivered preterm allowing modelling of different levels of prematurity visible in human infants. Piglets can also be used with or without being given colostrum to provide immunoglobulins (Shulman, Reference Shulman1993). Work with colostrum-deprived (i.e. receiving no colostrum and so immumocompromised) minipiglets was first carried out by Mehrazar and Kim (Reference Mehrazar and Kim1988) who used germfree colostrum-deprived piglets delivered 3 to 5 days preterm to study the ontogeny of the immune system. Borum (Reference Borum1993) developed their work by using non-germfree (i.e. born and housed in non-sterile conditions), colostrum-deprived pigs as a model for studies into human TPN. This model has been further developed by Sangild et al. (Reference Sangild, Petersen, Schmidt, Elnif, Petersen and Buddington2002), with the use of colostrum-deprived piglets with limited immunity gained via an infusion of maternal serum; they also delivered the piglets at 93% of gestation (i.e. 8 days preterm) with a good degree of success. Pigs in a germfree (i.e. sterile) environment can be administered TPN successfully for up to 21 days (Mehrazar and Kim, Reference Mehrazar and Kim1988; Borum, Reference Borum1993).

Apart from anything else, piglets have a large enough body size to allow the same TPN methods and procedures to be used as in humans and for regular blood sampling without significant diminution of the total blood volume. At the same time, their body size is small enough to limit experimental costs (Burrin, Reference Burrin2001). Pigs have the advantage of a large litter size and 2 to 3 litters per sow per year (Pond and Mersmann, Reference Pond and Mersmann2001) making repeat experiments easily achievable, and reducing, to some extent, genetic influences. A typical TPN solution (energy content 1.67 kJ/ml) is usually administered with 20% lipid emulsion (8.36 MJ/l) (Hyde et al., Reference Hyde, Amusquivar, Laws, Corson, Geering, Lean, Putet, Dodds, Herrera and Clarke2008). One factor important to emphasize is that premature piglets, as with premature human babies, are prone to fluid retention and so infusion rates must be altered accordingly (Hyde et al., Reference Hyde, Amusquivar, Laws, Corson, Geering, Lean, Putet, Dodds, Herrera and Clarke2008) by gradually stepping up the infusion rates over the first 24 h of life. Infusion rates also differ between piglets delivered by Caesarean section (112 days of gestation) and vaginally at term (115 days) (Hyde et al., Reference Hyde, Amusquivar, Laws, Corson, Geering, Lean, Putet, Dodds, Herrera and Clarke2008 and Reference Hyde, Griffin, Herrera, Byrne, Clarke and Kemp2010).

Over the last two decades the newborn piglet has proved to be a good model for studying: (i) calcium metabolism (Draper et al., Reference Draper, Yuen and Whyte1991); (ii) growth and development of the gut and gut barrier function (Burrin et al., Reference Burrin, Shulman, Storm and Reeds1991; Van Aerde et al., Reference Van Aerde, Keelan, Clandinin and Thomson1997; Kansagra et al., Reference Kansagra, Stoll, Rognerud, Niinikoski, Ou, Harvey and Burrin2003); (ii) the immune system (Mehrazar and Kim, Reference Mehrazar and Kim1988); (iv) amino acid metabolism (House et al., Reference House, Pencharz and Ball1997 and Reference House, Pencharz and Ball1998); and (v) non-alcoholic fatty liver disease (Hyde et al., Reference Hyde, Perkins, Laws, Dodds, Symmons, Geering, Litten, Corson, Lean and Clark2005) in TPN-fed human infants. More recently, the effects of the fatty acid composition of the lipid emulsion on fatty acid profiles of tissues from Caesarean-sectioned delivered piglets has shown that the fatty acid composition of key tissues mirrors the fatty acid profile of the lipid emulsion supplied (Amusquivar et al., Reference Amusquivar, Sánchez, Hyde, Laws, Clarke and Herrera2008); this could have important consequences, particularly for neurodevelopment, in premature human babies.

Postnatal digestive physiology

The piglets’ GI tract has a high degree of anatomical and physiological similarity to that of the human infant (Siggers et al., Reference Siggers, Siggers, Boye, Thymann, Molbak, Leser, Jensen and Sangild2008), and their protein and lipid metabolism are comparable with humans (Canavan et al., Reference Canavan, Flecknell, New, Alberti and Home1997; Davis et al., Reference Davis, Suryawan, Orellana, Nguyen and Fiorotto2008). Therefore, pigs are an accepted and extensively used model for specific types of nutritional studies (Moughan and Rowan, Reference Moughan and Rowan1989; Spurlock and Gabler, Reference Spurlock and Gabler2008). For example, pigs have been used to study protein digestion in human infants (Darragh and Moughan, Reference Darragh and Moughan1995; Lin et al., Reference Lin, Mahan, Wu and Kim2009); digestibility of dietary amino acids (Rowan et al., Reference Rowan, Moughan, Wilson, Maher and Tasman1994; Wu, Reference Wu1998; Reeds and Burrin, Reference Reeds and Burrin2000) and the effect of infant milk formulas on organ development and digestive enzyme activities (Moughan et al., Reference Moughan, Pedraza, Smith, Williams and Wilson1990 and Reference Moughan, Birtles, Cranwell, Smith and Pedraza1992). A review of the pig as an experimental model for elucidating the mechanisms governing dietary influence on mineral absorption has recently been published by Patterson et al. (Reference Patterson, Lei and Miller2008), and so this aspect of digestion will not be considered herein.

Another area of interest is the effect of dietary fats on the growth and development of the human infant, particularly the brain (Innis, Reference Innis2004). Piglets undergo a growth spurt in the brain at a similar time as that observed in humans; the porcine brain is relatively well developed at birth, having achieved 38% of its adult brain weight (Pond et al., Reference Pond, Boleman, Fiorotto, Ho, Knabe, Mersmann, Savell and Su2002). This growth spurt is accompanied by subcutaneous lipid accumulation and research shows that the fatty acid profile of the nutrients supplied to the developing piglet postnatally can influence the fatty acid profile of the brain and other tissues (Amusquivar et al., Reference Amusquivar, Sánchez, Hyde, Laws, Clarke and Herrera2008; Hyde et al., Reference Hyde, Amusquivar, Laws, Corson, Geering, Lean, Putet, Dodds, Herrera and Clarke2008). Hence, the piglet is an excellent model for the studies on fatty acid metabolism during the perinatal period (Purvis et al., Reference Purvis, Clandinin and Hacker1982; Laws et al., Reference Laws, Litten, Laws, Lean, Dodds and Clarke2009).

In addition to the physiological changes associated with the transition from the intrauterine to extrauterine environment, pig and human infants both are also prone to diarrhoea. Diarrhoea during this phase of development is usually associated with rotavirus and Escherichia coli, although Coronavirus (Transmissible Gastroenteritis) and Clostridium perfringens type C also cause enteritis in pigs. The beneficial effect of the probiotic lactic acid bacteria in human or animal health has interested scientists over the last century. Probiotics are essentially ‘living microorganisms which favourably influence the health of the host by improving the indigenous microflora’ (Fuller, Reference Fuller1989). The World Health Organization (WHO, 2002) defines them as ‘live microorganisms which, when administered in adequate quantities, confer a health benefit to the host’. Positive health effects may also be obtained by ‘prebiotics’ – non-digestible substances that increase the growth or the activity of specific microorganisms in the gastro-intestinal tract. Potential effects of prebiotics and probiotics include increased nutritional value of food due to improved digestibility and absorption, improvement of the immune system, and prevention of intestinal tract infections, cancer, atherosclerosis and osteoporosis (Ziemer and Gibson, Reference Ziemer and Gibson1998; Gill and Guarner, Reference Gill and Guarner2004).

Necrotizing enterocolitis (NEC) is a medical condition primarily seen in premature infants in which portions of the bowel undergo necrosis, and this condition has been investigated using Caesarean-delivered preterm pigs as a model for premature human infants (Sangild, Reference Sangild2006). A probiotic-containing formula can reduce the severity of NEC in this model. Probiotic administration immediately after birth appears to promote the colonization of a beneficial commensal microbiota capable of limiting the formula-induced mucosal atrophy, dysfunction, and pathogen load in preterm neonates, thereby reducing the incidence and severity of NEC (Siggers et al., Reference Siggers, Siggers, Boye, Thymann, Molbak, Leser, Jensen and Sangild2008).

Weaning piglets at an age of 3 to 5 weeks changes the flora, morphology and function of the porcine intestine (Nabuurs, Reference Nabuurs1998). The effectiveness of feeding probiotic (Bifidobacterium lactis HN019 concentration 10⁸ cfu/mL) against naturally acquired enteric infectious diseases caused by rotavirus and E. coli has been studied using a piglet model (Shu et al., Reference Shu, Qu and Gill2001). Results suggest that diarrhoea was significantly reduced in probiotic-administered weanlings. It is of interest to note that the reduction in the incidence and severity of diarrhoea in neonatal and weanling pigs receiving probiotics is comparable, further supporting the pig as a good model for neonatal GI infection in human infants (Figure 3).

Figure 3

The incidence and severity of diarrhoea as influenced by breast (colostrum) milk (BR), milk formula (FORM) and milk formula with probiotics (FORM+) in caesarean-delivered preterm pigs and weaned pigs orally administered either a placebo (PRO−) or probiotic (PRO+). Values are presented as means for incidence & mean ± s.e.m. for severity. FORM+ = Bifidobacterium animalis and Lactobacillus: L. acidophilus, L. casei, L. pentosus, L. plantarum. PRO+ = Bifidobacterium lactis HN019 (10⁹ colony-forming units (cfu)/piglet per day). Caesarean-delivered preterm pigs data adapted from Siggers et al. (Reference Siggers, Siggers, Boye, Thymann, Molbak, Leser, Jensen and Sangild2008). Weaned pigs data adapted from Shu et al. (Reference Shu, Qu and Gill2001).

Alternative animal models have been used, for example probiotics have been shown to reduce bacterial infection in the acute pancreatitis rat model, the major advantage of this model is that it resembles the human to such an extent that bacteriologic results, reaction to treatment and disease course can all be predicted (Van Minnen et al., Reference Van Minnen, Timmerman, Lutgendorff, Verheem, Harmsen, Konstantinov, Smidt, Visser, Rijkers, Gooszen and Akkermans2007). Supplementation with Bifidobacterium infantis in the rat model resulted in intestinal colonization and a significant reduction in intestinal colonization and a significant reduction in the incidence of necrotising enterocolitis in comparison to the controls (Caplan et al., Reference Caplan, Miller-Catchpole, Kaup, Russell, Lickerman, Amer, Xiao and Thomson1999). However, due to the small size of the rat, the tissues samples obtained are also small and often require pooling.

It is well documented that diet modulates immune functions in different ways and affects host resistance to infections. In addition to the essential nutrients, non-digestible carbohydrates such as inulin (IN) and oligofructose (OF) modulate the systemic immune system (Seifert and Watzl, Reference Seifert and Watzl2007). IN and OF are classified as prebiotics, which occur naturally as plant storage carbohydrates in vegetables, cereals and fruits. Results from human intervention studies suggest that the intake of IN and OF has beneficial effects on the gut-associated lymphoid tissue. Both porcine and rat models have demonstrated similar anti-inflammatory changes in the gut-associated lymphoid tissue as those observed in humans and give more insight into the immune tissue- specific effects of IN and OF (Roller et al., Reference Roller, Rechkemmer and Watzl2004; Seifert and Watzl, Reference Seifert and Watzl2007). More recently an in vitro dynamic model to simulate porcine ileal digestion (Meunier et al., Reference Meunier, Manzanilla, Anguita, Denis, Pérez, Gasa, Cardot, Garcia, Moll and Alric2008) has been developed, but the potential benefits of using a similar system to investigate the effects of probiotics/prebiotics remains to be established.

Bones and osteoporosis

Osteoporosis is a multigenic complex disorder. Though the mouse and rat are used as experimental models for human osteoporosis, the pig bone remodelling cycle is histologically more similar to human than the rat or mouse. Moreover, livestock genomics have many advantages over model organisms and human studies for complex trait dissection (Onteru et al., Reference Onteru, Fan and Rothschild2008). In addition to other features mentioned elsewhere in this review, the reproductive cycle of the pig is similar in duration to the human (18 to 21 days) and is continuous, also like the human (Turner, Reference Turner2001).

A syndrome of spontaneous vertebral fracture has been reported in the pig, a rarity in the animal world (Spencer, Reference Spencer1979), pigs are large enough to receive prosthetic implants, withstand serial bone biopsies and large volumes of blood sampling (Turner, Reference Turner2001). The rate of bone removal and deposition of the trabecular and cortical bone is also similar to that observed in humans (Mosekilde et al., Reference Mosekilde, Weisbrode, Safron, Stills, Jankowsky, Ebert, Danielsen, Søgaard, Franks, Stevens, Paddock and Boyce1993b). A porcine model has also been used to evaluate the initial fixation strength of a biodegradable interference screw in anterior cruciate ligament reconstruction using a bone-patellar tendon-bone graft (Seil et al., Reference Seil, Rupp, Krauss, Benz and Kohn1998; Walsh et al., Reference Walsh, Wijdicks, Parker, Hapa and LaPrade2009).

Findings suggest that the Sinclair Si minipig fed a 0.75% calcium restricted diet may be a good model to study the bone remodelling and perimenopausal bone loss in women when an ovariectomy is performed (Mosekilde et al., Reference Mosekilde, Weisbrode, Safron, Stills, Jankowsky, Ebert, Danielsen, Søgaard, Franks, Stevens, Paddock and Boyce1993a and Reference Mosekilde, Weisbrode, Safron, Stills, Jankowsky, Ebert, Danielsen, Søgaard, Franks, Stevens, Paddock and Boyce1993b; Boyce et al., Reference Boyce, Ebert, Youngs, Paddock, Mosekilde, Stevens and Gunderson1995). Lafage et al. (Reference Lafage, Balena, Battle, Shea, Seedor, Klein, Hayes and Rodan1995) have also used minipigs to investigate bone active agents such as sodium fluoride, assessed on bone quality and remodelling. Dynamic ¹⁸F-fluoride ion positron-emission tomography (PET) has demonstrated that porcine bone loss after total gastrectomy is related to a high-turnover bone disease without associated changes in bone blood flow. It is thought that the increased bone metabolism observed in minipigs is probably related to an elevated parathyroid hormone secretion, thus maintaining serum calcium homeostasis at the expense of the bone mineral content. Normalizing bone metabolic activity by the specific bone mass increases the sensitivity in the detection of osteopenic high turnover bone diseases. Therefore, the combination of quantitative computed tomography and ¹⁸F-fluoride ion PET seems to be the method of choice for the classification of metabolic bone diseases and for monitoring treatment effects quantitatively (Piert et al., Reference Piert, Zittel, Jahn, Stahlschmidt, Becker and Machulla2003).

Dental research

Since the tooth structure of the Prestige World Genetics (PWG) Micro-pig^® closely resembles that of human more than any other species, the pigs are valuable in dental research. Adult stem cells originated from pulp show that new dental material can be created and implanted. As the size of the Micro-Pig^® jawbone is as thick as humans, these pigs are frequently used in dental surgery, especially in the testing of dental implants and the use of dental stem cells to grow new teeth (PWG Genetics Korea, 2006; Mantesso, Reference Mantesso2008).

By examining the ontogeny of porcine enamel our understanding of human enamel deposition and maturation has been significantly improved (Robinson et al., Reference Robinson, Kirkham, Weatherell, Richards, Josephsen and Fejerskov1987 and Reference Robinson, Kirkham, Weatherell, Richards, Josephsen and Fejerskov1988; Kirkham et al., Reference Kirkham, Robinson, Weatherell, Richards, Fejerskov and Josephsen1988). Work has been completed on implants in pigs to consider loading, however caution needs to be used here as porcine bone remodels slightly faster than human bone. That noted this work provides a valuable insight into micro damage strains following implants (Ko et al., Reference Ko, Swift, DeLong, Douglas, Kim, An, Chang and Huang2002 and Reference Ko, Douglas, DeLong, Rohrer, Swift, Hodges, An and Ritman2003). A novel device for a bite force measurement system in a porcine model has recently been described (Bousdras et al., Reference Bousdras, Cunningham, Ferguson-Pell, Bamber, Sindet-Pedersen, Blunn and Goodship2006) which enables the biomechanical adaptation of the bone-implant interface to masticatory loads to be assessed with either natural dentition or single implant crowns.