Pancreatic endocrine function in newborn pony foals after induced or spontaneous delivery at term

Authors


Summary

Reasons for performing study: During the switch from parenteral to enteral nutrition at birth, endocrine glands such as the pancreas must assume a glucoregulatory role for the first time if the neonate is to survive the transition to extrauterine life.

Objectives: To determine the adaptations in pancreatic endocrine function during the neonatal period in term pony foals delivered by different methods.

Methods: By measuring insulin and glucagon concentrations, pancreatic α and β cell responses to exogenous glucose (0.5 g/kg bwt) and arginine (100 mg/kg bwt) and to endogenous muzzling for 90 min were determined periodically over the first 10 post natal days in foals born spontaneously (n = 8) or by induction of labour with oxytocin at term (n = 7).

Results: Pancreatic α and β cell function changed with post natal age in a manner related to the method of delivery. Induced foals had 2–3 fold greater β cell responses to exogenous glucose and arginine, despite similar glucose and α-amino nitrogen clearances compared with spontaneously delivered foals. Pancreatic β cell responses to glucose decreased by 50% while those to arginine doubled with increasing age in induced but not spontaneously delivered foals. In contrast, pancreatic α cell responses to arginine doubled with increasing age in foals born spontaneously but not by induction. These differences in pancreatic endocrine cell function with delivery method were associated with 2–3 fold higher cortisol levels in the induced foals and with differences in the absolute and age-related changes in basal concentrations of glucose, α-amino nitrogen and insulin.

Conclusions: Induced delivery leads to changes in pancreatic β cell sensitivity to glucose and/or tissue insulin resistance in association with persistent neonatal hypercortisolaemia.

Potential relevance: The altered post natal development of pancreatic endocrine function with induced delivery may compromise glucoregulation and adaptation to enteral nutrition in neonatal foals with potential consequences long after birth.

Introduction

At birth, provision of nutrients switches from a continuous placental source to an intermittent supply via the gastrointestinal tract (Sangild et al. 2000). During this transition from parenteral to enteral nutrition, maintenance of normoglycaemia depends on the glucoregulatory competence of endocrine glands, such as the pancreas. In mature horses, the pancreatic hormones, insulin and glucagon, vary in concentration with the composition and absolute intake of the diet and in response to physiological challenges to glucose homeostasis, such as fasting, pregnancy and exercise (Hall et al. 1982; Ralston and Baile 1982; Fowden et al. 1984a; Smyth et al. 1989; DePew et al. 1994; Sticker et al. 1995; Sticker et al. 1996). The endocrine pancreas is functional and responsive to stimuli before birth in all species studied to date including the horse (Fowden and Hill 2001). It secrets insulin in response to glucose and amino acids and has an important role in regulating fetal metabolism and growth (Fowden 1997). Pancreatic β cell sensitivity to glucose increases during late gestation in the horse and again during the first 10 days of post natal life in foals delivered naturally at full term (Fowden 1980; Fowden et al. 1982; Holdstock et al. 2004; Fowden et al. 2005). Less is known about perinatal development of pancreatic β cell sensitivity to amino acids, although arginine has been shown to stimulate insulin secretion in fetal horses and sheep during late gestation (Fowden 1980; Gresores et al. 1997; Fowden et al. 2005). When foals are delivered prematurely by induction of labour, their β cell sensitivity to glucose is poor immediately after birth compared with full-term neonates and is accompanied by progressive hypoglycaemia (Fowden et al. 1984a; Madigan 1991). Even when delivery is induced at full term, there are abnormalities in the insulin concentration profile during the 2 h after birth in foals that otherwise appear mature and normal behaviourally (Fowden et al. 1984a). However, to date, no studies have investigated the effect of inducing labour at term on development of pancreatic β cell function during the neonatal period.

Compared to pancreatic β cells, much less is known about the ontogeny of α cell function during the nutritional transition at birth. Glucagon is detectable in fetal plasma during late gestation in a number of species and rises in concentration in fetal horses towards term to peak at delivery (Fowden et al. 1999; Fowden and Hill 2001). Glucagon administration increases fetal glucose concentrations in late gestation in several species and has been shown to activate hepatic glucose production in fetal sheep near term (Apatu and Barnes 1991; Fowden and Forhead 2009). In newborn rats, sheep and pigs, glucagon concentrations are high immediately at birth and decrease over the next few days as suckling is established (Grajwer et al. 1977; Girard and Sperling 1983; Lepine et al. 1989). If lactation is slow in onset, as occurs in man or when neonates are fasted, glucagon concentrations continue to rise over the first 2–3 days of post natal life (Sperling et al. 1974; Kasser et al. 1981). Amino acids, such as arginine, are known to elicit a brisk α cell response in fetuses of sheep, pigs, horses and nonhuman primates during late gestation and in newborn human infants within 6 h of birth (Chez et al. 1974; Sperling et al. 1974; Fowden 1980; Bassett et al. 1982; Silver et al. 1986; Fowden et al. 1999). However, their effectiveness as glucagon secretogogues in the neonatal period during the establishment of nutritive suckling remains largely unknown in any species. Nor is it clear whether induction of delivery affects pancreatic α cell function in the neonate.

In many species, maturation of fetal tissues in preparation for extrauterine life depends on activation of the hypothalamic–pituitary–adrenal (HPA) axis and the prepartum rise in fetal cortisol concentrations (Fowden et al. 1998). Compared with other species, this cortisol surge occurs relatively late in gestation and continues into the first few hours after birth in the horse (Fowden and Silver 1995). Although little is known about the maturational effect of cortisol on the endocrine pancreas (Fowden and Hill 2001), cortisol concentrations during the perinatal period may influence the adaptations in pancreatic endocrine function to extrauterine life. This study, therefore, examined pancreatic α and β cell function in foals during the critical period of glucoregulatory adaptation in the first 10 days ex utero after either spontaneous or induced delivery at term. It also measured the circulating concentrations of cortisol and the catecholamines, adrenaline and noradrenaline, known regulators of fetal insulin and glucagon secretion (Fowden and Hill 2001), in the foals over this period.

Methods

Animals

A total of 15 newborn pony foals were used in this study. Eight were born spontaneously at full term (gestational age range 326–345 days), while the remainder were delivered after induction of labour using i.v. administration of oxytocin (20 iu) on the basis of milk electrolyte concentrations. The oxytocin was given on the day when the declining milk sodium concentration had crossed the rising milk potassium concentration, at a value of about 40 mmol/l, which normally occurs within 24–48 h of natural delivery (Ousey et al. 1984; Ousey 2004). Delivery occurred at mean ± s.e. 43 ± 7 min after oxytocin administration (n = 7). Mares delivered unassisted and were assigned randomly to either induced or natural labour. All had normal placentae. Shortly after birth, a blood sample was taken from the jugular vein of the foal by venipuncture to measure mean erythrocyte cell volume (MCV), total white blood cell count (WBC) and the ratio of neutrophil to lymphocyte concentrations (N:L ratio) as indices of maturity (Rossdale et al. 1984; Ousey et al. 2004). Foals were considered mature when they met 2 of the following 3 criteria: MCV <45 fl, WBC >6 x 109/l and N:L ratio >2 (Ousey et al. 2004). They were weighed at birth and then daily for the next 10 days. They were housed with their mothers in individual boxes and suckled ad libitum except during the tests of pancreatic endocrine function when they were muzzled for 90 min. All experimental procedures, including the induction of labour, were carried out under the Animals (Scientific Procedures) Act 1986.

Experimental procedures

Between 6 and 10 h after birth, a long-term polyurethane catheter (Milacath)a was inserted into the jugular vein of each foal under local anaesthesia (Intra-Epicaine)b and left in situ for the 10 day experimental period (Fig 1). Pancreatic endocrine responses were measured over 48 h at 3 time periods after birth (Fig 1. Period 1, Days 1 and 2; Period 2, Days 5 and 6; Period 3, Days 9 and 10 ex utero). After muzzling for 30 min, either glucose (0.5 g/kg bwt, 40% solution;b Days 1, 5 and 9) or arginine (100 mg/kg bwt, L-arginine hydrochloridec in 20 ml 0.9% w/v NaCl; Days 2, 6 and 10) was given i.v. over 5 min to the foal. Blood samples (5 ml) were taken immediately before (0 min) and at 5, 15, 30 and 60 min after administration of the solutions (Fig 1). On Days 2, 6 and 10, the foal was also muzzled and blood samples (5 ml) taken immediately before (0 min) and at 30 min intervals for 90 min after muzzling beginning about 3 h before arginine administration (Fig 1). At the end of each period of sampling, the muzzle was removed and the foals were allowed to suckle ad libitum. Blood samples were divided into ice cold tubes containing heparin (insulin, α-amino nitrogen), EDTA (glucose, cortisol), EDTA plus the protease inhibitor aprotonin (glucagon) or EDTA plus glutathione (catecholamines) depending on the hormones to be measured in the sample. After centrifugation at 4°C, the plasma was stored at -20°C until analysis. The 0 min samples taken immediately before glucose or arginine administration were used to provide the basal concentrations of metabolites and hormones on each of the post natal days of study (Fig 1).

Figure 1.

Schematic diagram of the experimental protocol showing the relationship between the period of muzzling and the administration of glucose and arginine to newborn foals over the first 10 days of post natal life after delivery either spontaneously or by induction. Timing of blood sampling is shown with thin arrows while administration of glucose or arginine is shown with broad arrows. Details of the doses are given in the text (Silver & Fowden, 1994). The 0 min sample used to provide the basal concentrations of metabolites and hormones on each of the post natal days studied are indicted with a B.

Biochemical analyses

Metabolite concentrations: Plasma glucose and α-amino nitrogen concentrations were determined colorimetrically as described previously (Fowden et al. 1982, 1999). Plasma α-amino nitrogen concentrations are a measure of total amino acid levels and, hence, provide an index of the changes in arginine concentration after its exogenous administration.

Hormone concentrations: Plasma insulin concentrations were measured on a 1235 AutoDELFIA automated immunoassay systemd using a 2 step time-resolved fluorimetric assay validated for use with equine plasma (Holdstock et al. 2004). Interassay coefficients of variation were 2.0–3.0% over the concentration range from 29 to 705 pmol/l. The limit of detection of the insulin assay was 1.3 pmol/l and the assay was linear to approximately 1100 pmol/l. Plasma glucagon concentrations were determined by radioimmunoassay using a commercial kit for porcine pancreatic glucagone validated previously for equine plasma (Fowden et al. 1999). The interassay coefficient of variation was 7.0% and the minimum detectable quantity of glucagon was 20–30 pg/ml. Plasma cortisol concentrations were also measured by radioimmunoassay as published previously (Rossdale et al. 1982). The interassay coefficient of variation was 8.0% and the minimum detectable quantity of cortisol was 1.5 ng/ml. Plasma adrenaline and noradrenaline concentrations were measured by high-pressure liquid chromatography (HPLC) as described previously (Silver et al. 1986). Briefly, plasma samples were prepared by absorption of 300–500 µl onto acid washed alumina and 20 µl samples of the 100 µl perchloric acid eluate were injected onto the column. Dihydroxybenzylamine (DHBA) was added as internal standard to each plasma sample before absorption. Standard solutions of adrenaline and noradrenaline were also treated in the same way.The limits of sensitivity were 0.07 ng/ml for adrenaline and 0.05 ng/ml for noradrenaline. The intra- and interassay coefficients of variation were 5.8 and 7.3% for adrenaline and 5.0 and 6.2% for noradrenaline, respectively.

Data analysis

Means ± s.e. are used throughout. Statistical significance of the pancreatic endocrine responses were assessed by 2-way repeated measures ANOVA using the Holm-Sidak post hoc test and method of delivery and time, or day, as factors with time/day as the repeated measure (Sigmastat 2.5)f. When day of study or an interaction between method of delivery and day were identified as a significant influence, the data from the 2 modes of delivery were analysed separately by one-way ANOVA with or without repeated measures or by paired or unpaired t test, as appropriate. Hormone concentrations were log transformed when required. For each secretogogue, the areas under the curve (AUC) for the glucose, α-amino nitrogen, insulin and glucagon responses were calculated as the integrated plasma concentration after administration of glucose or arginine from 5 to 60 min above the baseline concentration at 0 min for all positive values. Biometric data were analysed by unpaired t test. Insulin data from some of the spontaneously delivered foals have been published previously (Holdstock et al. 2004). For all statistical analyses, significance was accepted when P<0.05.

Results

Foal characteristics

All foals stood and sucked within 1 h of birth with minimal assistance, irrespective of their method of delivery. All were classified as mature by the haematological criteria used in clinical practice (Table 1; Rossdale et al. 1984). Foals delivered by induction had a lower MCV at birth (Table 1). They also weighed more at birth than those born spontaneously and remained heavier 10 days later (Table 1). Both groups of foals gained 0.5–1.0 kg/day; the mean increment in bodyweight over the 10 day experimental period did not differ significantly with the method of delivery (Table 1).

Table 1. Mean ± s.e. values of bodyweight and haematological indices of maturity in newborn foals delivered either spontaneously (n = 8) or by induction (n = 7) at term
 SpontaneousInduced
  • Significantly different from the value in the spontaneously delivered foals (P<0.02, t test).

Bodyweight (kg)  
 Birth (Day 1)20.5 ± 1.024.8 ± 0.8
 Day 1027.4 ± 1.532.5 ± 0.8
 Increment Day 1–106.9 ± 1.07.7 ± 0.9
Haematological parameters  
 Mean erythrocyte volume (fl)44.6 ± 1.241.0 ± 0.8
 White blood cell count (x109/l)7.89 ± 0.936.82 ± 0.30
 Neutrophil:lymphocyte ratio3.1 ± 0.63.6 ± 0.5

Basal metabolite and hormone concentrations

There were no significant differences in the basal concentrations of plasma glucose between induced and spontaneously delivered foals during the first 10 days of life (Table 2). Plasma glucose concentrations did not change with post natal age in the spontaneously delivered foals but rose significantly from Day 1 to Day 2 and then remained higher than Day 1 values until Day 6 in the induced foals (Table 2). In contrast, plasma α-amino nitrogen concentrations were higher on Day 2 and Day 6 than at birth or on Days 9 and 10 in foals born spontaneously but not in those delivered by induction (Table 2). Plasma α-amino nitrogen concentrations were also higher in induced than spontaneously delivered foals on Days 1, 9 and 10 (Table 2). Consistent with the glucose concentrations, basal insulin levels were not significantly different between the 2 groups of foals on any post natal day, although values rose with increasing age in the induced but not spontaneously delivered foals (Table 2). Plasma glucagon concentrations fell progressively over the first 10 days of post natal life in both groups of foals and did not differ with the method of delivery at any age (Table 2). In contrast, neonatal treatment and post natal age were both significant factors in determining the cortisol concentration of the foals overall (P<0.01, 2-way ANOVA). Plasma cortisol concentrations were significantly higher in induced than spontaneously delivered foals throughout the experimental period (Table 2). However, in the induced foals, plasma cortisol levels decreased from Day 1 to Day 6 whereas, in the spontaneously delivered foals, levels were stable over the 10 day experimental period (Table 2). Neither treatment nor post natal age were significant influences on the plasma catecholamine concentrations (P>0.05, 2-way ANOVA, Table 2).

Table 2. Mean ± s.e. basal concentration of plasma glucose, α-amino nitrogen, insulin, glucagon, cortisol, adrenaline and noradrenaline after muzzled for 30 min with respect to post natal age in newborn foals delivered either spontaneously (n = 8) or by induction (n = 7)
 Method of deliveryDay 1Day 2Day 5Day 6Day 9Day 10
  1. Values within rows with different superscripts are significantly different from each other (P<0.05, one-way repeated measures ANOVA). Significantly different from the value in the spontaneously delivered foals at the same age (P<0.01, 2-way repeated measures ANOVA).

GlucoseSpontaneous7.65 ± 0.678.54 ± 0.597.64 ± 0.207.91 ± 0.447.22 ± 0.367.62 ± 0.68
 (mmol/l)Induced5.92 ± 0.58a8.80 ± 0.28b8.08 ± 0.27ab8.68 ± 0.75b7.80 ± 0.48ab7.76 ± 0.47ab
α-amino nitrogenSpontaneous2.40 ± 0.24a3.83 ± 0.26b3.53 ± 0.21b3.64 ± 0.33ab3.11 ± 0.26ab3.46 ± 0.30ab
 (mmol/l)Induced3.82 ± 0.543.50 ± 0.283.79 ± 0.194.12 ± 0.354.15 ± 0.444.85 ± 0.30
InsulinSpontaneous97 ± 26152 ± 27128 ± 35153 ± 29117 ± 28161 ± 35
 (pmol/l)Induced69 ± 8a126 ± 20bc114 ± 40ab195 ± 57c129 ± 16bc191 ± 20c
GlucagonSpontaneous1456 ± 400a727 ± 152ab729 ± 82b695 ± 114bc590 ± 96bc383 ± 51c
 (pg/ml)Induced2057 ± 640a1000 ± 355ab600 ± 204b693 ± 243b897 ± 264ab651 ± 147b
CortisolSpontaneous36.8 ± 8.119.7 ± 3.025.3 ± 3.521.1 ± 4.424.7 ± 4.426.7 ± 3.4
 (ng/ml)Induced116.4 ± 28.7a67.1 ± 12.6ab74.8 ± 16.9ab56.2 ± 13.1b72.9 ± 10.9ab70.4 ± 17.1ab
AdrenalineSpontaneous43 ± 1753 ± 1858 ± 1854 ± 1563 ± 1543 ± 10
 (pg/ml)Induced69 ± 26110 ± 3286 ± 2553 ± 1977 ± 2573 ± 15
NoradrenalineSpontaneous218 ± 20222 ± 20305 ± 134165 ± 27162 ± 21176 ± 16
 (pg/ml)Induced265 ± 95315 ± 75259 ± 63208 ± 57210 ± 63201 ± 54

Pancreatic β cell responses

Exogenous administration of glucose stimulated a rapid release of insulin in all foals, irrespective of their age or method of delivery (Fig 2). Glucose concentrations peaked at the end of the 5 min infusion at all ages and returned to basal or below basal values by 30 min on Days 5 and 9 but remained elevated at 60 min on Day 1 (Fig 2a). The area under the glucose curve was, therefore, greater on Day 1 than subsequently in both groups of foals and did not differ between groups at any age (Fig 3a). Insulin concentrations peaked 5–15 min after glucose administration in all foals irrespective of age and, with the exception of the induced foals on Day 1, fell to basal or below basal values by 60 min (Fig 2b). The increment in the insulin concentration at the peak value was greater in induced than spontaneously delivered foals at all 3 ages and, on Day 1, remained greater throughout the 60 min sampling period (Fig 2b). Neonatal treatment and post natal age were both significant factors in determining the area under the insulin curve (P<0.02, 2-way ANOVA). This area was significantly greater in the induced than spontaneously delivered foals and decreased from Day 1 to Day 5 in the induced but not spontaneously delivered group of animals (Fig 3b).

Figure 2.

Mean ± s.e. change in the plasma concentrations of (a) glucose and (b) insulin above baseline values (0 min) in response to glucose administration on Days 1, 5 and 9 post partum in pony foals delivered either spontaneously (filled symbols, n = 8) or by induction of labour (open symbols, n = 7) at term. *significant increment in concentration above baseline (P<0.05, paired t test). †significantly different from the change in concentration observed in the spontaneously delivered foals (P<0.01, 2-way repeated measures ANOVA).

Figure 3.

Mean ± s.e. area under the response curve for (a) glucose and (b) insulin to glucose administration Days 1, 5 and 9 post partum in pony foals delivered either spontaneously (filled columns, n = 8) or by induction of labour (open columns, n = 7) at term. ns = not significant (P>0.05, 2-way repeated measures ANOVA). ‡significant difference with method of delivery (P<0.01, 2-way repeated measures ANOVA). †significantly different from the corresponding value in the spontaneously delivered foals (P<0.01, 2-way repeated measures ANOVA). Within each method of delivery, columns with different letters as superscripts are significantly different from each other (P<0.05, one-way repeated measure ANOVA). AUC = area under the curve.

Exogenous administration of arginine also evoked a brisk release of insulin in both groups of foals at all ages, although the increment in plasma insulin tended to be smaller and more transient than the corresponding responses to glucose administration, especially in foals born spontaneously (Fig 4). The increment in plasma α-amino nitrogen and the area under the α-amino nitrogen curve were unaffected by the age or method of the delivery of the foals (Figs 4a, 5a). There was an interaction between delivery method and age in determining the β cell response to arginine (P<0.01, 2-way ANOVA). On Day 10 only, the increment in plasma insulin and the area under the insulin curve were greater in the induced than spontaneously delivered foals (Figs 3b, 4b). The area under the insulin curve was also significantly greater on Day 10 than Days 2 or 6 in the induced foals (Fig 5b). This trend was not seen in the spontaneously delivered group (Fig 4b).

Figure 4.

Mean ± s.e. change in the plasma concentrations of (a) α-amino nitrogen, (b) insulin and (c) glucagon from baseline values (0 min) in response to arginine on Days 1, 5 and 9 post partum in pony foals delivered either spontaneously (filled symbols, n = 8) or by induction of labour (open symbols, n = 7) at term. *significant increment in concentration above baseline (P<0.05, paired t test). †significantly different from the change in concentration observed in the spontaneously delivered foals (P<0.01, 2-way repeated measures ANOVA).

Figure 5.

Mean ± s.e. area under the response curve for (a) α-amino nitrogen, (b) insulin and (c) glucagon to arginine administration on Days 1, 5 and 9 post partum in pony foals delivered either spontaneously (filled columns, n = 8) or by induction of labour (open columns, n = 7) at term. ns = not significant (P>0.05, 2-way repeated measures ANOVA). ‡significant interaction between age and method of delivery (P<0.01, 2-way repeated measures ANOVA). †significantly different from the corresponding value in the spontaneously delivered foals (P<0.01, 2-way repeated measures ANOVA). Within each method of delivery, columns with different letters as superscripts are significantly different from each other (P<0.05, one-way repeated measure ANOVA). AUC = area under the curve.

Muzzling led to a significant fall in the plasma glucose concentration in both groups of foals at all 3 ages (Table 3). The decrement in glucose after 90 min of muzzling was significantly less on Day 10 than Day 1 in both groups of foals (P<0.01, 2-way ANOVA) and, on Day 10, was significant only in the induced foals (Table 3). Plasma α-amino nitrogen concentrations also fell significantly during muzzling at all 3 ages in foals delivered spontaneously but not in those induced to deliver (Table 3). Plasma insulin levels decreased in parallel with the fall in plasma glucose in both groups of foals (Table 3). The mean decrements in plasma insulin during muzzling did not differ with age or method of delivery (P>0.05, 2-way ANOVA, Table 3).

Table 3. Mean ± s.e. concentrations of plasma glucose, α-amino nitrogen, insulin and glucagon before muzzling (0 min) and the mean change in concentration with respect to time after muzzling (Δ) of newborn foals delivered spontaneously (n = 8) and by induction (n = 7) on Days 2, 6 and 10 of post natal life
 TimeGlucose mmol/lα-amino nitrogen mmol/lInsulin pmol/lGlucagon pg/ml
SpontaneousInducedSpontaneousInducedSpontaneousInducedSpontaneousInduced
  1. *Significant change from 0 min value at each age (P<0.05, paired t test). Significantly different from value in the corresponding group of spontaneously delivered foals (P<0.02, t test).

Day 20 min9.87 ± 0.699.45 ± 0.453.84 ± 0.293.15 ± 0.27124 ± 35143 ± 24605 ± 167916 ± 256
 Δ+30 min-1.33 ± 0.47*-0.65 ± 0.28-0.51 ± 0.230.16 ± 0.22-52 ± 17*-44 ± 12*-45 ± 57-259 ± 227
 Δ+60 min-2.81 ± 0.60*-1.59 ± 0.57*-0.67 ± 0.23*0.21 ± 0.40-93 ± 27*-94 ± 21*-117 ± 74-222 ± 130
 Δ+90 min-3.70 ± 0.67*-2.88 ± 0.52*-0.76 ± 0.25*-0.09 ± 0.26-100 ± 31*-121 ± 26*-102 ± 46-248 ± 248
Day 60 min8.91 ± 0.388.82 ± 0.413.54 ± 0.303.46 ± 0.29140 ± 39239 ± 81584 ± 95586 ± 176
 Δ+30 min-1.01 ± 0.38*-0.15 ± 0.97-0.56 ± 0.12*-0.03 ± 0.15-46 ± 13*-120 ± 49*13 ± 3088 ± 123
 Δ+60 min-1.47 ± 0.52*-1.77 ± 0.36*-0.62 ± 0.26*-0.64 ± 0.27-81 ± 28*-165 ± 50*-15 ± 2535 ± 129
 Δ+90 min-1.85 ± 0.39*-1.73 ± 0.32*-0.73 ± 0.24*-0.45 ± 0.32-102 ± 31*-174 ± 50*-32 ± 4966 ± 109
Day 100 min7.75 ± 0.278.14 ± 0.493.51 ± 0.334.82 ± 0.30155 ± 31204 ± 38500 ± 85461 ± 121
 Δ+30 min-0.13 ± 0.53-0.38 ± 0.22-0.85 ± 0.27*-0.16 ± 0.53-83 ± 20*-82 ± 15*-70 ± 2830 ± 107
 Δ+60 min-1.13 ± 0.19*-0.89 ± 0.42-0.99 ± 0.18*-1.02 ± 0.48-104 ± 21*-71 ± 26*-104 ± 31-94 ± 73
 Δ+90 min-0.61 ± 0.26-1.30 ± 0.46*-1.05 ± 0.23*-0.79 ± 0.37-107 ± 27*-103 ± 19*-130 ± 4024 ± 102

Pancreatic α cell responses

Exogenous administration of arginine stimulated a small but transient increase in the plasma glucagon concentration in both groups of foals at all ages studied (Fig 4c). Plasma α-amino nitrogen concentrations were maximal at the end of 5 min arginine infusion and remained elevated until 15 min in all animals irrespective of age (Fig 4a). The peak increment in plasma glucagon occurred at the end of the 5 min infusion and did not differ in magnitude either between the 2 groups of foals or with age (Fig 4b). Plasma glucagon concentrations were maintained above basal values for 60 min only in the spontaneously delivered foals on Day 10 (Fig 3b). Consequently, there was an interaction between age and method of delivery in determining the pancreatic α cell response to arginine (P<0.02, 2-way ANOVA, Fig 4c). The area under the glucagon curve was greater at Day 10 than Day 2 or Day 6 in foals born spontaneously but not in those delivered by induction (Fig 5b). Glucagon concentrations were not measured in response to exogenous glucose administration and were unaffected by muzzling for 90 min, irrespective of age or method of delivery (Table 3).

Discussion

The results demonstrate that there are developmental changes in both pancreatic α and β cell function in pony foals during the critical period of glucoregulatory adaptation in the first 10 days after birth. The temporal pattern of these changes and the magnitude of the pancreatic endocrine responses to specific secretogogues differed between foals born spontaneously and those delivered by induction of labour at term. Pancreatic β cell responses to exogenous glucose and arginine decreased and increased, respectively, with increasing age in induced foals, whereas they showed no trend with age in foals born spontaneously. The magnitude of the insulin increment was also greater in the induced foals, particularly in response to exogenous glucose. In contrast, the method of delivery had no effect on the β cell responses to declining endogenous levels of glucose caused by muzzling of the foal. Pancreatic α cell responses to arginine, on the other hand, were unaffected by the method of delivery and increased with increasing age in foals born spontaneously but not in those delivered by induction. Collectively, these findings indicate that induction of delivery in the foal has important consequences for the neonatal adaptation in pancreatic endocrine function essential for the onset of tight glucoregulatory control after birth.

The greater insulin response to glucose seen in the induced foals may have been due to increased β cell sensitivity to glucose and/or to decreased tissue sensitivity to the actions of insulin as there were no differences in glucose clearance between the induced and spontaneously delivered foals at any age. The absence of basal hyperinsulinaemia in the induced foals and the similarity in the insulin response to muzzling suggest that differences in sensitivity of β cells to glucose and of peripheral tissues to insulin between the 2 groups of foals may only be physiologically relevant at the higher glucose and insulin concentrations, respectively. The causes of these changes in induced foals remain unclear. The increased birth weight of the induced foals may have been a contributory factor as greater β cell responses to glucose have been observed without changes in glucose clearance in newborn pony foals overgrown by embryo transfer into a Thoroughbred mare (Forhead et al. 2004). Since insulin is an important fetal growth hormone (Fowden and Forhead 2009), the enhanced β cell response of the induced foals may pre-date birth and hence, serendipitously, be unrelated to the method of delivery. However, it is more likely that the differences in neonatal insulin secretion with delivery method are related to the high cortisol levels of the induced foals, particularly as there were no significant differences in plasma catecholamine concentrations with the method of delivery.

Cortisol is known to cause insulin antagonism in mature animals including the horse (McCue 2002; Pivonella et al. 2005). It induces hyperglycaemia and protein catabolism, which in turn, lead to compensatory hyperinsulinaemia and, eventually, to type II diabetes in the horse and to β cell exhaustion and human type I diabetes (McCue 2002; Pivonella et al. 2005; Krikorian and Khan 2010). In the current study, the hypercortisolaemia of the induced foals was associated with higher α-amino nitrogen concentrations and age-related increases in the basal glucose and insulin concentrations as well as with the ability to maintain α-amino nitrogen concentrations during muzzling, in contrast to the observations in spontaneously delivered foals with lower cortisol levels. The finding that the area under the glucose curve after exogenous glucose administration was greatest on Day 1, irrespective of the method of delivery, is also consistent with cortisol causing insulin resistance as cortisol levels are known to be higher in the 6–12 h period preceding glucose administration on Day 1 than on subsequent days in both induced and spontaneously delivered foals (Silver et al. 1984; Ousey et al. 2004). Although glucocorticoids impair pancreatic endocrine development prenatally (Blondeau et al. 2001; Shen et al. 2003), their action in causing insulin resistance post natally may increase β cell proliferation in the neonatal pancreas, particularly as neogenesis of pancreatic endocrine cells can occur throughout life in the horse (Furuoka et al. 1989). Certainly, during pregnancy, insulin resistance is associated with β cell proliferation and enhanced glucose-stimulated insulin secretion in several species, including the mare (Frienkel 1964; Fowden et al. 1984b; Nielsen and Serup 1998). Indeed, increased β cell proliferation in response to prolonged hypercortisolaemia may explain, in part, the emergence of the heightened insulin response to arginine in induced foals 10 days after birth. However, there was no increment in the insulin response to exogenous glucose with increasing age of the induced foals, which suggests that insulin sensitivity also increases with age after birth in line with any increase in β cell mass. The heightened insulin responses to glucose and arginine in the induced foals may, therefore, reflect both increased β cell mass and tissue insulin resistance relative to spontaneously delivered foals at the same post natal age.

In common with other species (Girard and Sperling 1983), glucagon concentrations were high immediately after birth and decreased thereafter in both induced and spontaneously delivered foals. Neonatal α cells were also responsive to exogenous arginine but not to endogenous changes in glucose concentrations in both groups of foals. Similar α cell insensitivity to glycaemic changes has been observed in mature horses in response to feeding and glucose administration (Ralston and Baile 1982). In contrast to the β cells, the method of delivery appeared to have little effect on α cell function as basal glucagon concentrations and responses to arginine were similar in the 2 groups of foals. There was an increase in the glucagon response to arginine with increasing age in foals born spontaneously but not by induction, which may reflect developmental changes in the endocrine cell composition of the islets in response to the neonatal hypercortisolaemia seen after induced delivery.

In summary, the normal pattern of neonatal pancreatic endocrine development appears to be altered by induction of labour and the resulting hypercortisolaemia of the newborn foal. Why neonatal cortisol concentrations are persistently high in induced foals is unclear. Development of the HPA axis occurs very close to term in the horse and may be altered post natally when the final stages of prepartum maturation are circumvented by inducing birth 24–48 h earlier than naturally intended (Fowden and Silver 1995). Ontogenic changes in HPA function continue in the days after birth and are known to be influenced by the neonatal environment with consequences for both basal and stimulated cortisol concentrations in the foal (Rossdale et al. 1982; Silver et al. 1984; Hart et al. 2009). There may, therefore, be changes in the hypothalamus, pituitary, adrenal or in the peripheral tissues involved in controlling the plasma binding capacity and clearance of cortisol in the induced foals. Whatever the cause, the high cortisol levels and the changes in pancreatic endocrine function observed in these foals are likely to impair neonatal glucoregulation and the adaptation to enteral nutrition with potential metabolic sequelae long after birth.

Conflicts of interest

No conflicts of interest have been declared.

Source of funding

This work was funded by the Horserace Betting Levy Board.

Acknowledgements

We would like to thank the Horserace Betting Levy Board for their financial support and all the many members of the Departments of Clinical Veterinary Medicine and Physiology, Development and Neuroscience who have helped with these studies.

Manufacturers' addresses

a Mila International, Erlanger, Kentucky, USA.

b Arnolds Veterinary Products, Shrewsbury, Shropshire, UK.

c Sigma, St Louis, Missouri, USA.

d PerkinElmer, Waltham, Massachusetts, USA.

e ICN Biochemicals Inc, Carson, California, USA.

f Systat Software Inc., Point Richmond, California, USA.

Ancillary