Enteroinsular axis response to carbohydrates and fasting in healthy newborn foals

Abstract Background The enteroinsular axis (EIA) comprises intestinal factors (incretins) that stimulate insulin release after PO ingestion of nutrients. Glucose‐dependent insulinotropic polypeptide (GIP) and glucagon‐like peptide‐1 (GLP‐1) are the main incretins. The EIA has not been investigated in healthy neonatal foals but should be important because energy demands are high in healthy foals and dysregulation is frequent in sick foals. Objectives and Hypothesis To evaluate the EIA response to carbohydrates or fasting in newborn foals. We hypothesized that incretin secretion would be higher after PO versus IV carbohydrate administration or fasting. Animals Thirty‐six healthy Standardbred foals ≤4 days of age. Methods Prospective study. Blood was collected before and after a PO glucose test (OGT; 300, 500, 1000 mg/kg), an IV glucose test (IVGT; 300, 500, 1000 mg/kg), a PO lactose test (OLT; 1000 mg/kg), and fasting. Foals were muzzled for 240 minutes. Blood was collected over 210 minutes glucose, insulin, GIP, and GLP‐1 concentrations were measured. Results Only PO lactose caused a significant increase in blood glucose concentration (P < .05). All IV glucose doses induced hyperglycemia and hyperinsulinemia. Concentrations of GIP and GLP‐1 decreased until foals nursed (P < .05), at which time rapid increases in glucose, insulin, GIP, and GLP‐1 concentrations occurred (P < .05). Conclusions and Clinical Importance Healthy newborn foals have a functional EIA that is more responsive to milk and lactose than glucose. Non‐carbohydrate factors in mare's milk may be important for EIA activity. Constant exposure of intestinal cells to nutrients to maintain EIA activity could be relevant to management of sick foals. Foals can be fasted for 4 hours without experiencing hypoglycemia.


| INTRODUCTION
The enteroinsular axis (EIA) comprises intestinal factors (incretins) that stimulate insulin release after PO ingestion of nutrients. Glucosedependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1) are the main incretins. Incretins are secreted by enteroendocrine cells (GIP by K cells in the small intestine; GLP-1 by L cells in the distal small intestine and colon) in response to the PO intake of nutrients (carbohydrates, fats, and amino acids). They bind to their respective cell membrane receptors in pancreatic β-cells to enhance insulin secretion and facilitate glucose disposal. [1][2][3][4] In humans, it has been estimated that GIP and GLP-1 account for 50%-70% of insulin secretion after ingestion of a meal. 5 Incretins also inhibit pancreatic glucagon and somatostatin secretion, maintain β-cell mass, delay gastric emptying, and promote satiety. [1][2][3][6][7][8][9] Insulin is considered a central hormone for energy regulation. It increases cellular glucose uptake, stimulates glycogenesis, promotes lipogenesis, and decreases lipolysis and proteolysis. Although blood glucose is a major stimulus for insulin release in the horse, GLP-1 and GIP contribute to the overall insulin response in healthy and insulindysregulated ponies and horses after ingestion of soluble carbohydrates. [10][11][12][13][14] A strong association between active GLP-1 (aGLP-1) and insulin response to PO nonstructural carbohydrates was documented in ponies. 10 The same study indicated that 22.7% of the variation in insulin concentrations was attributable to variations in aGLP-1 concentrations. 10 Problems of energy dysregulation are frequent in critically ill foals, with up to 70% of foals presenting to a neonatal intensive care unit having a blood glucose concentration outside of the reference range. 15 Both hypoglycemia and hyperglycemia have been associated with nonsurvival in critically ill equine neonates. [15][16][17] A recent study in critically ill foals found that hypoglycemia, hypertriglyceridemia, as well as increased glucagon and decreased insulin concentrations were common findings. 16 Dysregulation of various endocrine factors that contribute to energy metabolism, including growth hormone, ghrelin, insulin-like growth factor-1, adrenocorticotropic hormone, cortisol, leptin, and thyroid hormones has been documented in hospitalized foals. 15,[18][19][20][21] Information describing gastrointestinal endocrine factors that regulate metabolic activity in healthy and sick equine neonates is minimal but relevant based on recent information on the importance of the EIA in human patients suffering from type 2 diabetes mellitus [1][2][3]22 and insulin dysregulation in adult horses. 10,11,23 Considering that energy disturbances and gastrointestinal disorders are common in critically ill foals, but information on intestinal factors that regulate pancreatic endocrine function is lacking, our goal was to investigate the response of the EIA (GIP, GLP-1, and insulin) in healthy newborn foals exposed to both PO and IV glucose, PO lactose, and fasting. We hypothesized that incretin secretion in response to PO carbohydrates would be higher than the same dose of glucose administered IV or when compared to a similar period of fasting. We also hypothesized that this incretin response would be linked proportionately to insulin concentrations. Understanding the biology of the EIA in healthy foals could have clinical implications in the management of foals with energy dysregulation and gastrointestinal disease.

| Experimental design
Thirty-six Standardbred foals ≤4 days of age, owned by a private breeding farm, were included in this prospective, randomized study.
Foals were considered healthy based on physical examination findings, normal CBC, and serum immunoglobulin G (IgG) concentration (>800 mg/dL). Testing took place over two 6-week periods from March to April during the 2017 (n = 17) and 2018 (n = 19) foaling seasons. Each foal was randomly assigned to the dextrose, lactose, or fasted experimental group. Foals receiving dextrose (n = 24) were randomly assigned to PO or IV route of administration of a low, medium, or high dose in a crossover design, with the alternative route of administration occurring the next day. Foals in the lactose and fasted groups were only sampled on day 1. Care was taken to minimize stress on these foals. Foals remained confined to stalls with their mares between sampling times.
At 60 minutes before initiation of the experimental period, foals were muzzled and an IV catheter (SURFLO EFTE IV Catheter 14G × 2 00 , Terumo Medical Corp, Somerset, New Jersey) was placed in the jugular vein using local anesthesia. Foals were manually restrained, and no sedative medications were administered at any point. Foals were muzzled for 240 minutes (−60 to 180 minutes).

| Oral lactose test (OLT)
Foals given lactose (Millipore Sigma, St. Louis, Missouri) as a 20% solution in water at a dosage of 1000 mg/kg (n = 6) by nasogastric intubation over 1 minute.

| Fasted group
A group of foals (n = 6) did not receive any sugar enterally or parenterally and remained muzzled and unable to nurse for the duration of the study (240 minutes).

| Sample analysis
Blood glucose concentrations were measured immediately after collection using a portable glucometer (AlphaTRAK 2 blood glucose monitoring system, Zoetis, Parsippany, New Jersey) previously validated for horses. 24 Commercially available ELISA kits previously validated for horses were used to measure plasma total GIP (EZHGIP-54K, Millipore Sigma) and plasma total GLP-1 (EZGLP1T-36 K, Millipore Sigma) concentrations. 10 Plasma insulin concentrations were measured using a human-specific ELISA (07M-60102, MP Biomedicals, Solon, Ohio) that had linearity at dilutions up to 1:8, inter-and intraassay coefficients of variation of <10% for equine samples, a working range of 1-300 μIU/mL, and a detection limit of 0.75 μIU/mL.

| Data analysis
Data sets were tested for normality using the Shapiro-Wilk normality test and were not normally distributed. Therefore, median and interquartile ranges were calculated. Comparisons among groups were car-

| Study population
Thirty-six Standardbred foals ≤4 days of age were used. The median age of foals at the time of study participation was 24 hours (range, 8-96 hours). Fourteen foals <24 hours of age were included in the study. Three each were used in the 500 mg/kg, 1000 mg/kg and lactose study groups, whereas only 1 foal <24 hours of age was included in the 300 mg/kg group and 4 foals were included in the fasted group. The median IgG concentration was 1700 mg/dL (range, 1295-1957 mg/dL). Twenty-three of 36 foals (64%) were fillies and 13/36 (36%) were colts.

| Blood glucose
Median baseline blood glucose concentration for all foals was 140 mg/dL (range, 123-160 mg/dL; Table 1). The PO administration of 300, 500, or 1000 mg/kg of dextrose did not induce a significant increase in blood glucose concentration at any time point during the testing period when compared to baseline results. In the OGT-300 group, median glucose concentration at 180 minutes was significantly lower than at time 0 (P < .01). Foals in the IVGT-300, IVGT-500, and IVGT-1000 groups had significant increases in blood glucose concentrations at 5-15 minutes compared to time 0 ( Figure 1 and Table 1; P < .01). Foals in the OLT group had a significant increase in blood glucose concentration at 30 minutes compared to time 0 ( Figure 1; P < .05). Fasted foals had a significant decrease in blood glucose concentration from 60 to 180 minutes (P < .05). However, glucose concentrations remained within the reference range, foals were active, and no foal showed signs of hypoglycemia. After foals were allowed to nurse, a significant increase in glucose concentration was evident within 15 minutes (between 195 and 210 minutes; Figure 1; P < .05).
When changes in blood glucose concentrations were adjusted to baseline results, the maximum concentration (which was observed at variable time points) increased by 1. T A B L E 1 Median and interquartile range (IQR) data for blood glucose and insulin in healthy newborn foals administered glucose orally (300, 500, and 1000 mg/kg) and IV (300, 500, and 1000 mg/kg), lactose (1000 mg/kg), and fasted

| Glucagon-like peptide 1 (GLP-1)
Baseline GLP-1 for all foals was 113 pM (range, 57.30-173.5 pM; Table 2). The only group that experienced a statistically significant The glucose-AUC and GIP-AUC were statistically different among study groups (Table 3). Foals in the OGT-300, OGT-1000, and IVGT-1000 groups had significantly higher glucose-AUC compared to fasted foals (P < .05). Insulin-AUC was not different among groups, except for the OLT group, which had a higher AUC compared to the fasted group (P < .05). For GIP, the AUC was not different between treated and fasted foals, but, when comparing equivalent PO and IV doses, the AUC for PO glucose was larger (P < .05). No significant difference for GIP-AUC was found between fasted foals and the other study groups.

| Access to free choice nursing
Access to free choice nursing resulted in a rapid and significant increase in blood glucose (34.6% at 195 minutes and 76% at 210 minutes),

| DISCUSSION
In our study, using PO and IV dextrose, PO lactose, and fasting, we documented that healthy equine neonates have a functional EIA. We also found that the response of the EIA in equine neonates in the immediate postpartum period is highly variable. To our knowledge, ours is the first study to investigate the EIA in healthy newborn foals.
Although the EIA response to lactose was the most evident, we secretion, but IV administration did not. 10,14 In contrast to the minimal incretin response to PO glucose or lactose, once foals were allowed to nurse, a rapid and significant increase in both GIP and GLP-1 concentrations occurred over their respective results at baseline and at 180 minutes. This finding indicates that the EIA in newborn foals is functional and highly responsive to nutrients contained in mare's milk, other than those administered in our study.
Human newborn infants showed a similar GLP-1 response after feeding milk. 28 Mare's milk 1-4 weeks postpartum is composed of approximately 2% fat, 3% protein, and 6% lactose. 29   Another point of interest regarding neonatal endocrinology is the >5-fold increase in GLP-1 and GIP concentrations in the foals of our study compared to results reported using the same assays for adult ponies and horses. [10][11][12][13] This finding indicates that equine enteroendocrine cells have a high capacity to produce incretins in the early neonatal period, perhaps as an evolutionary adaptation to unique components of mare's milk (eg, fatty acids, carbohydrates, and amino acids) and the need for rapid glucose disposal. It could also be that newborn foals have decreased DPP-4 activity compared to horses.
Similar to foals and horses, it has been shown that human infants have higher resting GLP-1 concentrations than adults. 28 The dynamics of incretin secretion over time in healthy foals remain to be investigated but preliminary work from our laboratory indicates that 3-day-old used (50% dextrose versus 20% lactose). Pancreatic β-cell response to glucose is low immediately postpartum compared to foals 5-7 days of age, 36,37 whereas the foals of our study were <4 days of age. A narrower age range of foals would have been ideal but foal access was dictated by the farm. It would be valuable to investigate incretin dynamic changes that occur in the first week after birth. The nutritional management of the mares was not considered in our study. Doing so may be warranted in future investigations because high starch diets fed to gestating mares have been shown to influence glucose and insulin dynamics in their offspring. 38 Considering that under natural conditions foals do not consume pure glucose, in addition to the marked EIA activation observed when foals nursed compared to after PO glucose or lactose, it will be important to evaluate other substrates (eg, amino acids, fats) to better characterize the response of the EIA in equine neonates.
Plasma aGLP-1 (GLP-1  amide and GLP-1 ), the bioactive form of GLP-1 that exists before degradation by DPP-4 into total GLP-1, was not measured in our foals but should be considered in future studies. In people, only 10%-15% of aGLP-1 enters systemic circulation and reaches the pancreas and other organs. 39 Therefore, measuring total GLP-1 (aGLP-1 and metabolites) gives a better estimation of intestinal L-cell secretion of GLP-1. 39 Additionally, quantification of catecholamines and cortisol is warranted in future studies because stress from handling and muzzling potentially could increase these hormones and affect the results of this type of study.
In conclusion, we documented that healthy newborn foals have a functional EIA. Although activation of the EIA was minimal in response to PO glucose and lactose at doses up to 1000 mg/kg, rapid and significant increases in GIP and GLP-1 concentrations were noted when foals were allowed to resume nursing ad libitum. Future research on the EIA in foals should focus on other substrates that may stimulate incretin release and their potential therapeutic implications.