| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ANIMAL GROWTH, PHYSIOLOGY, AND REPRODUCTION |
Equine Science Center, Department of Animal Sciences, Rutgers, The State University of New Jersey, New Brunswick 08901
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Key Words: Adiponectin • Diurnal Variation • Ghrelin • Horse • Leptin
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
), impaired glucose tolerance (Hoffman et al., 2003), and other maladies (Lewis, 1995). Conversely, prolonged negative energy balance is related to problems with decreased physical activity, growth rate, and milk production (Lewis, 1995). In the past decade, peripheral metabolic signals such as ghrelin, adiponectin, and leptin have been shown to play a role in energy balance, either through their effects on feed intake, fatty acid metabolism, and/or glucose and insulin regulation (Havel, 2001; Gale et al., 2004). More importantly, these humoral substances may serve as indicators of energy sufficiency, providing information regarding the current state of energy balance. In horses, leptin decreases in response to feed withdrawal (McManus and Fitzgerald, 2000) and also is correlated with BCS (Buff et al., 2002). Adiponectin and ghrelin, however, have not been characterized in horses. Therefore, more data on the physiology of these mediators of energy balance are needed because being either underweight or overweight can be detrimental to the health and performance of horses (Lewis, 1995).
Consequently, the objective of the present study was to measure ghrelin, adiponectin, and leptin concurrently over a 24-h period in Standardbred mares. The acylated form of ghrelin, known as "active ghrelin," was measured because it is considered the biologically active form (Kojima et al., 1999). It was hypothesized that active ghrelin, adiponectin, leptin, glucose, and insulin would demonstrate a diurnal variation entrained to feeding. Characterization of the 24-h profile of these peripheral signals is needed to establish appropriate timing of blood sampling in more complex studies of the control of energy homeostasis in horses.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Twelve unfit Standardbred mares (11 ± 2 yr of age; BW = 500 ± 75 kg) were studied. Horses were housed indoors in 3-m x 3-m stalls at the Rutgers University Equine Research Facility (New Brunswick, NJ) during the experimental period. They were gradually accustomed to being housed in stalls over a 4-wk period before the experiment. The mares were acclimated to the natural light-dark cycle for late October in New Jersey; however, low-intensity light (one light at far end of barn) was used overnight during the experiment to facilitate blood sampling. The experimental diet consisted of grass hay (approximately 10 to 12 kg per horse; as-fed basis) and a commercially pelleted grain mix (Brown’s Feeds, Pennington, NJ; approximately 3 kg per horse; as-fed basis) split into two feedings at 0730 and 1530. The hay portions were large enough for all horses to have hay throughout the entire 24-h period, whereas the grain portions were split equally between the feedings. The lapsed time between feedings was chosen based on its similarity to modern equine management practices and the fact that horses used in the present study had been accustomed to these feeding intervals for several years. The diet was balanced to meet the NRC (1989) requirements for horses at maintenance. The concentrate contained 0.64 Mcal/kg DE, 18.4% CP, 18.0% ADF, 34.5% NDF, 5.4% lignin, 38.2% nonfibrous carbohydrate, 14.3% sugar, and 4.6% crude fat (DM basis; Dairy One Forage Lab, Ithaca, NY). The hay contained 0.37 Mcal/kg DE, 7.1% CP, 42.8% ADF, 64.7% NDF, 8.9% lignin, 21.2% nonfibrous carbohydrate, 12.2% sugar, and 2.2% crude fat (DM basis). All mares maintained body mass within 2 to 3% for 2 wk before the start of the study and had BCS of 5 to 7 (on scale of 1 to 9, where 1 = emaciated to 9 = extremely obese; Henneke et al., 1983). A plain salt block and fresh water were available ad libitum. The Rutgers University Institutional Animal Care Review Board approved all methods and procedures used in this experiment.
Experimental Protocol
Before the experiment, catheters were inserted percutaneously into the left jugular vein, using sterile technique and lidocaine anesthesia. Blood sampling commenced 2 h after catheterization (at 1230) and continued every 2 h for 24 h (1430, 1630, 18301430, 1630, 2030, 2230, 0030, 0230, 0430, 0630, 0830, 1030, and 1230). More frequent blood sampling (every 20 min) occurred 1 h before and after each feeding time to document more acute changes in hormone concentrations around the feedings (1430, 1450, 1510, 1530, 1550, 1610, and 1630; Figure 1). Approximately 30 mL of blood was collected into a syringe and then transferred into three pre-chilled, 10-mL evacuated tubes. Tubes with heparin anticoagulant were used for analysis of glucose and insulin samples, whereas tubes with EDTA anticoagulant were used for analysis of ghrelin, leptin, and adiponectin samples. All blood tubes (Becton Dickinson, Inc., Franklin Lakes, NJ) were immediately centrifuged at 1,500 x g for 15 min, aliquoted, and frozen at –80°C.
|
Plasma active ghrelin and adiponectin concentrations were measured using commercial RIA kits (Linco Research, St. Louis, MO). The adiponectin kit was previously validated for use in horses (Kearns et al., 2005), and linearity and parallelism were established using horse plasma to validate partially the active ghrelin kit used in this study. In the absence of purified equine active ghrelin and adiponectin, results are expressed as human equivalents (HE) of immunoreactive (ir) ghrelin and adiponectin. The intra-assay CV for active ghrelin and adiponectin kits was <10%. The active ghrelin kit utilized 125I-labeled ghrelin with a specific activity of 302 µCi/µg, a guinea pig anti-ghrelin serum, and goat anti-guinea pig IgG serum in the precipitating reagent. Samples were run in duplicate and counted for 1 min in a gamma counter (Packard Instrument Company, Meridien, CT). Data from the manufacturer indicated a specificity of 100% for human, rat, and canine active ghrelin and <0.1% for des-octanylghrelin (total ghrelin). Parallelism of the ghrelin assay kit was established using a serial dilution of horse plasma and the ghrelin standard from the assay kit (Figure 2). Plasma leptin concentrations were measured using a commercial, multi-species leptin kit (Linco Research) previously validated for use in horses (McManus and Fitzgerald, 2000) and demonstrating an intra-assay CV of 8.5%. Plasma insulin and cortisol concentrations were determined using a commercial solid-phase RIA kit (Coat-a-Count, Diagnostic Products Corp., Los Angeles, CA) previously validated for use in horses (Freestone et al., 1991), and plasma glucose concentrations were determined via colorimetric kits (kit number 635, Sigma-Aldrich, St. Louis, MO). The intra-assay CV for the insulin, cortisol, and glucose assays each were <3%. All samples were run in duplicate and within one assay.
|
Body weights were measured using an electronic scale. Body condition scoring was performed by two independent, trained technicians and determined by palpation of body fat using the Henneke et al. (1983) rating system.
Statistical Analyses
Results are expressed as means ± standard error of the mean (±SEM). A one-way ANOVA for repeated measures was used to determine changes in ir-active ghrelin HE, ir-adiponectin HE, leptin, and cortisol. A two-way ANOVA for repeated measures was used to determine time and morning vs. afternoon glucose and insulin responses to the meals, based on a subset of the data from 60 min before to 180 min after feeding. An a priori level of statistical significance was set at P < 0.05 for all tests. Posthoc differences were determined using a Student-Newman-Keul test (Sigma Stat, SPSS, Inc., Chicago, IL).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
demonstrates both linearity and parallelism when serial dilutions of equine ir-active ghrelin HE in plasma are plotted against the standard curve generated from serial dilution of purified standard. Two samples of equine plasma added to 20 and 50 pg/mL of purified active ghrelin standard resulted in 97 and 118% recovery, respectively, of ghrelin standard.
Plasma concentrations of ir-active ghrelin HE were increased (P < 0.001) in the afternoon at 1550, 20 min after the 1530 feeding commenced. This increased concentration was different (P < 0.05) than concentrations of ghrelin at 1230, 2230, 0230, 0830, and 1030 (Figure 3A). Plasma leptin concentrations were greater (P < 0.05) at 1550 (20 min after feeding) and were less (P < 0.05) at 1230, 0630, 0650, 0810, and 0830 (Figure 3B) vs. the other time points examined. Plasma ir-adiponectin HE concentrations varied with time (P = 0.021); however, multiple comparison testing did not elucidate any time points that were different than other samples (Figure 3C). Plasma cortisol concentrations were greater (P < 0.05) during the morning hours than at other times during the 24-h period (Figure 3D).
|
and 5).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
Interestingly, despite two identical meals being fed to the horses on this study, the hormonal and biochemical responses to each meal differed, and demonstrably greater plasma glucose and insulin concentrations were observed after the morning feeding. Reasons for these distinctly different responses may be due to the timing of the meals or some other unknown factors.
The significant peak in plasma ghrelin concentration at 1550 (approximately 20 min after meal initiation) is not in agreement with studies conducted in humans (Cummings et al., 2001, 2004) and sheep (Sugino et al., 2002a,b) that demonstrated a peak in ghrelin before feeding. In these species, ghrelin concentrations increased before a meal and decreased rapidly (within 1 h) after feeding (Cummings et al., 2001; Sugino et al., 2002a,b). The reason for the discrepancy between the human and sheep data and the response observed in the horses of the present study is not clear; however, horses may have a slower ghrelin response to meal anticipation than other species. Another explanation could be that the feeding regimen of feeding free-choice forage in the present study did not allow for substantial meal anticipation. In support of this explanation, a study of ad libitum-fed sheep reported low and relatively constant concentrations of total ghrelin over a 24-h period (Sugino et al., 2002b). Another explanation for the species-related difference may be that active and not total ghrelin was measured in the present study. Ghrelin is only biologically active in its acylated, active form (Kojima et al., 1999), and it has been hypothesized that there may be conditions under which discrepancies arise between measurements of total and active ghrelin in plasma (Ariyasu et al., 2002). The studies demonstrating changes in ghrelin pre- and postfeeding in humans and sheep measured total ghrelin (Cummings et al., 2001; Sugino et al., 2002a,b).
The decreases in ir-active ghrelin HE concentrations from 1550 to 2230 could be attributed to increases in insulin concentration caused by the feeding. This speculation is supported by data from several studies (Saad et al., 2002; Flanagan et al., 2003; Murdolo et al., 2003) of other species that demonstrated an inverse relationship between insulin and ghrelin.
In addition to the previously mentioned observations, a more intensive study of diurnal variation of ghrelin in humans determined that lean individuals showed a nocturnal increase in plasma ghrelin concentration between 0000 and 0600 (Yildiz et al., 2004). The present study did not find a similar nocturnal increase; however, such an increase might have been missed because of less frequent 2-h sampling intervals compared with the human study that used a very rapid sampling frequency (i.e., every 7 min). The half-lives of total and active ghrelin are proposed to be approximately 27 to 31 min and 9 to 13 min, respectively (Akamizu et al., 2004). Therefore, substantial changes in concentration of the hormone could have been missed with the 2-h sampling intervals used in the present study. Another possible contributing factor that may explain species-related differences is that the human subjects were fed three distinct meals and an evening snack, whereas the horses of the present study were fed twice per day and the quantity of hay offered was sufficient to last until the next scheduled feeding time. The continual fermentation of forage in the hindgut of the horse may cause relatively consistent ghrelin concentrations (as discussed subsequently).
The humoral responses to the morning feeding were quite different compared with the afternoon feeding. Ghrelin concentrations did not increase before or after the morning meal (0730), despite the longer overnight period before this feeding. The lack of significant increase may be due to the horses having large amounts of grass hay given at both feeding times, basically providing free-choice access to the hay for the 24-h period. Perhaps the constant fermentation of forage in the hindgut, with its production and absorption of VFA, prevents a regular meal entrained rise in ghrelin as observed in other species. Furthermore, compared with the three distinct meals given to humans in another study (Cummings et al., 2001), only two meals of grain were fed in the present study, along with free-choice hay. This difference could have contributed in part to the apparent lack of a meal-feeding effect. As mentioned previously, sheep fed ad libitum have relatively constant concentrations of total ghrelin (Sugino et al., 2002b). Hence, the herbivore horse may be similar to the ruminant sheep in its regulation of ghrelin. Additionally, ir-active ghrelin HE did not significantly decrease after the morning feeding in this study, despite significantly greater insulin concentrations in response to the morning feeding vs. the afternoon feeding. The reason behind this failure to find a difference is unclear at this time, but it opens the possibility that physiological concentrations of plasma insulin may not be involved in the regulation of plasma ghrelin in horses.
Several studies have shown that increases in plasma leptin can be attributed to feeding in humans (Schoeller et al., 1997), sheep (Marie et al., 2001), and horses (Cartmill et al., 2003). Hence, the significant afternoon increase in leptin may be entrained to the morning feeding. This interpretation is supported by Wagner et al. (2000) that documented a time relationship between changes in plasma insulin and leptin concentrations in humans such that increases in plasma insulin concentration precede increases in leptin concentration by 6 h. Similarly, Cartmill et al. (2003) showed that increases in plasma leptin concentration occur 8 h after increases in insulin because of a single morning meal in horses. This delayed leptin response was similar to the response observed in the present study, where plasma leptin concentration increased approximately 7 h after an observed increase in plasma insulin concentration. Therefore, as supported by the results of Cartmill et al. (2003), it is possible that the increased plasma leptin concentration noted in the afternoon could be attributed to the morning feeding, with its subsequent insulin response.
In contrast to data from several studies in humans (Sinha et al., 1996; Licinio et al., 1998), rodents (Saladin et al., 1995), and horses (Piccione et al., 2004), which demonstrated a significant overnight increase in plasma leptin concentrations, similar results were not observed in the present study. The reason for this difference is not clear; however, it may be related to the fact that the horses in the present study received free-choice hay and only very small grain supplementation, as opposed to larger bolus meals. Interestingly, sheep that graze and ruminate continuously were found to have no nocturnal increases in plasma leptin concentration (Marie et al., 2001). Although the horse is not a ruminant, the role of grazing and hindgut fermentation supports speculation that a more stable continuous uptake of energy rather than fluctuations associated with bolus feeding could be responsible for the relative stability of plasma leptin concentrations.
Interestingly, leptin did not increase significantly after the morning feeding, as it did with the afternoon feeding. If feeding and its subsequent increases in insulin cause a significant increase in leptin concentrations approximately 8 h after insulin peaks, as noted in the study by Cartmill et al. (2003), then the afternoon meal also should have caused an increase in leptin concentrations at approximately 0330; however, this did not occur in the present study. Perhaps the increase in leptin only in the afternoon is in part due to the greater glucose and insulin responses found after the morning vs. the afternoon meal or the longer period between the afternoon and morning meal. It is possible that there is a threshold of insulin concentrations that must be attained to affect leptin concentrations, in which the insulin response to the morning feeding was able to induce a response, whereas the response to the afternoon feeding was too low.
As expected, plasma glucose and insulin concentrations increased in response to the afternoon feeding. The explanation for the lack of a substantial increase in glucose and insulin may be related to the feeding protocol used in this experiment. We have observed larger changes in glucose (30%) and insulin (200%) when identical grain was fed alone. Nonetheless, in those experiments, the grain was given to the same test subjects without hay accompaniment (M. E. Gordon and K. H. McKeever, unpublished data). However, the grain meals in the present study were fed shortly (within a few minutes) after hay was given, which is similar to actual feeding practices used by horse owners. Thus, the concurrent ingestion of the high-fiber hay might have slowed the absorption of glucose in the gastrointestinal tract. Such blunted glucose responses have been observed when fiber-supplemented diets have been given to humans and other mammals (Kay, 1982). Another factor that could have contributed to the blunted glucose/insulin responses could have been the amount of time between the feedings. The morning and afternoon feedings given in the present study were separated by only 8 h. Ralston (2002) hypothesized that the time between feedings and possibly the fasting that occurs before a feeding can affect glucose and insulin responses. Regardless, the blunted glucose/insulin responses observed in the present study are an important finding, especially for the management of horses with obesity and/or insulin resistance that may benefit from controlled glucose responses to feeding.
The significant increase in both glucose and insulin observed in the horses in the present study after the morning feeding was expected. Similar to the findings of S. L. Ralston and H. F. Hintz (unpublished data, Ithaca, NY), changes in glucose and insulin were different between the two feedings. The differential response may be, in part, due to the influence of cortisol concentrations (Plat et al., 1996) and/or a circadian rhythm of insulin secretion (Boden et al., 1996). Both scenarios may explain the responses observed in the horses studied in the present experiment, where cortisol concentrations were found to be greater in the morning. The greater cortisol concentrations in the morning are also in agreement with results from a study by Zolovick et al. (1966), which recorded a diurnal variation in this hormone.
Finally, the present study found no evidence of diurnal variation in ir-adiponectin HE concentrations. This finding agrees with the results of one study conducted in humans that found no diurnal variation in adiponectin in either lean or obese subjects (Yildiz et al., 2004), but it is in opposition to another human study that showed a diurnal variation with a significant decrease in adiponectin at night (Gavrila et al., 2003). Both human studies employed frequent sampling techniques (7 to 15 min) compared with the longer interval used in the present experiment. More research is needed to characterize the diurnal profile of adiponectin further.
In conclusion, ghrelin concentration was greatest after feed ingestion in horses, but only in the afternoon. Therefore, time of day and meal timing should be noted when measuring blood concentration of ghrelin in horses. Additionally, leptin may be entrained to meal feeding that exhibits a larger glucose/insulin response, and time of day also should be noted when measuring plasma leptin. Furthermore, high-fiber hay and diets formulated to use only small amounts of concentrate feeds may help control glucose/insulin responses in horses. Cortisol and time of day may interact with feedings to increase glucose and insulin responses to a meal and should be noted, especially when testing insulin sensitivity in horses.
|
Implications |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
1 Correspondence: 84 Lipman Drive (phone: 732-932-9390; fax: 732-932-6996; e-mail: mckeever{at}aesop.rutgers.edu).
Received for publication January 28, 2005. Accepted for publication June 21, 2005.
|
Literature Cited |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionImplicationsLiterature Cited |
|---|
This article has been cited by other articles:
|
|
 |
|
  P. R. Buff, B. R. Spader, C. D. Morrison, and D. H. Keisler Endocrine responses in mares undergoing abrupt changes in nutritional management J Anim Sci, October 1, 2006; 84(10): 2700 - 2707. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
