University of Groningen Glucose Homeostasis and Insulin ... · Glucose Homeostasis and Insulin Resistance in veal calves Studies on the effects of age, nutritional modulations and

University of Groningen

Glucose Homeostasis and Insulin Resistance in veal calvesPantophlet, Andre Jonatan

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2018

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Pantophlet, A. J. (2018). Glucose Homeostasis and Insulin Resistance in veal calves: Studies on theeffects of age, nutritional modulations and the applicability of metabolic profiling techniques. [Groningen]:Rijksuniversiteit Groningen.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 07-07-2020

https://www.rug.nl/research/portal/en/publications/glucose-homeostasis-and-insulin-resistance-in-veal-calves(c25d56b2-a7f3-479a-9f41-7cd281fbc57a).html



Glucose Homeostasis and Insulin Resistance in veal calvesStudies on the effects of age, nutritional modulations

and the applicability of metabolic profiling techniques

Andre J. Pantophlet

The studies presented in this thesis were conducted within the framework of the Dutch Carbohydrate Competence Center (CCC2 WP21), and was part of a multidisciplinary collaboration between the VanDrie Group, Tereos Starch & Sweeteners Europe, Wageningen University and the University Medical Center Groningen.

The publication of this thesis was financially supported by the University of Groningen (RUG), Graduate School of Medical Sciences (GSMS) and the University Medical Center Groningen (UMCG).

ISBN: 978-94-6375-181-0 Cover and layout design: Iliana Boshoven-Gkini | www.AgileColor.comPrinted by: Ridderprint | www.ridderprint.nl

Copyright © 2018 by Andre Jonatan Pantophlet. All rights reserved. No part of this thesis may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without permission of the author and when appropriate, the publisher holding the copyrights of the published articles.

Glucose Homeostasis and Insulin Resistance in veal calves

Studies on the effects of age, nutritional modulations and the applicability of metabolic profiling techniques

Proefschrift

ter verkrijging van de graad van doctor aan deRijksuniversiteit Groningen

op gezag van derector magnificus prof. dr. E. Sterken

en volgens besluit van het College voor Promoties.

De openbare verdediging zal plaatsvinden op

Maandag 19 november 2018 om 12.45 uur

door

Andre Jonatan Pantophlet

geboren op 26 februari 1986te Willemstad, Curaçao

PromotorProf. dr. R.J. Vonk

CopromotoresDr. M.G. PriebeDr. J.J.G.C. van den Borne

Beoordelingscommissie Prof. dr. G. van DijkProf. dr. L. DijkhuizenProf. dr. J. Glatz

Contents

CHAPTER 1General Introduction

7

CHAPTER 2Substantial replacement of lactose with fat in a high-lactose milk replacer diet increases liver fat accumulation but does not affect insulin sensitivity in veal calves

27

CHAPTER 3:The use of metabolic profiling to identify insulin resistance in veal calves

47

CHAPTER 4:Lactose in milk replacer can partly be replaced by glucose, fructose, or glycerol without affecting insulin sensitivity in veal calves

63

CHAPTER 5: Insulin sensitivity in calves decreases substantially during the first3 months of life and is unaffected by weaning or fructo-oligosaccharideSupplementation

81

CHAPTER 6: Short communication: Supplementation of fructo-oligosaccharides does not improve insulin sensitivity in heavy veal calves fed different sources of carbohydrates

99

CHAPTER 7: Discussion and Conclusions

107

Summary 117Samenvatting 121List of peer reviewed publications 125Curriculum Vitae 127Acknowledgements 129List of abbreviations 131

CHAPTER 1

General Introduction

“Research is creating new knowledge.” - Neil Armstrong -

CHAPTER 1

8

1

1.1 PhD project outline

Insulin resistance is a key factor in the development of type 2 diabetes, which has become a major health issue worldwide. Problems with insulin resistance and glucose metabolism have not only been observed in human, but also in various animal species including cats, dogs, horses and calves. Studies in these animals can provide important information, which may also be applied in the human situation. Several genetic and environmental factors, such as diet, aging and stress, have been implicated with the development of insulin resistance. Modulations in the diet could therefore potentially prevent the development of insulin resistance and type 2 diabetes.

Milk-fed calves raised for white veal production (i.e. veal calves) are fed a milk replacer (MR), roughage and concentrates. The vast majority of the digestible nutrient intake (60-70%) originates from the MR, which contains high amounts of lactose and fat (~45% and 35% of the digestible energy intake, respectively). Heavy veal calves (> 4 months of age) often develop problems with glucose homeostasis (Hostettler-Allen et al., 1994; Hugi et al., 1997), which lead to urinary glucose excretion (Vicari et al., 2008; Labussiere et al., 2009b) and hepatic steatosis (Gerrits et al., 2008), and could possibly develop into pre-diabetes. This might be attributed to the persistently high intakes of lactose from the MR (Hugi et al., 1998). Replacement of the lactose in the MR by other energy sources may improve glucose homeostasis and insulin resistance, which may stimulate efficient use of nutrients for growth processes and improve (metabolic) health. In addition, the price of lactose is subject to extremely large fluctuations, therefore providing also an economic incentive to replace lactose by alternative energy sources. Furthermore, studies in calves may increase our insights in the underlying mechanisms involved in the development of insulin resistance.

Therefore, in this project, the main aims were: 1) to increase the understanding of the pathology of insulin resistance in veal calves and to enable the development of feeding strategies to reduce the incidence of insulin resistance, and 2) to find a suitable substitution for a substantial part of the lactose in MR diets, without compromising nutrient digestion/fermentation and metabolic health in veal calves. The current PhD thesis focuses on the first aim. A second thesis has been devoted to the second aim (Gilbert, 2015).

The project was conducted within the framework of the Dutch Carbohydrate Competence Center (CCC), and was part of a multidisciplinary collaboration between the VanDrie Group, Tereos Starch & Sweeteners Europe, Wageningen University and the University Medical Center Groningen.

General introduction

9

1

1.2 Glucose metabolism and insulin resistance

1.2.1 Glucose homeostasis

Glucose is one of the most important energy sources for the body. The homeostasis of blood glucose is a complex process involving several tissues and receptors, hormones and co-regulators (Barros et al., 2009). In the fasted state, endogenous glucose production (predominately by the liver) ensures that sufficient glucose is available for body tissues and especially the brain. In the fed state, glucose is mainly derived from exogenous sources. Glucose homeostasis is maintained by pancreatic insulin, specifically produced by the islets of Langerhans. The blood glucose level stimulates pancreatic β-cells to secrete insulin. When blood glucose levels rises, insulin is secreted to stimulate glucose uptake primarily in skeletal muscle and adipose tissues. Also, insulin promotes glycogenesis (storage of glucose as glycogen) and inhibits glucose production via glycogenolysis (breakdown of glycogen to glucose) and gluconeogenesis (biosynthesis of glucose from non-carbohydrate precursors). In addition, insulin also promotes de novo lipogenesis from glucose (Stanfield, 2012; Röder et al., 2016). The actions of insulin on different tissues are summarized in Figure 1.1.

When blood glucose levels drop below a certain level, glucagon, a counter-regulatory hormone to insulin produced by pancreatic α-cells, is secreted to promote glycogenolysis and gluconeogenesis, to help maintain glucose homeostasis (Röder et al., 2016).

Several other hormones, such as the incretin hormones glucose-dependent insulinotropic polypeptide (GIP) and glucagon-like peptide-1 (GLP-1), amylin, adrenalin, cortisol, leptin, adiponectin, insulin-like growth factor and growth hormone, are known to be involved in the regulation of glucose homeostasis (Sherwin et al., 1980; Denroche et al., 2012; Eelderink et al., 2012a; Eelderink et al., 2012b; Hayes et al., 2014; Pantophlet et al., 2016). This illustrates the complexity of glucose homeostasis.

1.2.2 Insulin signaling pathways for glucose transport stimulation

One of the most important actions of insulin is the stimulation of glucose uptake, which occurs through a cascade of signaling events (Figure 1.2). Initially, insulin binds to the insulin receptor, a protein embedded in the cell membrane consisting of two extracellular α subunits and two transmembrane β subunits (Kanzaki, 2006). Binding induces a conformational change that activates tyrosine kinase activity of the β subunits. At least two discrete intracellular signaling pathways have been identified (Bryant et al., 2002; Watson et al., 2004; Kanzaki, 2006).

CHAPTER 1

10

1

Figure 1.1 | Actions of insulin on target tissues. Figure adapted from (Stanfield, 2012).

At first, a number of intracellular proteins, which include insulin receptor substrates (IRSs), growth factor receptor-bound protein 2 associated-binding protein 1 (GAB1) and Shc are phosphorylated (activated), providing binding sites for downstream signaling molecules. Among them, phosphoinositide 3-kinase (PI3K) has a major role because it activates Protein kinase B (PKB; also known as AKT) and Protein kinase C (PKC), which leads to the translocation of glucose transporter type 4 (GLUT4) to the cell membrane to promote glucose uptake. Downstream targets of these kinases resulting in GLUT4 translocation are unclear. In addition, this pathway also stimulates glycogen synthesis and lipogenesis (Fukushima et al., 2010; Guo, 2014).

In the second pathway, activation (phosphorylation) of the insulin receptor leads to the phosphorylation of the proto-oncogene Casitas b-lineage lymphoma (c-Cbl), which is in a complex with the adaptor protein CAP. This Cbl-CAP complex then translocates to lipid rafts at the cell membrane, and recruits a complex of Crk, an adaptor protein, and C3G, a guanine nucleotide exchange protein, into lipid rafts. C3G activates the GTP-binding protein family, TC10, which leads to the translocation of GLUT4 to the cell membrane. The downstream signaling steps of this TC10-mediated translocation of GLUT4 are not yet clear (Bryant et al., 2002).

1.2.3 Glucose uptake

The transport of glucose across the cell membrane is important in reducing blood glucose levels, maintaining glucose homeostasis and yielding cells with sufficient energy. Glucose is a polar molecule which cannot pass the bilayer lipid cell membrane on its own. Glucose transport is facilitated by glucose transporters (GLUTs), which allow transport of glucose down its concentration gradient. To date, a total of 14 GLUTs (i.e. GLUT 1-14) have been identified (Carvalho et al., 2011; Pyla et al., 2013), each with different kinetics and efficiency of glucose (and hexose) transport, and with tissue-specific distribution, which is species dependent (Thorens and Mueckler, 2010).


11

1

Also, the expression of these transporters is differentially regulated. Therefore, each transporter plays a different role in the promotion of glucose uptake in various tissues and the regulation glucose homeostasis.

GLUT1-4 are the most well established GLUTs (Thorens and Mueckler, 2010). GLUT1 is widely distributed over various tissues, in humans most abundant in erythrocytes and in the brain. It stimulates glucose uptake independently of insulin, and is also capable of transporting other hexoses such as mannose and galactose. GLUT2 is the major transporter in hepatocytes. It is also expressed in the intestinal mucosa, renal tubules, pancreatic β-cells and in the brain. Its affinity for glucose is low compared to GLUT1. It is also capable of transporting mannose, galactose and fructose. GLUT3 is most abundant in the brain, but is also expressed in other tissues, in a species-specific matter. It has a high affinity for glucose compared to GLUT1, it operates independent of insulin, and is also capable of transporting galactose, mannose, maltose and xylose. GLUT4 is the main transporter in insulin-sensitive cells and tissues, such as skeletal muscles cells, heart cells and adipose tissue, and is therefore extremely important in the regulation of post-prandial glucose homeostasis (see previous section). Its affinity to glucose is similar to that of GLUT1. For further reading on the characteristics and tissue-specific distribution of the different GLUTs, please see (Zhao and Keating, 2007; Thorens and Mueckler, 2010).

1.2.4 Insulin resistance

Insulin is important for the promotion of glucose uptake/utilization in skeletal muscle, adipose tissue and various vital organs. Insulin resistance is defined as a decreased ability of insulin to promote glucose uptake in insulin-sensitive cells and tissues. Pronounced (i.e., chronic) hyperglycemia, hyperinsulinemia and glucosuria are therefore indications of insulin resistance (Petersen and Shulman, 2006; Ye, 2013). Insulin resistance is often distinguished into hepatic insulin resistance and peripheral insulin resistance. Hepatic insulin resistance refers to the failure of insulin to adequately suppress hepatic endogenous glucose production, whereas peripheral insulin resistance refers to failure of insulin to adequately promote glucose uptake in peripheral tissues (e.g. skeletal muscle and adipose tissues). The causes of insulin resistance are not clear, but insulin resistance is considered a multi-factorial disorder, resulting from nutritional, physiological, genetic and environmental factors (Pedersen, 1999; Li et al., 2013; Roberts et al., 2013).

CHAPTER 1

12

1

Figure 1.2 | Insulin signaling pathways for glucose transport (Grusovin and Macaulay, 2003). At least two pathways have been implicated in insulin-stimulated glucose transport. In the first pathway, insulin binds to the insulin receptor, resulting in tyrosine phosphorylation and docking of insulin receptor substrates (IRSs). IRS provides binding site for several proteins including phosphoinositide 3-kinase (PI3K), which activates Protein kinase B (PKB) and Protein kinase C (PKC), resulting in the translocation of glucose transporter type 4 (GLUT4) to the cell membrane to promote glucose uptake. In the second pathway, insulin binds to the insulin receptor, resulting in tyrosine phosphorylation and docking of Casitas b-lineage lymphoma (Cbl) which is in a complex with the adaptor protein CAP. This Cbl-CAP complex translocates to lipid rafs, which recruits the adaptor protein Crk and C3G, a guanine nucleotide exchange protein. G3G activates the GTP-binding protein family TC10, which leads to translocation of GLUT4 to promote glucose uptake. The downstream signaling steps of this TC10-mediated GLUT4 translocation are not yet clear.

Insulin sensitivity can be influenced by diet composition. Chronic excessive energy consumption promotes insulin resistance through various processes, which include the stimulation of insulin secretion (which can lead to hyperinsulinemia), triglycerides synthesis and fat accumulation (Bessesen, 2001; Wilcox, 2005).

Over the last two decades, effects of several macronutrients on insulin resistance have been studied extensively. Especially high dietary saturated fat intake has been associated with insulin resistance (Wilcox, 2005; Weickert, 2012). Not all types of dietary fat negatively affect insulin sensitivity (Gadgil et al., 2013). Omega-3 fatty acids, for example, have shown to have a positive effect on insulin sensitivity, whereas saturated and trans fatty acids have been associated with insulin resistance (Rivellese et al., 2002). The fatty acid composition might play an important role on the long-term development of insulin resistance, via effects on cell membrane composition (as fatty acids are important components of cell membranes), cell signaling and membrane fluidity.


13

1

For carbohydrates, the glycemic index (GI) has been introduced in an attempt to rank carbohydrates-containing diets based their glycemic response in blood. Postprandial hyperglycemia and hyperinsulinemia have been associated with the development of insulin resistance. Therefore, diets with a low GI may reduce the risk of development of insulin resistance. Evidence on fructose (low GI), however, has shown that this is not always the case, as high-fructose diets have also been associated with insulin resistance (Bessesen, 2001; Balakumar et al., 2016). Many low GI diets are rich in fiber. Dietary fibers indirectly affect insulin secretion and action through effects on transit time, gut mobility, gastrointestinal hormone secretion (e.g. GIP and GLP-1) and colonic fermentation products (Wilcox, 2005). Starches that are not digested in the small intestine but fermented in the colon (called resistant starch) have been associated with decreased postprandial glucose (low GI) and insulin response, and improved whole-body insulin sensitivity (Robertson et al., 2005; Johnston et al., 2010). Pre-biotic fibers such as short-chain fructo-oligosaccharides (scFOS) selectively modulate the composition of the gut microbiota (Respondek et al., 2013). The supplementation of these fibers has been associated with improved glucose metabolism and insulin sensitivity in various animal species (Kaufhold et al., 2000; Respondek et al., 2008; Respondek et al., 2011; Respondek et al., 2013). The exact mechanisms are however not known and are currently under investigation.

Some dietary amino acids have an insulinotrophic effect and thus stimulate insulin secretion, which in turn stimulates glucose uptake. However, some proteins also stimulate glucagon secretion and gluconeogenesis. There is increasing evidence suggesting that long-term high-protein intake may lead to insulin resistance and type-2 diabetes (Sluijs et al., 2010; Mei et al., 2014). Especially branched-chain amino acids (i.e., leucine, iso-leucine and valine) have been associated with insulin resistance in human and rodents (Lynch and Adams, 2014; Yoon, 2016). The role of branched-chain amino acids in the development of insulin resistance is unclear, as it is not known whether branched-chain amino acids are causative agents or biomarkers of insulin resistance (and type-2 diabetes). It is been hypothesized that branched-chain amino acids mediate activation the mammalian target of rapamycin complex (mTOR) 1, which affects insulin signaling at early stage (Lynch and Adams, 2014). Further research is needed.

1.2.5 Physiological causes of insulin resistance

There are several physiological factors that may be involved in the development of insulin resistance. First, a decrease in the number of insulin receptors in skeletal muscle cells, and adipose tissue may cause insulin resistance (Accili et al., 1989; Taylor et al., 1990). In humans and rodents the number of insulin receptors decreases with age (Pagano et al., 1981; Torlińska et al., 2000). Nutritional factors may also affect the number of insulin receptors. For example, a decreased number of insulin receptors in skeletal muscle cells was found in insulin resistant calves fed a high lactose MR diet,

CHAPTER 1

14

1

compared to a standard lactose MR diet (Hugi et al., 1998). Hyperinsulinemia affects the affinity (but not the number) of insulin receptor in skeletal muscle cells (Soman and DeFronzo, 1980). Second, a reduction in the number of GLUT4 or impaired translocation of GLUT4 in skeletal muscle cells and adipose tissue may also cause insulin resistance. In a study with rats, translocation of GLUT4 in skeletal muscle cells and adipose tissue was reduced in insulin resistant rats (Cartee et al., 1993). A mice study showed that disruption of GLUT4 selectivity in skeletal muscle cells induces glucose intolerance and insulin resistance (Zisman et al., 2000). Third, defects in post-receptor insulin signaling can cause insulin resistance (Miura et al., 2001). For example, when insulin binds to its receptor it activates IRSs, which in turn activates the PI3K/AKT pathway ultimately resulting in translocation of GLUT4 to promote glucose uptake. PI3K/AKT, however, also activates the mTOR/S6 kinase pathway, which in turn causes phosphorylation and degradation of IRSs, and impairs insulin signaling. Therefore, hyperactivation of mTOR can cause insulin resistance (Blagosklonny, 2013). The mTOR/S6 kinase pathway is also activated by macronutrients. In humans and rodents, glucose, fatty acids and amino acids have all shown to activate mTOR, possibly resulting in insulin resistance. Therefore, low caloric diets may reduce the risk of developing insulin resistance. Several other physiological factors are also implicated in insulin resistance, which include inflammation and oxidative stress (Hotamisligil, 2006; Shoelson et al., 2006; Park et al., 2009).

1.3 Methods for assessing insulin sensitivity and glucose homeostasis

Several methods are available to assess insulin sensitivity/resistance and glucose homeostasis in humans and animals. These methods can be divided into direct and indirect methods. Each method has its own advantages and disadvantages. Factors such as invasiveness, reproducibility and validity as well as costs, the necessary expertise and qualified personnel play an important role when choosing a suitable method. In this section we will discuss the most frequently used methods.

1.3.1 Euglycemic hyperinsulinemic clamp

The euglycemic-hyperinsulinemic clamp technique, originally developed by DeFronzo et al. (1979), is considered as the “golden standard” for the assessment of insulin sensitivity (Muniyappa et al., 2008). This method is usually implemented after an overnight fasting period of at least 12 h, to achieve constant glucose turnover, and to prevent disturbances from pancreatic insulin production and exogenous glucose derived from the diet. Two catheters are placed, one for glucose and one for insulin infusion. First, insulin is infused to increase blood insulin to hyperinsulinemic levels (typically ~ 100 mU/L), which inhibits pancreatic insulin production. The infusion is kept constant for 2 to 4 h. The infusion of insulin causes a drop in blood glucose


15

1

levels. To maintain euglycemic levels glucose is infused. By measuring the glucose level regularly (5-15 min), the glucose infusion rate can be adjusted to reach a steady-state. The glucose infusion rate at steady-state (corrected for body weight) divided by the blood insulin level at steady is a measure of whole-body insulin sensitivity. The higher the amount of glucose infusion needed to compensate for glucose clearance from the blood, the more sensitive the body tissues are to insulin. The main advantage of the euglycemic-hyperinsulinemic clamp technique is that it directly measures whole-body glucose disposal at a given level of insulinemia under steady-state conditions. Also, the approach is conceptually straightforward. The main limitations of this technique are that it is time-consuming, expensive and labor-intensive and requires experienced personnel. Hence, for large studies (e.g. epidemiological and large clinical or non-clinical studies) this technique is not feasible (Buchanan et al., 2010).

1.3.2 Frequently sampled intravenous glucose tolerance test

The frequently sampled intravenous glucose tolerance test, developed by Bergman et al. (1979), is considered the “silver standard” for the assessment of insulin sensitivity. The test is an indirect method for the assessment of insulin sensitivity, and uses blood glucose and insulin levels obtained during a frequently sampled intravenous glucose tolerance test to calculate insulin sensitivity. After overnight fasting, an intravenous bolus of glucose (~ 0.3 g/kg body weight) is infused at t= 0 min. At t= 20 min an intravenous bolus of insulin (~ 0.03 IU/kg body weight) is infused. Blood samples are taken between t= -10 and 180 min, and glucose and insulin concentrations are determined. These data are subsequently subjected to minimal model analysis, which is a mathematical model that integrates glucose–insulin kinetics and their relationship. The model is embedded in the computer program MINMOD, and generates an index of insulin sensitivity (Pacini and Bergman, 1986; Boston et al., 2003). The main advantage of the frequently sampled intravenous glucose tolerance test is that, in addition to the insulin sensitivity index, several other parameters can be estimated, which include glucose effectiveness and β–cell function (insulin secretion). Compared to the euglycemic-hyperinsulinemic clamp it is slightly less labor-intensive, less expensive, does not require steady-state conditions, or experienced personnel to maintain constant insulin infusion. The main limitations of this technique are that it is labor-intensive, as it requires multiple blood sampling over a 3-h period, and also a special software package is needed. The model itself oversimplifies the physiology of glucose homeostasis (Muniyappa et al., 2008).

1.3.3 Insulin suppression test

The insulin suppression test is another direct method for the assessment of insulin sensitivity, and was originally developed by Shen et al. (1970). After overnight fasting, somatostatin (250 g/h) is intravenously infused to suppress endogenous insulin production. Simultaneously, glucose and insulin are infused at ~ 0.06 g/kg body weight/min and 0.05 IU/min, respectively, for 180 min at constant rate. Blood samples

CHAPTER 1

16

1

are taken every 30 min for the first 150 min for the determination of glucose and insulin. Between 150-180 min steady-state is assumed. During steady-state blood is collected every 10 min. The steady-state plasma insulin level is generally similar (but not always) among subjects. Therefore, the steady-state plasma glucose level gives an estimation of insulin sensitivity. A higher steady-state plasma glucose level indicates lower tissue (but not whole-body) insulin sensitivity. The main advantage of the insulin suppression test is that it directly measures tissue insulin sensitivity (no mathematical models needed) and is highly reproducible (Workeneh et al., 2010). Also, it is less labor-intensive and less technically demanding than the euglycemic-hyperinsulinemic clamp (as it does not require variable infusions). The main limitations of this technique are that it is expensive, requires experienced personnel (although less technically demanding than the euglycemic-hyperinsulinemic clamp), and there is a risk of hypoglycemia in insulin-sensitive subjects (Muniyappa et al., 2008). Also, similar to the euglycemic-hyperinsulinemic clamp this technique is not feasible for large studies.

1.3.4 Oral glucose/meal tolerance test

The oral glucose/meal tolerance test is commonly used to assess glucose homeostasis. In a clinical setting it is frequently used to diagnose glucose intolerance and type 2 diabetes (Tuomilehto, 2002). After overnight fasting, a standard glucose load or standard meal is given, and blood samples are collected at baseline and in regular intervals after consumption for the determination of blood glucose and insulin. These determinations are used to assess glucose homeostasis. Glucose and insulin dynamics in the oral/meal tolerance test mimic physiological conditions more closely than the euglycemic-hyperinsulinemic clamp, insulin suppression or the frequently sampled intravenous glucose tolerance test. Also, the oral/meal tolerance test is cost-effective and simple to execute. Although the test provides information useful for assessing glucose homeostasis/tolerance, it does not assess insulin sensitivity per se. Also, reproducibility can be a problem as gastric emptying, glucose absorption, splanchnic glucose uptake and incretin effects can vary significantly, even within the same individual. In cases when only insulin sensitivity is of interest, simple surrogate indexes of insulin sensitivity/resistance can be a good option.

1.3.5 Simple surrogate indexes of insulin sensitivity/resistance

The simple surrogate indexes of insulin sensitivity/resistance can be divided in two categories; 1) indexes derived from steady-state conditions, and 2) indexes derived from dynamic tests.

Indexes derived from steady-state conditions are determined from a single blood sample collected after overnight fasting. During fasting, blood glucose is homeostatically maintained within normal ranges by an equilibration between hepatic glucose production, pancreatic insulin secretion, and whole-body glucose uptake. Therefore, the relationship between fasting blood glucose and insulin levels reflects insulin


17

1

sensitivity/resistance. Steady-state indexes of insulin sensitivity/resistance include the homeostasis model assessment (Matthews et al., 1985) and the quantitative insulin sensitivity check index (Katz et al., 2000). Both methods are similar, as they are based on the same physiological principle, with the quantitative insulin sensitivity check index being the inverse logarithm of the homeostasis model assessment. Both methods have proven to be practical and useful in epidemiological studies. Furthermore, both methods have shown to correlate well with the euglycemic-hyperinsulinemic clamp in several studies (Katz et al., 2000; Chen et al., 2003; Tam et al., 2012). These indexes are simple, inexpensive, minimally invasive and can be applied in almost every setting.

Other surrogate indexes use information derived from dynamic tests such as the oral glucose or meal tolerance test. These indexes include the Stumvoll index (Stumvoll et al., 2000), Matsuda index (Matsuda and DeFronzo, 1999), Gutt index (Gutt et al., 2000), Avignon index (Avignon et al., 1999) and the Belfiore index (Belfiore et al., 1998), and have been validated against the euglycemic-hyperinsulinemic clamp in humans in several studies (Matsuda and DeFronzo, 1999; Gutt et al., 2000; Mari et al., 2001). For animals, however, the validity of these indexes has not been assessed extensively. Regardless, poor reproducibility of the oral glucose/meal tolerance test due to variability gastric emptying, glucose absorption, splanchnic glucose uptake and incretin effects limits the use of these indexes in practice. Also, fasting indexes are usually preferred because they are more cost-effective and less labor-intensive than dynamic tests.

1.4 Metabolomics

Metabolomics (metabonomics or metabolic profiling) is the analysis of all small-molecule metabolites (typically ≤ 1000 m/z) in body fluids, tissues and cells. In contrast to genes and proteins, the metabolome reflects true biological endpoints of a condition (Figure 1.3). Using highly advanced analytical techniques, such as Mass spectrometry and Nuclear magnetic resonance spectroscopy coupled to (ultra)-High-performance liquid chromatography, Gas chromatography or Capillary electrophoresis, metabolomics focuses on identifying and monitoring alterations in metabolite profiles as a result of pathology, nutritional interventions, drug interventions, genetic or environmental factors. Its potential has been demonstrated in various studies (Jansson et al., 2009; Ametaj et al., 2010; Jung et al., 2013). In humans, several markers of insulin resistance have been discovered using this approach (Gall et al., 2010). These include, for example, α-hydroxybutyrate, linoleoyl-glycerophosphocholine, glycine, creatine and branched chain amino acids. Based on 26 of these marker metabolites a model (Quantose algorithm) was developed that can predict insulin resistance (Cobb et al., 2013).

The main advantage of metabolomics is that one can obtain information of hundreds of metabolites from a very limited volume sample. This information can be used to discover and highlight metabolites and pathways that are related to a patho-

CHAPTER 1

18

1

physiological development for example. Metabolomics can be used in many fields, such as biomarker discovery, diagnostic research, nutritional research, toxicology and drug discovery. However, it is relatively expensive and requires special apparatus and experienced personnel. In addition, the metabolome is complex both in terms of chemical diversity (i.e., polarity, solubility, molecular weight, volatility, etc.) and dynamic range (varying in several orders of magnitude). Also, the number of metabolites present in different organisms is still unknown. For humans, for example, more than 8,000 metabolites have been identified so far according to the human metabolome database (Wishart et al., 2013). In the plant kingdom, over 200,000 metabolites are estimated (Weckwerth, 2003; Saito and Matsuda, 2010). Due to the great dynamic range and chemical diversity it is currently infeasible to analyze all metabolites using one analytical platform. Therefore, in practice, multiple analytical platforms are used when feasible. This increases the power of metabolomics. However, even so, analyzing the complete metabolome remains a great challenge.

Figure 1.3 | Schematic representation of the “omics” cascade. The metabolome reflects true biological endpoints and is therefore closest to the phenotype.


19

1

1.5 Veal calf production and nutrition

Meat from young calves (< 8 months of age) is called veal. The most common type of veal originates from calves which are typically up to six months of age, and have been fed mainly on milk replacer which is typically low in iron. The meat is therefore sometimes called ‘white veal’. Each year, around six million calves are raised for veal in the EU-28 countries (Source: Eurostat), with the majority raised for ‘white veal’ production.

In practice, male dairy and female calves are purchased from dairy farms at approximately 14 days of age and are transported to specialized gathering facilities, were they are mixed and re-grouped. Then, calves are transported to specialized veal farms where they are housed individually during the first 4-6 weeks to allow individual monitoring. Thereafter, calves are group-housed until slaughter age (approximately seven months in the Netherlands).

Veal calves, in contrast to ruminating (dairy) calves, are fed MR until slaughter, to allow a high average daily weight gain as well as a typical meat quality (characterized by its paleness and tenderness). The MR contains highly digestible nutrients such as ~50% lactose, ~20% crude fat and ~20% crude protein. In this composition, energy is mainly provided by lactose and crude fat (~45% and 35% of metabolizable energy intake). Although it is nowadays mandatory to feed veal calves with solid feed (i.e., concentrates and roughages), carbohydrates and fat from MR still contribute approximately 75% to the digestible energy intake (Labussiere et al., 2009a). Upon ingestion of the MR, the esophageal groove closes, allowing the MR to by-pass the reticulorumen (Abe et al., 1979). As a result, the MR flows directly into the abomasum, and from there into the small intestine (Figure 1.4). Once in the small intestine, nutrients are enzymatically hydrolyzed and absorbed, like in monogastric animals. In contrast, calves raised for beef production or breeding, between 4-6 months of age are predominantly fed concentrates and roughages (Blum and Harmon, 1999). These solids enter the reticulorumen, where they are fermented by the microbial population. Therefore, these calves operate as true ruminants.

Figure 1.4 | Schematic representation of the milk flow in milk-fed calves. In milk-fed calves, the esophageal groove closes, allowing the milk to by-pass the recticulorumen and flow directly into the abomasum. Source: Merrick Animal Nutrition, Inc.

CHAPTER 1

20

1

1.6 Glucose homeostasis and insulin resistance in veal calves

Veal calves often show disturbances in glucose homeostasis and insulin sensitivity. These problems have been identified in heavy veal calves (4-6 months of age; in the second phase of the fattening period), and are characterized by high incidences of hyperglycemia, hyperinsulinemia and urinary glucose excretion (Wijayasinghe et al., 1984; Palmquist et al., 1992; Hostettler-Allen et al., 1994; Hugi et al., 1997; Hugi et al., 1998). Also, hepatic steatosis (excessive accumulation of fat in the liver) has also been reported (Gerrits et al., 2008). In heavy veal calves, blood glucose and insulin increase excessively after feeding and remain high, not returning to basal levels even after 4-5 h (Hugi et al., 1997; Vicari et al., 2008). In healthy humans, blood glucose and insulin typically return to basal levels within 2-3 h after feeding (Eelderink et al., 2012b). The symptoms observed in veal calves resemble those commonly observed in humans suffering from type 2 diabetes.

The causes behind disturbed glucose homeostasis and insulin sensitivity in veal calves are not clear. The etiology of insulin resistance in veal calves may be complex and multifactorial (Hostettler-Allen et al., 1994; Gerrits et al., 2008). The problems with glucose homeostasis and insulin sensitivity in veal calves appear to be age-dependent (Hugi et al., 1997). These problems have been identified especially in calves at end of their fattening period. This could perhaps be due to the high lactose content in veal calf diets. One study in veal calves has suggested that insulin sensitivity is modulated by supplemental lactose, in an age-dependent manner (Hugi et al., 1998). Decreasing the amount of lactose in veal calf diets will lead to reduced postprandial glucose and insulin responses, which could potentially reduce (or prevent) problems with glucose homeostasis and insulin sensitivity in veal calves. Apart from lactose, also the high fat content in veal calf diets might be causing problems with glucose homeostasis and insulin sensitivity. In humans, a high fat intake was found to be independently correlated with insulin resistance. Weaned non-ruminant animal species (e.g. pigs) do not suffer from insulin resistance, despite their much higher glucose (but lower fat) intake than veal calves and humans, which indicates that interactions between dietary fatty acids and glucose may be crucial for disturbing glucose metabolism in veal calves. Another factor that could affect glucose homeostasis and insulin sensitivity in veal calves is the discrepancy between the diet of veal calves and their ontogenetic background. In nature, calves between 4 and 6 months of age are grazing and plant fragments are fermented in the rumen along with the production of short-chain fatty acids as a major energy source. Therefore, veal calves, which are ontogenetic ruminants, may not be equipped with the genetic capacity to keep dealing with large amounts of dietary lactose adequately, when they become older.


21

1

1.7 Scope of this thesis

The development of insulin resistance is not only a problem in human, but also in various animal species and might lead to type 2 diabetes.In this project, we focused on the understanding of changes in glucose homeostasis and the development of insulin resistance in veal calves, and on quantifying the influence of several dietary modulations.

We studied in particular:1. The age-related development of insulin resistance 2. The effects of partial replacement of the lactose in calf MR by other energy

sources, such as glucose, fructose and glycerol, on glucose homeostasis and insulin resistance

3. The effects of dietary scFOS on glucose homeostasis and insulin resistance 4. The applicability of metabolic profiling techniques to find metabolites

(biomarkers) and pathways related to insulin resistance

The following research questions were addressed:

1. Are insulin sensitivity and glucose homeostasis in veal calves affected by the replacement of a substantial amount of the lactose in a calf milk replacer by fat?Lactose and fat are the main energy sources in veal calf nutrition. However, the possible contribution of these dietary energy sources to a deteriorated glucose homeostasis and insulin resistance is currently unknown. Standardized studies in which lactose and fat are exchanged (iso-energetically) may reveal the contribution of these dietary energy sources to the development of insulin resistance in calves. Therefore, an experiment was designed to compare effects of a high-lactose vs. a high-fat MR on glucose homeostasis and insulin sensitivity in heavy veal calves (Chapter 2).

2. Can plasma metabolomic profiling techniques be used to study and identify insulin resistance in veal calves?Approximately 50% of heavy veal calves on a high-lactose or high-fat MR diet develop insulin resistance. Therefore, it is worthwhile to study the patho-physiological mechanisms behind insulin resistance, and detect biomarkers that perhaps can be used to detect insulin resistance (or a decrease in insulin sensitivity) at an early stage. By detecting insulin resistance at an early stage, management and feeding strategies could perhaps be adopted to prevent this problem. Therefore, using metabolic profiling techniques, we attempted to discover pathways and markers that are associated with insulin resistance in veal calves (Chapter 3).

CHAPTER 1

22

1

3. Are insulin sensitivity and glucose homeostasis in veal calves affected by the replacement of a substantial amount of the lactose in a calf milk replacer by glucose, fructose or glycerol?Glucose, fructose and glycerol are possible alternatives to lactose. Especially fructose and glycerol are expected to positively affect glucose homeostasis, and possibly improve insulin sensitivity. Therefore, an experiment was designed to study the effects of replacing a substantial amount of the dietary lactose in the MR by glucose, fructose or glycerol on glucose homeostasis and insulin sensitivity in veal calves (Chapter 4).

4. Is the age-dependent development of insulin resistance influenced by a strong contrast in feeding strategy (i.e., weaning compared to milk replacer feeding)?Insulin sensitivity decreases in calves during the first months of life. Yet, it remains unclear whether the decrease in insulin sensitivity can be influenced by strong contrast in feeding strategy (i.e., prolonged MR feeding vs. progressive weaning), or is explained by the ontogenetic development of calves. Therefore, an experiment was designed to assess age-related and diet-induced (i.e., MR only vs. progressive weaning) changes in glucose homeostasis and insulin sensitivity in calves during the first three months of life (Chapter 5).

5. Can supplementation of scFOS improve insulin sensitivity and glucose homeostasis in young veal calves?Studies in various animal species have shown that dietary scFOS improved whole body insulin sensitivity. In dogs and horses with obesity, for example, an increase in insulin sensitivity was measured after feeding FOS for a period of 6 weeks. Whether FOS supplementation can prevent the decrease/or improve insulin sensitivity in veal calves is not known. Therefore, an experiment was designed to assess the effect of FOS supplementation on insulin sensitivity and glucose homeostasis in veal calves during their first three months of life (Chapter 5 and 6).


23

1

1.8 References

Abe, M., T. Iriki, K. Kondoh, and H. Shibui. 1979. Effects of nipple or bucket feeding of milk-substitute on rumen by-pass and on rate of passage in calves Br. J. Nutr. 41:175-181.

Accili, D., C. Frapier, L. Mosthaf, C. McKeon, S. C. Elbein, M. A. Permutt, E. Ramos, E. Lander, A. Ullrich, and S. I. Taylor. 1989. A mutation in the insulin receptor gene that impairs transport of the receptor to the plasma membrane and causes insulin-resistant diabetes. EMBO J. 8:2509-2517.

Ametaj, B. N., Q. Zebeli, F. Saleem, N. Psychogios, M. J. Lewis, S. M. Dunn, J. Xia, and D. S. Wishart. 2010. Metabolomics reveals unhealthy alterations in rumen metabolism with increased proportion of cereal grain in the diet of dairy cows. Metabolomics 6:583-594.

Avignon, A., C. Boegner, D. Mariano-Goulart, C. Colette, and L. Monnier. 1999. Assessment of insulin sensitivity from plasma insulin and glucose in the fasting or post oral glucose-load state. Int. J. Obes. Relat. Metab. Disord. 23.

Balakumar, M., L. Raji, D. Prabhu, C. Sathishkumar, P. Prabu, V. Mohan, and M. Balasubramanyam. 2016. High-fructose diet is as detrimental as high-fat diet in the induction of insulin resistance and diabetes mediated by hepatic/pancreatic endoplasmic reticulum (ER) stress. Mol. Cell. Biochem. 423:93-104.

Barros, R. P. A., C. Gabbi, A. Morani, M. Warner, and J. A. ke Gustafsson. 2009. Participation of ERα and ERβ in glucose homeostasis in skeletal muscle and white adipose tissue. Am. J. Physiol. Endocrinol. Metab. 297:E124–E133.

Belfiore, F., S. Iannello, and G. Volpicelli. 1998. Insulin sensitivity indices calculated from basal and OGTT-induced insulin, glucose, and FFA levels. Mol. Genet. Metab. 63:134-141.

Bergman, R. N., Y. Z. Ider, C. R. Bowden, and C. Cobelli. 1979. Quantitative estimation of insulin sensitivity. Am. J. Physiol. 236:E667-E677.

Bessesen, D. H. 2001. The role of carbohydrates in insulin resistance. The Journal of nutrition 131:2782S-2786S.Blagosklonny, M. V. 2013. TOR-centric view on insulin resistance and diabetic complications: perspective for

endocrinologists and gerontologists. Cell Death Dis. 12:e964.Blum, J. W. and D. L. Harmon. 1999. Endocrine and metabolic aspects in milk-fed calves. Domest. Anim.

Endocrinol. 17:219-230.Boston, R. C., D. Stefanovski, P. J. Moate, A. E. Sumner, R. M. Watanabe, and R. N. Bergman. 2003. MINMOD

Millennium: a computer program to calculate glucose effectiveness and insulin sensitivity from the frequently sampled intravenous glucose tolerance test. Diabetes Technol. Ther. 5:1003-1015.

Bryant, N. J., R. Govers, and D. E. James. 2002. Regulated transport of the glucose transporter GLUT4. Nature 3:267-277.

Buchanan, T. A., R. M. Watanabe, and A. H. Xiang. 2010. Limitations in Surrogate Measures of Insulin Resistance. J. Clin. Endocrinol. Metab. 95:4874-4876.

Cartee, G. D., C. Briggs-Tung, and E. W. Kietzke. 1993. Persistent effects of exercise on skeletal muscle glucose transport across the life-span of rats. J. Appl. Physiol. 75:972-978.

Carvalho, K. C., I. Isabela, W. Cunha, R. M. Rocha, F. R. Ayala, M. M. Cajaíba, M. D. Begnami, R. S. Vilela, G. R. Paiva, R. G. Andrade, and F. A. Soares. 2011. GLUT1 expression in malignant tumors and its use as an immunodiagnostic marker. Clinics 66:965-972.

Chen, H., G. Sullivan, L. Q. Yue, A. Katz, and M. J. Quon. 2003. QUICKI is a useful index of insulin sensitivity in subjects with hypertension. Am. J. Physiol. Endocrinol. Metab. 284:E804-E812.

Cobb, J., W. E. Gall, K. Adam, P. Nakhle, E. Button, J. Hathorn, K. Lawton, M. Milburn, R. Perichon, M. Mitchell, A. Natali, and E. Ferrannini. 2013. A novel fasting blood test for insulin resistance and prediabetes. J. Diabetes Sci. Technol. 7:100-110.

DeFronzo, R. A., J. D. Tobin, and R. Andres. 1979. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am. J. Physiol. 237:E214-E223.

Denroche, H. C., F. K. Huynh, and T. J. Kieffer. 2012. The role of leptin in glucose homeostasis. J Diabetes Investig. 3:115-129.

Eelderink, C., T. C. W. Moerdijk-Poortvliet, H. Wang, M. Schepers, T. Preston, T. Boer, R. J. Vonk, H. Schierbeek, and M. G. Priebe. 2012a. The glycemic response does not reflect the In vivo starch digestibility of fiber-rich wheat products in healthy men. J. Nutr. 142:258-263.

Eelderink, C., M. Schepers, T. Preston, R. J. Vonk, L. Oudhuis, and M. G. Priebe. 2012b. Slowly and rapidly digestible starchy foods can elicit a similar glycemic response because of differential tissue glucose uptake in healthy men. Am. J. Clin. Nutr. 96:1017-1024.

CHAPTER 1

24

1

Fukushima, A., K. Loh, S. Galic, B. Fam, B. Shields, F. Wiede, M. L. Tremblay, M. J. Watt, S. Andrikopoulos, and T. Tiganis. 2010. T-cell protein tyrosine phosphatase attenuates STAT3 and insulin signaling in the liver to regulate gluconeogenesis. Diabetes 59:1906-1914.

Gadgil, M. D., L. J. Appel, E. Yeung, C. A. M. Anderson, F. M. Sacks, and E. R. Miller. 2013. The effects of carbohydrate, unsaturated fat, and protein intake on measures of insulin sensitivity. Diabetes Care. 36:1132-1137.

Gall, W. E., K. Beebe, K. A. Lawton, K. P. Adam, M. W. Mitchell, P. J. Nakhle, J. A. Ryals, M. V. Milburn, M. Nannipieri, S. Camastra, A. Natali, and E. Ferrannini. 2010. alpha-Hydroxybutyrate is an early biomarker of insulin resistance and glucose intolerance in a nondiabetic population. PLoS One 5:e10883.

Gerrits, W. J. J., J. J. G. C. van den Borne, and J. W. Blum. 2008. Low-dietary protein intake induces problems with glucose homeostasis and results in hepatic steatosis in heavy milk-fed calves. Domest. Anim. Endocrinol. 35:121-129.

Gilbert, M. S. 2015. Replacing lactose from calf milk replacers : effects on digestion and post-absorptive metabolism. Vol. PhD Thesis. Wageningen University, Wageningen, The Netherlands.

Grusovin, J. and S. L. Macaulay. 2003. Snares for GLUT4-mechanisms directing vesicular trafficking of GLUT4. Front. Biosci. 8:d620-d641.

Guo, S. 2014. Insulin signaling, resistance, and the metabolic syndrome: insights from mouse models into disease mechanisms. The Journal of endocrinology 220:T1-T23.

Gutt, M., D. C.L., S. B. Spitzer, M. M. Llabre, M. Kumar, E. M. Czarnecki, N. Schneiderman, J. S. Skyler, and J. B. Marks. 2000. Validation of the insulin sensitivity index [ISI(0,120)]: comparison with other measures. Diabetes Res. Clin. Pract. 47:177-184.

Hayes, M. R., E. G. Mietlicki-Baase, S. E. Kanoski, and B. C. De Jonghe. 2014. Incretins and Amylin: Neuroendocrine Communication between the Gut, Pancreas, and Brain in Control of Food Intake and Blood Glucose. Annu. Rev. Nutr. 34:237-260.

Hostettler-Allen, R. L., L. Tappy, and J. W. Blum. 1994. Insulin resistance, hyperglycemia, and glucosuria in intensively milk-fed calves. J. Anim. Sci. 72:160-173.

Hotamisligil, G. S. 2006. Inflammation and metabolic disorders. Nature 444:860-867.Hugi, D., R. M. Bruckmaier, and J. W. Blum. 1997. Insulin resistance, hyperglycemia, glucosuria, and

galactosuria in intensively milk-fed calves: dependency on age and effects of high lactose intake. J. Anim. Sci. 75:469-482.

Hugi, D., L. Tappy, H. Sauerwein, R. M. Bruckmaier, and J. W. Blum. 1998. Insulin-dependent glucose utilization in intensively milk-fed veal calves is modulated by supplemental lactose in an age-dependent manner. J. Nutr. 128:1023-1030.

Jansson, J., B. Willing, M. Lucio, A. Fekete, J. Dicksved, J. Halfvarson, C. Tysk, and P. Schmitt-Kopplin. 2009. Metabolomics reveals metabolic biomarkers of crohn’s disease. PLoS One 4:e6386.

Johnston, K. L., E. L. Thomas, J. D. Bell, G. S. Frost, and M. D. Robertson. 2010. Resistant starch improves insulin sensitivity in metabolic syndrome. Diabet. Med. 27:391-397.

Jung, J., S. H. Kim, H. S. Lee, G. S. Choi, Y. S. Jung, D. H. Ryu, H. S. Park, and G. S. Hwang. 2013. Serum metabolomics reveals pathways and biomarkers associated with asthma pathogenesis. Clin. Exp. Allergy 43:425-433.

Kanzaki, M. 2006. Insulin receptor signals regulating GLUT4 translocation and actin dynamics. Endocr. J. 53:267-293.

Katz, A., S. S. Nambi, K. Mather, A. D. Baron, D. A. Follmann, G. Sullivan, and M. J. Quon. 2000. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J. Clin. Endocrinol. Metab. 85:2402-2410.

Kaufhold, J., H. M. Hammon, and J. W. Blum. 2000. Fructo-oligosaccharide supplementation: effects on metabolic, endocrine and hematological traits in veal calves. J. Vet. Med. A Physiol. Pathol. Clin. Med. 47:17-29.

Labussiere, E., S. Dubois, J. van Milgen, G. Bertrand, and J. Noblet. 2009a. Effect of solid feed on energy and protein utilization in milk-fed veal calves. J. Anim. Sci. 87:1106-1119.

Labussiere, E., G. Maxin, S. Dubois, J. van Milgen, G. Bertrand, and J. Noblet. 2009b. Effect of feed intake on heat production and protein and fat deposition in milk-fed veal calves. Animal 3:557-567.

Li, L., X. Li, W. Zhou, and J. L. Messina. 2013. Acute psychological stress results in the rapid development of insulin resistance. J. Endocrinol. 217:175-184.


25

1

Lynch, C. J. and S. H. Adams. 2014. Branched-chain amino acids in metabolic signalling and insulin resistance. Nat. Rev. Endocrinol. 10:723-736.

Mari, A., G. Pacini, E. Murphy, B. Ludvik, and J. J. Nolan. 2001. A model-based method for assessing insulin sensitivity from the oral glucose tolerance test. Diabetes Care. 24:539–548.

Matsuda, M. and R. A. DeFronzo. 1999. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care. 22:1462-1470.

Matthews, D. R., J. P. Hosker, A. S. Rudenski, B. A. Naylor, D. F. Treacher, and R. C. Turner. 1985. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 28:412-419.

Mei, S., X. Yang, H. Guo, H. Gu, L. Zha, J. Cai, X. Li, Z. Liu, B. J. Bennett, L. He, and W. Cao. 2014. A small amount of dietary carbohydrate can promote the HFD-induced insulin resistance to a maximal level. PLoS One 9:e100875.

Miura, T., W. Suzuki, E. Ishihara, I. Arai, H. Ishida, Y. Seino, and K. Tanigawa. 2001. Impairment of insulin-stimulated GLUT4 translocation in skeletal muscle and adipose tissue in the Tsumura Suzuki obese diabetic mouse: a new genetic animal model of type 2 diabetes. Eur. J. Endocrinol. 145:785-790.

Muniyappa, R., S. Lee, H. Chen, and M. J. Quon. 2008. Current approaches for assessing insulin sensitivity and resistance in vivo: advantages, limitations, and appropriate usage. Am. J. Physiol. Endocrinol. Metab. 294:E15-E26.

Pacini, G. and R. N. Bergman. 1986. MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsivity from the frequently sampled intravenous glucose tolerance test. Comput. Methods Programs Biomed. 23:113-122.

Pagano, G., M. Cassader, A. Diana, E. Pisu, C. Bozzo, F. Ferrero, and G. Lenti. 1981. Insulin resistance in the aged: the role of the peripheral insulin receptors. Metabolism 30:46-49.

Palmquist, D. L., J. Doppenberg, K. L. Roehrig, and D. J. Kinsey. 1992. Glucose and insulin metabolism in ruminating and veal calves fed high and low fat diets. Domest. Anim. Endocrinol. 9:233-241.

Pantophlet, A. J., S. Wopereis, C. Eelderink, R. J. Vonk, J. H. Stroeve, S. Bijlsma, L. van Stee, I. Bobeldijk, and M. G. Priebe. 2016. Metabolic profiling reveals differences in plasma concentrations of arabinose and xylose after consumption of fiber-rich pasta and wheat bread with differential rates of systemic appearance of exogenous glucose in healthy men. J. Nutr. [Epub ahead of print].

Park, K., M. Gross, D. H. Lee, P. Holvoet, J. H. Himes, J. M. Shikany, and D. R. Jacobs. 2009. Oxidative stress and insulin resistance: the coronary artery risk development in young adults study. Diabetes Care. 32:1302-1307.

Pedersen, O. 1999. Genetics of insulin resistance. Exp. Clin. Endocrinol. Diabetes 107:113-118.Petersen, K. F. and G. I. Shulman. 2006. Etiology of insulin resistance. Am. J. Med. 119:S10-S16.Pyla, R., N. Poulose, J. Y. Jun, and L. Segar. 2013. Expression of conventional and novel glucose transporters,

GLUT1, -9, -10, and -12, in vascular smooth muscle cells. Am. J. Physiol., Cell Physiol. 304:C574-C589.Respondek, F., P. Gerard, M. Bossis, L. Boschat, A. Bruneau, S. Rabot, A. Wagner, and J. C. Martin. 2013.

Short-chain fructo-oligosaccharides modulate intestinal microbiota and metabolic parameters of humanized gnotobiotic diet induced obesity mice. PLoS One 12:e71026.

Respondek , F., K. Myers, T. L. Smith, A. Wagner, and R. J. Geor. 2011. Dietary supplementation with short-chain fructo-oligosaccharides improves insulin sensitivity in obese horses. J. Anim. Sci. 89:77-83.

Respondek, F., K. S. Swanson, K. R. Belsito, B. M. Vester, A. Wagner, L. Istasse, and M. Diez. 2008. Short-chain fructooligosaccharides influence insulin sensitivity and gene expression of fat tissue in obese dogs. J. Nutr. 138:1712-1718.

Rivellese, A. A., C. De Natale, and S. Lilli. 2002. Type of dietary fat and insulin resistance. Ann. N. Y. Acad. Sci. 967:329-335.

Roberts, C. K., L. A. Hevener, and R. J. Barnard. 2013. Metabolic syndrome and insulin resistance: underlying causes and modification by exercise training. Compr. Physiol. 3:1-58.

Robertson, M. D., A. S. Bickerton, A. L. Dennis, H. Vidal, and K. N. Frayn. 2005. Insulin-sensitizing effects of dietary resistant starch and effects on skeletal muscle and adipose tissue metabolism. Am. J. Clin. Nutr. 82:559-567.

Röder, P. V., B. Wu, Y. Liu, and W. Han. 2016. Pancreatic regulation of glucose homeostasis. Exp. Mol. Med. 48:e219.

Saito, K. and F. Matsuda. 2010. Metabolomics for functional genomics, systems biology, and biotechnology. Annu. Rev. Plant Biol. 61:462-489.

CHAPTER 1

26

1

Shen, S. W., G. M. Reaven, and J. W. Farquhar. 1970. Comparison of impedance to insulin-mediated glucose uptake in normal subjects and in subjects with latent diabetes. J. Clin. Invest. 49:2151-2160.

Sherwin, R. S., H. Shamoon, R. Hendler, L. Saccà, N. Eigler, and M. Walesky. 1980. Epinephrine and the regulation of glucose metabolism: effect of diabetes and hormonal interactions. Metabolism 29:1146-1154.

Shoelson, S. E., S. Lee, and A. B. Goldfire. 2006. Inflammation and insulin resistance. J. Clin. Invest. 161:1793-1801.

Sluijs, I., J. W. Beulens, D. L. van der A, A. M. Spijkerman, D. E. Grobbee, and Y. T. van der Schouw. 2010. Dietary intake of total, animal, and vegetable protein and risk of type 2 diabetes in the European Prospective Investigation into Cancer and Nutrition (EPIC)-NL study. Diabetes Care. 33:43-48.

Soman, V. R. and R. A. DeFronzo. 1980. Direct evidence for downregulation of insulin receptors by physiologic hyperinsulinemia in man. Diabetes 29:159-163.

Stanfield, C. L. 2012. The endocrine system: Regulation of energy metabolism and growth. Pages 603-630 in Principles of Human Physiology. 5 ed. K. Churchman, S. Scharf, C. C. Jones, J. McElroy, A. Williams, D. Cogan, and N. Tabor, ed. Pearson Education, London, UK.

Stumvoll, M., A. Mitrakou, W. Pimenta, T. Jenssen, H. Yki-Jarvinen, T. Van Haeften, W. Renn, and J. Gerich. 2000. Use of the oral glucose tolerance test to assess insulin release and insulin sensitivity. Diabetes Care. 23:295-301.

Tam, C. S., W. Xie, W. D. Johnson, W. T. Cefalu, L. M. Redman, and E. Ravussin. 2012. Defining Insulin Resistance From Hyperinsulinemic-Euglycemic Clamps. Diabetes Care. 35:1605-1610.

Taylor, S. I., T. Kadowaki, H. Kadowaki, D. Accili, A. Cama, and C. McKeon. 1990. Mutations in insulin-receptor gene in insulin-resistant patients. Diabetes Care. 13:257-279.

Thorens, B. and M. Mueckler. 2010. Glucose transporters in the 21st Century. Am. J. Physiol. Endocrinol. Metab. 298:E141-E145.

Torlińska, T., P. Maćkowiak, L. Nogowski, T. Hryniewiecki, H. Witmanowski, M. Perz, E. M dry, and K. W. Nowak. 2000. Age dependent changes of insulin receptors in rat tissues. J. Physiol. Pharmacol. 51:871-881.

Tuomilehto, J. 2002. Point: a glucose tolerance test is important for clinical practice. Diabetes Care. 25:1880-1882.

Vicari, T., J. J. G. C. van den Borne, W. J. J. Gerrits, Y. Zbinden, and J. W. Blum. 2008. Separation of protein and lactose intake over meals dissociates postprandial glucose and insulin concentrations and reduces postprandial insulin responses in heavy veal calves. Domest. Anim. Endocrinol. 34:182-195.

Watson, R. T., M. Kanzaki, and J. E. Pessin. 2004. Regulated membrane trafficking of the insulin-responsive glucose transporter 4 in adipocytes. Endocr. Rev. 25:177-204.

Weckwerth, W. 2003. Metabolomics in systems biology. Annu. Rev. Plant Biol. 54:669-689.Weickert, M. O. 2012. Nutritional modulation of insulin resistance. Scientifica 2012:424780.Wijayasinghe, M. S., N. E. Smith, and R. L. Baldwin. 1984. Growth, health, and blood glucose concentrations

of calves fed high-glucose or high-fat milk replacers. J. Dairy Sci. 67:2949-2956.Wilcox, G. 2005. Insulin and Insulin Resistance. Clin. Biochem. Rev. 26:19-39.Wishart, D. S., T. Jewison, A. C. Guo, M. Wilson, C. Knox, Y. Liu, Y. Djoumbou, R. Mandal, F. Aziat, E. Dong,

S. Bouatra, I. Sinelnikov, D. Arndt, J. Xia, P. Liu, F. Yallou, T. Bjorndahl, R. Perez-Pineiro, R. Eisner, F. Allen, V. Neveu, R. Greiner, and A. Scalbert. 2013. HMDB 3.0—The Human Metabolome Database in 2013. Nucleic Acids Res. 41:D801-D807.

Workeneh, B., F. Abbasi, and G. Reaven. 2010. Fasting urine pH is independent of insulin sensitivity. Am. J. Clin. Nutr. 91:586-588.

Ye, J. 2013. Mechanisms of insulin resistance in obesity. Front. Med. 7:14-24.Yoon, M. S. 2016. The emerging role of branched-chain amino acids in insulin resistance and metabolism.

Nutrients 8:E405.Zhao, F. and A. F. Keating. 2007. Functional properties and genomics of glucose transporters. Curr. Genomics.

8:113-128.Zisman, A., O. D. Peroni, E. D. Abel, M. D. Michael, F. Mauvais-Jarvis, B. B. Lowell, J. F. Wojtaszewski, M. F.

Hirshman, A. Virkamaki, L. J. Goodyear, C. R. Kahn, and B. B. Kahn. 2000. Targeted disruption of the glucose transporter 4 selectively in muscle causes insulin resistance and glucose intolerance. Nat. Med. 6:924-928.

CHAPTER 2

Substantial replacement of lactose with fat in a high-lactosemilk replacer diet increases liver fat accumulation but does

not affect insulin sensitivity in veal calves

A.J. Pantophlet,1 W.J.J. Gerrits,2 R.J. Vonk,3 and J.J.G.C. van den Borne2

Adapted from Journal of dairy science 2016; 99 (12):10022–10032

1Department of Pediatrics, Center for Liver, Digestive and Metabolic Diseases, University Medical Centre Groningen, the Netherlands;

2Animal Nutrition Group, Wageningen University, Wageningen, the Netherlands; 3Centre for Medical Biomics, University Medical Center Groningen, Groningen, the Netherlands

"An investment in knowledge pays the best interest"- Benjamin Franklin -

CHAPTER 2

28

2

Abstract

In veal calves, the major portion of digestible energy intake originates from milk replacer (MR), with lactose and fat contributing approximately 45 and 35%, respectively. In veal calves older than 4 months, prolonged high intakes of MR may lead to problems with glucose homeostasis and insulin sensitivity, ultimately resulting in sustained insulin resistance, hepatic steatosis, and impaired animal performance. The contribution of each of the dietary energy sources (lactose and fat) to deteriorated glucose homeostasis and insulin resistance is currently unknown. Therefore, an experiment was designed to compare the effects of a high-lactose and a high-fat MR on glucose homeostasis and insulin sensitivity in veal calves. Sixteen male Holstein-Friesian calves (120 ± 2.8 kg of BW) were assigned to either a high-lactose (HL) or a high-fat (HF) MR for 13 consecutive weeks. After at least 7 weeks of adaptation, whole-body insulin sensitivity and insulin secretion were assessed by euglycemic-hyperinsulinemic and hyperglycemic clamps, respectively. Postprandial blood samples were collected to assess glucose, insulin, and triglyceride responses to feeding, and 24-h urine was collected to quantify urinary glucose excretion. At the end of the trial, liver and muscle biopsies were taken to assess triglyceride contents in these tissues. Long-term exposure of calves to HF or HL MR did not affect whole-body insulin sensitivity (averaging 4.2 ± 0.5 × 10−2 [(mg/kg·min)/(μU/mL)]) and insulin secretion. Responses to feeding were greater for plasma glucose and tended to be greater for plasma insulin in HL calves than in HF calves. Urinary glucose excretion was substantially higher in HL calves (75 ± 13 g/d) than in HF calves (21 ± 6 g/d). Muscle triglyceride content was not affected by treatment and averaged 4.5 ± 0.6 g/kg, but liver triglyceride content was higher in HF calves (16.4 ± 0.9 g/kg) than in HL calves (11.2 ± 0.7 g/kg), indicating increased hepatic fat accumulation. We conclude that increasing the contribution of fat to the digestible energy intake from the MR from 20 to 50%, at the expense of lactose does not affect whole-body insulin sensitivity and insulin secretion in calves. However, a high-lactose MR increases postprandial glucose and insulin responses, whereas a high-fat MR increases fat accumulation in liver but not muscle.

Keywords: veal calves, lactose, fat, insulin sensitivity

High-lactose vs. high-fat diet

29

2

2.1 Introduction

Veal calves are fed milk replacer (MR) and solid feed (consisting of concentrates and roughages). Despite the tendency in recent years to increase the amount of solid feed, the vast majority (60-70%) of the digestible energy intake originates from MR. After feeding, MR bypasses the calf ’s rumen and directly enters the abomasum because of closure of the esophageal groove. Milk replacer typically contains highly digestible nutrients such as lactose, fat, and proteins, which provide approximately 45%, 35%, and 20% of the digestible energy intake, respectively.

Prolonged intakes of high amounts of MR have been shown to induce problems with glucose homeostasis and insulin sensitivity in heavy (> 4 months of age) veal calves, characterized by high incidences of hyperglycemia and hyperinsulinemia and increased urinary glucose excretion (Hostettler-Allen et al., 1994; Hugi et al., 1997; Pantophlet et al., 2016a). These problems may result in metabolic and pro-inflammatory diseases as seen in humans (Hotamisligil, 2006; Shoelson et al., 2006) and in hepatic steatosis (Gerrits et al., 2008).

Dietary factors contributing to the disturbed glucose homeostasis and insulin sensitivity in heavy calves have been studied (Hugi et al., 1997, 1998; Pantophlet et al., 2016a), and results indicate that high amounts of lactose may be a factor. Ingesting high amounts of lactose in only 2 daily meals and for a prolonged period (i.e., ~6 months of life) deviates from the ontogenetic background of calves. In nature, calves between 4 and 6 months of age are grazing, and feedstuffs from plant origin are fermented in the rumen, along with short-chain fatty acids being produced as a major energy source. Thus, in nature, a gradual shift occurs from glucose and long-chain fatty acids from milk as main energy sources to short-chain fatty acids originating from rumen fermentation during the first months of the calf ’s life. In general, problems with glucose metabolism and insulin sensitivity appear to be age dependent in veal calves (Hugi et al., 1997, 1998; Pantophlet et al., 2016b). Heavy veal calves produce very little fatty acids from glucose (Roehrig et al., 1988; van den Borne et al., 2006), and ingestion of large quantities of glucose perturbs their glucose homeostasis for a substantial period after feeding. This circumstance could explain why a high lactose intake negatively affects glucose metabolism and insulin sensitivity in veal calves (Hugi et al., 1997, 1998) and leads to significant amounts of glucose being excreted in urine (Hugi et al., 1997; Pantophlet et al., 2016a).

Alternatively, the high dietary fat content in MR for calves could affect glucose homeostasis and insulin sensitivity. The composition of the digestible energy in veal calves (i.e., high fat and high carbohydrate content) resembles that of the adult Western human diet (Schwarz et al., 2003), and such high dietary fat intake has consistently been associated with the development of insulin resistance (Randle et al., 1963; Storlien et al., 1996; Frayn, 2003; Müller and Kersten, 2003). In rodents, fat-induced problems with insulin sensitivity may operate via several mechanisms, including

CHAPTER 2

30

2

triglyceride accumulation in skeletal muscle and adipocytes, which impairs GLUT4 translocation (Storlien et al., 1996), and reduction of the number of insulin receptors in adipocytes (Harris, 1992; Harris and Kor, 1992). However, weaned non-ruminant animal species, such as pigs and rodents, do not commonly develop insulin resistance, although they consume higher amounts of glucose (but less fat) than veal calves. We therefore hypothesize that, apart from species differences, interactions between fatty acids and glucose may play a role in perturbing glucose homeostasis and inducing the development of insulin resistance in veal calves.

Standardized studies in which lactose and fat are exchanged (iso-energetically) may reveal the contribution of the dietary energy source to the development of insulin resistance in calves. The objective of the current study was therefore to assess effects of a large iso-energetic exchange of lactose and fat intake on insulin sensitivity in veal calves.

2.2 Materials and Methods

2.2.1 Animals and housing

Sixteen male Holstein-Friesian calves (120 ± 2.8 kg of BW; 99 ± 2.0 days old) were purchased. During the first 6 weeks of the 13-week study, calves were housed in pens of 4 calves each (2 m2 per calf) that were fitted with a wooden slatted floor and galvanized fencings. Calves were then transferred to metabolic cages (dimension: 0.80 × 1.8 m) for the next 7 weeks. During this period, several measurements were performed (see Experimental Procedures). Ventilation occurred by ceiling fans, and illumination was by natural light and artificial (fluorescent lamps) light between 0700 and 1900 h. Temperature was controlled at 18°C and humidity at 65%.

Experimental procedures complied with the Dutch Law on Experimental Animals and the ETS123 (Council of Europe 1985 and the 86/609/EEC Directive) and were approved by the Animal Care and Use Committee of Wageningen University.

2.2.2 Experimental design, diets, and feeding

Calves were assigned to either a high-lactose diet (HL; n = 8) or a high-fat diet (HF; n = 8), and to 1 of 8 blocks (pairs of calves) with 1 HL calf and 1 HF calf per block. Body weight and age did not differ between treatments at the start of the trial. Because of health problems in 2 HF calves, block 7 consisted of 2 HL calves, and block 8 (with the 2 remaining HF calves) was not included in the insulin sensitivity, insulin secretion, postprandial blood metabolites, and urinary glucose excretion measurements. Between treatments (Table 2.1), fat and lactose were exchanged iso-energetically based on digestible energy. Energy values of 39.0 kJ/g fat and 16.5 kJ/g lactose and ileal digestibilities of 95% for fat and 94% for lactose were assumed (Hof, 1980). High-lactose diet calves received 25% more feed than HF calves to obtain iso-energetic and


31

2

iso-nitrogenous feeding strategies. Apart from fat and lactose, the daily allowance of protein and micronutrients was similar for the 2 treatments. The feeding level for both MR diets was 2.25 times the digestible energy requirements for maintenance, which was set at 473 kJ/kg0.75 per day (Van Es et al., 1967). At the start of the study, calves received a commercial MR. Introduction of the experimental MR occurred gradually within the first 3 days of the study. The feeding schedule was adjusted according to the calf ’s metabolic weight (kg0.75), which was measured weekly. Milk replacer was prepared at a concentration of 167 g of MR/L and provided in a bucket at a temperature of 40–42°C at 0800 and 1630 h. Milk replacer refusals (maximum of 15 min after feeding) were weighed and recorded after each feeding. During the first 6 weeks, calves were tethered with a chain during the MR feeding episodes, allowing individual feeding.

In addition, each calf received 9 g of DM solid feed per kg0.75 per day. On a DM basis, the solid feed consisted of 50% concentrates, 25% wheat straw, and 25% maize silage. Solid feed was provided per pen during group housing and per individual calf during individual housing in metabolic cages. Solid feed was provided once daily after feeding MR at 1630 h. Solid feed refusals were weighed and recorded before the morning MR feeding. Throughout the study, calves were allowed ad libitum access to water.

Table 2.1 | Composition of the high-lactose (HL) and high-fat (HF) milk replacers1

Diet Diet

HL HF HL HF

Ingredients (g/kg) Nutrients (g/kg)

Whey 396 38.0 Crude protein 184 230

Whey protein concentrate 35% 348 188 Crude fat 94 300

Whey protein concentrate 75% - 150 Crude fiber 0.01 0.01

Delactosed whey - 22.2 Crude ash 65 67

Lactose 90.0 - Moisture 21 26

Fat-filled whey palm / coconut2 62.4 275 Lactose 591 337

Fat-filled whey lard2 70.0 238 Fe (mg/kg) 72.8 56.1

Fat-filled whey coconut2 20.0 62.5 Gross energy (MJ/kg) 17.9 22.6

Fat blend2 5.0 5.0

Calcium carbonate 4.0 9.2

Mono calcium phosphate - 8.1

Magnesium oxide 0.6 1.4

DL-methionine 0.5 0.6

Premix 3.2 3.4 1 HL calves received 25% more feed than HF calves to obtain iso-energetic and iso-nitrogenous feeding strategies.2 Total fat composition: lard 40%, palm oil 40%, coconut oil 20%.

CHAPTER 2

32

2

2.2.3 Experimental procedures

For all calves (i.e., all 8 blocks), fasting blood samples were taken in week 1 (before treatment), 6, and 12. Liver and muscle tissue samples were taken at slaughter in week 13. Other measurements were taken during seven 1-week measurement periods (week 7–13) with 1 block (i.e., 1 HF and 1 HL calf for block 1-6 and 2 HL calves for block 7) per period (n = 14 calves). These blocks were staggered to allow the laborious measurements and housing in respiration chambers. On day 1 of each measurement period, a euglycemic-hyperinsulinemic clamp was conducted to measure insulin sensitivity. Urine was collected quantitatively from day 2 to 5 for the determination of urinary glucose excretion, postprandial blood samples were taken on day 4 to measure postprandial concentrations of metabolites and hormones, and a hyperglycemic clamp was conducted on day 6 to measure insulin secretion.

2.2.3.1 Euglycemic-hyperinsulinemic clamp Whole-body insulin sensitivity was assessed by the euglycemic-hyperinsulinemic clamp technique. Euglycemic-hyperinsulinemic clamps were performed at 202 ± 2.9 kg of BW for HL calves and at 206 ± 3.6 kg of BW for HF calves.

Semi-permanent catheters (Careflow, Becton Dickinson, Alphen aan den Rijn, the Netherlands) were inserted in both jugular veins under local anesthesia. The double-lumen catheter in the left jugular vein was used for glucose and insulin infusion and the catheter in the right vein was used for blood sampling during the experimental period. Vicryl suturing (Ethicon, Somerville, NJ) was used to attach the catheter to the skin. The catheter was extended with a 3-layer extension tube to allow infusion and blood sampling.

Calves were fasted for 15 h, and the morning feeding was omitted to achieve a steady glucose turnover rate. Prior starting the 4-h clamp study, 3 blood samples of 5 mL each were taken at 20-min intervals. The average glucose concentration from these samples was used to define the basal blood glucose concentration, which was the target glucose concentration during the study. At 1030 h, a priming dose of insulin (Actrapid 100 IE/ mL, Novo Nordisk, Denmark) at 2.1 mU/kg of BW per min was infused into the left catheter within 5 min to rapidly increase the plasma insulin concentration, immediately followed by a continuous infusion of insulin at a rate of 1 mU/kg of BW per min. This continuous infusion was maintained for a period of 4 h (plasma insulin levels ~135 mU/L). At 5 min after starting the insulin infusion, continuous glucose infusion (20% glucose solution for intravenous infusion; B. Braun, Melsungen, Germany) was initiated to maintain the basal plasma glucose concentration (i.e., infusion rate depended on the clearance rate). Syringe infusion pumps were used to control infusion rates of glucose and insulin separately, and infusion rates were continuously recorded.

During the clamp study, 0.3-mL blood samples were taken from the right jugular vein catheter at 10-min intervals during the first 2 h and at 15-min intervals during the second 2 h. Plasma glucose concentration was measured in each sample using


33

2

Precision Xtra Plus test strips in combination with the Precision Exceed Sensor (Abbott, Weesp, the Netherlands). For analyses of insulin and glucose, 5-mL blood samples were collected at 30-min intervals and transferred into tubes with heparin (insulin) or sodium fluoride (glucose), which were stored on crushed ice. Samples were centrifuged (1,500 × g for 12 min), and plasma was harvested and stored at −20°C pending further analyses. Catheters were flushed with heparinized saline after each sample to prevent blood coagulation in the catheter.

Depending on the changes in plasma glucose level, the glucose infusion rate during the euglycemic-hyperinsulinemic clamp was adjusted if necessary to maintain a constant, basal plasma glucose level during insulin infusion. Glucose disposal (M-value) was defined as the average glucose infusion rate at steady state divided by BW. Whole-body insulin sensitivity was defined as the M-value divided by the average plasma insulin level at steady state (M/I-value).

2.2.3.2 Postprandial blood glucose, insulin, and triglyceride concentrationsFor assessing metabolic responses to feeding, 8-mL blood samples were taken at −15, 30, 60, 90, 120, 150, 180, 240, 300, 360, and 420 min relative to feeding the experimental MR at 0800 h. Blood was transferred into tubes with heparin (insulin, triglycerides) or sodium fluoride (glucose), which were stored on crushed ice. Samples were centrifuged (1,500 × g for 12 min), and plasma was harvested and stored at −20°C pending analyses. Catheters were flushed with heparinized saline after each sample to prevent blood coagulation in the catheter.

The ΔCmax (maximum concentration minus fasting concentration), Tmax (time to maximum concentration), and the incremental area under the curve (iAUC0−7h) were calculated for all compounds. The iAUC was calculated using the trapezoid method (Le Floch et al., 1990).

2.2.3.3 Hyperglycemic clamp A 3-h hyperglycemic clamp was conducted to obtain a proxy for insulin secretion under glucose-stimulated conditions. During the first 10 min of the clamp, plasma glucose was primed by infusing a 20% glucose solution (B. Braun) intravenously until a glucose level of 10 mmol/L was reached. Thereafter, blood glucose was measured in 10-min intervals (Precision Extra Plus test strips; Abbott) and the rate of glucose infusion was adapted to maintain the high blood glucose level. In addition, 8-mL blood samples were taken at 30-min intervals and transferred into tubes with heparin (insulin) or sodium fluoride (glucose), which were stored on crushed ice. Samples were centrifuged (1,500 × g for 12 min) and plasma was harvested and stored at −20°C pending analyses of glucose and insulin. Catheters were flushed with heparinized saline after each sample to prevent blood coagulation in the catheter.

Insulin secretion was assessed by calculating the total insulin response (iAUC0−3h). The iAUC was calculated using the trapezoid method (Le Floch et al., 1990).

CHAPTER 2

34

2

2.2.3.4 Urine collection Harnesses for the fecal collection bags were attached 5 days before the start of the measurement period to allow separate collection of feces and urine. Urine was quantitatively collected during 4 consecutive days (measurement period day 2-5) in plastic bags that were harnessed to the calves. Urine was collected in a pit containing 750 mL of 4.5 M sulfuric acid, and samples were taken and stored at −20°C pending glucose analysis.

2.2.3.5 Fasting blood concentrations of insulin and metabolites After a 15-h fasting period, blood was collected in week 1 (before assignment to experimental diet), 6, and 12 from the jugular vein by venipuncture. The 20 mL sample was centrifuged (1,500 × g for 12 min), and plasma was harvested and stored at −20°C pending analysis of glucose, insulin, HDL- and LDL-cholesterol, triglycerides, malondialdehyde, and non-esterified fatty acids (NEFA). The quantitative insulin sensitivity check index (QUICKI), which is another index of insulin sensitivity, was calculated from the fasting plasma glucose and insulin concentrations (Chen et al., 2003; Muniyappa et al., 2008). This index was calculated using the following formula:

2.2.3.6

QUICKI = 1

(log( glucose mgdL

)+ log(insulin mUL

))

Liver and muscle biopsies In week 13, calves were euthanized by intravenous injection of sodium pentobarbital. The liver and musculus rectus femoris were dissected and sampled. Both tissues were snap-frozen within 15 min and subsequently stored at −80°C pending analysis of triglyceride concentrations.

2.2.4 Laboratory analyses

Plasma glucose was analyzed on an Architect ci8200 analyzer using the hexokinase method (Abbott Laboratories, Chicago, IL). The within- and between-run coefficient of variation were ≤2%. Plasma insulin was analyzed using a Coat-a-Count radioimmunoassay kit (Siemens Healthcare Diagnostics, Erlangen, Germany). The within- and between-run coefficient of variation were ≤5% and ≤7%, respectively. Analysis of cholesterol and triglycerides in plasma were performed on a Roche-Hitachi Modular automatic analyzer using enzymatic colorimetric methods (Roche Diagnostics, Almere, the Netherlands). The within- and between-run coefficient of variation were ≤2% for both analyses. Plasma malondialdehyde was measured with an isocratic Varian HPLC from Chromosystems (Munich, Germany). A 10-cm C18 cartridge (Varian) was used with a flow rate of 1.0 mL/min at ambient temperature (25°C). Fluorescence detection occurred with excitation at 515 nm and emission at 553 nm. The within- and


35

2

between-run coefficient of variation were ≤5% and ≤11%, respectively. Plasma NEFA was analyzed with a NEFA-C kit (Wako Chemicals, Neuss, Germany) using a Cobas Mira Plus automatic analyzer (Roche, Basel, Switzerland). The within- and between-run coefficient of variation were ≤ 3%.

Triglyceride concentrations in muscle and liver tissues were analyzed using an enzymatic colorimetric method (GPO-PAP). Slices of 100 to 200 mg of tissues were cut, and 5 mL of a 25% KOH solution was added to saponify triacylglycerols into triglycerides, according to the method of Eggstein and Kreutz (1966). Samples were ground using a homogenizer (Ultra-turrax T8, IKA Labortechnik, Staufen, Germany). After 20 min in a 70°C water bath, triglycerides were converted to glycerol using 1 mL of a 2.5 mol/L HClO4 solution. After homogenization, tubes were centrifuged during 5 min at 1,500 × g, and the supernatant was harvested and colored using a Liquicolor triglyceride kit (Human Diagnostics Worldwide, Wiesbaden, Germany). The glycerol concentration was spectrophotometrically measured at 546 nm (Humalyzer 3500, Human). The within- and between-run coefficient of variation were ≤6% and ≤8%, respectively.

2.2.5 Statistical analysis

SPSS (version 22; IBM Corp., Armonk, NY) statistical software was used for all statistical analyses. Data are presented as means ± SEM. Whole-body insulin sensitivity, insulin secretion, urinary glucose excretion, and data derived from postprandial blood concentrations (i.e., ΔCmax, Tmax, and iAUC0−7h) were tested for treatment and block effects by ANOVA using the GLM (Univariate) procedure. Treatment and block were used as factors, and calf was the experimental unit.

Liver and muscle triglyceride concentrations were tested for treatment effect by ANOVA using the GLM (Univariate) procedure. Treatment was used as the factor, and calf was the experimental unit. Fasting plasma glucose, insulin, cholesterol, triglycerides, malondialdehyde and NEFA concentrations, and QUICKI (measured longitudinally) were tested for time, treatment, and time × treatment effects using the mixed effects model procedure. Time, treatment, and their interaction were used as fixed terms, and time as a repeated variable within calf. Only values at weeks 6 and 12 were used as dependent variables. Because calves were on the commercial MR during the initial measurement (values week 1), analysis was performed on delta values (i.e., values at week 6 or 12 minus values at week 1). Based on fit statistics (Akaike and Bayesian information criteria), the heterogeneous first-order autoregressive covariance structure was used for all models. When a significant interaction was found, partial tests between per time point were performed. A Bonferroni correction was used to correct for multiple comparisons.

Normality of the model residuals was assessed by visual inspection. Non-normal distributed data were (log) transformed to obtain normality. P-values < 0.05 were considered significant, and P-values < 0.10 were considered a trend toward significance.

CHAPTER 2

36

2

2.3 Results

Two HF calves had major deteriorations in health (and growth) performance and were excluded from the trial. Therefore, block 7 consisted of 2 HL calves (see Material and Methods). However, none of the outcomes was affected by block; hence, block effects are not reported. Feed refusal was negligible (i.e., only a few incidences) during the trial in both groups. The average daily gain, measured over the 13-week trial was 1,366 ± 36 g/d and did not differ between dietary treatments.

2.3.1 Insulin sensitivity and insulin secretion

Whole-body insulin sensitivity, measured by the euglycemic–hyperinsulinemic clamp, varied from 1.5 to 8.3 [(mg/kg·min)/(μU/mL)] between calves and did not differ (P > 0.05) between dietary treatments (Figure 2.1). Pancreatic β-cell function was assessed by quantification of insulin secretion during the hyperglycemic clamp. Data were available for 12 instead of 14 calves (6 HL and 6 HF), because some technical difficulties occurred with blood sampling for 2 HL calves (not enough blood samples collected). Plasma glucose concentrations reached maximum between 60 and 90 min for both treatment groups, whereas insulin reached maximum concentrations at 120 min (Figure 2.2). The total insulin secretion (iAUC0−3h) did not differ between dietary treatments (Figure 2.3).

Figure 2.1 | Whole-body insulin sensitivity in veal calves fed a high-lactose (HL) milk replacer (n=8) or a high-fat (HF) milk replacer (n=6). Milk replacer compositions are given in Table 2.1. ISclamp = insulin sensitivity derived from a euglycemic-hyperinsulinemic clamp. Error bars represent SEM and symbols () refer to individual calves. Insulin sensitivity did not differ between dietary treatments.

2.3.2 Development of fasting blood metabolites and hormones

The change in fasting glucose, insulin, NEFA, and triglycerides over time did not differ between HL and HF calves (Table 2.2; time × treatment interaction, P > 0.05). HDL- and LDL-cholesterol concentrations over time differed between treatments (time × treatment interaction, P < 0.05 for both). The levels were higher (P < 0.05) in HF calves than in HL calves, both in week 6 and 12. Fasting malondialdehyde concentrations tended to differ over time between treatments (time × treatment interaction, P < 0.10). The change in QUICKI, derived from the fasting glucose and insulin concentrations, did not differ (P > 0.05) between treatments.


37

2

2.3.3 Postprandial blood glucose, insulin, and triglycerides

Data were available for 11 instead of 14 calves (6 HL and 5 HF), because 1 HF calf did not consume the entire meal on the day of measurement, and some technical difficulties occurred with blood sampling for 2 HL calves (not enough blood samples collected). The postprandial blood glucose, insulin, and triglycerides responses are shown in Figure 2.4. Fasting blood insulin and triglyceride concentrations did not differ (P > 0.05) between treatment groups (Table 2.3). Fasting glucose concentrations tended to be higher (P < 0.10) in HF calves than in HL calves. The Tmax values for glucose, insulin, and triglycerides did not differ (P > 0.05) between treatments. For glucose and insulin, ΔCmax was higher (P < 0.05) in HL calves, whereas ΔCmax for triglycerides was higher in HF calves. For glucose, iAUC0−7h for glucose was higher (P < 0.05) in HL calves than in HF calves. For insulin, the iAUC0−7h tended to be higher (P < 0.10) in HL calves than in HF calves. For triglycerides, iAUC0−7h was higher (P < 0.05) in HF calves than in HL calves.

Figure 2.2 | Plasma glucose (A) and insulin (B) concentrations during a 3 hour-hyperglycemic clamp in veal calves that were adapted to a high-lactose (, n=6) milk replacer or a high-fat (, n=6) milk replacer. Error bars represent SEM. Milk replacer compositions are given in Table 2.1. The total insulin secretion is depicted in Figure 2.3.

2.3.4 Urinary glucose excretion

Urinary glucose excretion was substantially higher (P < 0.001) in HL calves (75 ± 13 g/d) than in HF calves (21 ± 6 g/d). Liver and Muscle Fat Liver triglyceride (fat) content was greater (P < 0.05) in HF calves (16.4 ± 0.9 g/kg) than in HL calves (11.2 ± 0.7 g/kg). Muscle triglyceride (fat) content did not differ (P > 0.05) between dietary treatments and was 6.9 ± 1.5 and 6.2 ± 0.4 (g/kg) for HL and HF calves, respectively.

2.3.5 Liver and muscle fat

Liver triglyceride (fat) content was greater (P < 0.05) in HF calves (16.4 ± 0.9 g/kg) than in HL calves (11.2 ± 0.7 g/kg). Muscle triglyceride (fat) content did not differ (P > 0.05) between dietary treatments and was 6.9 ± 1.5 and 6.2 ± 0.4 (g/kg) for HL and HF calves, respectively.

CHAPTER 2

38

2

Figure 2.3 | Insulin secretion (assessed during a 3 hour-hyperglycemic clamp) in veal calves fed a high-lactose (HL; n=6) or a high-fat (HF; n=6) milk replacer. Milk replacer compositions are given in Table 2.1. Error bars represent SEM and symbols () refer to individual calves. Total insulin secretion did not differ between treatments.

Figure 2.4 | Plasma glucose (A), insulin (B) and triglycerides (C) responses in veal calves fed (at time = 0) a high-lactose milk replacer (, N=6) or a high-fat milk replacer (, N=5). Milk replacer compositions are given in Table 2.1. Error bars represent SEM. Calculated parameters and statistics are given in Table 2.3.


39

2

Table 2.2 | Fasting plasma metabolite and hormone concentrations in veal calves fed either a high-lactose (HL) milk replacer (n=8) or a high-fat (HF) milk replacer (n=6) diet. The milk replacer compositions are given in Table 1.

Item

Treatment (Trt)P-value

HL HF

Week 6 Week 12 Week 6 Week 12 Time Trt Time x Trt

Fasting plasma concentration1

Glucose (mmol/L) 5.1 ± 0.2 4.9 ± 0.2 5.8 ± 0.1 5.6 ± 0.2 0.102 0.132 0.847

Insulin (mU/L) 11.8 ± 3.4 9.9 ± 3.5 5.1 ± 2.4 7.5 ± 1.8 0.917 0.893 0.209

HDL cholesterol (mmol/L) 0.8 ± 0.1a 1.3 ± 0.1x 1.5 ± 0.1b 1.6 ± 0.1y <0.001 <0.001 0.002

LDL cholesterol (mmol/L) 0.4 ± 0.1a 0.6 ± 0.1x 0.9 ± 0.2a 1.6 ± 0.2y <0.001 <0.001 0.010

Triglycerides (mmol/L) 0.09 ± 0.01 0.08 ± 0.01 0.11 ± 0.03 0.10 ± 0.01 0.321 0.263 0.784

NEFA (μmol/L) 281 ± 35 239 ± 58 245 ± 45 295 ± 147 0.960 0.763 0.543

Malondialdehyde (μg/L) 5.5 ± 0.2 5.7 ± 0.4 6.3 ± 0.5 5.1 ± 0.2 0.092 0.278 0.072

QUICKI2 0.357±0.014 0.346±0.014 0.341±0.007 0.349±0.013 0.636 0.835 0.298a,b, x,y Different superscripts indicate differences between groups at experimental week 6 (a, b) or week 12 (x, y) (P < 0.05). 1 HDL = high-density lipoprotein; LDL = low-density lipoprotein; NEFA = nonesterified fatty acids. 2 QUICKI = quantitative insulin sensitivity check index (see Material and Methods for details).

Table 2.3 | Postprandial responses (means ± SEM) of glucose, insulin, and triglycerides in veal calves fed either a high-lactose (HL) milk replacer diet (n = 6) or a high-fat (HF) milk replacer diet (n = 5).

Treatment

Item1 HL HF P-value

Plasma glucose concentration

Fasting values (mmol/L) 5.0 ± 0.2 5.4 ± 0.2 0.071

Δcmax (mmol/L) 4.7 ± 0.7 1.6 ± 0.3 0.011

Tmax (min) 75 ± 23 78 ± 28 0.739

iAUC0-7h 1032 ± 224 369 ± 71 0.029

Plasma insulin concentration

Fasting values (mU/L) 11.4 ± 4.0 7.8 ± 2.7 0.478

ΔCmax (mU/L) 398 ± 82 137 ± 41 0.024

Tmax (min) 170 ± 13 144 ± 29 0.405

iAUC0-7h x 103 79 ± 22 26 ± 6 0.073

Plasma triglycerides concentration

Fasting values (mmol/L) 0.06 ± 0.01 0.10 ± 0.03 0.211

ΔCmax (mmol/L) 0.20 ± 0.05 0.49 ± 0.12 0.036

Tmax (min) 150 ± 29 168 ± 22 0.646

iAUC0-7h 41 ± 2 83 ± 13 0.0051ΔCmax = maximum concentration minus fasting concentration; Tmax = time to maximum concentration;iAUC0−7h = incremental area under the curve. The milk replacer compositions are given in Table 1.

CHAPTER 2

40

2

2.4 Discussion

2.4.1 Insulin Sensitivity and Insulin Secretion

Whole-body insulin sensitivity values were in agreement with results obtained in calves of similar age (Sternbauer et al., 1998a,b; Sternbauer and Luthman, 2002). Cut-off M/I-values for defining insulin resistance in humans during a euglycemic–hyperinsulinemic clamp (cutoff values for calves have not been defined until now) are set at values < 4.4 to 5.0 (Huan et al., 2010; Lind et al., 1995) Compared with these values, our results, with 50% of the calves having M/I-values < 4.4, suggest that our calves were low in insulin sensitivity. This finding corresponds with those from other veal calf studies (Hostettler-Allen et al., 1994; Hugi et al., 1997; Pantophlet et al., 2016a). Increasing the contribution of fat to the digestible energy intake from the MR from 20 to 50% (at the expense of lactose) did not differentially affect whole-body insulin sensitivity, suggesting that either none or both of these dietary energy sources are equally responsible for the low insulin sensitivity observed in veal calves. These results are in contrast to Palmquist et al. (1992), who studied the effects of high-fat (18%) and low-fat (10%) diets on the development of insulin sensitivity in 8- to 16-week-old veal calves. They found that the high-fat diet did not lead to a decrease in insulin sensitivity, whereas the low-fat diet did lead to a decrease. Their measurements of whole-body insulin sensitivity were, however, performed in the fed state (4 h after feeding), when plasma glucose and insulin levels are still elevated, whereas our measurements were performed in the fasting state. Also, the glucose utilization rate, which was used as a measure for insulin sensitivity during an intravenous glucose tolerance test, was not corrected for the total insulin response in their study. When correcting their data for the insulin response (glucose utilization rate/AUCinsulin), a < 1% difference in insulin sensitivity remains, suggesting that in their study too, fat intake did not substantially affect insulin sensitivity. The absence of a differential effect of dietary source (fat vs. lactose) on whole-body insulin sensitivity could be explained by the low insulin sensitivity in calves older than 3 months (Pantophlet et al., 2016b). In a previous study, we found that replacing 33% of the lactose in MR by glucose, fructose, or glycerol affected postprandial glucose and insulin homeostasis but had no effect on whole-body insulin sensitivity (Pantophlet et al., 2016a). In young calves (< 3 months), insulin sensitivity decreases substantially during the first weeks of life (Stanley et al., 2002; Bach et al., 2013), and the decrease is not affected by extreme contrasts in feeding strategy (e.g., weaning vs. MR feeding; Pantophlet et al., 2016b), indicating that the reduction in insulin sensitivity in calves is dominated by their ontogenetic development. Nonetheless, a higher level of MR feeding may still reduce insulin sensitivity in early life of calves (Yunta et al., 2015). In the current study, insulin secretion (at identical plasma glucose levels) did not differ between treatments groups, indicating that the dietary energy source does not affect the capacity of calves to respond to glucose. This finding corresponds with previous findings in veal calves that were fed a high-lactose vs. a control MR (Hugi et al., 1998).


41

2

Remarkably, compared with glucose, the response of insulin to a MR meal or glucose infusion is delayed in calves (i.e., Tmax insulin > Tmax glucose), as seen during the hyperglycemic clamp and meal response in our study and other calf studies (Hostettler-Allen et al., 1994; Pantophlet et al., 2016a,b). This delay is, however, not found in humans, rats, and pigs (where Tmax insulin ≤ Tmax for glucose; Seki et al., 2005; Eelderink et al., 2012; Souza da Silva et al., 2014). The reason for the delayed insulin response in calves is unclear and warrants further study.

2.4.2 Postprandial blood glucose, insulin, and triglyceride homeostasis and glucosuria

The maximal postprandial blood glucose concentrations (ΔCmax) and cumulative increase in plasma glucose during the postprandial period (iAUC0−7h) were higher for HL calves than for HF calves, which was expected because the diet contained significantly larger amounts of lactose. Correspondingly, maximal insulin concentrations were higher for HL calves and the iAUC0−7h for plasma insulin tended to be higher. Similarly, the more pronounced postprandial plasma triglyceride response and iAUC0−7h for triglycerides in HF calves can be associated with the larger amounts of dietary fat.

Urinary glucose excretion was higher in HL calves than in HF calves, which corresponds to the higher postprandial blood glucose response in HL calves. Glucose excretion is frequently observed in veal calves (Hugi et al., 1997; Vicari et al., 2008; Pantophlet et al., 2016a). A significant correlation (r = 0.78; P < 0.01) was found between the urinary glucose excretion and iAUC0−7h for plasma glucose, confirming the association between blood glucose and urinary glucose excretion. Urinary glucose excretion occurs when the renal threshold for glucose reabsorption is exceeded. This threshold varies strongly among individuals, but average values of 8.3 to 11.1 mmol/L have been reported for calves (Wijayasinghe et al., 1984; Scholz and Hoppe, 1987; Hostettler-Allen et al., 1994). Hence, based on these values, we did not expected urinary glucose excretion in HF calves. In comparison, calves that were fed a MR in which 33% of the lactose was replaced by fructose had higher postprandial plasma glucose levels (peak ~8.1 vs. 6.8 mmol/L for HF calves), but only a negligible amount of urinary glucose excretion was found (~3.6 vs. 21.0 g/d for HF calves; Pantophlet et al., 2016a). The presence of glucose in urine of HF calves cannot be explained with our data, but it might be related to possible (long-term) effects of high dietary fat on renal function.

2.4.3 Fasting blood glucose, insulin, triglyceride, NEFA, cholesterol, and malondialdehyde

Fasting blood glucose and insulin concentrations did not change over time and did not differ between treatment groups. These results are in agreement with Hostettler-Allen et al. (1994), but in contrast to Palmquist et al. (1992), Hugi et al. (1997), and Pantophlet et al. (2016a), who found an age-related increase in fasting blood insulin concentrations. The QUICKI (an index of insulin sensitivity calculated from fasting

CHAPTER 2

42

2

blood glucose and insulin concentrations) was not affected by treatment, which is in agreement with the results on whole-body insulin sensitivity obtained from the euglycemic-hyperinsulinemic clamp.

Fasting NEFA concentrations, which are associated with fat catabolism, were not altered during the trial. These finding are in agreement with previous calf studies (Hostettler-Allen et al., 1994; Hugi et al., 1997). Fasting NEFA concentrations also did not differ between dietary treatments (HL vs. HF), indicating that exchanging a part of the dietary lactose with fat does not induce fat catabolism in the fasting state.

Fasting HDL- and LDL-cholesterol concentrations were higher in HF calves than in HL calves during the whole trial period. High amounts of fat intake are known to increase blood cholesterol concentrations (Hostettler-Allen et al., 1994; Siri-Tarino et al., 2010), which might explain the higher cholesterol concentrations in HF calves. In humans, the triglyceride/HDL cholesterol ratio is used to characterize dyslipidemia, which is associated with insulin resistance (Li et al., 2008; Bitzur et al., 2009). Whether this relationship is also true for calves is unknown. Nevertheless, triglyceride/HDL cholesterol ratios were low (< 1), suggesting that the current study provided no evidence of dyslipidemia.

Malondialdehyde estimates the level of lipid peroxidation and is a marker of oxidative stress. (Vlková and Celec, 2009). Oxidative stress has been associated with insulin resistance and type 2 diabetes (Tangvarasittichai, 2015). During our study, however, fasting plasma malondialdehyde concentrations were not correlated with QUICKI (i.e., whole-body insulin sensitivity; data not shown), suggesting that oxidative stress does not influence whole-body insulin sensitivity in veal calves.

2.4.4 Liver and Muscle Triglyceride Content

Liver triglyceride concentrations in our study were lower than in the veal calf study by Gerrits et al. (2008), who reported values between 25.5 and 42.5 g/kg (vs. 9.2 and 20.0 g/kg in the current trial). Liver triglyceride concentrations were 46% higher in HF calves than in HL calves. Excessive fat accumulation in the liver (i.e., hepatic steatosis) is a common metabolic disorder in humans and animals (Bogin et al., 1986), and it has also been linked to insulin resistance and obesity (Gerrits et al., 2008; Fabbrini and Magkos, 2015). Hence, controlling hepatic fat accumulation may improve metabolic health. Triglyceride concentrations in the rectus femoris muscle were similar to previously reported concentrations in the rectus abdominis muscle (5.4 and 7.6 kg/g) and lower than in the semitendinosus muscle (13.5 and 23.9 kg/g) in veal calves (van den Borne et al., 2007). In our trial, muscle triglyceride concentrations did not differ between dietary treatments, which suggests that the dietary source (lactose vs. fat) is unlikely to affect intramuscular fat accumulation in veal calves. Neither the liver nor the muscle triglyceride (i.e., fat) concentrations correlated with QUICKI measured near the end of the trial (i.e., week 12; data not shown), which suggests that the amount of fat in the liver and muscle does not directly influence whole-body insulin sensitivity in veal calves.


43

2

2.5 Conclusions

Increasing the contribution of fat to the digestible energy intake from the MR from 20 to 50% at the expense of lactose decreased postprandial glucose and insulin homeostasis and reduced urinary glucose losses. Whole-body insulin sensitivity and insulin secretion were not, however, affected by dietary energy source (i.e., high-lactose vs. high-fat MR). The high-fat MR increased liver fat accumulation.

2.6 Acknowledgments

The authors thank the Dutch Technology Foundation STW, which is part of the Netherlands Organisation for Scientific Research (NWO) and is partly funded by the Ministry of Economic Affairs, and the Product Board Animal Feed (The Hague, the Netherlands) for financially supporting this research. Sven Alferink, Harma Berends, Tamme Zandstra (Animal Nutrition Group, Wageningen University, the Netherlands), Caroline Breit-Corbière (Ecole d’Ingénieurs de Purpan, Toulouse, France), Etienne Labussière (INRA-UMR Pegase, Saint-Gilles, France), and Maria Shipandeni (Department of Animal Science, University of Namibia, Windhoek, Namibia) are gratefully acknowledged for their technical and scientific contributions.

CHAPTER 2

44

2

2.7 References

Bach, A., L. Domingo, C. Montoro, and M. Terre. 2013. Short communication: Insulin responsiveness is affected by the level of milk replacer offered to young calves. J. Dairy Sci. 96:4634–4637.

Bitzur, R., H. Cohen, Y. Kamari, A. Shaish, and D. Harats. 2009. Triglycerides and HDL cholesterol stars or second leads in diabetes? Diabetes Care. 32 (Suppl 2):373-377.

Bogin, E., Y. Avidar, and M. Merom. 1986. Biochemical changes in liver and blood during liver fattening in rats. J. Clin. Chem. Clin. Biochem. 24:621-626.

Chen, H., G. Sullivan, L. Q. Yue, A. Katz, and M. J. Quon. 2003. QUICKI is a useful index of insulin sensitivity in subjects with hypertension. Am. J. Physiol. Endocrinol. Metab. 284:E804-E812.

da Silva, S., D. Haenen, S. J. Koopmans, G. J. Hooiveld, G. Bosch, J. E. Bolhuis, B. Kemp, M. Müller, and W. J. J. Gerrits. 2014. Effects of resistant starch on behaviour, satiety-related hormones and metabolites in growing pigs. Animal 8:1402-1411.

Eelderink, C., M. Schepers, T. Preston, R. J. Vonk, L. Oudhuis, and M. G. Priebe. 2012. Slowly and rapidly digestible starchy foods can elicit a similar glycemic response because of differential tissue glucose uptake in healthy men. Am. J. Clin. Nutr. 96:1017-1024.

Eggstein, M. and F. H. Kreutz. 1966. Eine neue Bestimmung der Neutralfette im Blutserum und Gewebe. Klin. Wochenschr.:262-273.

Fabbrini, E. and F. Magkos. 2015. Hepatic steatosis as a marker of metabolic dysfunction. Nutrients 7:4995-5019.

Frayn, K. N. 2003. The glucose-fatty acid cycle: a physiological perspective. Biochem. Soc. Trans. 31:1115-1119.Gerrits, W. J. J., J. J. G. C. van den Borne, and J. W. Blum. 2008. Low-dietary protein intake induces problems

with glucose homeostasis and results in hepatic steatosis in heavy milk-fed calves. Domest. Anim. Endocrinol. 35:121-129.

Harris, R. B. S. 1992. Adipocyte insulin responsiveness in female Sprague-Dawley rats fed a low fat diet containing a fat-mimetic carbohydrate. J. Nutr. 122:1802-1810.

Harris, R. B. S. and H. Kor. 1992. Insulin insensitivity is rapidly reversed in rats by reducing dietary fat from 40 to 30% of energy. J. Nutr. 122:1811-1822.

Hof, G. 1980. An investigation into the extent to which various dietary components, particularly lactose, are related to the incidence of diarrhoea in milk-fed calves. Vol. Doctoral Thesis. Wageningen Agricultural University, Wageningen.


Hotamisligil, G. S. 2006. Inflammation and metabolic disorders. Nature 444:860-867.Huan, Y., S. DeLoach, C. Daskalakis, S. R. Dunn, K. Sharma, and B. Falkner. 2010. Regulation of transforming

growth factor-b1 by insulin in prediabetic African Americans. Kidney Int. 78:318-324.Hugi, D., R. M. Bruckmaier, and J. W. Blum. 1997. Insulin resistance, hyperglycemia, glucosuria, and



Le Floch, J. P., P. Escuyer, E. Baudin, D. Baudon, and L. Perlemuter. 1990. Blood glucose area under the curve. methodological aspects. Diabetes Care. 13:172-175.

Li, C., E. S. Ford, Y. X. Meng, A. H. Mokdad, and G. M. Reaven. 2008. Does the association of the triglyceride to high-density lipoprotein cholesterol ratio with fasting serum insulin differ by race/ethnicity? Cardiovasc. Diabetol. 7:4.

Lind, L., C. Berne, and H. Lithell. 1995. Prevalence of insulin resistance in essential hypertension. J. Hypertens. 13:1457-1462.

Müller, M. and S. Kersten. 2003. Nutrigenomics: goals and strategies. Nat. Rev. Genet. 4:315-322.Muniyappa, R., S. Lee, H. Chen, and M. J. Quon. 2008. Current approaches for assessing insulin sensitivity

and resistance in vivo: advantages, limitations, and appropriate usage. Am. J. Physiol. Endocrinol. Metab. 294:E15-E26.

Palmquist, D. L., J. Doppenberg, K. L. Roehrig, and D. J. Kinsey. 1992. Glucose and insulin metabolism in ruminating and veal calves fed high and low fat diets. Domest. Anim. Endocrinol. 9:233-241.


45

2

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, H. Roelofsen, M. G. Priebe, and R. J. Vonk. 2016a. Lactose in milk replacer can partly be replaced by glucose, fructose or glycerol without affecting insulin sensitivity in veal calves. J. Dairy Sci. 99:3072-3080.

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, M. G. Priebe, and R. J. Vonk. 2016b. Insulin sensitivity in calves decreases substantially during the first three months of life and is unaffected by weaning or fructo-oligosaccharide supplementation. J. Dairy Sci. in press.

Randle, P. J., P. B. Garland, C. N. Hales, and E. A. Newsholme. 1963. The glucose fatty-acid cycle. Its role in insulin sensitivity and the metabolic disturbances of diabetes mellitus. Lancet 1:785-789.

Roehrig, K., K. E. Nestor Jr, and D. L. Palmquist. 1988. ATP citrate lyase activity in liver and adipose tissue of veal or ruminating calves (Bos taurus). Comp. Biochem. Physiol. B 90:147-149.

Scholz, H. and S. Hoppe. 1987. Renal glucose loss after intravenous glucose-infusion in calves. Dtsch. Tierarztl. Wochenschr. 94:473-476.

Schwarz, J. M., P. Linfoot, D. Dare, and K. Aghajanian. 2003. Hepatic de novo lipogenesis in normoinsulinemic and hyperinsulinemic subjects consuming high-fat, low-carbohydrate and low-fat, high-carbohydrate isoenergetic diets. Am. J. Clin. Nutr. 77:43-50.

Seki, T., R. Nagase, M. Torimitsu, M. Yanagi, Y. Ito, M. Kise, A. Mizukuchi, N. Fujimura, K. Hayamizu, and T. Ariga. 2005. Insoluble fiber is a major constituent responsible for lowering the post-prandial blood glucose concentration in the pre-germinated brown rice. Biol. Pharm. Bull. 8:1539-1541.


Siri-Tarino, P. W., Q. Sun, F. B. Hu, and R. M.K. 2010. Saturated fatty acids and risk of coronary heart disease: modulation by replacement nutrients. Curr Atheroscler Rep. 12:384-390.

Stanley, C. C., C. C. Williams, B. F. Jenny, J. M. Fernandez, H. G. Bateman, W. A. Nipper, J. C. Lovejoy, D. T. Gantt, and G. E. Goodier. 2002. Effects of feeding milk replacer once versus twice daily on glucose metabolism in holstein and jersey calves. J. Dairy Sci. 85:2335-2343.

Sternbauer, K., J. Luthman, A. Hänni, and S. O. Jacobsson. 1998a. Clenbuterol-induced insulin resistance in calves measured by hyperinsulinemic, euglycemic clamp technique. Acta Vet. Scand. 39:281-289.

Sternbauer, K., J. Luthman, and S. O. Jacobsson. 1998b. Flumethasone-induced insulin resistance in calves. Zentralbl. Veterinarmed. A 45:441-443.

Sternbauer, K. and L. Luthman. 2002. Insulin Sensitivity of Heifers on Different Diets. Acta Vet. Scand. 43:107-114.

Storlien, L. H., L. A. Baur, A. D. Kriketos, D. A. Pan, G. J. Cooney, A. B. Jenkins, G. D. Calvert, and L. V. Campbell. 1996. Dietary fats and insulin action. Diabetologia 39:621-631.

Tangvarasittichai, S. 2015. Oxidative stress, insulin resistance, dyslipidemia and type 2 diabetes mellitus. World J Diabetes 6:456-480.

van den Borne, J. J. G. C., G. E. Lobley, M. W. A. Verstegen, J. M. Muijlaert, S. J. J. Alferink, and W. J. J. Gerrits. 2007. Body fat deposition does not originate from carbohydrates in milk-fed calves. J. Nutr. 137:2234-2241.

van den Borne, J. J. G. C., M. W. A. Verstegen, S. J. J. Alferink, F. H. M. van Ass, and W. J. J. Gerrits. 2006. Synchronizing the availability of amino acids and glucose decreases fat retention in heavy preruminant calves. J. Nutr. 136:2181-2187.

Van Es, A. J. H., H. J. Nijkamp, E. J. Van Weerden, and K. K. Van Hellemond. 1967. Energy, carbon and nitrogen balance experiments with veal calves. Pages 197-201 in Energy Metabolism of Farm Animals. K. L. Blaxter, J. Kielanowski, and G. Thorbek, ed. Oriel Press Newcastle-upon-Tyne, UK.

Vicari, T., J. J. G. C. van den Borne, W. J. J. Gerrits, Y. Zbinden, and J. W. Blum. 2008. Separation of protein and lactose intake over meals dissociates postprandial glucose and insulin concentrations and reduces postprandial insulin responses in heavy veal calves. Domest. Anim. Endocrinol. 34:182-195.

Vlková, B. and P. Celec. 2009. Does Enterococcus faecalis contribute to salivary thiobarbituric acid-reacting substances? In Vivo 23:343-345.

Wijayasinghe, M. S., N. E. Smith, and R. L. Baldwin. 1984. Growth, health, and blood glucose concentrations of calves fed high-glucose or high-fat milk replacers. J. Dairy Sci. 67:2949-2956.

Yunta, C., M. Terré, and A. Bach. 2015. Short- and medium-term changes in performance and metabolism of dairy calves offered different amounts of milk replacers. Livest Sci 181:249-255.

CHAPTER 3

The use of metabolic profiling to identify insulin resistance in veal calves

“When you can measure what you are speaking about, and express it in numbers, you know something about it”

- Lord Kevin -

A.J. Pantophlet,1 H. Roelofsen,3 M.P. de Vries,1,3 W.J.J. Gerrits,2 J.J.G.C. van den Borne,2 and R.J. Vonk3

Adapted fromPLoS One 2017; 12 (6): e0179612



CHAPTER 3

48

3

Abstract

Heavy veal calves (4-6 months old) are at risk of developing insulin resistance and disturbed glucose homeostasis. Prolonged insulin resistance could lead to metabolic disorders and impaired growth performance. Recently, we discovered that heavy Holstein-Friesian calves raised on a high-lactose or high-fat diet did not differ in insulin sensitivity, that insulin sensitivity was low and 50% of the calves could be considered insulin resistant. Understanding the patho-physiological mechanisms underlying insulin resistance and discovering biomarkers for early diagnosis would be useful for developing prevention strategies. Therefore, we explored plasma metabolic profiling techniques to build models and discover potential biomarkers and pathways that can distinguish between insulin resistant and moderately insulin sensitive veal calves. The calves (n = 14) were classified as insulin resistant (IR) or moderately insulin sensitive (MIS) based on results from a euglycemic-hyperinsulinemic clamp, using a cut-off value (M/I-value < 4.4) to identify insulin resistance. Metabolic profiles of fasting plasma samples were analyzed using reversed phase (RP) and hydrophilic interaction (HILIC) liquid chromatography–mass spectrometry (LC-MS). Orthogonal partial least square discriminant analysis was performed to compare metabolic profiles. Insulin sensitivity was on average 2.3x higher (P <0.001) in MIS than IR group. For both RP-LC-MS and HILIC-LC-MS satisfactory models were built (R2Y > 90% and Q2Y > 66%), which allowed discrimination between MIS and IR calves. A total of 7 and 20 metabolic features (for RP-LC-MS and HILIC-LC-MS respectively) were most responsible for group separation. Of these, 7 metabolites could putatively be identified that differed (P < 0.05) between groups (potential biomarkers). Pathway analysis indicated disturbances in glycerophospholipid and sphingolipid metabolism, the glycine, serine and threonine metabolism, and primary bile acid biosynthesis. These results demonstrate that plasma metabolic profiling can be used to identify insulin resistance in veal calves and can lead to underlying mechanisms.

Key words: veal calves, insulin resistance, metabolomics, biomarkers

Metabolic profiling of insulin resistance

49

3

3.1 Introduction

Veal calves are fed milk replacer (MR), roughage and concentrates. A large portion (60-70%) of the digestible nutrient intake originates from MR. The MR contains large amounts of lactose and fat, approximately 45% lactose and 20% fat on DM basis. Persistently high intakes of lactose and fat may lead to dysregulations in glucose homeostasis, which are characterized by a high incidence of hyperglycemia, hyperinsulinemia and glucosuria. These problems have been identified in heavy (4-6 months old) veal calves (Hostettler-Allen et al., 1994; Hugi et al., 1997; Pantophlet et al., 2016c). In addition, a substantial decrease in insulin sensitivity is observed in calves during the first months of life (Stanley et al., 2002; Pantophlet et al., 2016b). In a recent study with heavy veal calves raised on a high-fat or high-lactose MR diet we observed that insulin sensitivity values were low (averaging 4.2 ± 0.5 x 10−2 [(mg/(kg*min))/(μU/mL)]), and 50% of the calves develop insulin resistance (when comparing insulin sensitivity values with human cut-off values for defining insulin resistance; Pantophlet et al., 2016a). In order to prevent the development of insulin resistance, it is of importance to understand the patho-physiological mechanisms of insulin resistance and to identify early biomarkers of decreased insulin sensitivity. By detecting decreased insulin sensitivity at an early stage, management and feeding strategies could be developed to prevent the development insulin resistance. Therefore, we investigated the applicability of metabolomic profiling techniques to identify insulin resistance in veal calves.

Metabolomics focuses on the analysis of the metabolome together with pattern recognition techniques to highlight and monitor metabolic changes related to disease status or nutritional intervention (Berger, 2013; Pantophlet et al., 2016d). Its potential has been demonstrated in the diagnosis of several metabolic diseases (Gowda et al., 2008; Zhang et al., 2013; Graham et al., 2015). In the current exploratory study we applied metabolic profiling, to build models to discover potential biomarkers and pathways related to insulin resistance in veal calves.

3.2 Material and Methods


Sixteen male Holstein-Friesian calves (120 ± 2.8 kg BW; 99 ± 2.0 days old) were purchased and housed at the experimental facilities of Wageningen University. During the first 6 weeks of the 13-week study, calves were housed in pens of 4 calves each (2 m2 per calf), which were fitted with a wooden slatted floor and galvanized fencings. Then, calves were transferred to metabolic cages (dimension: 0.80 x 1.8 m) for the next 7 weeks, during which whole-body insulin sensitivity was measured (see experimental procedures). Ventilation occurred by ceiling fans, and illumination by natural light

CHAPTER 3

50

3

and artificial (fluorescent lamps) light between 0700 and 1900 h. Temperature and humidity were controlled at 18˚C and 65% respectively.

The study was conducted in 2011. Experimental procedures complied with the Dutch Law on Experimental Animals, and the ETS123 (Council of Europe 1985 and the 86/609/EEC Directive) and were approved by the Animal Care and Use Committee of Wageningen University.

3.2.2 Experimental design, diets and feeding

A detailed description of the experimental design, diets and feeding were described previously (Pantophlet et al., 2016a). Briefly, calves were assigned to either a high-lactose diet (HL; n = 8) or a high-fat diet (HF; n = 8), and to 1 of 8 blocks (pairs of calves) with one HL calf and one HF calf per block. Due to health problems in two HF calves, block seven consisted of two HL calves and block 8 (with the two remaining HF calves) was not included in the whole-body insulin sensitivity and metabolomic profiling measurements. Lactose and fat were exchanged iso-energetically between treatments based on digestible energy. MR was fed on individual basis twice a day (0800 and 1630 h). In addition, solid feed was provided per pen when calves were housed in groups and per individual calf when housed separately on metabolic cages. Solid feed was provided once a day. Calves had ad libitum access to drinking water throughout the study. At end of the study, calves were euthanized by an intravenous injection of sodium pentobarbital.


A detailed description of the experimental procedures were given elsewhere (Pantophlet et al., 2016a). In short, whole-body insulin sensitivity was assessed by the euglycemic-hyperinsulinemic clamp technique in seven consecutive weeks (i.e., experimental week 7-13; 1 block per week). Semi-permanent catheters (Careflow, Becton Dickinson, Alphen aan den Rijn, The Netherlands) were inserted in both jugular veins. Calves were fasted for 15 h (morning feed omitted) to achieve a steady glucose turnover rate prior to the measurements. Before starting the 4-h clamp study, three 5 mL blood samples were taken from -40 to -10 min (before infusion) to determine basal plasma glucose concentrations. At start of the clamp, a priming dose of insulin of 2.1 mU/kg BW/min (Actrapid 100 IE/mL, Novo Nordisk, Denmark) was infused into the left jugular vein catheter, within 5 min, to rapidly increase the plasma insulin concentration. Then, the rate of insulin infusion was decreased and maintained at 1 mU/kg BW/min for a period of 4 h (plasma insulin levels ~135 mU/L). At t = 5 min glucose (20% glucose solution for intravenous infusion; B. Braun, Melsungen, Germany) was continuously infused to maintain basal plasma glucose concentration, hence the infusion rate was adjusted to the glucose clearance rate.

During the clamp study, 0.3 mL blood samples were taken from the catheter in 10-min and 15-min intervals during 0-2 h and 2-4 h respectively. In these samples,


51

3

plasma glucose concentrations were measured using Precision Xtra Plus test strips in combination with the Precision Exceed Sensor (Abbott, Weesp, The Netherlands).

In addition, 5 mL blood samples were taken in 30-min intervals for the analysis of plasma glucose and insulin concentrations. Blood was collected in sodium fluoride vacutainer tubes for glucose and in heparin vacutainer tubes for insulin (BD diagnostics, Breda, The Nether- lands). Samples were centrifuged (1,500 x g for 12 min) and plasma was harvested and stored at -20˚C until analysis.

Plasma glucose was analyzed on an Architect ci8200 analyzer using the hexokinase method (Abbott Laboratories, Chicago, IL, USA) and plasma insulin was analyzed using a Coat-a-Count radioimmunoassay kit (Siemens Healthcare Diagnostics, Erlangen, Germany). The within- and between-run coefficients of variation for glucose were ≤ 2%. The within and between-run coefficients of variation for insulin were ≤ 5% and ≤ 7%, respectively.

The glucose infusion rate (GIR) was adjusted (depending on the changes in plasma glucose level) to maintain a constant, basal plasma glucose level during insulin infusion. Glucose disposal (M-value) was defined as the average GIR at steady state divided by BW. Whole-body insulin sensitivity was defined as the M-value divided by the average plasma insulin level at steady state (M/I-value).

3.2.4 Metabolomic profiling

Plasma metabolic profiling was performed using reversed phase (RP) and hydrophilic interaction (HILIC) liquid chromatography–mass spectrometry (LC-MS). These techniques are complementary, with RP-LC-MS able to separate and detect non-polar to weakly polar metabolites and HILIC-LC-MS able to separate and detect weakly polar to polar metabolites. LC-MS was performed using a UFLC Prominence system (Shimadzu, Kyoto, Japan) coupled to a high-resolution LTQ-Orbitrap XL mass spectrometer (Thermo Fisher Scientific, Bremen, Germany), equipped with an Ion Max electrospray source. Analyses were performed in both positive and negative ionization mode. Mass spectrometric data was acquired in the centroid mode over the range of 100-800 m/z at a resolution of 60,000 at m/z 400. Low-resolution collision induced dissociation fragmentation data using the LTQ (MS/MS) was also acquired to facilitate compound identification in a TOP-n data dependent acquisition.

3.2.4.1 Sample preparationPlasma samples were allowed to thaw at 4˚C for 6 h. Then, 800 μL of a methanol/acetonitrile/acetone (1:1:1 v/v) solution was added to 200 μL plasma each study sample. The mixture was gently vortexed at 4˚C for 15 min and centrifuged at 12,500 x g for 10 min at 4˚C. Then, 800 μL of the supernatant was evaporated to dryness under a gentle stream of nitrogen at 30˚C. The residue was reconstituted in 100 μL methanol and 300 μL elution solvent A (see below for composition) for RP and in 400 μL elution solvent B (see below for composition) for HILIC. In addition, a quality control (QC) sample

CHAPTER 3

52

3

and a blank sample were prepared. The QC sample was prepared by mixing 100 μL of each study sample (to represent the biochemical diversity of the study samples), and processed identical to the study samples. For the blank sample Milli-Q water was used. Sample processing was identical to the study samples.

3.2.4.2 Sequence of injectionThe analytical run started with the blank sample (injected 3 times for background subtraction), followed by the QC sample (injected 6 times for column conditioning; not used for data analysis). Then, the study samples were injected in random order. The QC sample was injected again after every 3 study samples (and at the end of the run) to calculate the analytical precision for each metabolic feature.

3.2.4.3 Reversed-phase chromatographyFor reversed-phase chromatography a Kinetex C18 column (100 mm × 2.1 mm, 2.6 μm particles) with a SecurityGuard column (2.1 mm × 2 mm, 2 μm particles) was used (Phenomenex, Torrance, CA, USA). The column temperature was set at 35˚C and the autosampler temperature was 5˚C. The gradient elution solvents were A; 95:5 water-acetonitrile (v/v), containing 5 mM ammonium formate and 0.1% formic acid (v/v), and B; 95:5 acetonitrile-water, containing 5 mM ammonium formate and 0.1% formic acid. The gradient (A:B, v/v) was as follows: an isobaric period at 98:2 for 5 min, followed by a linear gradient from 98:2 to 2:98 in 25 min, then held at 2:98 for 5 min, followed by a linear gradient change from 2:98 to 98:2 in 1 min, then held at 98:2 for 5 min. The flow rate was 0.2 mL/min.

3.2.4.4 HILIC chromatography For HILIC chromatography a Kinetex HILIC column (100 mm × 2.1 mm, 2.6 μm particles) with a SecurityGuard column (2.1 mm × 2 mm, 2 μm particles) was used (Phenomenex, Torrance, CA, USA). The column temperature was set at 35˚C and the autosampler temperature was 5˚C. The gradient elution solvents were A; 95:5 water- acetonitrile (v/v), containing 5 mM ammonium formate and 0.1% formic acid (v/v), and B; 95:5 acetonitrile-water, containing 5 mM ammonium formate and 0.1% formic acid. The gradient (A:B, v/v) was as follows: an isobaric period at 5:95 for 5 min, followed by a linear gradient from 5:95 to 50:50 in 25 min, then held at 50:50 for 5 min, followed by a linear gradient change from 50:50 to 5:95 in 1 min, then held at 5:95 for 5 min. The flow rate was 0.2 mL/min.

3.2.4.5 Mass spectrometryThe electrospray MS settings for both RP and HILIC were as followed: spray voltage 4.5 kV for positive ionization mode (3 kV for negative mode) and the capillary temperature was set at 250˚C for positive ionization mode (250˚C for negative mode). Nitrogen sheath gas and auxiliary gas were set at 25 and 15 arbitrary units, respectively.


53

3

3.2.5 Data processing and statistical analysis

The raw LC-MS data were processed with Sieve 2.2 (Thermo Scientific) using the default settings except for the minimal signal to noise ratio for peak detection, which was set at 5. Metabolomic features with a coefficient of variation (of the normalized peak area) of the QC samples > 25% were excluded from the dataset. The spectral data was then exported to Excel and results of the positive and negative mode analyzed with the same technique (i.e., RP or HILIC) were merged. After data processing, a multivariate analysis was conducted using SIMCA-P (Umetrics, Sweden). The data was pareto-scaled and subjected to orthogonal projection to latent structures discriminant analysis (OPLS-DA). The quality and reliability of the models were assessed by R2Y, representing the explained variation described by the model, and Q2Y, representing the predictive power of the model (based on the default 7-round cross validation procedure used in SIMCA-P). Permutation tests (n = 100) were performed to assess the robustness of the models. Also, a CV-ANOVA was calculated to assess the reliability of the models.The variable importance in the projection (VIP) was used to identify the metabolic features that most significantly contributed to the clustering of groups within the OPLS-DA models (Xia et al., 2009). Metabolic features with a VIP > 2.0 were considered important. Also, an independent t-test was performed (using SPSS version 22, IBM, SPSS Inc., Chicago, IL) on all metabolic features with a VIP > 2.0 to highlight which of these metabolic features also differ at univariate level between groups. A P-value of < 0.05 was considered significant. Metabolic features with VIP > 2.0 and P < 0.05 were considered potential biomarkers.

3.2.6 Metabolite identification and pathway analysis

Metabolite identification was performed on the potential biomarkers, and was achieved by comparing spectral data (exact mass and MS/MS) with data available from the human metabolome database (http://hmdb.ca/) and the METLIN database (http://metlin.scripps.edu/). When available, in-house, putative ID’s were confirmed by comparison with authentic standards (retention time, exact mass and MS/MS).

Pathway analysis was performed on the (putatively) identified biomarkers to highlight pathways that were disturbed due to the reduced insulin sensitivity. Metaboanalyst (version 2.0), a web-based program that uses the KEGG (http://www.genome.jp/kegg/) pathway database (Xia et al., 2012) was used for analysis. The Bos Taurus library was chosen for pathway analysis.

CHAPTER 3

54

3

3.3 Results

3.3.1 Insulin sensitivity

Insulin sensitivity ranged from 1.5 to 8.3 x 10−2 (mg/(kg*min))/(μU/mL) between calves and was not differentially affected by dietary treatment (P > 0.05; Pantophlet et al., 2016a). 50% of all calves had an insulin sensitivity (M/I-value) < 4.4 and thus were considered insulin resistant (IR). Other calves were classified as moderately insulin sensitive (MIS). Whole-body insulin sensitivity differed substantially (P < 0.001) between IR and MIS calves (Table 3.1) and ranged from 2.1-3.8 in IR calves and 4.4-8.2 in MIS calves.

3.3.2 Metabolomic profiling

Metabolomic profiling of IR vs. MIS calves was performed on the fasting plasma samples collected on the day of the clamp study. The models obtained using OPLS-DA are shown in Figure 3.1. A total of 247 metabolic features were detected using the C18 RP LC-MS. The C18 model clearly distinguished MIS calves from IR calves. The non-orthogonal component of this model explained 92% of the variation (R2Y = 0.92). The predictive power of the model, measured by seven-fold cross validation was 66% (Q2Y = 0.66). The CV-ANOVA P-value was 0.03, indicating that the differences between the two groups were significant.

Table 3.1 | Characteristics of insulin resistant (IR) vs. moderately insulin sensitive (MIS) veal calves

IR MIS P-value4

High-Fat diet, n 4 2 -

High-Lactose diet, n 3 5 -

Age, days 169±10 165±15 0.750

BW, kg 248±7 241±6 0.483

Insulin1 (mU/L) 135±5 133±3 0.640

M-value2 (mg/BW/min) 3.4±0.3 7.7±0.6 < 0.001

M/I-value3 x 10-2 [mg/(kg*min)) / (μU/ml)] 2.6±0.3 5.8±0.5 < 0.0011 Plasma insulin concentration at steady state during a euglycemic-hyperinsulinemic clamp. 2 M-value = glucose disposal derived from a euglycemic-hyperinsulinemic clamp. 3 M/I-value = insulin sensitivity derived from a euglycemic-hyperinsulinemic clamp. Calves with a M/I-value <4.4 were considered insulin resistant. M/I-valuesIR ranged from 2.1-3.8 and M/I-valuesMIS ranged from 4.4-8.2; x 10-2 [mg/(kg*min)) / (μU/ml)]. 4 P-value was calculated from independent T-test.

Furthermore, permutation tests (n = 100) were performed to assess the robustness of the model. The validation plots confirmed that the model was valid and unlikely obtained by chance, as the permuted R2 and Q2 data were lower that original values, and Q2 had a negative intercept. A total of 625 metabolic features were detected using the HILIC LC-MS. The HILIC model could also distinguish MIS calves from IR calves. This


55

3

model explained 90% of the variation (R2Y = 0.90). The predictive power was 73% (Q2Y = 0.73) and the CV-ANOVA P-value was 0.01. Permutation tests (n = 100) confirmed that the model was valid and unlikely obtained by chance. Both the permuted R2 and Q2 data were lower that original values, and Q2 had a negative intercept.

A total of 7 and 20 metabolic features had a variable importance in the projection > 2.0 in the C18 and HILIC model, respectively (Supporting information 1). Of these metabolites, a total of 1 and 11 metabolic features differed (P < 0.05) between insulin sensitivity groups in the C18 and HILIC model, respectively (Table 3.2). Seven of these metabolic features decreased in IR calves and five increased. A total of 7 metabolites could be (putatively) identified. The chromatographic response of these metabolites (as a measure of the plasma concentration) is given in Figure 3.2.

3.3.3 Pathway analysis

Pathway analysis was performed on the 7 putatively identified metabolites using Metaboanalyst to highlight pathways possibly associated with insulin resistance in veal calves. The 7 putatively identified metabolites are involved in 4 pathways; the glycerophospholipid metabolism, sphingolipid metabolism, glycine, serine and threonine metabolism, and primary bile acid biosynthesis (Table 4.3).

Table 3.2 | Marker metabolites of insulin resistance found in OPLS-DA models of HILIC and C18 LC-MS plasma metabolic profiling of veal calves

m/z VIP1 P-value2 Metabolites3 Chemical class Fold change (IR/MIS)6

520.339 8.274 0.005 Lysophosphatidylcholine (18:2) Lysophospholipids 1.45

703.574 5.474 0.013 Sphingomyelins 0.70

104.107 4.484 0.002 Choline Cholines 1.19

204.123 3.654 0.031 Acetylcarnitine Acyl carnitines 0.50

185.127 3.614 0.005 Sphingomyelins 1.15

813.682 3.574 0.002 Sphingomyelin (d18:1/24:1) Sphingomyelins 0.49

498.288 2.915 0.017 Taurochenodeoxycholic acid Bile acids 0.44

811.668 2.474 0.001 Sphingomyelins 0.56

258.110 2.464 0.041 Glycerophosphorylcholine Glycerophosphorylcholines 1.50

815.698 2.184 0.000 Sphingomyelin (d18:0/24:1) Sphingomyelins 0.33

787.668 2.104 0.001 Sphingomyelins 0.95

564.330 2.024 0.034 1.341 Variable importance in the projection (VIP) obtained from OPLS-DA models with a threshold of ≥2.0.2 P-value was calculated from independent samples T-test. Threshold was set at P ≤0.05.3 (Putative) identification based on human metabolome database and METLIN database search combined with MS/MS fragmentation analysis, and in some cases, comparison with authentic standards (when available in house).4 VIP obtained from the HILIC OPLS-DA model.5 VIP obtained from the C18 OPLS-DA model.6 IR = Insulin resistant calves. MIS = moderately insulin sensitive calves.

CHAPTER 3

56

3

Figure 3.1 | OPLS-DA score plots from plasma metabolic profiles of moderately insulin sensitive and insulin resistant veal calves. The white triangles represent moderately insulin sensitive veal calves (n = 7) and black triangles represent insulin resistant veal calves (n = 7). The models were obtained using C18 LC-MS (A) and HILIC LC-MS (B) blood plasma metabolic profiling. R2Y, which is the variation described by the models was 92 and 90% for the C18 and HILIC model, respectively. Q2Y, which describes how accurately the models can predict class membership, was 66 and 73% for the C18 and HILIC model, respectively.

Table 3.3 | Metabolic pathway analysis of the putative identified marker metabolites of insulin resistance (Table 3.1) found in blood plasma of veal calves

Metabolic pathway1 Marker metabolite Trend2

Glycerophospholipid metabolism Lysophosphatidylcholine (18:2) ↑

Choline ↑

Glycerophosphorylcholine ↑

Sphingolipid metabolism Sphingomyelin (d18:1/24:1) ↓

Sphingomyelin (d18:0/24:1) ↓

Glycerine, serine and threonine metabolism Choline ↑

Primary bile acid biosynthesis Taurochenodeoxycholic acid ↓ 1 Metabolic pathway analysis was performed using MetaboAnalyst version 2.0. 2 ↑ and ↓ indicate that the marker was increased or decreased in insulin resistant claves compared with moderately insulin sensitive calves.


57

3

Figure 3.2 | Plasma response of the (putatively) identified biomarkers of insulin resistance in veal calves. The chromatographic peak area is a measure of the blood plasma concentration. MIS = moderately insulin sensitive veal calves (n = 7). IR = insulin resistant veal calves (n = 7). Further details of these two groups are given in Table 3.1. Error bars represent SEM.

CHAPTER 3

58

3

3.4 Discussion

Plasma metabolic profiling techniques have been applied to build models to find biomarkers and pathways that can identify insulin resistant veal calves and distinguish these calves from moderately insulin sensitive calves. To the best of our knowledge, this is the first time that metabolic profiling has been applied on veal calves to study insulin resistance. Satisfactory models (Q2Y = 66 and 73% for C18 and HILIC, respectively) were developed, that could clearly identify insulin resistant veal calves, and which could possibly be used in early diagnosis. The predictive powers of these models are slightly lower compared to human metabolic profiling studies (Q2Y = 76-93%; Zhang et al., 2009; Lucio et al., 2010). This might be attributed to multiple factors: 1] the small number of calves used in this study 2] possible differences in the metabolic profiling techniques used, and 3] possible differences in degree of the experimental contrasts in insulin sensitivity between human and calf studies. In veal calves, insulin sensitivity decreases substantially within the first weeks of life (Pantophlet et al., 2016b), which leads to smaller contrasts in insulin sensitivity in later life. One possible source of variation that can be excluded from subsequent studies is the use of multiple dietary treatments. Despite the fact that insulin sensitivity was not differentially affected by the dietary treatments (dietary treatments almost balanced out between classification groups), it could be that certain metabolites that are more strongly affected by dietary treatment. Nonetheless, the potential biomarkers found in this study were not affected by dietary treatment (as assessed by independent t-test; P-values > 0.05). In subsequent studies, it might be beneficial to restrict feeding to the standard (commercial) lactose MR diet. Another source of variation that can be excluded from subsequent studies is the possible effect of age on insulin sensitivity. In our study, insulin sensitivity was measured within a period of 7 weeks (2 calves per week). The possible effect of age on insulin sensitivity, however, was balanced out between classification groups, as age did not differ between groups. In subsequent studies, insulin sensitivity should be measured at the same time (day). Additionally, it would be interesting to measure insulin sensitivity in time (i.e., with age) in the same set of calves to study the discovered biomarkers and their association with the development of insulin resistance with age.

A M/I cut-off value of 4.4 was used to discriminate between insulin resistant and moderately insulin sensitive calves. This value was based on cut-off values for defining insulin resistance in humans, because cut-off values for calves are not established (Pantophlet et al., 2016a). A different cut-off value would have perhaps let to discovery of additional/other biomarkers (and pathways). Nevertheless, the cut-off value used in this study let to clear discrimination between insulin sensitivity groups (i.e., a low vs. moderate group). Therefore, the discovered biomarkers are related to differences in insulin sensitivity levels.

In our study, not all potential biomarkers could be identified. This is a well-known bottleneck of untargeted MS metabolic profiling techniques (Bowen and Northen, 2010; Lynn et al., 2015). Future studies should also consider including additional identification


59

3

techniques such as nuclear magnetic resonance spectroscopy.Interestingly, pathway analysis of the putatively identified potential biomarkers revealed

multiple disturbances in the glycerophospholipid and the sphingolipid metabolism. To the best of our knowledge, this is the first time that these pathways (and potential biomarkers) have been associated with insulin resistance in veal calves. This demonstrates the power of metabolic profiling in identifying markers and pathways that may be important in understanding the development of insulin resistance in calves. In dairy cows, both pathways have recently been associated with reduced insulin sensitivity (Humer et al., 2016). In humans, these pathways have frequently been associated with insulin resistance and type 2 diabetes (Straczkowski et al., 2004; Nolan et al., 2006; Larsen and Tennagels, 2014; Bergman et al., 2015; Meikle and Summers, 2016). A previous study has shown that different sphingolipids associate either positively or negatively with insulin resistance (Bergman et al., 2015). In human and rodents, sphingomyelins (a type of sphingolipid) patches on β-cells and predicts insulin secretory capacity (Kavishwar and Moore, 2013). Decreased glucose tolerance and insulin secretion have been observed in sphingomyelin synthase 1 knockout mice (Yano et al., 2011; Li et al., 2012). Our data also show that not all sphingomyelins are negatively associated with insulin resistance. The mechanisms behind the different associations warrant further study. Glycerophospholipids have also been associated with insulin resistance and type 2 diabetes. In human studies, both positive and negative associations have been found for metabolites related to glycerophospholipid metabolism (Motley et al., 2002; Gall et al., 2010; Suhre et al., 2010; Rauschert et al., 2016). In our study a positive association was found. Glycerophospholipids are major components of cell membranes. Disturbances in membrane glycerophospholipid metabolism could influence insulin secretion via alteration of the physico-chemical properties of the membrane (Nolan et al., 2006). However, clear mechanisms behind the associations of metabolites related to glycerophospholipid metabolism have not been identified yet. Future (mechanistic) studies on the development of insulin resistance in calves should apply a targeted lipidomic approach that specifically focuses on metabolites related to the glycerophospholipid and the sphingolipid metabolism.

3.5 Conclusion

Based on plasma metabolic profiling satisfactory models were developed that are capable of distinguishing veal calves differing in insulin sensitivity (i.e., moderate vs. insulin resistant/ extremely low insulin sensitive calves). Several metabolic alterations (potential biomarkers) were observed between the moderate and low insulin sensitive calves. These alterations were related to the glycerophospholipid metabolism, sphingolipid metabolism, glycine, serine and threonine metabolism, and primary bile acid biosynthesis. Future studies should be performed to study these pathways and biomarkers in early life (i.e., neonatal calves) and their association with the development of insulin resistance with age in veal calves.

CHAPTER 3

60

3

3.6 Acknowledgements

The authors thank Dutch Technology Foundation STW, which is part of the Netherlands Organisation for Scientific Research (NWO) and which is partly funded by the Ministry of Economic Affairs, and the Product Board Animal Feed (The Hague, The Netherlands) for financially supporting this research.

3.7 Supporting information

Supporting Information Table 1 | Metabolic features (with VIP > 2.0) found in OPLS-DA models of HILIC and C18 LC-MS plasma metabolic profiling of moderately insulin sensitive and insulin resistant veal calves

Metabolic feature m/z Mode1 VIP2

1 520.339 HILIC 8.272 703.574 HILIC 5.47

3 104.107 HILIC 4.48

4 524.370 HILIC 3.70

5 204.123 HILIC 3.65

6 185.127 HILIC 3.61

7 496.343 HILIC 3.59

8 524.374 HILIC 3.59

9 813.682 HILIC 3.57

10 498.289 HILIC 3.47

11 464.282 HILIC 2.97

12 522.355 HILIC 2.57

13 414.302 HILIC 2.53

14 811.668 HILIC 2.47

15 258.110 HILIC 2.46

16 432.311 HILIC 2.34

17 815.698 HILIC 2.18

18 116.071 HILIC 2.14

19 787.668 HILIC 2.10

20 564.330 HILIC 2.02

1 520.340 C18 6.79

2 496.339 C18 4.59

3 524.370 C18 4.53

4 789.619 C18 3.32

5 522.355 C18 3.14

6 498.288 C18 2.917 464.283 C18 2.00

1 Chromatographic mode used. HILIC = Hydrophilic interaction chromatography and C18 = reversed phase C18 chromatography.2 VIP = Variable importance in the projection, obtained from the OPLS-DA models


61

3

3.8 References

Berger, R. D. 2013. A review of applications of metabolomics in cancer. Metabolites 3:552-574.Bergman, B. C., J. T. Brozinick, A. Strauss, S. Bacon, A. Kerege, H. H. Bui, P. Sanders, P. Siddall, M. S. Kuo, and

L. Perreault. 2015. Serum sphingolipids: relationships to insulin sensitivity and changes with exercise in humans. Am. J. Physiol. Endocrinol. Metab. 309:E398-E408.

Bowen, B. P. and T. R. Northen. 2010. Dealing with the unknown: metabolomics and metabolite atlases. J. Am. Soc. Mass Spectrom. 21:1471-1476.

Gall, W. E., K. Beebe, K. A. Lawton, K. P. Adam, M. W. Mitchell, P. J. Nakhle, J. A. Ryals, M. V. Milburn, M. Nannipieri, S. Camastra, A. Natali, and E. Ferrannini. 2010. alpha-Hydroxybutyrate is an early biomarker of insulin resistance and glucose intolerance in a nondiabetic population. PLoS One 5:e10883.

Gowda, G. A. N., S. Zhang, H. Gu, V. Asiago, N. Shanaiah, and D. Raftery. 2008. Metabolomics-based methods for early disease diagnostics: A review. Expert Rev Mol Diagn. 8:617-633.

Graham, S. F., O. P. Chevallier, C. T. Elliott, C. Hölscher, J. Johnston, B. McGuinness, P. G. Kehoe, A. P. Passmore, and B. D. Green. 2015. Untargeted metabolomic analysis of human plasma indicates differentially affected polyamine and L-arginine metabolism in mild cognitive impairment subjects converting to Alzheimer’s disease. PLoS One 10:e0119452.


Hugi, D., R. M. Bruckmaier, and J. W. Blum. 1997. Insulin resistance, hyperglycemia, glucosuria, and galactosuria in intensively milk-fed calves: dependency on age and effects of high lactose intake. J. Anim. Sci. 75:469-482.

Humer, E., A. Khol-Parisini, B. U. Metzler-Zebeli, L. Gruber, and Q. Zebeli. 2016. Alterations of the lipid metabolome in dairy wows experiencing excessive lipolysis early postpartum. PLoS One 11:e0158633.

Kavishwar, A. and A. Moore. 2013. Sphingomyelin patches on pancreatic beta-cells are indicative of insulin secretory capacity. J Histochem Cytochem. 61:910-919.

Larsen, P. J. and N. Tennagels. 2014. On ceramides, other sphingolipids and impaired glucose homeostasis. Mol Metab. 3:252-260.

Li, Z., Y. Fan, J. Liu, Y. Li, C. Huan, H. H. Bui, M. S. Kuo, T. S. Park, G. Cao, and X. C. Jiang. 2012. Impact of sphingomyelin synthase 1 deficiency on sphingolipid metabolism and atherosclerosis in mice. Arterioscler. Thromb. Vasc. Biol. 32:1577-1584.

Lucio, M., A. Fekete, C. Weigert, B. Wägele, X. Zhao, J. Chen, A. Fritsche, H. U. Häring, E. D. Schleicher, G. Xu, P. Schmitt-Kopplin, and R. Lehmann. 2010. Insulin sensitivity is reflected by characteristic metabolic fingerprints--a Fourier transform mass spectrometric non-targeted metabolomics approach. PLoS One 5:e13317.

Lynn, K. S., M. L. Cheng, Y. R. Chen, C. Hsu, A. Chen, T. M. Lih, H. Y. Chang, C. J. Huang, M. S. Shiao, W. H. Pan, T. Y. Sung, and W. L. Hsu. 2015. Metabolite identification for mass spectrometry-based metabolomics using multiple types of correlated ion information. Anal. Chem. 87:2143-2151.

Meikle, P. J. and S. A. Summers. 2016. Sphingolipids and phospholipids in insulin resistance and related metabolic disorders. Nat. Rev. Endocrinol. [Epub ahead of print].

Motley, E. D., S. M. Kabir, C. D. Gardner, K. Eguchi, G. D. Frank, T. Kuroki, M. Ohba, T. Yamakawa, and S. Eguchi. 2002. Lysophosphatidylcholine inhibits insulin-induced Akt activation through protein kinase C-alpha in vascular smooth muscle cells. Hypertension 39:508-512.

Nolan, C. J., M. S. Madiraju, V. Delghingaro-Augusto, M. L. Peyot, and M. Prentki. 2006. Fatty acid signaling in the beta-cell and insulin secretion. Diabetes 55:S16-S23.

Pantophlet, A. J., W. J. J. Gerrits, R. J. Vonk, and J. J. G. C. van den Borne. 2016a. Substantial replacement of lactose with fat in a high-lactose milk replacer diet increases liver fat accumulation but does not affect insulin sensitivity in veal calves. J. Dairy Sci. 99:10022-10032.

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, M. G. Priebe, and R. J. Vonk. 2016b. Insulin sensitivity in calves decreases substantially during the first 3 months of life and is unaffected by weaning or fructo-oligosaccharide supplementation. J. Dairy Sci. 99:7602-7610.

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, H. Roelofsen, M. G. Priebe, and R. J. Vonk. 2016c. Lactose in milk replacer can partly be replaced by glucose, fructose or glycerol without affecting insulin sensitivity in veal calves. J. Dairy Sci. 99:3072-3080.

CHAPTER 3

62

3

Pantophlet, A. J., S. Wopereis, C. Eelderink, R. J. Vonk, J. H. Stroeve, S. Bijlsma, L. van Stee, I. Bobeldijk, and M. G. Priebe. 2016d. Metabolic profiling reveals differences in plasma concentrations of arabinose and xylose after consumption of fiber-rich pasta and wheat bread with differential rates of systemic appearance of exogenous glucose in healthy men. J. Nutr. [Epub ahead of print].

Rauschert, S., O. Uhl, B. Koletzko, F. Kirchberg, T. A. Mori, R. C. Huang, L. J. Beilin, C. Hellmuth, and W. H. Oddy. 2016. Lipidomics reveals associations of phospholipids with obesity and insulin resistance in young adults. J. Clin. Endocrinol. Metab. 101:871-879.

Stanley, C. C., C. C. Williams, B. F. Jenny, J. M. Fernandez, H. G. Bateman, W. A. Nipper, J. C. Lovejoy, D. T. Gantt, and G. E. Goodier. 2002. Effects of feeding milk replacer once versus twice daily on glucose metabolism in Holstein and Jersey calves. J. Dairy Sci. 85:2335-2343.

Straczkowski, M., I. Kowalska, A. Nikolajuk, S. Dzienis-Straczkowska, I. Kinalska, M. Baranowski, M. Zendzian-Piotrowska, Z. Brzezinska, and J. Gorski. 2004. Relationship between insulin sensitivity and sphingomyelin signaling pathway in human skeletal muscle. Diabetes 53:1215-1221.

Suhre, K., C. Meisinger, A. Döring, E. Altmaier, P. Belcredi, C. Gieger, D. Chang, M. V. Milburn, W. E. Gall, K. M. Weinberger, H. W. Mewes, M. Hrabé de Angelis, H. E. Wichmann, F. Kronenberg, J. Adamski, and T. Illig. 2010. Metabolic footprint of diabetes: a multiplatform metabolomics study in an epidemiological setting. PLoS One 5:e13953.

Xia, J., R. Mandal, I. V. Sinelnikov, D. Broadhurst, and D. S. Wishart. 2012. MetaboAnalyst 2.0—a comprehensive server for metabolomic data analysis. Nucleic Acids Res. 40:W127-W133.

Xia, J., N. Psychogios, N. Young, and D. S. Wishart. 2009. MetaboAnalyst: a web server for metabolomic data analysis and interpretation. Nucleic Acids Res. 37:W652-W660.

Yano, M., K. Watanabe, T. Yamamoto, K. Ikeda, T. Senokuchi, M. Lu, T. Kadomatsu, H. Tsukano, M. Ikawa, M. Okabe, S. Yamaoka, T. Okazaki, H. Umehara, T. Gotoh, W. J. Song, K. Node, R. Taguchi, K. Yamagata, and Y. Oike. 2011. Mitochondrial dysfunction and increased reactive oxygen species impair insulin secretion in sphingomyelin synthase 1-null mice. J. Biol. Chem. 286:3992-4002.

Zhang, X., Y. Wang, F. Hao, X. Zhou, X. Han, H. Tang, and L. Ji. 2009. Human serum metabonomic analysis reveals progression axes for glucose intolerance and insulin resistance statuses. J. Proteome Res. 8:5188-5195.

Zhang, X., L. Xu, J. Shen, B. Cao, T. Cheng, T. Zhao, X. Liu, and H. Zhang. 2013. Metabolic signatures of esophageal cancer: NMR-based metabolomics and UHPLC-based focused metabolomics of blood serum. Biochim. Biophys. Acta 1832:1207-1216.

CHAPTER 4

Lactose in milk replacer can partly be replaced by glucose, fructose, or glycerol without affecting insulin sensitivity

in veal calves

“Discovery consists of seeing what everybody has seen, and thinking what nobody has thought”

- Albert Szent-Gyorgyi -

A.J. Pantophlet,1 M.S. Gilbert,2 J.J.G.C. van den Borne,2 W.J.J. Gerrits,2 H. Roelofson,3 M.G. Priebe,3 and R.J. Vonk3

Adapted from Journal of dairy science 2016; 99 (4):3072-3080



CHAPTER 4

64

4

Abstract

Calf milk replacer (MR) contains 40 to 50% lactose. Lactose strongly fluctuates in price and alternatives are desired. Also, problems with glucose homeostasis and insulin sensitivity (i.e., high incidence of hyperglycemia and hyperinsulinemia) have been described for heavy veal calves (body weight >100 kg). Replacement of lactose by other dietary substrates can be economically attractive, and may also positively (or negatively) affect the risk of developing problems with glucose metabolism. An experiment was designed to study the effects of replacing one third of the dietary lactose by glucose, fructose, or glycerol on glucose homeostasis and insulin sensitivity in veal calves. Forty male Holstein-Friesian (body weight = 114 ± 2.4 kg; age = 97 ± 1.4 d) calves were fed an MR containing 462 g of lactose/kg (CON), or an MR in which 150 g of lactose/kg of MR was replaced by glucose (GLU), fructose (FRU), or glycerol (GLY). During the first 10 days of the trial, all calves received CON. The CON group remained on this diet and the other groups received their experimental diets for a period of 8 weeks. Measurements were conducted during the first (baseline) and last week of the trial. A frequently sampled intravenous glucose tolerance test was performed to assess insulin sensitivity and 24 h of urine was collected to measure glucose excretion. During the last week of the trial, a bolus of 1.5 g of [U-13C] substrates was added to their respective meals and plasma glucose, insulin, and 13C-glucose responses were measured. Insulin sensitivity was low at the start of the trial and remained low [1.2 ± 0.1 and 1.0 ± 0.1 (mU/L)−1 × min−1], and no treatment effect was noted. Glucose excretion was low at the start of the trial (3.4 ± 1.0 g/d), but increased in CON and GLU calves (26.9 ± 3.9 and 43.0 ± 10.6 g/d) but not in FRU and GLY calves. Postprandial glucose was higher in GLU, lower in FRU, and similar in GLY compared with CON calves. Postprandial insulin was lower in FRU and GLY and similar in GLU compared with CON calves. Postprandial 13C-glucose increased substantially in FRU and GLY calves, indicating that calves are able to partially convert these substrates to glucose. We concluded that replacing one third of lactose in MR by glucose, fructose, or glycerol in MR differentially influences postprandial glucose homeostasis but does not affect insulin sensitivity in veal calves.

Key words: veal calves, fructose, glycerol, insulin sensitivity, glucose homeostasis

Glucose, fructose, glycerol and insulin sensitivity

65

4

4.1 Introduction

Veal calves are fed milk replacer (MR), roughage, and concentrates. Despite the tendency to increase the amounts of roughage and concentrates in the diet, the vast majority (60–70%) of the digestible nutrient intake originates from MR. Upon closure of the esophageal groove, MR bypasses the rumen and flows directly into the abomasum. Lactose is the predominant, if not the only, carbohydrate source in MR. Calf MR commonly contains approximately 45% lactose, which is efficiently digested and absorbed from the calf intestinal lumen (Burt and Irvine, 1970; Coombe and Smith, 1974).

However, the commercial availability of lactose (or whey) for feed applications is limited and not constant, resulting in large fluctuations in raw material prices. This provides an economic incentive for MR manufacturers to replace lactose by alternative energy sources.

Importantly, a prolonged high intake of lactose, combined with substantial amounts of fat, has been associated with impaired glucose homeostasis. Hyperglycemia, hyperinsulinemia, and insulin resistance have been observed in veal calves in the second phase of the fattening period (Hostettler-Allen et al., 1994; Hugi et al., 1997). Such metabolic problems may eventually result in diabetes and (pro)-inflammatory stress, as demonstrated in humans (Hotamisligil, 2006; Shoelson et al., 2006), and in hepatic steatosis (Gerrits et al., 2008).

Starch or starch-based products, such as maltodextrins, are the most obvious alternatives for lactose. These products are widely available and are also attractive from an economic perspective. However, we recently demonstrated (Gilbert et al., 2015a) that calves have difficulties digesting starch-based products from MR diets, probably due to low activities of α-amylase and maltase in the small intestine. Nonetheless, the vast majority of starch does not reach the end of the small intestine, which can probably be explained by fermentation (Gilbert et al., 2015a,b).

Apart from starch-based products, glucose, fructose, and glycerol may also replace lactose in MR. Partly replacing lactose by fructose and glycerol may beneficially affect postprandial glucose homeostasis. These substrates have lower glycemic (and insulinemic) responses than lactose (Foster-Powell et al., 2002). It is believed that a lower glycemic (and insulinemic) response is beneficial for health, especially in subjects with impaired glucose metabolism (Howlett and Ashwell, 2008). In humans, fructose and glycerol are almost completely absorbed and metabolized by the liver (Grunnet and Lundquist, 1967; Schaefer et al., 2009; Sun and Empie, 2012). Therefore, the effects of these substrates on postprandial glucose homeostasis will likely depend on the rate and extent of conversion of these substrates to glucose by the liver. In humans, fructose is only partly (29-51%) converted to glucose by the liver (Sun and Empie, 2012); whether this is also the case for veal calves is not clear. Compared with fructose and glycerol, glucose may lead to higher glycemic (and insulinemic) responses than lactose (Foster-Powell et al., 2002), and thus may negatively affect postprandial glucose homeostasis.

CHAPTER 4

66

4

Whether prolonged exposure to these substrates (as partial replacer of lactose in veal calves) also affects insulin sensitivity is not known.

The objective of the current study was, therefore, to study the effects of partial replacement of dietary lactose by glucose, fructose, and glycerol on glucose homeostasis and insulin sensitivity in veal calves. The effects on energy and protein utilization for growth were also assessed and described elsewhere (Gilbert et al., 2016).



Forty male Holstein-Friesian calves were housed at the research facility of the Department of Animal Sciences at Wageningen University. At start of the trial calves were 97 ± 1.4 days of age and weighted 114 ± 2.4 kg (means ± SEM).

Calves were housed in groups, except for the last 6 days of the pre-experimental period (i.e., first 10 days of the trial) and the last 14 days of the trial. During these periods calves were housed individually in metabolic cages (dimensions = 0.80 × 1.8 m). During group housing, calves were housed in pens (5 calves/pen), which were fitted with wooden slatted floors and galvanized fencing. Per calf, 2 m2 was available. Ventilation occurred by ceiling fans, and illumination by natural light and artificial (fluorescent lamps) light between 0630 and 1730 h. The average temperature and humidity were 18.5 ± 0.4°C and 69.5 ± 1.2%, respectively (means ± SEM). The experimental procedures were approved by the Animal Care and Use Committee of Wageningen University.


Calves were fed 2.0 times the ME requirements for maintenance, which was set at 460 kJ/kg of metabolic BW per day (Van Es et al., 1967). Individual BW was measured weekly and the feeding rate was adjusted accordingly.

The trial consisted of a pre-experimental period (first 10 days) and an experimental period of 55 days. During the pre-experimental period, all calves received the control MR diet, which contained 462 g of lactose/kg of MR. The composition of the MR is given in Table 4.1. Thereafter, calves were assigned to 1 of 4 dietary treatments. The control group (CON) remained on the control MR diet; in the other groups 150 g of lactose (per kg of MR) was replaced by iso-energetic amounts of either glucose (GLU; Tereos Syral, Marckolsheim, France), fructose (FRU; Tate & Lyle Europe, Boleraz, Slovakia) or glycerol (GLY; Triconor Distribution BV, Soest, the Netherlands). All calves remained on their respective diets for a period of 55 days. The introduction of the lactose replacers occurred gradually, by increasing the lactose replacement by 50 g/kg of MR every 3 days.

In addition to MR, each calf received 10 g of DM of solid feed per kilogram of metabolic BW per day. The solid feed consisted of 80% concentrates and 20% wheat


67

4

straw (based on DM). The concentrates were composed of 279 g/kg of barley, 458 g/kg of corn, 205 g/kg of lupines, 24 g/kg of palm oil, and 34 g/kg of premix.

Milk replacer was fed on individual basis, at a concentration of 140 g of MR/L and supplied at a temperature of ~42°C. The concentration increased to 160 g/L of water when the MR volume ≥9.0 L. Solid feed was provided per pen (5 calves/pen) during group housing, and per individual calf when the calves were housed in metabolic cages. Calves were allowed ad libitum access to water. Feeding took place twice a day, at 0630 and 1530 h, via buckets. Milk replacer was supplied first, followed by solid feed. Milk replacer refusals (max 60 min after feeding) were weighed and recorded twice a day before solid feed supply, whereas solid feed refusals were weighed and recorded once a day before the morning MR feeding.


The measurements were concentrated during the pre-experimental period (first 10 days of trial; measurement period 1) and last 7 days of the trial (measurement period 2).

4.2.3.1 Frequently sampled intravenous glucose tolerance test A frequently sampled intravenous glucose tolerance test (FSIGTT) was performed during both measurement periods (on day 9 and 65 of the trial). Five (measurement period 1) and 12 days (measurement period 2) before the test, calves were moved to metabolic cages and were prepared with a central venous catheter (Careflow, Becton Dickinson, Alphen a/d Rijn, the Netherlands) in their jugular vein, for glucose and insulin infusion, and blood sampling.

All calves were fasted overnight for 16 to 19 h before the FSIGTT. At 0 min, an intravenous glucose bolus of 0.3 g/kg of BW (20% glucose solution, B. Braun, Oss, the Netherlands) was administered within 2 min, followed by an intravenous insulin bolus of 0.03 IU/ kg of BW (100 IU/mL of solution, Insuman Rapid, Sanofi-Aventis, Gouda, the Netherlands) at 20 min (administered within 1 min). Blood samples were collected from the jugular catheter at −8, −4, 2, 4, 6, 8, 10, 12, 14, 16, 19, 22, 25, 30, 35, 40, 50, 60, 75, 90, 120, 150, and 180 min relative to the intravenous glucose bolus. Blood samples were transferred immediately into 6-mL lithium-heparin vacutainer tubes (Becton Dickinson, Breda, the Netherlands) and stored on ice. Plasma was collected, after centrifugation (1,516 × g, 10 min), within 1.5 h after blood sampling and stored at −20°C until analysis of plasma glucose and insulin concentrations. In addition, plasma triglycerides and high-density and low-density lipoprotein (HDL and LDL, respectively) cholesterol concentrations were analyzed in the fasting plasma samples.

CHAPTER 4

68

4

Table 4.1 | Composition of the control (CON) milk replacer

Ingredients g/kg Nutrients g/kg DM

Basal milk replacer

Delactosed whey powder 244.5 Dry matter3, g/kg 966

Calcium formate 7.5 Crude protein 205

Coconut oil 39.2 Crude fat 209

Lard 72.5 Crude ash 74

Tallow 72.5 Lactose4 478

Lecithin 5.9 Fe, mg/kg DM 49.9

Ricinoleate emulsion 5.9 Gross energy, MJ/kg 19.8

Premix1 10

Whey protein concentrate 391.1

Methionine 1.1

Additional

Lactose2 1501 Premix (per kg): crude protein, 0.7 g; starch, 5.0 g; crude ash, 1.5 g; Ca, 17 mg; P, 7.5 mg; Na, 0.7 mg; K, 7.3 mg; Cl, 13 mg; Mg, 0.5 g; Fe, 45 mg; Cu, 8.0 mg; Zn, 0.1 g; Mn, 43 mg; Se, 0.3 mg; I, 1.0 mg; Vitamin A, 25,013 IU; Vitamin D3, 4,002 IU; Vitamin E, 150 IU; Vitamin K3, 2.1 mg; Vitamin C, 0.3 g; Vitamin B1, 8.2 mg; Vitamin B2, 10 mg; Vitamin B3, 35 mg; Vitamin B5, 18 mg; Vitamin B6, 10 mg; Vitamin B12, 0.1 mg; biotin, 0.2 mg; folate, 0.7 mg; choline 0.4 g. 2 For the glucose, fructose and glycerol groups, 150 g lactose per kg MR was replaced iso-energetically by glucose, fructose and glycerol, respectively.3 Dry matter of the basal milk replacer. 4 Calculated content.

The insulin sensitivity index was calculated according to Bergman’s minimal model approach, using MinMod Millennium (MinMod Inc., Los Angeles, CA; version 6.0.2), a computer-based software for the calculation of insulin sensitivity from FSIGTT data (Pacini and Bergman, 1986; Boston et al., 2003). Insulin sensitivity derived by this method (ISminmod) encompasses both peripheral and hepatic insulin sensitivity (Bergman et al., 1979). Therefore, another index of insulin sensitivity, the quantitative insulin sensitivity index (QUICKI), was calculated from the fasting plasma glucose and insulin levels collected during the FSIGTT. The QUICKI primary reflects hepatic insulin sensitivity (Chen et al., 2003; Muniyappa et al., 2008). This index was calculated using the following formula:

QUICKI = 1

(log( glucose mgdL

)+ log(insulin mUL

))


69

4

4.2.3.2 Urine collectionPrior to catheterization (for the FSIGTT), calves were shaved (especially the tail region) and harnessed. Plastic bags were connected to the harness to collect feces. Clean urine was quantitatively collected for a period of 3 and 5 days, during measurement period 1 and 2, respectively. The urine was collected via funnels in buckets, which were placed under the cages. Each bucket contained 1,500 mL of a 25% sulfuric acid solution (BOOM, Meppel, the Netherlands). At the end of the collection period, the amount of urine was weighed and an aliquot of 100 mL was transferred into a urine collection cup and frozen at −20°C until analysis of glucose, catecholamines, and cortisol.

4.2.3.3 Postprandial blood glucose, 13C-glucose, and insulin concentrations On day 61 of the trial (measurement period 2), a bolus of 1.5 g of [U-13C] enriched substrates (99% isotopic enrichment; Cambridge Isotope Laboratories, Tewksbury, MA) was added to the morning MR meal. Hence, GLU calves received [U-13C] glucose, FRU calves received [U-13C] fructose, and GLY calves received [U-13C] glycerol, whereas CON calves did not receive 13C-enriched substrates. Solid feed was not provided during this (enriched) test meal.

Blood samples were collected from the jugular vein at −10, 15, 30, 60, 120, 180, 240, 300, and 360 min relative to MR feeding. Blood samples were transferred into 6-mL lithium heparin Vacutainer tubes (Becton Dickinson) and stored on ice. Plasma was collected, after centrifugation (1,516 × g, 10 min), within 1.5 h after blood sampling and was stored at −20°C until analysis. Plasma glucose and insulin concentrations as well as 13C enrichment in plasma glucose were measured. For glucose and insulin, the maximum concentration minus fasting concentration (ΔCmax), time to maximum concentration, and the incremental area under the curve (iAUC0–6h) were calculated. For 13C enrichment in plasma glucose, only the iAUC0–6h was calculated. The iAUC0–6h

was calculated using the trapezoid method (Le Floch et al., 1990).


Plasma glucose, triglycerides, and HDL and LDL cholesterol were measured on a Roche-Hitachi Modular automatic analyzer (Roche Diagnostics, Basel, Switzerland) using enzymatic colorimetric methods. The within- and between-run coefficient of variation were ≤2% for all analysis. Insulin was measured using a bovine ELISA kit (Mercodia, Uppsala, Sweden). The within- and between-run coefficient of variation were ≤5.6 and 8.2%, respectively.

Catecholamines in urine were measured using an online solid phase extraction-liquid chromatography/tandem mass spectrometry method previously described by de Jong et al. (2010). Cortisol was measured using the online solid phase extraction-liquid chromatography/tandem mass spectrometry method described by Jones et al. (2012). This method was originally devolved for saliva samples but was successfully applied on urine samples. The 13C-to-12C ratio in plasma glucose was measured using a modified

CHAPTER 4

70

4

version of the gas chromatography-combustion-isotope ratio mass spectrometry method previously described by Vonk et al. (2000a, b). The modifications are described in detail elsewhere (Eelderink et al., 2012). A calibration curve was used to calculate the molar percentage excess (7 standards; range = 0.01–2.00 molar percent excess) of the measured samples. Then, the molar percentage excess was multiplied by the plasma glucose concentration to obtain Δ[13C]-glucose concentrations.

4.2.5 Statistical analysis

The SPSS (Version 22, IBM Corp., Armonk, NY) statistical software was used for all statistical analyses. Data are presented as means ± SEM. The effects of treatment on insulin sensitivity, urinary glucose excretion, and fasting blood levels of glucose, insulin, cholesterol, triglycerides, catecholamines, and cortisol were tested by ANOVA using the GLM (Univariate) procedure. Treatment was used as factor and calf was the experimental unit. The final values (measurement period 2) were used as a dependent variable, with their respective initial values (pre-experimental; measurement period 1) as a covariate. Growth and data derived from postprandial blood concentrations (i.e., ΔCmax, time to maximum, and iAUC0–6h) were also tested for treatment effects using the same procedure. However, no covariates were included in the model.

Normality of the studentized residuals was assessed by visual inspection. Non-normally distributed data were (log) transformed to obtain normality. P-values < 0.05 were considered significant and P-values < 0.10 were considered a trend toward significance. When treatment effects were significant, the Šidák (1967) method was used to correct for pairwise comparisons.

4.3 Results

No difference in feed refusals was noted between groups. However, 1 calf in the GLU group was excluded from the trial due to ruminal drinking. The ADG, measured over the 65-day trial period, was, on average, 1.19 ± 0.03 kg/d and did not differ between treatments.

4.3.1 Insulin sensitivity

Initial values for ISminmod and QUICKI (measured during the pre-experimental period) were 1.2 ± 0.1 × 10−4 [(mU/L)−1 × min−1] and 0.371 ± 0.001, respectively. These values decreased by 18 and 10%, respectively, on average, during the trial, but these changes were not affected by dietary treatment (Table 4.2).


71

4

4.3.2 Fasting blood metabolites and hormones

Fasting plasma glucose concentrations (collected during the FSIGTT) were greater (P < 0.05) for FRU calves (5.8 mmol/L) than for GLU calves (5.3 mmol/L; Table 4.2). Fasting plasma insulin concentrations, however, were not affected by treatment but increased by 78% during the trial. Fasting triglyceride concentrations were not affected by treatment but decreased by 64%. Fasting HDL and LDL cholesterol concentrations were not affected by treatment.

4.3.3 Postprandial glucose, insulin, and 13C enrichment in plasma glucose

Postprandial blood glucose and insulin concentrations were measured after 51 days of adaptation to the experimental diets. On this day, fasting glucose and insulin concentrations were measured again for calculation of the postprandial parameters (Table 4.3). The treatment effects did not differ from the fasting measurements conducted during the FSIGTT (Table 4.2). Directly after feeding, plasma glucose, and insulin concentrations increased (Figure 4.1), but the time to the maximum plasma concentrations of glucose and insulin did not differ between treatments. The postprandial increase in plasma glucose concentration (ΔCmax) differed between treatments, with the levels being higher (P < 0.05) in GLU calves than in FRU and GLY calves. The ΔCmax for insulin also differed between treatments, with the levels being higher (P < 0.05) in GLU and CON calves than in FRU calves. The iAUC0–6h for glucose was higher (P < 0.05) for GLU calves than the other calves, and higher (P < 0.05) for CON calves than for FRU calves. The iAUC0–6h for insulin was higher (P < 0.05) for GLU and CON calves than for FRU and GLY calves. The iAUC0–6h for the Δ[13C]-glucose did differ between treatments, with GLU > GLY > FRU (Figure 4.2; Table 4.3).

4.3.4 Urinary glucose excretion

During the pre-experimental period, urinary glucose excretion was low (3.4 ± 1.0 g/d). At the end of the trial (i.e., after 55 days of adaptation to the experimental diets), the glucose excretion was significantly (P < 0.01) higher for CON and GLU calves (26.9 ± 3.9 and 43.0 ± 10.6 g/d, respectively; Table 4.2), compared with the FRU and GLY calves.

4.3.5 Urinary excretion of catecholamines and cortisol

The urinary excretion of 4 stress-related markers (adrenaline, noradrenaline, dopamine, and cortisol) was measured during this trial. Urinary excretion of these stress-related markers was not affected by dietary treatments (Table 4.2).

CHAPTER 4

72

4

Table 4.2 | Insulin sensitivity, fasting plasma metabolite and hormone concentrations, and the urinary excretion of glucose and stress-related markers (means ± SEM) in veal calves fed a control milk replacer (CON, 462 g lactose/kg milk replacer; n=10) or a milk replacer in which 150 g of lactose per kg milk replacer was replaced by iso-energetic amounts of glucose (GLU; n=9), fructose (FRU; n=10) or glycerol (GLY; n=10), for a period of 55 days (measurement period 2).

Item1 Treatment

CON GLU FRU GLY P-value

Insulin sensitivity

ISminmod x 10-4, ((mU/L)-1 x min-1) 1.04±0.21 1.23±0.23 0.83±0.14 0.75±0.11 0.702

QUICKI x 101 3.30±0.09 3.39±0.1 3.39±0.07 3.43±0.08 0.203

Fasting plasma concentration

Glucose, mmol/L 5.7±0.1ab 5.3±0.1a 5.8±0.1b 5.6±0.1ab 0.016

Insulin, mU/L 12.7±2.2 10.8±1.3 9.8±1.8 11.7±2.1 0.786

Triglycerides, mmol/L 0.15±0.01 0.15±0.01 0.13±0.01 0.14±0.01 0.939

HDL-cholesterol, mmol/L 2.3±0.2 2.6±0.2 3.1±0.2 2.7±0.1 0.078

LDL-cholesterol, mmol/L 0.41±0.06 0.54±0.07 0.68±0.10 0.56±0.07 0.135

Urinary excretion

Glucose, g/day 26.9±3.9a 43.0±10.6a 3.6±1.0b 6.3±1.7b <0.001

Adrenaline, µg/kg BW/day 0.07±0.01 0.05±0.01 0.06±0.01 0.05±0.01 0.251

Noradrenaline, µg/kg BW/day 0.68±0.08 0.68±0.10 0.77±0.07 0.65±0.10 0.239

Dopamine, µg/kg BW/day 0.71±0.11 0.77±0.09 0.83±0.05 0.68±0.10 0.143

Cortisol, µg/kg BW/day 0.07±0.01 0.08±0.02 0.08±0.01 0.06±0.01 0.499a,b Different superscripts indicate pairwise differences (P < 0.05).1ISminmod = insulin sensitivity derived from MinMod Millenium (MinMod Inc., Los Angeles, CA); QUICKI = quantitative insulin sensitivity index.

Figure 4.1 | Plasma glucose (A) and insulin (B) responses in veal calves fed (at time=0) a control milk replacer (, 462 g lactose/kg milk replacer; n=10) or a milk replacer in which 150 g of lactose per kg milk replacer was replaced by iso-energetic amounts of glucose (; n=9), fructose (; n=10) or glycerol (; n=10), for a period of 51 days. Error bars represent SEM. Calculated parameters and statistics are given in Table 4.3.


73

4

Figure 4.2 | Changes in plasma Δ13C-glucose concentration in veal calves fed (at time=0) a control milk replacer (, 462 g lactose/kg milk replacer; n=10) or a milk replacer in which 150 g of lactose per kg milk replacer was replaced by iso-energetic amounts of glucose (; n=9), fructose (; n=6) or glycerol (; n=10), for a period of 51 days. On top of the milk replacer diets (except for the control group) 1.5 g of the corresponding U-13C substrate (i.e. 13C-glucose, fructose or glycerol) was added to their diets on day 46. Error bars represent SEM. Calculated parameters and statistics are given in Table 4.3.

Table 4.3 | Postprandial responses of plasma glucose and insulin, and 13C-enrichment in plasma glucose (means ± SEM) in veal calves fed a control milk replacer (CON, 462 g lactose/kg milk replacer; n=10) or a milk replacer in which 150 g of lactose per kg milk replacer was replaced by iso-energetic amounts of glucose (GLU; n=9), fructose (FRU; n=10) or glycerol (GLY; n=10), for a period of 51 days.

Item1 Treatment

CON GLU FRU GLY P-value

Plasma glucose concentration

Fasting values, mmol/L 5.4 ± 0.1ab 4.9 ± 0.1a 5.5 ± 0.2b 5.4 ± 0.2ab 0.024

ΔCMax, mmol/L 4.7 ± 0.5ab 6.3 ± 0.4a 3.5 ± 0.3b 3.7 ± 0.4b <0.001

Time to maximum, min 56 ± 9 77 ± 20 50 ± 17 54 ± 4 0.534

iAUC0-6h 761 ± 84b 1233 ± 110a 383 ± 77c 560 ± 91bc <0.001

Plasma insulin concentration

Fasting values, mU/L 17.0 ± 3.5 14.4 ± 1.0 10.3 ± 1.3 11.0 ± 1.5 0.077

ΔCMax, mU/L 909 ± 133a 831 ± 79a 478 ± 85b 569 ± 68a 0.008

Time to maximum, min 150 ± 10 140 ± 10 130 ± 16 126 ± 7 0.435

iAUC0-6h x 103 146 ± 21a 156 ± 11a 72 ± 11b 90 ± 9b <0.00113C enriched plasma glucose2

iAUC0-6h x 10-2 4.8 ± 1.03 887 ± 55a 151 ± 14b 250 ± 18c <0.001a–c Different superscripts indicate a pairwise differences (P < 0.05).1ΔCmax = maximum concentration minus fasting concentration; iAUC0–6h = incremental area under the curve.2 On top of the milk replacer diets (except for the control group) 1.5 g of the corresponding U-13C substrate (i.e., 13C-glucose, fructose or glycerol) was added to their diets on day 46.3 CON calves were not included in statistical analysis because they did not receive a 13C-enriched substrate.

CHAPTER 4

74

4

4.4 Discussion

4.4.1 Fructose and glycerol gluconeogenesis

The total postprandial blood glucose response (iAUC0–6h) was higher for GLU calves than for the other calves, and higher for CON calves than for FRU calves. In general, blood glucose levels reflect both the exogenous glucose influx, endogenous glucose production, as well as the glucose uptake into peripheral tissues (Priebe et al., 2010). Therefore, the differences in iAUC0–6h for glucose may be due to a limited ability of the liver to convert fructose (and glycerol) into glucose. To assess appearance of exogenous substrates into the systemic circulation and to obtain indications on hepatic conversion into glucose, each calf (except for the CON group) received 1.5 g of the corresponding U-13C substrate. The Δ[13C]-glucose levels increased in all groups (Figure 4.2) and, as expected, the iAUC0–6h differed between treatments, with GLU calves being substantially higher than FRU and GLY calves, and GLY being higher than FRU. This indicates that fructose and glycerol are (partly) converted to glucose in calves and may partly explain the differences observed in postprandial glucose concentrations. Nevertheless, the rates of conversion could not be calculated in our experiment, as the whole body glucose flux was not determined. In humans, between 29 and 50% of ingested fructose is converted to glucose within first 6 h (Delarue et al., 1993; Tran et al., 2010). In our experiment, the plasma 13C-glucose enrichment increased consistently within the first 6 h. This indicates that, for calves, an incomplete conversion of fructose to glucose occurs within the first 6 h after intake. Gluconeogenesis from glycerol has also been studied in humans, although less extensive than fructose; Massicotte et al. (2006) found that approximately 9% is converted to glucose within the first 2 h after ingestion. Glycerol is also a known glucogenic precursor in ruminants (i.e., cows and sheep), although conversion rates have not been quantified (Aschenbach et al., 2010; Werner Omazic et al., 2015).

Urinary glucose excretion was significantly higher in CON and GLU calves than in FRU and GLY calves, where the urinary excretion was negligible. Excretion of glucose via urine is a result of hyperglycemia (Chao, 2014; Wilding, 2014); this occurs when the renal threshold for glucose reabsorption is exceeded. In our experiment, the plasma glucose peak levels for the CON and GLU calves exceeded the renal threshold of 8.3 to 11.1 mmol/L in calves (Wijayasinghe et al., 1984; Scholz and oppe, 1987; Hostettler-Allen et al., 1994; Stanley et al., 2002), and therefore urinary excretion was expected in these groups. Glucosuria is frequently observed in milk-fed calves (Hostettler-Allen et al., 1994; Hugi et al., 1997; Kaufhold et al., 2000). Although urinary glucose excretion differed between treatments, insulin sensitivity did not differ. This indicates that urinary glucose excretion in calves is not associated with insulin resistance, but rather relates to exceeding the glucose renal threshold.

Based on the negative effect of GLU calves on postprandial glucose homeostasis (i.e., significantly greater blood glucose levels), and glycosuria for GLU and CON, compared with FRU and GLY calves, it could be expected that calves assigned to the first 2 treatments


75

4

would be more insulin resistant. Chronic postprandial hyperglycemia can lead to glucotoxicity, which can ultimately result in insulin resistance and diabetes (Campos, 2012); however, this was not confirmed by the insulin sensitivity measurements.

4.4.2 Development of insulin sensitivity

The absence of effects on insulin sensitivity may be explained by the already low insulin sensitivity in calves before the dietary intervention. The average insulin sensitivity before the intervention was 1.2 ± 0.1 × 10−4 [(mU/L)−1 × min−1], which is ~10× lower than in neonatal calves at 3 and 6 weeks of age (Stanley et al., 2002) and ~3 to 6× lower than in non-ruminants (Caumo et al., 2000; Stefanovski et al., 2011). Instead, values were more comparable to those in lactating cows (Stanley et al., 2002). Whether the lactose replacers could have an effect on younger calves (which could be more sensitive to insulin) requires further study.

Several factors may contribute to the (decreased) insulin sensitivity observed in veal calves. One factor that may contribute is the high fat content in conventional calf MR (~200 g/kg of MR). The composition of the digestible energy intake in veal calves MR (i.e., high fat and carbohydrate content) is similar to that of the adult western human diet (Schwarz et al., 2003; Cordain et al., 2005). In humans, high dietary fat intake was found to be associated with insulin resistance (Storlien et al., 1996; Vessby, 2000; Marshall and Bessesen, 2002). However, despite their higher glucose (but lower fat) intake, pigs and rats generally do not suffer from insulin resistance as much as veal calves. It may be speculated that, apart from the species difference, interactions between dietary fat and glucose may contribute to reduced insulin sensitivity in veal calves. Another factor that may contribute to the (decreased) insulin sensitivity observed in veal calves is the discrepancy between their diet and ontogenetic background. Insulin resistance seems to be age-dependent in veal calves (Hugi et al., 1997, 1998). In nature, calves between 4 and 6 months of age are grazing and plant fragments are fermented in the rumen along with the production of volatile fatty acids such as acetate, propionate, and butyrate as the main sources of energy. Therefore, the veal calf, which is an ontogenetic ruminant, may not be equipped with the genetic capacity to deal with the large amounts of lactose (e.g., the capacity to use the glucose as precursor for de novo fatty acid synthesis; Roehrig et al., 1988). This may explain why high lactose intake might induce insulin resistance in veal calves and why calves on conventional MR do not synthesize body fat from dietary carbohydrates (van den Borne et al., 2007). Nevertheless, whether the decrease in insulin sensitivity is diet-related or part of the genetic programming of calves (transition from pre-ruminant to ruminant) is not clear and needs to be investigated. Another factor that can modulate insulin sensitivity is environmental stress (Sato et al., 2011). Hugi et al. (1997) measured the urinary excretion of noradrenaline and dopamine in veal calves and found a significant increase with age. They concluded that the enhanced activity of the nervous system (stress) might have contributed to the decrease in insulin sensitvity. In our study, however, none of the stress-related markers increased with age and no

CHAPTER 4

76

4

treatment effect was noted. In addition, we performed Pearson bivariate correlation analysis between the measured insulin sensitivity values and the rates of urinary excretion of the different stress markers (data not shown). Significant correlations were not found between insulin sensitivity and any of the stress markers, suggesting that stress did not significantly affect insulin sensitivity in the current study.

4.4.3 Fasting blood glucose, insulin, triglycerides, and cholesterol

In our study, an increase in fasting plasma insulin levels was observed for all treatments, which (at constant fasting glucose levels) suggests a decrease in hepatic insulin sensitivity (Muniyappa et al., 2008). The increase in plasma insulin is in agreement with findings by Breier et al. (1988) and Hugi et al. (1997), but in contrast with Hostettler-Allen et al. (1994), who did not detect an age-related increase in fasting plasma insulin levels.

In humans, insulin resistance is also often associated with dyslipidemia, characterized by a high ratio between triglycerides and HDL cholesterol (Li et al., 2008; Bitzur et al., 2009); however, whether this is also true for calves is unknown. Nevertheless, in our trial the triglyceride-to-HDL cholesterol ratios were extremely low (ratios < 1) and decreased with age in all groups (lower triglyceride concentrations), suggesting no evidence of dyslipidemia in these calves.

4.5 Conclusions

Replacing 150 g of lactose/kg of MR with glucose, fructose, or glycerol does affect postprandial glucose homeostasis. Replacement with glucose leads to increased postprandial blood glucose levels, but similar blood insulin levels and urinary glucose losses. Fructose and glycerol are (partly) converted into glucose and lead to reduced postprandial glucose and insulin levels when lactose is replaced with fructose, and reduced insulin levels when replaced with glycerol. Both substrates do not lead to significant urinary glucose losses. Despite differences in postprandial glucose homeostasis, lactose replacement did not affect insulin sensitivity.

4.6 Acknowledgments

The authors thank Gerlof Reckman, Theo Boer, and Jeltje Kloosterman for their laboratory assistance (University Medical Centre Groningen), and the animal caretakers at Carus, the experimental facilities of Wageningen University. This project was jointly financed by the European Union, European Regional Development Fund and the Ministry of Economic Affairs, Agriculture and Innovation, Peaks in the Delta, the Municipality of Groningen, the Provinces of Groningen, Fryslan and Drenthe as well as the Dutch Carbohydrate Competence Center (CCC2 WP21). Financial support was also provided by Tereos Syral (Marckolsheim, France), VanDrie Group (Mjidrecht, the Netherlands), and Wageningen University (Wageningen, the Netherlands).


77

4

4.7 References

Aschenbach, J. R., N. B. Kristensen, S. S. Donkin, H. M. Hammon, and G. B. Penner. 2010. Gluconeogenesis in dairy cows: the secret of making sweet milk from sour dough. IUBMB life. 62:869-877.

Bergman, R. N., Y. Z. Ider, C. R. Bowden, and C. Cobelli. 1979. Quantitative estimation of insulin sensitivity. Am. J. Physiol. 236:E667-E677.

Bitzur, R., H. Cohen, Y. Kamari, A. Shaish, and D. Harats. 2009. Triglycerides and HDL cholesterol stars or second leads in diabetes? Diabetes Care. 32 (Suppl 2):373-377.

Boston, R. C., D. Stefanovski, P. J. Moate, A. E. Sumner, R. M. Watanabe, and R. N. Bergman. 2003. MINMOD Millennium: a computer program to calculate glucose effectiveness and insulin sensitivity from the frequently sampled intravenous glucose tolerance test. Diabetes Technol. Ther. 5:1003-1015.

Breier, B. H., P. D. Gluckman, and J. J. Bass. 1988. Plasma concentrations of insulin-like growth factor-I and insulin in the infant calf: ontogeny and influence of altered nutrition. J. Endocrinol. 119:43-50.

Burt, A. W. A. and S. M. Irvine. 1970. Carbohydrates in milk replacers for calves. J. Sci. Food Agric. 21:342-346.Campos, C. 2012. Chronic hyperglycemia and glucose toxicity: pathology and clinical sequelae. Postgrad.

Med. 124:90-97.Caumo, A., R. N. Bergman, and C. Cobelli. 2000. Insulin sensitivity from meal tolerance tests in normal

subjects: a minimal model index. J. Clin. Endocrinol. Metab. 85:4396-4402.Chao, E. C. 2014. SGLT-2 Inhibitors: A new mechanism for glycemic control. Clin. Diabetes 32:4-11.Chen, H., G. Sullivan, L. Q. Yue, A. Katz, and M. J. Quon. 2003. QUICKI is a useful index of insulin sensitivity

in subjects with hypertension. Am. J. Physiol. Endocrinol. Metab. 284:E804-E812.Coombe, N. B. and R. H. Smith. 1974. Digestion and absorption of starch, maltose and lactose by the

preruminant calf. Br. J. Nutr. 31:227-235.Cordain, L., S. B. Eaton, A. Sebastian, N. Mann, S. Lindeberg, B. A. Watkins, J. H. O’Keefe, and J. Brand-

Miller. 2005. Origins and evolution of the Western diet: health implications for the 21st century. Am. J. Clin. Nutr. 81:341-354.

de Jong, W. H., E. G. de Vries, B. H. Wolffenbuttel, and I. P. Kema. 2010. Automated mass spectrometric analysis of urinary free catecholamines using on-line solid phase extraction. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 878:1506-1512.

Delarue, J., S. Normand, C. Pachiaudi, M. Beylot, F. Lamisse, and J. P. Riou. 1993. The contribution of naturally labelled 13C fructose to glucose appearance in humans. Diabetologia. 36:338-345.

Eelderink, C., T. C. W. Moerdijk-Poortvliet, H. Wang, M. Schepers, T. Preston, T. Boer, R. J. Vonk, H. Schierbeek, and M. G. Priebe. 2012. The glycemic response does not reflect the In vivo starch digestibility of fiber-rich wheat products in healthy men. J. Nutr. 142:258-263.

Foster-Powell, K., S. H. A. Holt, and J. C. Brand-Miller. 2002. International table of glycemic index and glycemic load values: 2002. Am. J. Clin. Nutr. 76:5-56.


Gilbert, M. S., A. J. Pantophlet, H. Berends, A. M. Pluschke, J. J. G. C. van den Borne, W. H. Hendriks, H. A. Schols, and W. J. J. Gerrits. 2015a. Fermentation in the small intestine contributes substantially to intestinal starch disappearance in calves. J. Nutr. 145:1147-1155.

Gilbert, M. S., A. J. Pantophlet, J. J. G. C. van den Borne, W. H. Hendriks, H. A. Schols, and W. J. J. Gerrits. accepted. Replacing lactose with glucose, fructose or glycerol does not affect energy retention in veal calves. J. Dairy Sci.

Gilbert, M. S., J. J. G. C. van den Borne, H. Berends, A. J. Pantophlet, H. A. Schols, and W. J. J. Gerrits. 2015b. A titration approach to identify the capacity for starch digestion in milk-fed calves. Animal 9:248-257.

Grunnet, N. and F. Lundquist. 1976. Kinetics of glycerol kinases from mammalian liver and Candida mycoderma. Eur. J. Biochem. 3:76-84.


Hotamisligil, G. S. 2006. Inflammation and metabolic disorders. Nature 444:860-867.Howlett, J. and M. Ashwell. 2008. Glycemic response and health: summary of a workshop. Am. J. Clin. Nutr.

87:S212-S216.

CHAPTER 4

78

4



Jones, R. L., L. J. Owen, J. E. Adaway, and B. G. Keevil. 2012. Simultaneous analysis of cortisol and cortisone in saliva using XLC-MS/MS for fully automated online solid phase extraction. J. Chromatogr. B Analyt. Technol. Biomed. Life Sci. 881-882:42-48.



Li, C., E. S. Ford, Y. X. Meng, A. H. Mokdad, and G. M. Reaven. 2008. Does the association of the triglyceride to high-density lipoprotein cholesterol ratio with fasting serum insulin differ by race/ethnicity? Cardiovasc. Diabetol. 7:4.

Marshall, J. A. and D. H. Bessesen. 2002. Dietary fat and the development of type 2 diabetes. Diabetes Care. 25:620-622.

Massicotte, D., A. Scotto, F. Péronnet, H. M’Kaouar, M. Milot, and C. Lavoie. 2006. Metabolic fate of a large amount of 13C-glycerol ingested during prolonged exercise. Eur. J. Appl. Physiol. 96:322-329.


Pacini, G. and R. N. Bergman. 1986. MINMOD: a computer program to calculate insulin sensitivity and pancreatic responsivity from the frequently sampled intravenous glucose tolerance test. Comput. Methods Programs Biomed. 23:113-122.

Priebe, M. G., H. Wang, D. Weening, M. Schepers, T. Preston, and R. J. Vonk. 2010. Factors related to colonic fermentation of nondigestible carbohydrates of a previous evening meal increase tissue glucose uptake and moderate glucose-associated inflammation. Am. J. Clin. Nutr. 91:90-97.

Roehrig, K., K. E. Nestor, and D. L. Palmquist. 1988. ATP citrate lyase activity in liver and adipose tissue of veal or ruminating calves (Bos taurus). Comp. Biochem. Physiol. B 90:147-149.

Sato, M., S. Ishibashi, M. Higashimoto, Y. Kadota, T. Kawakami, and S. Suzuki. 2011. Early changes induced by environmental stresses in insulin sensitivity-related genes. Eur. J. Pharmacol. 668:472-476.

Schaefer, E. J., J. A. Gleason, and M. L. Dansinger. 2009. Dietary fructose and glucose differentially affect lipid and glucose homeostasis. J. Nutr. 139:1257-1262.

Scholz, H. and S. Hoppe. 1987. Renal glucose loss after intravenous glucose-infusion in calves. Dtsch. Tierarztl. Wochenschr. 94:473-476.

Schwarz, J. M., P. Linfoot, D. Dare, and K. Aghajanian. 2003. Hepatic de novo lipogenesis in normoinsulinemic and hyperinsulinemic subjects consuming high-fat, low-carbohydrate and low-fat, high-carbohydrate isoenergetic diets. Am. J. Clin. Nutr. 77:43-50.


Šidák, Z. 1967. Rectangular confidence regions for the means of multivariate normal distributions. J. Am. Stat. Assoc. 62:626-633.


Stefanovski, D., J. M. Richey, O. Woolcott, M. Lottati, D. Zheng, L. N. Harrison, V. Ionut, S. P. Kim, I. Hsu, and R. N. Bergman. 2011. Consistency of the Disposition Index in the Face of Diet Induced Insulin Resistance: Potential Role of FFA. PLoS One 6:e18134.

Storlien, L. H., L. A. Baur, A. D. Kriketos, D. A. Pan, G. J. Cooney, A. B. Jenkins, G. D. Calvert, and L. V. Campbell. 1996. Dietary fats and insulin action. Diabetologia. 39:621-631.

Sun, S. Z. and M. W. Empie. 2012. Fructose metabolism in humans - what isotopic tracer studies tell us. Nutr Metab (Lond). 9:89.


79

4

Tran, C., D. Jacot-Descombes, V. Lecoultre, B. A. Fielding, G. Carrel, K. A. Lê, P. Schneiter, M. Bortolotti, K. N. Frayn, and L. Tappy. 2010. Sex differences in lipid and glucose kinetics after ingestion of an acute oral fructose load. Br. J. Nutr. 104:1139-1147.

van den Borne, J. J. G. C., G. E. Lobley, M. W. A. Verstegen, J. M. Muijlaert, S. J. J. Alferink, and W. J. J. Gerrits. 2007. Body fat deposition does not originate from carbohydrates in milk-fed calves. J. Nutr. 137:2234-2241.

Van Es, A. J. H., H. J. Nijkamp, E. J. Van Weerden, and K. K. Van Hellemond. 1967. Energy, carbon and nitrogen balance experiments with veal calves. Pages 197-201 in Energy Metabolism of Farm Animals. K. L. Blaxter, J. Kielanowski, and G. Thorbek, ed. Oriel Press Newcastle-upon-Tyne, UK.

Vessby, B. 2000. Dietary fat and insulin action in humans. Br. J. Nutr. 83:S91-S96.Vonk, R. J., R. E. Hagedoorn, R. De Graaff, H. Elzinga, S. Tabak, Y. X. Yang, and F. Stellaard. 2000a. Digestion

of so-called resistant starch sources in the human small intestine. Am. J. Clin. Nutr. 72:432-438.Vonk, R. J., Y. Lin, H. A. Koetse, C. Huang, G. Zeng, H. Elzinga, J. Antoine, and F. Stellaard. 2000b. Lactose

(mal)digestion evaluated by the 13C-lactose digestion test. Eur. J. Clin. Invest. 30:140-146.Werner-Omazic, A., C. Kronqvist, L. Zhongyan, H. Martens, and K. Holtenius. 2015. The fate of glycerol

entering the rumen of dairy cows and sheep. J. Anim. Physiol. Anim. Nutr. (Berl) 99:258-264.Wijayasinghe, M. S., N. E. Smith, and R. L. Baldwin. 1984. Growth, health, and blood glucose concentrations

of calves fed high-glucose or high-fat milk replacers. J. Dairy Sci. 67:2949-2956.Wilding, J. P. H. 2014. The role of the kidneys in glucose homeostasis in type 2 diabetes: Clinical implications

and therapeutic significance through sodium glucose co-transporter 2 inhibitors. Metabolism 63:1228-1237.

CHAPTER 5

Insulin sensitivity in calves decreases substantially during the first 3 months of life and is unaffected by weaning or

fructo-oligosaccharide supplementation

“The most beautiful thing we can experience is the mysterious. It is the source of all true art and science”

- Albert Einstein-

A.J. Pantophlet,1 M.S. Gilbert,2 J.J.G.C. van den Borne,2 W.J.J. Gerrits,2 M.G. Priebe,3 and R.J. Vonk3

Adapted fromJournal of dairy science 2016; 99 (9): 7602-7611



CHAPTER 5

82

5

Abstract

Veal calves at the age of 4 to 6 months often experience problems with glucose homeostasis, as indicated by postprandial hyperglycemia, hyperinsulinemia, and insulin resistance. It is not clear to what extent the ontogenetic development of calves or the feeding strategy (e.g. prolonged milk replacer (MR) feeding) contributes to this pathology. The objective of this study was therefore to analyze effects of MR feeding, weaning, and supplementation of short-chain fructo-oligosaccharides (FOS) on the development of glucose homeostasis and insulin sensitivity in calves during the first 3 months of life. Thirty male Holstein-Friesian calves (18 ± 0.7 days of age) were assigned to 1 of 3 dietary treatments: the control (CON) group received MR only, the FOS group received MR with the addition of short-chain FOS, and the solid feed (SF) group was progressively weaned to SF. The CON and FOS calves received an amount of MR, which gradually increased (from 400 to 1,400 g/d) during the 71-day trial period. For the SF calves, the amount of MR increased from 400 to 850 g/d at day 30, and then gradually decreased, until completely weaned to only SF at day 63. The change in whole body insulin sensitivity was assessed by intravenous glucose tolerance tests. Milk tolerance tests were performed twice to assess changes in postprandial blood glucose, insulin, and nonesterified fatty acid responses. Whole-body insulin sensitivity was high at the start (16.7 ± 1.6 × 10−4 [μU/mL]−1), but decreased with age to 4.2 ± 0.6 × 10−4 [μU/mL]−1 at the end of the trial. The decrease in insulin sensitivity was most pronounced (~70%) between day 8 and 29 of the trial. Dietary treatments did not affect the decrease in insulin sensitivity. For CON and FOS calves, the postprandial insulin response was 3-fold higher at the end of the trial than at the start, whereas the glucose response remained similar. The SF calves, however, showed pronounced hyperglycemia and hyperinsulinemia at the end of the trial, although weaning did not affect insulin sensitivity. We conclude that whole body insulin sensitivity decreases by 75% in calves during the first 3 months of life. Weaning or supplementation of short-chain FOS does not affect this age-related decline in insulin sensitivity. Glucose homeostasis is not affected by supplementation of short-chain FOS in young calves, whereas postprandial responses of glucose and insulin to a MR meal strongly increase after weaning.

Key words: veal calf, weaned calf, insulin sensitivity, glucose homeostasis, fructo-oligosaccharide

Insulin sensitivity in young calves

83

5

5.1 Introduction

Veal calves are fed milk replacer (MR) and solid feed (SF; consisting of concentrates and roughages). The vast majority of ingested MR is directed into the abomasum through closure of the esophageal groove upon MR ingestion. Despite the tendency to increase the amounts of roughage and concentrates in veal calf diets, approximately 60 to 70% of the digestible nutrient intake originates from MR.

Milk replacer contains high amounts of lactose (~45%) and fat (~20%) on a DM basis. It has been shown that a prolonged intake of high levels of MR (hence large amounts of lactose and fat) may induce problems with glucose homeostasis and insulin sensitivity in heavy veal calves (> 4 months old), as characterized by high incidences of hyperglycemia and hyperinsulinemia (Hostettler-Allen et al., 1994; Hugi et al., 1997). These problems may ultimately result in (pro)inflammatory stress and metabolic diseases, as evident from human studies (Hotamisligil, 2006; Shoelson et al., 2006), and hepatic steatosis (Gerrits et al., 2008). A previous study showed that replacing 33% of the lactose in the MR by fructose or glycerol improved postprandial glucose homeostasis (i.e., lower glucose and insulin peaks), but not insulin sensitivity in calves (Pantophlet et al., 2016). More interestingly, that study also showed that calves at ~14 weeks of age are already relatively insensitive to insulin, compared with healthy non-ruminants (Caumo et al., 2000; Stefanovski et al., 2011). Another study revealed that insulin sensitivity in neonatal dairy calves (< 6 weeks of age) decreases from week 3 to 6, when calves are gradually weaned (Stanley et al., 2002). Moreover, Bach et al. (2013) reported that a greater level of MR allowance (8 vs. 4 L/d) had a negative effect on the development of insulin sensitivity in young calves (1–8 weeks of age) with ad libitum access to starter feed. Yet, it remains unclear whether the decrease in insulin sensitivity in young calves can be influenced by feeding strategy (prolonged MR feeding vs. progressive weaning), or is explained by the ontogenetic development of calves. In contrast to dairy calves, veal calves are commonly maintained at diets containing large amounts of MR. Preventing the decrease in insulin sensitivity in veal calves during early life may augment the efficiency of energy utilization for growth and possibly also metabolic health in later life. This requires more insight in the changes in insulin sensitivity in young calves and underlying mechanisms (age- or diet-related).

In addition, the reduction in insulin sensitivity in young calves may to some extent be prevented by supplementation of short-chain fructo-oligosaccharides (FOS). Studies in various animal species have shown that dietary short-chain FOS affect whole body insulin sensitivity. In dogs and horses with obesity, for example, an increase in insulin sensitivity was measured after feeding FOS for a period of 6 weeks (Respondek et al., 2008, 2011). The mechanisms behind the effects of short-chain FOS are poorly understood, but it has been hypothesized that short-chain FOS alters the intestinal microbiota composition, which directly or indirectly increases insulin sensitivity. In veal calves, oral short-chain FOS supplementation to a high-lactose MR decreased

CHAPTER 5

84

5

postprandial levels of plasma glucose and lactate, whereas insulin levels increased (Kaufhold et al., 2000). Whether this is due to changes in insulin sensitivity is not known, but if similar mechanisms operate in young calves then short-chain FOS supplementation may improve insulin sensitivity.

Therefore, the objective of this study was to assess age-related and diet-induced (i.e., MR only vs. MR+FOS vs. progressive weaning) changes in whole-body glucose homeostasis and insulin sensitivity in veal calves during the first 3 months of life.



Thirty male Holstein-Friesian calves were housed at experimental facility “De Haar” of VanDrie Group, the Netherlands. At start of the trial, calves were 18 ± 0.7 days of age and weighed 44 ± 0.3 kg (both mean ± SEM).

During the whole trial calves were housed individually, in 1.50 × 1.10 m pens fitted with a wooden slatted floor and galvanized fencings. Ventilation occurred by ceiling fans, and illumination by natural light and artificial (fluorescent lamps) light between 0600 and 1800 h. The average temperature and humidity were 18.2 ± 0.1°C and 75.3 ± 0.1%, respectively (both mean ± SEM).

Experimental procedures complied with the Dutch Law on Experimental Animals, and the ETS123 (Council of Europe 1985 and the 86/609/EEC Directive) and were approved by the Animal Care and Use Committee of Wageningen University.


Calves were randomly assigned to 1 of 3 treatment groups: the control group (CON, n = 10), fructo-oligosaccharide group (FOS, n = 10), or solid feed group (SF, n = 10). Calves were fed amounts of MR (see Table 5.1 for MR composition) according to practical feeding schemes (Table 5.2), which were based on the expected BW. Body weight was measured weekly to monitor for any major deviations from the expected BW gain. The CON and FOS calves received similar amounts of MR, which increased from 400 g/d at the start to 1,400 g/d at the end of the trial. In addition, FOS calves received short-chain FOS supplementation (Profeed P95, Beghin-Mei-ji, Marckolsheim, France), which increased gradually from 0.8 g/d on day 5 to 2.2 g/d at the end of trial (day 71), and was equally divided over the 2 daily feedings and provided with the MR. For SF calves, the amount of MR increased from 400 g/d at the start of trial to 850 g/d at day 39 of the trial. Subsequently, the amount of MR gradually decreased to 250 g/d at day 60. From day 63 these calves were completely weaned to only SF.

In addition to MR, all calves received SF, which consisted of 70.9% concentrate, 14.5% wheat straw, and 14.6% alfalfa (based on DM). For CON and FOS calves, the amount of SF increased from 0 to 376 g/d during the trial (Table 5.2), whereas this amount increased from 0 to 2,158 g/d for SF calves.


85

5

Milk was fed at a concentration of 125 g of MR/L reconstituted MR and supplied at a temperature of ~43°C. Calves were fed from buckets at 0630 and 1600 h. The MR refusals were weighed and recorded twice a day (maximum 60 min after feeding). After MR feeding, SF was provided in the buckets. The SF refusals were collected before the next MR feeding, in separate buckets, and weighed and recorded once a week. Calves were allowed ad libitum access to water.

Table 5.1 | Ingredient and nutrient composition of the milk replacer.

Ingredients g/kg Nutrients2 g/kg DM

Delactosed whey powder 149.7 Dry matter, g/kg 978

Whey 466.1 Crude protein (N x 6.25) 207

Whey protein concentrate 111.3 Crude fat 193

Soy protein concentrate 10.0 Crude ash 86

Soluble wheat protein 64.1 Lactose3 473

Coconut oil 49.2 Fe, mg/kg DM 45.9

Lard 52.5

Tallow 52.5

Lecithin 5.0

Emulsifier 5.0

Premix1 10

Calcium formate 8.4

Methionine 2.1

Lysine 10.0

Threonine 2.0

Calcium carbonate 2.51 Premix (per kg milk replacer): crude protein, 0.7 g; starch, 5.0 g; crude ash, 1.5 g; Ca, 16.6 mg; P, 7.5 mg; Na, 0.7 mg; K, 7.3 mg; Cl, 13.1 mg; Mg, 0.5 g; Fe, 44 mg; Cu, 8.0 mg; Zn, 110 mg; Mn, 43 mg; Se, 0.3 mg; I, 1.0 mg; Vitamin A, 25,013 IU; Vitamin D3, 4,002 IU; Vitamin E, 135 mg; Vitamin K3, 2.1 mg; Vitamin C, 0.3 g; Vitamin B1, 8.2 mg; Vitamin B2, 10.2 mg; Vitamin B3, 34.9 mg; Vitamin B5, 18 mg; Vitamin B6, 10 mg; Vitamin B12, 0.1 mg; biotin, 0.2 mg; folate, 0.7 mg; choline 0.4 g. 2 Values in g per kg dry matter unless stated otherwise.3 Calculated content.


5.2.3.1 Intravenous glucose tolerance testAn intravenous glucose tolerance test (IVGTT) was performed on day 8, 29, 50, and 71 of the trial. Prior to the first 3 IVGTT, all calves were fasted for 15 to 20 h. Before the last test, SF calves were fasted for a longer time period (20 to 30 h) to prevent strong interference of nutrient absorption from rumen fermentation.

CHAPTER 5

86

5

At least 1 h before the IVGTT, the calves were prepared with a central venous catheter (Intraflon 2, 14G 80 mm, Vygon Nederland BV, Valkenswaard, the Netherlands) in a jugular vein for glucose infusion and blood sampling. At t = 0 min, an intravenous glucose bolus of 0.18 g/kg of BW (20% glucose solution, Baxter, Utrecht, the Netherlands) was administered within 2 min. Blood samples were collected at t = −4, −2, 4, 8, 12, 18, 24, 30, 36, 45, and 60 min relative to the glucose bolus. Blood samples were transferred immediately into 6-mL lithium-heparin vacutainer tubes (Becton Dickinson, Breda, the Netherlands) and stored on ice. Plasma was collected, after centrifugation, within 45 min after blood sampling and stored at −20°C until analysis of plasma glucose and insulin.

The insulin sensitivity index (ISi) was calculated from the following equation, proposed by Tura et al. (2010):

/α=∆

KgISiAUCins T

where α is a scaling factor (α = 0.276), Kg is rate of glucose disappearance (slope of log glucose), ΔAUCins is the area under the curve (above basal) of insulin, calculated using the trapezoid method (Le Floch et al., 1990), and T is the time interval between 8 and 60 min (= 52 min) when Kg and ΔAUCins are computed.

Table 5.2 | Feeding schemes of veal calves fed either a milk replacer (MR) diet, a MR diet with the addition of short-chain fructo-oligosaccharides or calves which were progressively weaned to only solid feed.

Experimental week

Treatment

CON FOS SF

Milk powder1 (g/d)

Solid feed2 (g/d)

Milk powder1

(g/d)Solid feed2

(g/d)FOS3 (g/d)

Milk powder1 (g/d)

Solid feed2 (g/d)

1 400 0 400 0 - 400 0

2 450-600 59 450-600 59 0.78 450-575 59

3 600-700 135 600-700 135 1.05 575-625 187

4 725-900 198 725-900 198 1.21 650-800 438

5 900-1,000 248 900-1,000 248 1.52 800-850 319

6 1,000-1,050 277 1,000-1,050 277 1.64 850-800 510

7 1,050-1,150 374 1,050-1,150 374 1.76 800-700 882

8 1,150-1,225 245 1,150-1,225 245 1.83 700-550 983

9 1,250-1,288 438 1,250-1,288 438 2.03 500-2504 1,726

10 1,300-1,400 376 1,300-1,400 376 2.10 - 2,1581 Daily portion was equally divided over two feedings. The MR was fed at a concentration of 125 g MR/L. Values represent amounts at the beginning and end of the week, respectively.2 Solid feed consisted of 70.9% concentrates, 14.5% wheat straw and 14.6% alfalfa (based on DM).3 Daily portion was equally divided over two feedings and provided with the MR.4 Calves were weaned from the MR from day 6 during this week.


87

5

5.2.3.2 Milk tolerance test A milk tolerance test (MTT) was performed on day 22 and 64 of the trial. Prior to the first MTT, all calves were fasted for 16 to 21 h. Prior to the last MTT, SF calves were fasted for a longer time period (20 to 25 h).

All calves received 7.6 g of MR per kg of BW (t = 0 min). Blood was collected from a jugular vein by venipuncture, at t = −5, 30, 60, 120, 180, 240, and 360 min relative to feeding. Blood samples were transferred immediately into 4-mL lithium heparin vacutainer tubes (Becton Dickinson) for glucose and insulin analyses and stored on ice, and into 4-mL sodium fluoride/ potassium oxalate vacutainer tubes (Becton Dickinson) for non-esterified fatty acid (NEFA) analysis. Plasma was collected, after centrifugation, within 45 min after blood sampling and stored at −20°C until analysis.

For glucose and insulin, the ΔCmax (= maximum concentration – fasting concentration), time to maximum concentration, and incremental area under the curve (iAUC0–6h) were calculated. For NEFA, the ΔCmin (= fasting concentration − minimum concentration) and time to maximum concentration was calculated. The iAUC0–6h was calculated using the trapezoid method (Le Floch et al., 1990).


Plasma glucose was measured on a Roche-Hitachi Modular automatic analyzer (Roche Diagnostics, Basel, Switzerland) using an enzymatic colorimetric method. The within- and between-run CV was ≤2%. Insulin was analyzed using a bovine ELISA kit (Mercodia, Uppsala, Sweden). The within- and between-run CV was ≤5.6 and 8.2%, respectively. The NEFA was analyzed using a NEFA FS kit (Diasys, Holzheim, Germany). The within- and between-run CV was ≤2%.

5.2.5 Statistical analyses

The SPSS (version 22, IBM, SPSS Inc., Chicago, IL) statistical software was used for all statistical analyses. Data are presented as means ± SEM. Insulin sensitivity, glucose disappearance rate (Kg), and insulin response (AUCins) during the IVGTT were analyzed for time, treatment, and time × treatment effects using he mixed effects model procedure. Time, treatment, and their interaction were used as fixed terms, and time as repeated variable within calf. Based on fit statistics (Akaike and Bayesian information criteria), the heterogeneous first-order autoregressive covariance structure was used for all models. When a significant interaction was found, treatment effects were analyzed for each time point separately.

The fasting and postprandial blood parameters (i.e., ΔCmax, time to maximum/minimum, and iAUC0–6h) were also analyzed for time, treatment, and time × treatment effects using the mixed effects model procedure. Time, treatment, and their interaction were used as fixed terms, and time as repeated variable within calf. Based on fit statistics (Akaike and Bayesian information criteria), the heterogeneous first-order autoregressive covariance structure was used for all models. When a significant

CHAPTER 5

88

5

interaction was found, treatment effects were analyzed for each time point separately. Fasting insulin levels were not statistically evaluated because most values were between the limit of detection and the limit of quantification; hence, only means ± SEM are provided.

The effect of treatment on the average daily gain, calculated over the complete trial period (i.e., BW at end of the trial – BW at start of the trial divided by the length of the trial in days), was assessed using the GLM univariate procedure.

Normality of the model residuals was assessed by visual inspection. Non-normal distributed data were (log) transformed to obtain normality. P-values < 0.05 were considered significant. When significant, the Bonferroni method was used to correct for multiple comparisons.

5.3 Results

5.3.1 Feed intake and growth performance

Refusals of MR were negligible throughout the trial. The SF refusals, however, were high (45%) at the start of the trial but decreased to 6% at the end of the trial. Average daily BW gain, measured over the complete trial, was approximately 100 g lower (P = 0.006) in SF calves (643 ± 29 g/d) than in CON and FOS calves (743 ± 24 g/d and 740 ± 14 g/d, respectively).

5.3.2 Insulin sensitivity, glucose disappearance, and insulin response

Whole-body insulin sensitivity (ISi) decreased (P < 0.01) 4-fold in all groups during the trial (Figure 5.1). The decrease did not differ between groups (time × treatment interaction, P > 0.05). The rate of glucose disappearance (Kg) during the IVGTT decreased (P < 0.01) by 22% during the trial. The decrease did not differ between groups (time × treatment interaction, P > 0.05). The total insulin response (AUCins) to the intravenous glucose bolus increased (P < 0.01) 2.9-fold during the trial. However, the increase did not differ between groups (time × treatment interaction, P > 0.05).


89

5Figure 5.1 | Insulin sensitivity (A), rate of glucose disappearance (Kg; solid lines, B) and insulin response (AUCins; dashed lines, B) to an IVGTT in veal calves fed either a milk replacer (MR) diet (; n=10), a MR diet with the addition of short-chain fructo-oligosaccharides (; n=10), or calves which were progressively weaned to only solid feed (; n=10). Error bars represent SEM. Significant (P < 0.05) time effects were found for all 3 parameters, but no time × treatment interaction.

5.3.3 Fasting levels and postprandial responses of plasma glucose, insulin, and NEFA

Fasting levels and responses of plasma glucose, insulin, and NEFA to the MTT are shown in Figure 5.2 and Table 5.3. Compared with the start of the trial, fasting plasma glucose concentrations were higher (P < 0.001; Table 5.3) at end of the trial in CON and FOS calves, but lower (P < 0.01) in SF calves (time × treatment, P < 0.001). Fasting plasma insulin levels changed numerically for all treatment groups during the trial (significance could not be tested as most of the observations were between the limit of detection and the limit of quantification). Compared with the start of the trial, fasting plasma NEFA concentrations were higher at end of the trial in SF (P < 0.001) and FOS calves (P < 0.05), but did not change over time in CON calves (time × treatment

CHAPTER 5

90

5

interaction, P < 0.001). At end of the trial, fasting plasma NEFA concentrations were markedly higher (P < 0.001) in SF calves than in CON and FOS calves.

For SF calves, the time to maximum (Tmax) for plasma glucose after the MTT at the end of the trial was delayed by more than 100 min (P < 0.001) when compared with the start of the trial, whereas Tmax for plasma glucose was unaffected in CON and FOS calves (time × treatment interaction, P < 0.001). Consequently, at end of the trial, Tmax for plasma glucose was markedly larger (P < 0.001) for SF calves than for CON and FOS calves. For SF calves, Tmax for plasma insulin was delayed by more than 60 min (P < 0.01) at the end of the trial when compared with the start of the trial, whereas this was unaffected in CON and FOS calves. The time to minimum (Tmin) for plasma NEFA was delayed by almost 60 min (P < 0.01) when comparing measurements at the end and start of the trial for SF calves, whereas this was unaffected in CON and FOS calves (time × treatment interaction, P < 0.05). Consequently, at end of the trial, Tmin for plasma NEFA was markedly larger (P < 0.001) for SF calves than for CON and FOS calves.

For SF calves, ΔCmax for glucose after the MTT was larger (P < 0.01) at end than at the start of the trial, whereas ΔCmax for the CON and FOS calves remained unaffected (time × treatment interaction, P < 0.001). Consequently, at end of the trial, ΔCmax for glucose was larger (P < 0.01) for SF calves than for CON and FOS calves. For all treatment groups, ΔCmax for insulin after the MTT was larger (P < 0.05) at the end of the trial than at the start of the trial. For SF and FOS calves, ΔCmin for NEFA was larger (P < 0.001 and P < 0.05, respectively) at the end of the trial, whereas responses of plasma NEFA to the MTT were not affected in CON calves (time × treatment interaction, P < 0.001). At end of the trial, ΔCmin for NEFA was markedly larger (P < 0.001) for SF calves than for CON and FOS calves.

The AUC0–6h glucose after the MTT was larger (P < 0.001) at the end than at the start of the trial for SF calves, but not for CON and FOS calves (time × treatment interaction, P < 0.001). Consequently, at the end of the trial, AUC0–6h glucose was markedly larger (P < 0.001) for SF calves than for CON and FOS calves. For all treatment groups, AUC0–6h insulin was increased (P < 0.01) at the end compared with the start of the trial.


91

5

Figure 5.2 | Plasma glucose, insulin and NEFA responses in calves fed (time=0) a fixed amount of milk replacer (MR), according to their body weight (i.e., milk tolerance test). The tolerance test was performed on day 22 and 64 of the trial. During the trial calves were fed either a MR diet (; n=10), a MR diet with the addition of short-chain fructo-oligosaccharides (; n=10), or calves which were progressively weaned to only solid feed (; n=10). Error bars represent SEM. Calculated parameters and statistics are given in Table 5.3.

CHAPTER 5

92

5

Table 5.3 | Fasting levels and postprandial responses of plasma glucose and insulin, and NEFA (means ± SEM) in veal calves fed either a milk replacer (MR) diet (CON; n=10), a MR diet with the addition of short-chain fructo-oligosaccharides (FOS; n=10), or calves which were progressively weaned to only solid feed (SF; n=10).

Age (weeks)

Treatment (Trt)

CON FOS SF P-value

5 11 5 11 5 11 Time Time × Trt

Glucose

Fasting, mmol/L 4.7±0.2 5.8±0.3a 4.8±0.1 5.6±0.1a 4.5±0.2 3.9±0.2b 0.002 <0.001

TMax, min 66±10 51±6a 66±10 63±7a 60±8 162±16b <0.001 <0.001

ΔCMax, mmol/L 2.3±0.3 2.6±0.4a 2.0±0.4 2.7±0.4a 2.3±0.3 4.8±0.4b <0.001 <0.001

iAUC0-6h 396±60 413±100a 368±50 415±53a 457±90 1,294±113b <0.001 <0.001

Insulin

Fasting, mU/L 0.9±0.1 1.6±0.2 0.8±0.1 1.6±0.3 0.9±0.2 1.3±0.2 - -

TMax, min 78±9 99±14 102±19 107±8 87±11 150±18 0.009 0.062

ΔCMax, mU/L 60±11 132±29 76±27 182±39 32±9 145±30 <0.001 0.410

iAUC0-6h x 103 8.0±1.1 19±3.2 11±3 23±5 5.1±1.3 26±6 <0.001 0.112

NEFA

Fasting, μmol/L 253±14 285±32a 176±25 275±22a 215±37 542±44b <0.001 <0.001

TMin, min 45±9 57±14a 33±4 48±10a 39±5 97±15b <0.001 0.015

ΔCMin, μmol/L 208±16 228±30a 143±26 227±20a 176±34 445±47b <0.001 <0.001a-b Different superscripts indicate differences between groups at 11 weeks of age (P < 0.05).

5.4 Discussion

5.4.1 Growth performance and feed intake average

SF refusals were high at start of the trial (~45%), but decreased to ~8% at the end. Although the relative SF refusals (i.e., expressed as percentage of SF allowance) did not differ between the groups, the absolute amount of SF refusals was higher for SF calves. Therefore, the progressive increase in SF intake in SF calves was limited by their intake capacity during the trial, which resulted in slightly lower average daily BW gains for SF calves.

5.4.2 Decrease in insulin sensitivity

Insulin sensitivity decreased with age in all treatment groups; the glucose disappearance rate (Kg) decreased with age, whereas the total insulin response (AUCins) increased. Also, the time to peak insulin (response to glucose infusion) did not differ in age or between treatments (data not shown). The age-related decrease in insulin sensitivity found in the current trial is in agreement with previous studies performed on young calves (Stanley et al., 2002; Bach et al., 2013; Yunta et al., 2015). This is, however, the


93

5

first study in which this decrease has been assessed and compared in calves that were maintained on MR and calves that were progressively weaned during their first 3 months of life, indicating that the decrease in sensitivity is independent of feeding strategy. Interestingly, the biggest decrease in insulin sensitivity (~70%), which was ~90% of the total decrease during the trial period, occurred during the first 3 weeks, when calves were 3 to 6 weeks of age. During the first 3 weeks of the trial, the decrease in insulin sensitivity was not affected by weaning, which may be due to the relatively small differences in feed intake during this period (i.e., when effects may be expected due to the high insulin sensitivity). The age-related decrease in insulin sensitivity has also been observed in young lambs (Gelardi et al., 1999), rats (Torlińska et al., 2000; Yuan et al., 2011), and piglets (Bergeron et al., 2007). Several physiological factors may be involved in the age-related decrease in insulin sensitivity. First, the number of insulin receptors in skeletal muscle, liver, and adipose tissue tend to decrease with age. A study in newborn rats, for example, has shown that the number of high affinity insulin receptors in skeletal muscle substantially decreases between 5 days to 1 year of age (Torlińska et al., 2000). Also, a decreased number of skeletal muscle insulin receptors has been found in (insulin resistant) veal calves (> 4 months old) fed a high lactose diet compared with a standard lactose diet (Hugi et al., 1998), suggesting that the number of insulin receptors (and thus insulin sensitivity) is also influenced by nutritional factors. Second, a decrease in the number of insulin-stimulated glucose transporters (GLUT4) in skeletal muscle and adipose tissue may contribute to the age-related decrease in insulin sensitivity. For example, an age-related decrease has been reported in epitrochlearis muscles of 1 to 13-month-old rats (Cartee et al., 1993; Gulve et al., 1993). In neonatal goat and lambs (< 5 weeks of age), however, no difference in GLUT4 was found in adipose tissue and skeletal muscle (Trayhurn et al., 1993; Gelardi et al., 1999). Third, age-related changes in post-receptor insulin signaling proteins such as insulin receptor substrate-1 have been associated with the development of insulin resistance (Carvalho et al., 1996; Nagasaki et al., 2000). These proteins are influenced by nutrients such as glucose, amino acids, and fatty acids (Blagosklonny, 2013). Nevertheless, further research is required to determine whether a combination of these factors (i.e., insulin receptor, GLUT4, insulin signaling proteins), or possibly other factors are responsible for the strong age-related decrease in whole body insulin sensitivity in young calves.

In the current study, short-chain FOS did not affect the decrease in insulin sensitivity in young calves. To our knowledge, this is the first study that assessed effects of short-chain FOS on insulin sensitivity in young calves. Kaufhold et al. (2000) studied effects of short-chain FOS supplementation on glucose homeostasis in veal calves at 10 weeks of age and found lower glucose and higher insulin peaks after feeding MR, but effects on insulin sensitivity were not assessed. From our previous study, we learned, however, that differences in postprandial glucose homeostasis are not always associated with differences in insulin sensitivity in calves (Pantophlet et al., 2016). In non-ruminants,

CHAPTER 5

94

5

such as dogs and horses, short-chain FOS supplementation improved insulin sensitivity (Respondek et al., 2008, 2011). The absence of an effect of short-chain FOS on insulin sensitivity in young calves is not yet understood, but might be related to differences in the dosage administered (in our study calves received a relatively low dosage compared with, for example, horses), the form or period of administration, digestive or metabolic differences between ruminants and non-ruminants, or a combination of these. Explaining interspecies differences in the efficacy of short-chain FOS on insulin sensitivity, therefore, requires additional studies.

5.4.3 Postprandial responses of plasma glucose, insulin, and NEFA

For CON and FOS calves, fasting glucose levels increased with age. The increase is in agreement with Hugi et al. (1997), but not with Hostettler-Allen et al. (1994), who found no increase in fasting plasma glucose levels in calves from 5 to 13 weeks of age. Fasting insulin increased numerically with age in all treatment groups, which corresponds with an age-related increase in fasting insulin in heavy (> 4 months old) veal calves (Pantophlet et al., 2016). In contrast to CON calves, fasting plasma glucose decreased and NEFA levels increased substantially for the SF calves with age, indicating increased (energetic need for) fat metabolism in these calves. This can likely be explained by a relatively lower ME intake in calves after weaning, as the progressive increase in SF intake was limited by their intake capacity. Similarly, an increase in fat metabolism (and thus NEFA levels) to meet energy requirements after weaning was also seen in early-weaned lambs (Fennessya et al., 1972).

The total plasma insulin response (AUC0–6h insulin) after the MTT increased with age for CON and FOS calves, whereas the total plasma glucose response (AUC0–6h glucose) did not change. This corresponds with the age-related reduction in insulin sensitivity, as more insulin is needed to allow glucose disappearance from the systemic circulation. The AUC0–6h insulin is however much smaller than in heavy veal calves (Pantophlet et al., 2016). Interestingly, SF calves demonstrated hyperglycemia and hyperinsulinemia (i.e., high and broad glucose and insulin peaks not returning to basal levels) during the second MTT (i.e., post-weaning). The reason for this is unclear but might be attributed to several pre- and post-absorptive factors, as these calves might have adapted their digestion and metabolism due to adaptation to SF instead of MR. One factor that could contribute to the delayed glucose peaks is a difference in abomasal emptying rate between milk-fed and weaned calves. Abomasal emptying was shown to occur very rapidly after feeding in milk-fed calves (Labussière et al., 2014), but in weaned calves the emptying rate may be reduced due to an increased consistency of the digesta as affected by greater SF intake. As a consequence, digestion and absorption of carbohydrates from MR may have been delayed. Another explanation could be a decreased activity of lactase in the brush-border of the small intestine, which would shift glucose absorption to a more distal region of the small intestine. Previous studies have shown that lactase


95

5

activity in milk-fed calves is highest in the proximal small intestine (Le Huerou et al., 1992; Gilbert et al., 2015). Although weaning negatively affects small intestinal lactase activity in calves (Le Huerou et al., 1992), this has not yet been associated with a different site of digestion. In addition to the sustained glucose appearance in SF calves after the MTT, glucose clearance was likely also reduced in these calves (i.e., the high glucose peaks not returning to basal levels within 6 h after feeding). As insulin sensitivity did not differ between treatments, other underlying mechanisms may operate. First, first pass uptake of glucose by splanchnic tissues (mainly liver and small intestine) and portal-drained viscera (PDV) may decrease upon adaptation of these tissues to the lack of luminal lactose supply after weaning. In steers, for example, infusion of starch into the abomasum increases glucose utilization by the PDV when compared with ruminal infusion (Harmon et al., 2001). Therefore, splanchnic tissues of milk-fed calves may use more glucose in the first pass. Consequently, the postprandial MTT glucose response may have been higher in weaned calves as a combined result of reduced firstpass utilization and a reduced clearance rate of glucose. Second, the relative contribution of glucose transport by insulin-independent transporters (GLUT1) may be relatively more important in milk-fed than in weaned calves. The age-related development of GLUT1 and GLUT4 in milk-fed calves is not known, but Abe et al. (2001) reported that the number of insulin-dependent transporters (i.e. GLUT4) in skeletal muscle and adipose tissues of calves decreases after weaning, whereas the number of GLUT1 transporters is unaffected. Assuming that the insulin-sensitive GLUT4 also decreases with age in MR-fed calves (due to the age-related decrease in insulin sensitivity observed in both MR an weaned calves), it might be that, in contrast to weaned calves, glucose transport by insulin-independent transporters increases in MR calves to help facilitate glucose uptake, due to the prolonged high intakes of dietary lactose (i.e., high glucose load). Third, adaptation of the endocrine pancreas to the feeding strategy (e.g., roughage and concentrates vs. milk) or changes in hepatic glucose production (e.g., increased gluconeogenesis in ruminants) may also have contributed to the divergent metabolic responses to MR feeding in weaned calves.

5.5 Conclusions

Whole-body insulin sensitivity decreases by 75% in calves during the first 3 months of life, with the most pronounced decrease occurring within the first 6 weeks. This decrease in insulin sensitivity is not affected by weaning or by supplementation of short-chain FOS. Postprandial insulin increases with age for CON and FOS calves, whereas postprandial glucose remains unaffected, which is consistent with the decrease in insulin sensitivity in calves during early life. After weaning, SF calves show pronounced hyperglycemia and hyperinsulinemia after a milk tolerance test. The underlying mechanisms of this apparent mismatch are not yet known.

CHAPTER 5

96

5

5.6 Acknowledgments

The authors thank Karin Tholen-Cazemier, Gerlof Reckman, and Jeltje Kloosterman for their laboratory assistance (University Medical Centre Groningen, Groningen, the Netherlands), and the animal caretakers at “De Haar,” the experimental facilities of VanDrie Group. This project was jointly financed by the European Union, the European Regional Development Fund and the Ministry of Economic Affairs, Agriculture and Innovation, Peaks in the Delta, the Municipality of Groningen, the Provinces of Groningen, Fryslan, and Drenthe as well as the Dutch Carbohydrate Competence Center (CCC2 WP21). Financial support was also provided by Tereos Starch and Sweeteners Europe (Marckolsheim, France), VanDrie Group (Mijdrecht, the Netherlands), and Wageningen University (Wageningen, the Netherlands).


97

5

5.7 References

Abe, H., Y. Kawakita, K. Hodate, and M. Saito. 2001. Postnatal development of glucose transporter proteins in bovine skeletal muscle and adipose tissue. J. Vet. Med. Sci. 63:1071-1075.


Bergeron, K., P. Julien, T. A. Davis, A. Myre, and M. C. Thivierge. 2007. Long-chain n-3 fatty acids enhance neonatal insulin-regulated protein metabolism in piglets by differentially altering muscle lipid composition. J. Lipid Res. 48:2396-2410.

Blagosklonny, M. V. 2013. TOR-centric view on insulin resistance and diabetic complications: perspective for endocrinologists and gerontologists. Cell Death Dis. 12:e964.

Cartee, G. D., C. Briggs-Tung, and E. W. Kietzke. 1993. Persistent effects of exercise on skeletal muscle glucose transport across the life-span of rats. J. Appl. Physiol. 75:972-978.

Carvalho, C. R., S. L. Brenelli, A. C. Silva, A. L. Nunes, L. A. Velloso, and M. J. Saad. 1996. Effect of aging on insulin receptor, insulin receptor substrate-1, and phosphatidylinositol 3-kinase in liver and muscle of rats. Endocrinology 137:151-159.

Caumo, A., R. N. Bergman, and C. Cobelli. 2000. Insulin sensitivity from meal tolerance tests in normal subjects: a minimal model index. J. Clin. Endocrinol. Metab. 85:4396-4402.

Fennessya, P. F., M. R. Woodlocka, and K. T. Jaguscha. 1972. The effect of early weaning on the concentrations of non-esterified fatty acids and glucose in the plasma of lambs. New Zeal. J. Agr. Res. 15:802-807.

Gelardi, N. L., R. E. Rapoza, J. F. Renzulli, and R. M. Cowett. 1999. Insulin resistance and glucose transporter expression during the euglycemic hyperinsulinemic clamp in the lamb. Am. J. Physiol. 277:E1142-E1149.


Gilbert, M. S., A. J. Pantophlet, H. Berends, A. M. Pluschke, J. J. G. C. van den Borne, W. H. Hendriks, H. A. Schols, and W. J. J. Gerrits. 2015. Fermentation in the small intestine contributes substantially to intestinal starch disappearance in calves. J. Nutr. 145:1147-1155.

Gulve, E. A., E. J. Henriksen, K. J. Rodnick, J. H. Youn, and J. O. Holloszy. 1993. Glucose transporters and glucose transport in skeletal muscles of 1- to 25-mo-old rats. Am. J. Physiol. 264:E319-E327.

Harmon, D. L., C. J. Richards, K. C. Swanson, J. A. Howell, J. C. Mathews, A. D. True, G. B. Huntington, S. A. Gahr, and R. W. Russell. 2001. Influence of ruminal or postruminal starch on visceral glucose metabolism in steers. Pages 273-276 in Energy metabolism in animals. EAAP Publication No. 103. A. Chwalibog and K. Jakobsen, ed. Wageningen pers, Wageningen, the Netherlands.


Hotamisligil, G. S. 2006. Inflammation and metabolic disorders. Nature 444:860-867.Hugi, D., R. M. Bruckmaier, and J. W. Blum. 1997. Insulin resistance, hyperglycemia, glucosuria, and




Labussière, E., H. Berends, M. S. Gilbert, J. J. G. C. van den Borne, and W. J. J. Gerrits. 2014. Estimation of milk leakage into the rumen of milk-fed calves through an indirect and repeatable method. Animal 8:1643-1652.


Le Huerou, I., P. Guilloteau, C. Wicker, A. Mouats, J. A. Chayvialle, C. Bernard, J. Burton, R. Toullec, and A. Puigserver. 1992. Activity distribution of seven digestive enzymes along small intestine in calves during development and weaning. Dig. Dis. Sci. 37:40-46.

CHAPTER 5

98

5

Nagasaki, M., N. Nakai, Y. Oshida, Z. Li, M. Xu, M. Obayashi, T. Murakami, A. Yoshimura, N. Fujitsuka, Y. Shimomura, and Y. Sato. 2000. Exercise training prevents maturation-induced decreases in insulin receptor substrate-1 and phosphatidylinositol 3-kinase in rat skeletal muscle. Metabolism 49:954-958.

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, H. Roelofsen, M. G. Priebe, and R. J. Vonk. 2016. Lactose in milk replacer can partly be replaced by glucose, fructose or glycerol without affecting insulin sensitivity in veal calves. J. Dairy Sci. 99:3072-3080.





Stefanovski, D., J. M. Richey, O. Woolcott, M. Lottati, D. Zheng, L. N. Harrison, V. Ionut, S. P. Kim, I. Hsu, and R. N. Bergman. 2011. Consistency of the disposition Index in the face of diet induced insulin resistance: potential role of FFA. PLoS One 6:e18134.

Torlińska, T., P. Maćkowiak, L. Nogowski, T. Hryniewiecki, H. Witmanowski, M. Perz, E. M dry, and K. W. Nowak. 2000. Age dependent changes of insulin receptors in rat tissues. J. Physiol. Pharmacol. 51:871-881.

Trayhurn, P., M. E. Thomas, and J. S. Keith. 1993. Postnatal development of uncoupling protein, uncoupling protein mRNA, and GLUT4 in adipose tissues of goats Am. J. Physiol. 265:R676-R682.

Tura, A., S. Sbrignadello, E. Succurro, L. Groop, G. Sesti, and G. Pacini. 2010. An empirical index of insulin sensitivity from short IVGTT: validation against the minimal model and glucose clamp indices in patients with different clinical characteristics. Diabetologia. 53:144-152.

Yuan, Q., L. Chen, C. Liu, L. Xu, X. Mao, and C. Liu. 2011. Postnatal pancreatic islet β cell function and insulin sensitivity at different stages of lifetime in rats born with intrauterine growth retardation. PLoS One 6:e25167.

Yunta, C., M. Terré, and A. Bach. 2015. Short- and medium-term changes in performance and metabolism of dairy calves offered different amounts of milk replacers. Livest Sci 181:249-255.

CHAPTER 6

Short communication: Supplementation of fructo-oligosaccharides does not improve insulin sensitivity in

heavy veal calves fed different sources of carbohydrates

“The important thing in science is not so much to obtain new facts as to discover new ways of thinking about them”

- William Lawrence Bragg -

A.J. Pantophlet,1 M.S. Gilbert,2 J.J.G.C. van den Borne,2 W.J.J. Gerrits,2 H. and R.J. Vonk3

Journal of dairy science 2017; 100 (11): 9442-9446



CHAPTER 6

100

6

Abstract

Heavy veal calves (4–6 months old) often develop problems with insulin sensitivity. This could lead to metabolic disorders and impaired animal growth performance. Studies in various animal species have shown that the supplementation of short-chain fructo-oligosaccharides (scFOS) can improve insulin sensitivity. We therefore studied the effects of scFOS supplementation on insulin sensitivity in heavy veal calves. Forty male Holstein- Friesian calves (BW = 190 ± 2.9 kg; age = 162 ± 1.4 days at the start of the trial) were fed either a control milk replacer (MR) diet or a diet in which one-third of the lactose was replaced by glucose, fructose, or glycerol for 10 weeks prior to the start of the trial. At the start of the trial, calves were subjected to a frequently sampled intravenous glucose tolerance test to assess whole-body insulin sensitivity (muscle and hepatic insulin sensitivity). Calves within each dietary treatment group were ranked based on their insulin sensitivity value. Half of the calves received scFOS (12 mg/kg of BW) with the MR for 6 weeks (supplementation was equally distributed over the insulin sensitivity range). Subsequently, a second frequently sampled intravenous glucose tolerance test was conducted to assess the effect of scFOS. In addition, fasting plasma levels of glucose, insulin, triglycerides, and cholesterol were determined to calculate the quantitative insulin sensitivity check index and triglyceride:high-density lipoprotein cholesterol ratio (fasting indicators of insulin sensitivity). Whole-body insulin sensitivity was low at the start of the trial and remained low in all groups [1.0 ± 0.1 and 0.8 ± 0.1 (mU/L)−1 · min−1 on average, respectively]. Supplementation of scFOS did not improve insulin sensitivity in any of the treatment groups. The quantitative insulin sensitivity check index and the triglyceride:high-density lipoprotein cholesterol ratio also did not differ between scFOS and non-scFOS calves and averaged 0.326 ± 0.003 and 0.088 ± 0.004, respectively, at the end of the trial. We conclude that scFOS supplementation does not improve insulin sensitivity in heavy veal calves regardless of the carbohydrate composition of the MR. This is in contrast to other animals (e.g., dogs and horses), where scFOS supplementation did improve insulin sensitivity. The absence of an effect of scFOS might be related to the dosage or to metabolic differences between ruminants and non-ruminants. Increasing evidence indicates that dietary interventions in veal calves have little or no effect on insulin sensitivity, possibly because of low levels of insulin sensitivity.

Key words: veal calf, milk replacer, fructo-oligosaccharides, insulin sensitivity

Fructo-oligosaccharides and insulin sensitivity

101

6

6.1 Introduction

Veal calves are fed milk replacer (MR) and solid feed (roughage and concentrates) and are exposed to large quantities of lactose and fat via the MR. Prolonged high intakes of MR can induce problems with glucose homeostasis and insulin sensitivity in veal calves. These problems have been identified in heavy veal calves (4–6 months), characterized by a high incidence of hyperglycemia, hyperinsulinemia, and glucosuria (Hostettler-Allen et al., 1994; Hugi et al., 1997) and insulin resistance (Pantophlet et al., 2016a). In a recent study, we found that calves raised on a lactose MR diet, or diets in which one-third of the lactose was replaced by glucose, fructose, or glycerol, did not differ in insulin sensitivity and that insulin sensitivity was low (Pantophlet et al., 2016c). This could lead to metabolic disorders and impaired animal growth performance. Therefore, prevention strategies must be developed. Various animal studies have shown that dietary short-chain fructo-oligosaccharide (scFOS) supplementation can help prevent problems and improve whole-body insulin sensitivity (i.e., muscle and hepatic insulin sensitivity). In obese dogs and horses, for example, supplementation of dietary scFOS for a period of 6 weeks increased insulin sensitivity (Respondek et al., 2008; Respondek et al., 2011). In young veal calves (< 3 months old), supplementation of scFOS for 10 weeks did not improve insulin sensitivity or glucose homeostasis (Pantophlet et al., 2016b). In older veal calves (10–13 week old), however, scFOS supplementation did improve postprandial glucose homeostasis (Kaufhold et al., 2000). A decrease in postprandial response for glucose was observed, but not for insulin. Mechanisms were not reported. It is not clear whether the improved glucose homeostasis during scFOS supplementation in older calves is related to changes in insulin sensitivity. Therefore, the objective of this study was to assess the effects of scFOS supplementation on insulin sensitivity in heavy veal calves.

6.2 Material and Methods

Forty male Holstein-Friesian calves (BW = 190 ± 2.9 kg; age = 162 ± 1.4 days; mean ± SEM) were housed at the research facility of the Department of Animal Sciences at Wageningen University (Wageningen, the Netherlands). Calves were housed in groups (5 calves/pen) except for the first and last week of the trial. During these periods calves were housed individually in metabolic cages (0.80 × 1.8 m). Experimental procedures complied with the Dutch Law on Experimental Animals and the ETS123 (Council of Europe 1985 and the 86/609/EEC Directive) and were approved by the Animal Care and Use Committee of Wageningen University.

Prior to the start of the trial, the calves were fed either a control MR diet (n = 10) or diets in which one-third of the lactose in the MR was replaced by isoenergetic amounts of glucose (Tereos Syral, Marckolsheim, France; n = 10), fructose (Tate and Lyle Europe, Boleraz, Slovakia; n = 10), or glycerol (Triconor Distribution BV, Soest, the Netherlands;

CHAPTER 6

102

6

n = 10) for 10 weeks. All calves remained on their diets throughout the trial. A detailed description of the diets and feeding strategy is provided elsewhere (Pantophlet et al., 2016c). In short, calves were fed MR and solid feed (20% wheat straw and 80% concentrates) twice a day, at 0630 and 1530 h. The MR was fed on an individual basis (BW was measured weekly), and solid feed was provided per pen during group housing and per individual calf when the calves were housed in metabolic cages. Calves had ad libitum access to water.

Whole-body insulin sensitivity was assessed at the start and the end of the trial. The start values were used to assign calves to scFOS or no scFOS within each dietary treatment group (i.e., n = 5 for scFOS; n = 5 for no scFOS). The calves were ranked to evenly distribute the supplementation scFOS over the insulin sensitivity range in each group. The scFOS were added to the MR at a dose of 12 mg/kg of BW for 6 weeks. At the end of the 6-week period, whole-body insulin sensitivity was assessed again to study the effect of scFOS supplementation. One calf in the glucose group was excluded from all measurements due to ruminal drinking.

Whole-body insulin sensitivity was assessed using the frequently sampled intravenous glucose tolerance test. A detailed description of this experimental procedure is provided elsewhere (Pantophlet et al., 2016c). In short, a central venous catheter (Careflow, Becton Dickinson, Franklin Lakes, NJ) was inserted in the jugular vein for glucose and insulin infusion and blood sampling. Calves were fasted to achieve a steady glucose turnover rate before the test. At t = 0 min, an intravenous glucose bolus of 0.3 g/kg of BW (20% glucose solution; B. Braun, Oss, the Netherlands) was administered within 2 min followed by an intravenous insulin bolus of 0.03 IU/kg of BW (100 IU/mL solution; Insuman Rapid, Sanofi-Aventis, Gouda, the Netherlands) at t = 20 min (administered within 1 min). Blood samples were collected at t = −8, −4, 2, 4, 6, 8, 10, 12, 14, 16, 19, 22, 25, 30, 35, 40, 50, 60, 75, 90, 120, 150, and 180 min relative to the glucose bolus. The samples were centrifuged (1,516 × g for 10 min), and plasma was harvested for the analysis of plasma glucose and insulin concentrations. In addition, plasma triglycerides and high-density lipoprotein (HDL) cholesterol concentrations were analyzed in the fasting plasma sample collected at t = −8 min (at the start and the end of the trial). Whole-body insulin sensitivity was calculated according to Bergman’s minimal model approach using MinMod Millennium (version 6.0.2; MinMod Inc., Los Angeles, CA). In addition, another index of insulin sensitivity, the quantitative insulin sensitivity index (QUICKI; Muniyappa et al., 2008), was calculated from the fasting plasma glucose and insulin concentrations. Also, clean urine was quantitatively collected for a period of 5 and 3 days at the start and the end of the trial, respectively. A detailed description of the urine collection procedure is provided elsewhere (Pantophlet et al., 2016c). Urinary glucose and plasma glucose, triglycerides, and HDL cholesterol were measured on a Roche-Hitachi modular automatic analyzer (Roche Diagnostics, Basel, Switzerland) using enzymatic colorimetric assays. The within- and between-run coefficients of variation were ≤2% for all analyses. Insulin was measured using a bovine


103

6

ELISA kit (Mercodia, Uppsala, Sweden). The within- and between-run coefficients of variation were ≤5.6 and 8.2%, respectively.

Data were analyzed using SPSS (version 22; IBM, Armonk, NY). The effect of scFOS on whole-body insulin sensitivity, QUICKI, triglyceride:HDL cholesterol ratio, fasting plasma glucose and insulin, and urinary glucose excretion was tested by ANOVA using the GLM (Univariate) procedure. The overall effect of scFOS and the within-treatment effect (i.e., within the several diets) were tested. Supplementation with scFOS (i.e., scFOS vs. non-scFOS calves) was used as a factor, and calf was the experimental unit. The final values (i.e., measured at the end of the trial) were used as dependent variables, with their respective initial values as covariates. For the overall effects an scFOS × dietary treatment interaction was included. P-values < 0.05 were considered significant.

6.3 Results

The average daily BW gain, measured over the whole trial period, was on average 1.19 ± 0.02 kg/d and did not differ between dietary treatments or scFOS treatment. Feed refusal was negligible throughout the trial. Whole-body insulin sensitivity calculated using Min-Mod Millennium was low at the start of the trial (it did not differ between scFOS and non-scFOS calves) and remained low until the end [1.0 ± 0.1 and 0.8 ± 0.1 (mU/L)−1 · min−1 on average at the start and the end of the trial, respectively]. Supplementation of scFOS did not improve insulin sensitivity (P = 0.324; Table 6.1). During the trial, fasting plasma glucose levels decreased by 9%, whereas insulin levels increased by 25% (5.9 ± 0.1 mmol/L and 14.1 ± 1.3 mU/L on average, respectively, at the end of the trial). Short-chain fructo-oligosaccharides did not affect fasting plasma glucose or insulin levels (P = 0.100 and 0.204 for glucose and insulin, respectively; Table 6.1). The QUICKI, derived from the fasting plasma and insulin levels, did not improve by the supplementation of scFOS (P = 0.434; Table 6.1). Fasting plasma triglyceride decreased by 20%, whereas HDL cholesterol levels increased by 35% (0.109 ± 0.005 and 0.741 ± 0.054 mmol/L, respectively, at the end of the trial). The triglyceride:HDL cholesterol ratio (fasting indicator of insulin resistance) did not improve by the supplementation of scFOS (P = 0.191; Table 6.1). Urinary glucose excretion, which is a result of hyperglycemia (occurs when the threshold for glucose reabsorption is exceeded), was substantially higher (P < 0.01) in control and glucose calves compared with fructose and glycerol calves. Glucose excretion did not differ between scFOS and non-scFOS calves (P = 0.243; Table 6.1).

CHAPTER 6

104

6

Table 6.1 | Insulin sensitivity (ISminmod and QUICKI), triglyceride/HDL-cholesterol ratio, fasting glucose and insulin and urinary glucose excretion (means ± SEM; at end of the trial) in veal calves fed a control milk replacer (CON; n = 10) or a milk replacer in which one-third of the lactose was replaced by iso-energetic amounts of glucose (GLU; n = 9), fructose (FRU; n = 10), or glycerol (GLY; n = 10). Half of the calves within each treatment group (n = 5) received 12 mg/kg BW of short-chain fructo-oligosaccharides (scFOS) with the milk replacer for 6 weeks.

Item1Treatment

scFOS No-scFOS P-value P-value interaction

ISminmod x 10-4, ((mU/L)-1 x min-1)Overall 0.79±0.14 0.89±0.13 0.324 0.463 CON 0.74±0.14 1.01±0.34 0.291 GLU 0.71±0.27 0.83±0.10 0.852 FRU 1.36±0.39 1.00±0.27 0.505 GLY 0.44±0.10 0.62±0.10 0.130QUICKIOverall 0.324±0.004 0.329±0.005 0.434 0.734 CON 0.308±0.011 0.323±0.011 0.558 GLU 0.320±0.005 0.342±0.007 0.063 FRU 0.340±0.013 0.325±0.003 0.807 GLY 0.324±0.006 0.331±0.007 0.639Triglyceride/HDL-cholesterol ratioOverall 0.087±0.006 0.090±0.007 0.191 0.288 CON 0.089±0.012 0.106±0.012 0.090 GLU 0.095±0.009 0.103±0.026 0.160 FRU 0.067±0.016 0.079±0.003 0.482 GLY 0.091±0.013 0.077±0.009 0.430Fasting glucose, (mmol/L)Overall 5.1±0.1 5.0±0.1 0.100 0.069 CON 5.0±0.2 5.1±0.2 0.574 GLU 4.8±0.2 5.1±0.1 0.295 FRU 5.3±0.1 5.1±0.2 0.295 GLY 5.4±0.1 5.1±0.1 0.171Fasting insulin, (mU/L)Overall 15.5±2.2 12.7±1.3 0.204 0.397 CON 23.2±9.4 16.1±3.8 0.467 GLU 16.0±1.7 9.6±1.9 0.062 FRU 10.6±1.7 13.1±1.0 0.629 GLY 15.5±2.2 12.7±1.3 0.615Urinary glucose excretion, (g/d)Overall 48.5±14.2 31.8±8.6 0.243 0.435 CON 49.5±21.7 37.3±10.9 0.438 GLU 126.9±20.3 83.4±15.8 0.298 FRU 7.0±2.7 7.7±4.8 0.670 GLY 3.0±2.5 4.2±2.5 0.8771 ISminmod = insulin sensitivity derived from MinMod Millenium (MinMod Inc., Los Angeles, CA); QUICKI = quantitative insulin sensitivity check index; P-value interaction = P-value of scFOS x dietary treatment interaction.


105

6

6.4 Discussion

In contrast with observations in other species (Respondek et al., 2008, 2011), scFOS supplementation did not improve insulin sensitivity in veal calves. It cannot be excluded that the absence of an effect of scFOS is related to the administered dose, which in this study was based on meta-analysis of the dose-response effect of scFOS supplementation on feed conversion ratio in veal calves (Kaufhold et al., 2000; Tai et al., 2009; Grand et al., 2013), because of the lack of data on the dose–response effect of scFOS supplementation on insulin sensitivity. We assumed that the feed conversion ratio is related to glucose utilization and thus might be related to insulin sensitivity. Based on BW, the administered dose in this trial was approximately 3 to 4 times lower than in previous studies in horses (in which scFOS improved insulin sensitivity; Respondek et al., 2011) and veal calves (in which scFOS improved glucose homeostasis; Kaufhold et al., 2000). Future studies need to focus on the dose-response effect of scFOS supplementation on insulin sensitivity in veal calves, if any. In some non-ruminant species (e.g., dogs and horses), scFOS supplementation has been shown to improve insulin sensitivity. The absence of an effect of scFOS might therefore also be related to metabolic differences between ruminants and non-ruminants. In calves, insulin sensitivity rapidly decreases during the first months of life (Stanley et al., 2002) independently of the feeding strategy (Stanley et al., 2002; Pantophlet et al., 2016b), and although differences in glucose homeostasis are observed as a result of dietary interventions, generally little to no effect on insulin sensitivity is found (Bach et al., 2013; Pantophlet et al., 2016a,c). This might be related to the ontogenetic development of calves, as ontogenetic ruminants absorb little glucose from the intestinal tract and therefore may not have the genetic capacity to rapidly deal with large amounts of glucose. Consequently, absorbing large quantities of glucose from the MR disturbs glucose homeostasis in heavy veal calves. Mechanistic insight into species differences in the regulation of insulin sensitivity is required to obtain progress in this field. In conclusion, supplementation of scFOS to the MR did not improve insulin sensitivity in heavy veal calves.

6.5 Acknowledgements

The authors thank Gerlof Reckman and Jeltje Kloosterman for their laboratory assistance (University Medical Centre Groningen). This project was jointly financed by the European Union, European Regional Development Fund and the Ministry of Economic Af- fairs, Agriculture and Innovation, Peaks in the Delta, the Municipality of Groningen, the Provinces of Groningen, Fryslan, and Drenthe as well as the Dutch Carbohydrate Competence Center (CCC2 WP21). Financial support was also provided by Tereos Syral, VanDrie Group, and Wageningen University.

CHAPTER 6

106

6

6.6 References


Grand, E., F. Respondek, C. Martineau, J. Detilleux, and G. Bertrand. 2013. Effects of short-chain fructooligosaccharides on growth performance of preruminant veal calves. J. Dairy Sci. 96:1094-1101.





Pantophlet, A. J., W. J. J. Gerrits, R. J. Vonk, and J. J. G. C. van den Borne. 2016a. Substantial replacement of lactose with fat in a high-lactose milk replacer diet increases liver fat accumulation but does not affect insulin sensitivity in veal calves. J. Dairy Sci. 99:10022-10032.

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, M. G. Priebe, and R. J. Vonk. 2016b. Insulin sensitivity in calves decreases substantially during the first 3 months of life and is unaffected by weaning or fructo-oligosaccharide supplementation. J. Dairy Sci. 99:7602-7610.

Pantophlet, A. J., M. S. Gilbert, J. J. G. C. van den Borne, W. J. J. Gerrits, H. Roelofsen, M. G. Priebe, and R. J. Vonk. 2016c. Lactose in milk replacer can partly be replaced by glucose, fructose or glycerol without affecting insulin sensitivity in veal calves. J. Dairy Sci. 99:3072-3080.




Tai, X., X. Long, Z. Xiang, and F. Zuo. 2009. Effects of Fructooligosaccharides on Performance and Blood Biochemistry Index and Intestinal Mocosa Structure in Early-weaned Calves. Chinese J. Anim. Sci.:34-38.

CHAPTER 7

Discussion and Conclusions

“Basically, I’m not interested in doing research and I have never been. I’m interested in understanding, which is quite a different thing”

-David Blackwell-

CHAPTER 7

108

7

7.1 Introduction

Insulin resistance is not only a problem in humans, but also in various animal species. Veal calves, for example, often develop problems with glucose homeostasis and insulin sensitivity, which could develop into pre-diabetes. Avoiding these problems may stimulate efficient use of energy for growth processes and improve (metabolic) health. The objective of this thesis was to increase the understanding of the development of insulin resistance in veal calves, and to study the influence of several dietary modulations on the development of insulin resistance. In this last chapter, the different experimental chapters are discussed and general conclusions are provided.

7.2 Insulin resistance and age

In this project, over 250 measurements of insulin sensitivity were performed in three large-scale in vivo studies. Insulin sensitivity was assessed at different ages and when assigned to different diets. In the first study, the euglycemic-hyperinsulinemic clamp was used to measure insulin sensitivity in 27-week-old calves (n=14) on a high-lactose or high-fat milk replacer (MR) diet (Chapter 2). In the second study, the frequently sampled intravenous glucose tolerance test was used to measure and follow the development of insulin sensitivity in 40 calves at 15, 23 and 29 weeks of age, which were on a control MR diet, or MR diets in which one third of the lactose was replaced by glucose, fructose or glycerol, with and without the supplementation on fructo-oligosaccharides (Chapter 4 and 6). In the third study, an intravenous glucose tolerance test was used to measure and to follow the longitudinal development of whole-body insulin sensitivity in 30 calves at 3, 6, 9 and 12 week of age, which were on a commercial MR diet (with or without the supplementation of fructo-oligosaccharides) or calves which were progressively weaned (Chapter 5). All results obtained with the (frequently sampled) intravenous glucose tolerance tests (n ~240) during this project are given in Figure 7.1. In order to compare the results (in approximation), a scaling factor was used, which is described in Tura et al., (2010). Results obtained with the euglycemic-hyperinsulinemic clamp were not included because this test generates a different index compared to the intravenous glucose tolerance tests, and thus cannot be used to highlight age-related effects. Results from the intravenous glucose tolerance tests show that insulin sensitivity substantially decreases with age (~93%) in veal calves during the first months of life (i.e., from 3 to 29 weeks of age).

A decrease in whole-body insulin sensitivity was also evident from fasting plasma glucose and insulin data (when comparing results from 3 vs. 29 weeks of age). Fasting plasma glucose concentrations increased by ~46% (from 3.6±0.1 to 5.1±0.1 mmol/L), whereas fasting plasma insulin concentrations increased by ~1700% (from 0.8±0.2 to 14.1±1.3 mU/L). The strong increase in fasting plasma insulin concentrations shows that, as calves get older, more insulin is needed to maintain normal glucose homeostasis, indicating decreased insulin sensitivity.


109

7

An age-related decrease in insulin sensitivity has also been found in another calf study (Stanley et al., 2002) and studies in neo-natal lambs (Gelardi et al., 1999). The strong age-related decrease in insulin sensitivity (and apparent absence of nutritional effects) suggests that the decrease is probably related to the ontogenetic development of ruminants. In nature, young calves between 4 and 6 months of age are grazing and plant fragments are fermented in the rumen along with the production of short-chain fatty acids as a major energy source (and not glucose like most non-ruminants). Therefore, during the transition of calves from pre-ruminants to ruminants they may become insulin resistant, as they may not be equipped anymore with the genetic capacity to keep dealing with high amounts of glucose/lactose (from MR) quickly and efficiently. This conclusion is also supported by other calf studies (Hugi et al., 1997; Stanley, 2005).

Future research should focus on putative physiological factors involved in the development of insulin resistance and their relationship with age (transition) in calves. These include the number (and affinity) of insulin receptors, the number (and affinity) of glucose type-4 transporter and its translocation efficiency, and post-receptor insulin signaling pathways.

Figure 7.1 | The decrease in insulin sensitivity in calves with age (~240 measurements). Insulin sensitivity was assessed by either the frequently sampled intravenous glucose tolerance test (week 3-12; n = 30 calves) or an intravenous glucose tolerance test (week 15-29; n = 40 calves). A scaling factor, described by Tura et al., (2010) was used to compare the results of the different tests. Calves were fed several dietary treatments, which are described in Chapter 4, 5 and 6. Insulin sensitivity was not affected by dietary treatment.

7.3 Insulin sensitivty, glucose homeostasis, glucosuria and nutritional modulations

One of the objectives in this project was to evaluate the effect of several nutritional strategies on insulin sensitivity in veal calves. Veal calves (> 4 months of age) often develop problems with glucose homeostasis (Hostettler-Allen et al., 1994; Hugi et al., 1997), which is characterized by a high incidence of hyperglycemia, hyperinsulinemia and glucosuria. Aberrant glucose homeostasis could contribute to the development of

CHAPTER 7

110

7

insulin resistance (Tomás et al., 2002; Norris and Rich, 2012). Reducing postprandial glucose and insulin concentrations may reduce the risk of developing insulin resistance. The high incidence of hyperglycemia and hyperinsulinemia in veal calves is probably the result of high amounts of dietary lactose in the MR. In this project, dietary lactose was partially replaced by other energy sources, namely fructose, glucose or glycerol (Chapter 4). The effects of these substrates on glucose homeostasis, glucosuria and insulin sensitivity are summarized in Table 7.1.

FructoseFructose has a lower glycemic response than lactose, therefore partial replacement of lactose with fructose was expected to positively impact postprandial glucose homeostasis (i.e. lower glucose and insulin responses), and reduce (or even prevent) the decrease in insulin sensitivity in veal calves. Partial replacement with lactose did improve postprandial glucose homeostasis, but this did not result in a positive effect on insulin sensitivity. The improved postprandial glucose homeostasis resulted in negligible loss of glucose via urine (Figure 7.2).

GlucoseObviously, with the highest glycemic response, partial replacement with glucose was expected to negatively impact postprandial glucose homeostasis (i.e., higher glucose and insulin responses), and deteriorate insulin sensitivity in veal calves. Partial replacement with glucose indeed deteriorated postprandial glucose homeostasis and increased urinary glucose loss, but this did not result in reduced insulin sensitivity.

GlycerolSimilar to fructose, glycerol has a lower glycemic response than lactose, and therefore partial replacement of lactose with glycerol was expected to positively impact postprandial glucose homeostasis and, consequently, reduce (or even prevent) the decrease in insulin sensitivity. Partial replacement with glycerol improved postprandial glucose homeostasis, but this did not result in a positive effect on insulin sensitivity. The improved postprandial glucose homeostasis resulted in negligible loss of glucose via urine.

Overall, these results show that, in veal calves, an improved postprandial glucose homeostasis (i.e., reduced glycaemia and insulinemia) does not lead to improved insulin sensitivity on the long run. This is in contrast to what is generally observed in humans; where reduced postprandial glycaemia and insulinemia are generally associated with improved insulin sensitivity (Blaak et al., 2012; Dunstan et al., 2012; Visuthranukul et al., 2015). Also, these results show that postprandial glucose homeostasis deteriorates with age in veal calves fed a lactose-MR diet. This is evident from the age-related increase in glucosuria seen in the lactose-MR fed calves (Figure 7.2). Replacing part of the lactose with fructose or glycerol can reduce (or prevent) this deterioration.


111

7

Table 7.1 | The effect of partial replacement of lactose in the milk replacer by glucose, fructose or glycerol on postprandial glucose homeostasis, glucosuria and insulin sensitivity in veal calves.

Item

Glucose homeostasis Glucosuria Insulin sensitivity3

Treatment1

Lactose - - -

Glucose Deterioration Increase No Difference

Fructose Improvement Decrease2 No Difference

Glycerol Improvement Decrease2 No Difference1One third of the lactose was replaced by iso-energetic amounts of glucose, fructose or glycerol.2Negligible amounts of glucose in urine.3Measured after ~8 weeks of supplementation.

Figure 7.2 | Glucosuria in calves fed either a lactose milk replacer diet or diets in which one third of the lactose was replaced by glucose, fructose or glycerol. At 15 weeks of age all calves were on the lactose milk replacer diet. The black bars represent calves which were fed a lactose milk replacer diet. The grey, white and hatched bars represent calves in which part of the lactose was replaced by glucose, fructose and glycerol respectively. Glucosuria increased in the lactose and glucose-fed calves, but was negligible in the fructose and glycerol-fed calves.

Another nutritional factor studied in this project was the effect of dietary fat on insulin sensitivity in veal calves (Chapter 2). Calf MR does not only contain high amounts of dietary lactose but also high amounts of dietary fat. In non-ruminants, high dietary fat intake has consistently been associated with the development of insulin resistance (Storlien et al., 1996; Frayn, 2003; Müller and Kersten, 2003). Results from the current study showed that insulin sensitivity is not differentially affected by fat or lactose, which suggests that either none or both substrates may be responsible for the decrease in insulin sensitivity. Insulin sensitivity was, however, only directly measured once, at end of the study (i.e., using a direct measurement of insulin sensitivity). But the quantitative insulin-sensitivity check index, a surrogate index of insulin sensitivity

CHAPTER 7

112

7

derived from fasting plasma glucose and insulin concentrations, was calculated at multiple timepoints throughout the study. It did not change with age and did also not differ between fat- and lactose-fed calves. This suggests that both substrates do not affect insulin sensitivity in veal calves (older than 3 months of age).

In addition to dietary lactose and fat, the effect of progressive weaning on the development on insulin sensitivity in calves was studied in this project (Chapter 5). This feeding strategy more closely mimics what happens in nature in calves, and is in strong contrast to the feeding of veal calves, which are commonly maintained on diets containing large amounts of MR. The effects of progressive weaning vs. MR feeding on the development on insulin sensitivity was studied in calves during the first three months of life. Regardless of the feeding strategy, insulin sensitivity substantially decreased during this period, which suggests that the decrease is primarily related to the ontogenetic development of calves and that the MR itself does not negatively impact insulin sensitivity. The most pronounced decrease occurred in calves between weeks 3 to 6 weeks of age. The extent of the decrease was similar to another calf study (Stanley et al., 2002). During this period, however, the calves in our study were not yet completely weaned and the differences in feed intake were relatively small (i.e., both groups were on significant amounts of MR). Therefore, a possible impact of the MR on insulin sensitivity cannot be fully excluded.

In this project, also the effect short-chain fructooligosaccharides (scFOS) supplementation on insulin sensitivity was studied, as a means of improving or preventing the decrease insulin sensitivity found in calves (Chapter 5 and 6). Studies in dogs and horses have shown that supplementation of scFOS greatly improves insulin sensitivity (Respondek et al., 2008; Respondek et al., 2011). In the current project, the effect of scFOS supplementation was studied in both young (3 weeks of age; relative insulin sensitive) calves and older (23 weeks of age; relative insulin resistant) calves. Both studies showed that scFOS supplementation, at the given dosage, does not improve or prevent the decrease in insulin sensitivity found in veal calves. The absence of an effect of scFOS might be related to metabolic differences between ruminants and non-ruminants or to the given dosage. The dosage chosen in this project was based on the feed conversion ratio (= feed intake dived by weight gain within a particular time period) in veal calves fed MR plus scFOS, but was approximately 3 to 4 times lower than studies in non-ruminants (Respondek et al., 2008; Respondek et al., 2011). In the future, a dose-response study should be conducted to assess if scFOS supplementation in higher doses does affect/improve insulin sensitivity in veal calves or not. However, given the overall absence of nutritional effects on insulin sensitivity seen in this project, any effect of scFOS seems unlikely.

The results of this project show that, overall, nutritional modulations affect glucose homeostasis and glucosuria, but not insulin sensitivity in veal calves. The absence of nutritional effects is further evidence that the strong decrease in insulin sensitivity is predominantly related to the ontogenetic development of calves.


113

7

7.4 Metabolic alteriations related to insulin resistance

To study the development of insulin resistance in veal calves, several advanced metabolic profiling techniques were used in this project. These techniques may allow identification of metabolic pathway alterations related to insulin resistance and the discovery of potential biomarkers. In this project, 2 high-performance liquid chromatography-mass spectrometry (LC-MS) platforms (i.e., C18 LC-MS and HILIC LC-MS) were used. Metabolic profiling was performed on fasting plasma samples of moderately insulin sensitive vs. insulin resistant veal calves, which were either on a high-lactose or high-fat MR diet (Chapter 3). Satisfactory multivariate models were built that allowed discrimination between insulin sensitive vs. insulin resistant veal calves. Several metabolic alterations were found that were related to the glycerophospholipid metabolism, sphingolipid metabolism, glycine, serine and threonine metabolism, and primary bile acid biosynthesis. Multiple alterations were found in the glycerophospholipid and the sphingolipid metabolism. These pathways have also been associated with insulin resistance in dairy cows, humans and rodents. The mechanisms behind the associations are not clear, but disturbances in membrane glycerophospholipid metabolism (major components of cell membranes) for example could influence insulin secretion via alteration of the physico-chemical properties of the membrane.

The samples were also analyzed on 1H nuclear magnetic resonance spectroscopy (NMR) platform. However, with NMR it was not possible to build a discriminative multivariate model, probably because the number of metabolites detected was limited (n= 23).

In addition to multivariate analysis, univariate correlation analysis was performed to assess whether there are metabolites that correlate with individual values of insulin sensitivity. For NMR, correlation analysis was performed on all detected metabolites, whereas for LC-MS, correlation analysis was only performed on the metabolic features that most significantly contributed to the group discrimination (variable importance in the projection > 2.0) in multivariate models. The Benjamini-Hochberg correction was employed to adjust for multiple testing (Benjamini and Hochberg, 1995). For NMR, succinate positively correlated with insulin sensitivity (rpearson = 0.805; p = 0.002), and for LC-MS, taurochenodeoxycholic acid (rpearson = 0.728; p = 0.003 with C18 LC-MS and rpearson = 0.714; p = 0.004 with HILIC LC-MS) and acetylcarnitine (rpearson = 0.676; p = 0.008) positively correlated with insulin sensitivity. For LC-MS one metabolic feature with an m/z value of 204.123 negatively correlated with insulin sensitivity (rpearson = -0.774; p = 0.001), however, it could not be identified because none of the metabolomic database searches yielded any hits and de-novo metabolite identification using MS and MS/MS data did not generate any candidate compounds. The identified metabolites may be possible biomarkers of insulin resistance in calves and should be evaluated in subsequent studies. Overall, the results show that the use of multiple metabolomic

CHAPTER 7

114

7

platforms can increase the metabolome coverage and may lead to the discovery of additional metabolites and pathways related to insulin resistance. In future studies, the use of additional metabolic platforms (e.g., gas chromatography–mass spectrometry and capillary electrophoresis–mass spectrometry) should also be considered.

7.5 Overall conslusion

Whole-body insulin sensitivity in veal calves decreases substantially with age. Nutritional modulation, such as partial replacement of the lactose in the milk replacer by substrates with lower glycemic responses can improve postprandial glucose homeostasis (and reduce glucosuria), but does not affect the development of insulin sensitivity. Also, the supplementation of fructo-oligosaccharides, which improved insulin sensitivity in other species, did not improve or prevent the decrease in insulin sensitivity in calves. The absence of nutritional effects, in combination with the substantial age-related decrease in insulin sensitivity found in calves suggests that the development of insulin resistance is primarily related to the ontogenetic development of calves (as calves typically progress from pre-ruminants to ruminants). Potential ontogenetic (age-related) effects on physiological factors influencing insulin sensitivity (e.g., number or affinity of insulin receptor and glucose transporter type 4 or downstream signaling) remain unclear. Advanced metabolic profiling techniques revealed various pathways and metabolites related to insulin resistance in calves. Multiple alterations were found in the glycerophospholipid and the sphingolipid metabolism. These pathways have also been associated with insulin resistance in other species. The mechanisms behind the associations are not clear but could for example be related to alteration of the physico–chemical properties of the cell membrane, which could influence insulin secretion. The pathways and metabolites should be further exploited in future studies to better understand the metabolic changes associated with the age–related development of insulin resistance in calves.


115

7

7.6 References

Benjamini, Y. and Y. Hochberg. 1995. Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society 57:289–300.

Blaak, E. E., J. M. Antoine, D. Benton, I. Björck, L. Bozzetto, F. Brouns, M. Diamant, L. Dye, T. Hulshof, J. J. Holst, D. J. Lamport, M. Laville, C. L. Lawton, M. A., A. Nilson, S. Normand, A. A. Rivellese, S. Theis, S. S. Torekov, and S. Vinoy. 2012. Impact of postprandial glycaemia on health and prevention of diseas. Obes Rev 13:923–984.

Dunstan, D. W., B. A. Kingwell, R. Larsen, G. N. Healy, E. Cerin, M. T. Hamilton, J. E. Shaw, D. A. Bertovic, P. Z. Zimmet, J. Salmon, and N. Owen. 2012. Breaking up prolonged sitting reduces postprandial glucose and insulin responses. Diabetes Care. 35:976-983.

Frayn, K. N. 2003. The glucose-fatty acid cycle: a physiological perspective. Biochem. Soc. Trans. 31:1115-1119.Gelardi, N. L., R. E. Rapoza, J. F. Renzulli, and R. M. Cowett. 1999. Insulin resistance and glucose transporter

expression during the euglycemic hyperinsulinemic clamp in the lamb. Am. J. Physiol. 277:E1142-E1149.Hostettler-Allen, R. L., L. Tappy, and J. W. Blum. 1994. Insulin resistance, hyperglycemia, and glucosuria in

intensively milk-fed calves. J. Anim. Sci. 72:160-173.Hugi, D., R. M. Bruckmaier, and J. W. Blum. 1997. Insulin resistance, hyperglycemia, glucosuria, and


Müller, M. and S. Kersten. 2003. Nutrigenomics: goals and strategies. Nat. Rev. Genet. 4:315-322.Norris, J. M. and S. S. Rich. 2012. Genetics of glucose homeostasis implications for insulin resistance and

metabolic syndrome. Arterioscler. Thromb. Vasc. Biol. 32:2091-2096.Respondek , F., K. Myers, T. L. Smith, A. Wagner, and R. J. Geor. 2011. Dietary supplementation with short-

chain fructo-oligosaccharides improves insulin sensitivity in obese horses. J. Anim. Sci. 89:77-83.Respondek, F., K. S. Swanson, K. R. Belsito, B. M. Vester, A. Wagner, L. Istasse, and M. Diez. 2008. Short-

chain fructooligosaccharides influence insulin sensitivity and gene expression of fat tissue in obese dogs. J. Nutr. 138:1712-1718.

Stanley, C. C. 2005. Regulation of glucose metabolism in dairy cattle. in Dairy Science. Vol. PhD. Louisiana State University, US.


Storlien, L. H., L. A. Baur, A. D. Kriketos, D. A. Pan, G. J. Cooney, A. B. Jenkins, G. D. Calvert, and L. V. Campbell. 1996. Dietary fats and insulin action. Diabetologia. 39:621-631.

Tomás, E., Y. S. Lin, Z. Dagher, A. Saha, Z. Luo, Y. Ido, and N. B. Ruderman. 2002. Hyperglycemia and insulin resistance: possible mechanisms. Ann. N. Y. Acad. Sci. 967:43-51.

Tura, A., S. Sbrignadello, E. Succurro, L. Groop, G. Sesti, and G. Pacini. 2010. An empirical index of insulin sensitivity from short IVGTT: validation against the minimal model and glucose clamp indices in patients with different clinical characteristics. Diabetologia. 53:144-152.

Visuthranukul, C., P. Sirimongkol, A. Prachansuwan, C. Pruksananonda, and S. Chomtho. 2015. Low-glycemic index diet may improve insulin sensitivity in obese children. Pediatr. Res. 78:567-573.

Summary

117

Summary

Insulin resistance, an important risk factor for the development of type 2 diabetes, is a problem that has not only been identified in humans worldwide, but also in animals. Veal calves often develop problems with glucose homeostasis and insulin resistance, which are typically characterized by a high incidence of hyperglycemia and hyperinsulinemia. Avoiding these problems may stimulate efficient use of energy for health and growth. Studies in calves may also increase our understanding of the mechanisms behind the development of insulin resistance. The objective of this thesis was to study the effects of age and dietary factors (modulations) on the development of insulin resistance in veal calves. Metabolic profiling was performed to find markers of insulin resistance and the associated metabolic networks.

Veal calves are fed a milk replacer (MR) that contains high amounts of lactose and fat. Both lactose and fat have been associated with insulin resistance in other species. In Chapter 2, the contribution of dietary lactose and fat on the development of insulin resistance in heavy veal calves were studied. Sixteen male Holstein-Friesian calves (99 days of age) were assigned to either a high-lactose or a high-fat MR diet (n=8 per treatment) for 13 weeks (lactose and fat exchanged iso-energetically). Postprandial plasma glucose was higher in high-lactose fed calves compared to high-fat fed calves, whereas postprandial plasma insulin only tended to be higher in high-lactose fed calves. The high plasma glucose led to increased urinary glucose excretion in high-lactose fed calves. Fasting plasma glucose and insulin did not differ between the dietary treatment groups. Insulin sensitivity and insulin secretion were not differentially affected by dietary treatment, which suggests that either none or both dietary energy sources equally affect insulin sensitivity. Results from chapter 2 showed that heavy veal calves are relativity insensitive to insulin and 50% of the calves could be considered insulin resistant.

In Chapter 3, liquid chromatography–mass spectrometry metabolic profiling techniques were used to study the patho-physiological mechanisms underlying insulin resistance in calves and to discover potential biomarkers for early diagnosis. The calves (described in chapter 2) were classified as either insulin resistant or moderately insulin sensitive, based on their insulin sensitivity values. Metabolic profiling was performed in the fasted state. Satisfactory models were build which allowed discrimination between insulin resistant and moderately insulin sensitive calves. The metabolic alterations associated to insulin resistance were related to the glycerophospholipid metabolism, sphingolipid metabolism, glycine, serine and threonine metabolism, and primary bile acid biosynthesis. Multiple alterations were found in the glycerophospholipid and the sphingolipid metabolism. These pathways have also been associated with insulin resistance in dairy cows, human and rodents. The mechanisms behind the associations are not clear, but disturbances in membrane glycerophospholipid metabolism (major components of cell membranes) for example could influence insulin secretion via

Summary

118

alteration of the physico-chemical properties of the membrane. In future veal calf studies, these pathways and associated biomarkers, and their role in the development of insulin resistance, should be studied in early life (from birth on).

Prolonged intake of high amounts of lactose has been associated with impaired glucose homeostasis, which could ultimately cause insulin resistance. Partial replacement of lactose by other dietary substrates with lower glycemic (and insulinemic) responses may reduce the risk of developing disturbances in glucose homeostasis and insulin sensitivity. In Chapter 4, the effect of partial replacement of dietary lactose by fructose, glycerol or glucose on glucose homeostasis and insulin sensitivity was studied. Forty male Holstein-Friesian calves (97 days of age) were assigned to either a lactose control MR or a MR in which one third of the lactose was replaced by isoenergetic amounts of fructose, glycerol or glucose (n=10 per treatment) for 8 weeks. Postprandial plasma glucose was higher in glucose-fed calves compared to the other calves and lower in fructose-fed calves compared to the control calves. Postprandial plasma insulin was lower in the fructose and glycerol-fed calves compared to the other calves. Postprandial plasma glucose response exceeded the renal threshold for glucose in glucose-fed calves and control calves, resulting in urinary glucose excretion. Conversion of fructose and glycerol to glucose was confirmed by an increase in 13C enrichment of plasma glucose after feeding [U-13C] fructose and [U-13C] glycerol, respectively. Fasting plasma glucose and insulin in glucose, fructose and glycerol-fed calves did not differ from control calves. Measurement of stress markers showed that in calves insulin sensitivity is not influenced by stress. Despite improved postprandial glucose homeostasis in fructose-fed calves and glycerol-fed calves, insulin sensitivity was not differently affected by the dietary treatments, was already low at start of the study (at ~14 weeks of age) and remained low.

Results from chapter 4 showed that young calves are already relatively insensitive to insulin, compared to healthy non-ruminants. It is not clear to what extent diet-induced effects or the ontogenetic development of calves contribute to the development of insulin resistance. In Chapter 5, the effects of MR feeding, weaning, and MR feeding plus short-chain fructo-oligosaccharides (scFOS) supplementation on the development of glucose homeostasis and insulin sensitivity were studied in calves during the first 3 months of life (i.e. age-related and diet-induced effects). Thirty male Holstein-Friesian calves (18 days of age) were assigned to either a lactose control MR, a lactose MR with the addition of scFOS or progressively weaned to solid feed (n=10 per treatment; period of 10 weeks). Short-chain fructo-oligosaccharides were included because studies in various animal species showed that scFOS supplementation can improve insulin sensitivity. Postprandial plasma glucose did not differ with age for the MR and MR+scFOS calves, but postprandial plasma insulin substantially increased with age. Supplementation of scFOS did not improve postprandial glucose homeostasis. Weaned calves suffered from pronounced hyperglycemia and hyperinsulinemia when fed MR again after weaning. Insulin sensitivity decreased by 75% during the first 3 months of life, and weaning and

Summary

119

scFOS supplementation did not affect the age-related decrease in insulin sensitivity. This suggests that the decrease in insulin sensitivity is primarily related to ontogenetic development of calves (and not on nutritional effects).

In chapter 5 the effect of scFOS supplementation on insulin sensitivity was studied in young relatively insulin sensitive calves. In Chapter 6, the effect of scFOS was studied in older relatively insulin resistant calves. At the end of the study described in chapter 4 scFOS was introduced to the diets of half of the calves (162 days of age) in each treatment group, for 6 weeks. Supplementation of scFOS did not improve insulin resistance in any of the treatment groups. These results are in strong contrast to results obtained in non-ruminants, and might be related to the dosage or metabolic differences between ruminants and non-ruminants.

The studies reported in this thesis demonstrate that insulin sensitivity in calves decreases rapidly with age (from birth on) and remains low thereafter. Nutritional modulations can improve postprandial glucose homeostasis (and reduce urinary glucose loss), but will not prevent loss of insulin sensitivity. The rapid age-related decrease in insulin sensitivity, and apparent absence of effects of nutrition suggests that the development of insulin resistance in calves is primary related to the ontogenetic development of calves (as calf progress from pre-ruminants to ruminants). Possible ontogenetic effects on physiological factors influecing insulin sensitivity (e.g. number or affinity of insulin receptor and glucose transporter type 4) remain unclear. Using advanced metabolic profiling several potential biomarkers and metabolic pathways were identified that could be associated with the development of insulin resistance in calves, and also in other species. Further application of these techniques may increase our understanding of the mechanisms behind the development of this metabolic disorder.

Samenvatting

121

Samenvatting

Insuline resistentie, een risicofactor voor het ontwikkelen van type 2 diabetes, is niet alleen een probleem bij mensen, maar komt ook frequent voor bij dieren. Vleeskalveren ontwikkelen vaak problemen met glucose homeostase en insuline resistentie, welke gepaard gaan met een hoge incidentie van hyperglycemie en hyperinsulinemie. Het voorkomen van insuline resistentie bij vleeskalveren zou de efficiënte benutting van nutriënten (voor groei en gezondheid) kunnen verbeteren. Studies in kalveren zou ook onze algemene kennis betreffende de mechanismen achter de ontwikkeling van insuline resistentie kunnen vergroten. Het doel van deze thesis was om de effecten van leeftijd en diverse voedingsstrategieën op de ontwikkeling van insuline resistentie in kalveren te onderzoeken. Metabolomics technieken werden gebruikt om markers van insuline resistentie te vinden en geassocieerde metabole routes te identificeren.

Vleeskalveren krijgen normaal melkvervangers (MR) als voeding, die een grote hoeveelheid lactose en vet bevat. Zowel lactose en vet worden geassocieerd met insuline resistentie in andere diersoorten. In Hoofdstuk 2 werd de bijdrage van lactose en vet aan de ontwikkeling van insuline resistentie in zware vleeskalveren onderzocht. Zestien mannelijke Holstein-Friesian kalveren (99 dagen oud) kregen of een hoog-lactose of hoog-vet MR dieet (n=8 per dieet) voor 13 weken (lactose werd iso-energetisch uitgewisseld met vet). Postprandiale plasma glucose response was hoger in kalveren die een hoog-lactose dieet kregen, vergeleken met de kalveren die een hoog-vet dieet kregen. De postprandiale insuline respons vertoonde een trend iets hoger te zijn in de kalveren met hoog-lactose dieet. Voor kalveren die een hoog-lactose dieet kregen leidde de hoge postprandiale plasma glucose niveaus tot een verhoogde excretie van glucose via de urine. Er waren geen verschillen in basale plasma glucose en insuline niveaus tussen de groepen. Insuline gevoeligheid en insuline excretie werden niet verschillend beïnvloed door de beide diëten. Dit suggereert dat of geen van beide energiebronnen (dus lactose en vet) insuline gevoeligheid beïnvloeden in kalveren, of dat beide energiebronnen insuline gevoeligheid in dezelfde mate beïnvloeden. Uit de resultaten van hoofdstuk 2 bleek dat zware vleeskalveren een lage insuline gevoeligheid hebben en dat 50% zelfs als insuline resistent gecategoriseerd zou kunnen worden.

In Hoofdstuk 3 werden verschillende metabolomics technieken gebruikt om de patho-fysiologische mechanismen achter de ontwikkeling van insuline resistentie bij kalveren te onderzoeken en om vroege biomarkers voor insuline resistentie te vinden. Deze technieken maken gebruik van vloeistofchromatografie en massaspectrometrie. De kalveren beschreven in hoofdstuk 2 werden geclassificeerd als insuline resistent of matig insuline gevoelig. De classificering werd gedaan op basis van de gemeten waarde voor insuline gevoeligheid. Metabolomics werd toegepast in nuchtere toestand (dus niet postprandiaal). Robuste modellen werden gebouwd die onderscheid konden maken tussen insuline resistent en matig insuline gevoelige kalveren. De gevonden metabole verschillen (geassocieerd met insuline resistentie) zaten in de volgende

Samenvatting

122

metabole routes: 1) de metabolisme van glycerofosfolipiden, 2) de metabolisme van sfingolipiden, 3) de metabolisme van glycine, serine en threonine en 4) de synthese van primaire galzouten. In de metabole routes van glycerofosfolipiden en sfingolipiden werden meerdere verschillen gevonden. Deze metabole routes worden in mensen, melkkoeien en knaagdieren ook geassocieerd met insuline resistentie. De mechanismen hierachter achter zijn nog niet volledig duidelijk. Veranderingen in metabolisme van glycerofosfolipiden in het membraan, bijvoorbeeld, zouden de fysisch-chemische eigenschappen van het membraan kunnen beïnvloeden en daardoor invloed kunnen hebben op insuline secretie. In vervolgstudies moeten de gevonden metabole routes (en biomarkers) en hun rol in de ontwikkeling van insuline resistentie verder onderzocht worden in jonge (pasgeboren) kalveren.

Langdurige consumptie van grote hoeveelheden lactose word geassocieerd met verslechterde glucose homeostase, wat uiteindelijk kan resulteren in insuline resistentie. Het deels vervangen van de lactose door substraten met een lagere glycemische respons zou het risico tot verslechterde glucose homeostase en het ontwikkelen van insuline resistentie kunnen verminderen. In Hoofdstuk 4 werden de effecten van het (deels) vervangen van lactose door fructose, glycerol en glucose op glucose homeostase en insuline gevoeligheid onderzocht. Veertig mannelijke Holstein-Friesian kalveren (97 dagen oud) kregen of controle MR of een MR waar een derde van de lactose vervangen werd door iso-energetische hoeveelheden van fructose, glycerol of glucose (n= 10 per dieet), voor 8 weken. Postprandiale plasma glucose was hoger in de kalveren die glucose gekregen hebben, vergeleken met de andere kalveren. Postprandiale plasma glucose was hoger in de kalveren die fructose gekregen hebben, vergeleken de controle kalveren. Postprandiale plasma insuline was lager in de kalveren die fructose of glycerol gekregen hebben, vergeleken met de andere kalveren. Voor controle kalveren en de kalveren die glucose gekregen hebben werd de nierdrempel voor glucose overschreden doordat de postprandiale plasma glucose niveaus te hoog waren. Dit leidde tot excretie van glucose via de urine. Kalveren kunnen fructose en glycerol (deels) omzetten naar glucose. Dit werd bevestigd door de gevonden toename in de 13C verrijking van plasma glucose na het geven van [U-13C] fructose en [U-13C] glycerol, respectievelijk. De basale (nuchtere) plasma glucose en insuline niveaus in kalveren die glucose, fructose of glycerol kregen hebben verschilde niet van de controle kalveren. Stress markers werden gemeten en uit de resultaten bleek dat de mate van stress in kalveren geen invloed heeft op insuline gevoeligheid. Ondanks de verbeterde postprandiale glucose homeostase in kalveren die fructose en glycerol gekregen hebben werden er geen verschillen gevonden in insuline gevoeligheid tussen de verschillende groepen. Insuline gevoeligheid was al laag aan het begin van de studie (kalveren ~ 14 weken oud) en bleef laag.

Uit de resultaten van hoofdstuk 4 bleek dat jonge kalveren al relatief ongevoelig zijn voor insuline, vergeleken met gezonde mensen. Het is niet duidelijk in hoeverre voedingseffecten of de ontogenetische ontwikkeling van kalveren bijdragen aan de ontwikkeling van insuline resistentie. In Hoofdstuk 5 werden de effecten van het voeren

Samenvatting

123

van MR, het spenen en het voeren van MR plus korte-keten fructo-oligosacchariden (scFOS) op de ontwikkeling van glucose homeostase en insuline resistentie onderzocht gedurende de eerste 3 maanden na geboorte. Dertig mannelijke Holstein-Friesian kalveren (18 dagen oud) werden of geleidelijk gespeend (naar alleen vast voer), of kregen een controle lactose MR of een controle lactose MR plus scFOS (voor een periode van 10 weken; n=10 per dieet). Korte-keten fructo-oligosacchariden werden onderzocht omdat studies in diverse diersoorten lieten zien dat scFOS suppletie insuline gevoeligheid kan verhogen. Voor de kalveren die MR en MR plus scFOS gekregen hebben veranderden postprandiale plasma glucose spiegels niet met de leeftijd. De postprandiale plasma insuline respons, daarentegen, nam sterk toe met leeftijd. De suppletie van scFOS leidde niet tot een verbetering van de postprandiale glucose homeostase. Gespeende kalveren hadden last van extreme hyperglycemie en hyperinsulinemie wanneer ze na het spenen weer MR gevoerd kregen (tijdens een melk tolerantie test). Insuline gevoeligheid nam gedurende de eerste 3 maanden met 75% af. Spenen of de suppletie van scFOS had geen effect of de sterke afname in insuline gevoeligheid. Dit suggereert dat de ontwikkeling van insuline resistentie in kalveren hoofdzakelijk gerelateerd is aan de ontogenetische ontwikkeling van kalveren (en niet aan voedingseffecten).

In hoofdstuk 5 werd het effect van scFOS suppletie op insuline gevoeligheid in jonge relatief insuline gevoelige kalveren (< 3 maanden oud) onderzocht. In Hoofdstuk 6, werd het effect van scFOS op oudere, relatief insuline resistente kaveren onderzocht. Aan het einde van het experiment beschreven in hoofdstuk 4 werd scFOS geïntroduceerd in de diëten van de helft van de kalveren (162 dagen oud) in elke groep (voor een periode van 6 weken). De suppletie van scFOS leidde in geen enkele groep tot een verbeterde insuline gevoeligheid. Dit resultaat staat in sterk contrast met studies in niet-herkauwers, en kan gerelateerd zijn aan de gegeven dosering of aan metabole verschillen tussen herkauwers en niet- herkauwers.

De studies in deze thesis geven aan dat insuline gevoeligheid in kalveren sterk daalt vanaf de geboorte en daarna laag blijft. Voedingsaanpassingen kunnen leiden tot verbeterde postprandiale glucose homeostase en verminderde glucoseverlies via de urine, maar lijken de ontwikkeling van insuline resistentie niet te voorkomen. De leeftijdsgerelateerde afname van insuline gevoeligheid in combinatie met de afwezigheid van voedingseffecten suggereert dat de ontwikkeling van insuline resistentie hoofdzakelijk gerelateerd is aan de ontogenetische ontwikkeling van kalveren (de ontwikkeling tot herkauwers). Mogelijke ontogenetische effecten op fysiologische factoren gerelateerd aan insuline gevoeligheid (b.v. het aantal of de affiniteit van de insuline receptor en de glucose transporter type 4) zijn nog onduidelijk. Met behulp van geavanceerde metabolomics technieken werden verschillende biomarkers en metabole routes ontdekt die mogelijk gerelateerd zijn aan insuline resistentie in kalveren (en mogelijk ook in andere species ). Het verder toepassen van metabolomics technieken zou de kennis van de mogelijke mechanismen achter de ontwikkeling van insuline resistentie kunnen vergroten.

List of peer reviewed scientific publications

125

List of peer reviewed scientific publications

Pantophlet, A.J., M.S. Gilbert, J.J.G.C. van den Borne, W.J.J. Gerrits, H. Roelofsen, M.G. Priebe, and R.J. Vonk, 2016. Lactose in milk replacer can partly be replaced by glucose, fructose or glycerol without affecting insulin sensitivity in veal calves. The Journal of Dairy Science 99:3072-3080.

Pantophlet, A.J., M.S. Gilbert, J.J.G.C. van den Borne, W.J.J. Gerrits, H. Roelofsen, M.G. Priebe, and R.J. Vonk, 2016. Insulin sensitivity in calves decreases substantially during the first 3 months of life and is unaffected by weaning or fructo-oligosaccharide supplementation. The Journal of Dairy Science 99:7602-7610.

Pantophlet, A.J., W.J.J. Gerrits, R.J. Vonk, and J.J.G.C. van den Borne, 2016. Substantial replacement of the lactose in a high lactose milk replacer diet by fat increases liver fat accumulation, but does not affect insulin sensitivity in veal calves. The Journal of Dairy Science 99:10022-10032.

Pantophlet, A.J*., S.Wopereis*, C.Eelderink*, R.J. Vonk, J.H.M. Dijk-Stroeve, S.Bijlsma, L. van Stee, I. Bobeldijk, and M.G. Priebe, 2017. Metabolic profiling reveals differences in plasma concentrations of arabinose and xylose after consumption of fiber-Rich pasta and wheat bread with differential rates of systemic appearance of exogenous glucose in healthy men. The Journal of Nutrition 147:152-160

Pantophlet, A.J., H. Roelofsen, M.P. de Vries, W.J.J. Gerrits, J.J.G.C. van den Borne and R.J. Vonk, 2017. The use of metabolic profiling to identify insulin resistance in veal calves. PLoS One 12: e0179612

Pantophlet, A.J., M.S. Gilbert, W.J.J. Gerrits and R.J. Vonk, 2017. Short Communication: Supplementation of fructo-oligosaccharides does not improve insulin sensitivity in heavy veal calves fed different sources of carbohydrates. The Journal of Dairy Science 100:9442–9446.

Gilbert, M. S., A. J. Pantophlet, H. Berends, A. M. Pluschke, J. J. G. C. van den Borne, W. H. Hendriks, H. A. Schols, and W. J. J. Gerrits, 2015. Fermentation in the small intestine contributes substantially to intestinal starch disappearance in calves. The Journal of Nutrition 145:1147-1155

Gilbert, M. S., J. J. G. C. van den Borne, H. Berends, A. J. Pantophlet, H. A. Schols, and W. J. J. Gerrits. 2015b. A titration approach to identify the capacity for starch digestion in milk-fed calves. Animal 9:248-257

Gilbert, M. S., A. J. Pantophlet, J.J.G.C. van den Borne, W.H. Hendriks, H.A Schols, and W.J.J. Gerrits, 2016. Effects of replacing lactose from milk replacer by glucose, fructose, or glycerol on energy partitioning in veal calves. The Journal of Dairy Science 99:1121-1132

*Equal contribution

Curriculum vitae

127

Curriculum vitae

Andre Pantophlet was born on February 26th 1986 in Willemstad, Curaçao. He graduated from the secondary school (VWO) in 2004 at Peter Stuyvesant College, Willemstad, Curaçao. Thereafter, he moved to the Netherlands and studied analytical chemistry at the University of Applied Science Leiden. After obtaining his BASc he continued with his MSc education at the VU University Amsterdam, where he specialized in (bio-) analytical sciences. He performed his research internship at the department of Chemistry and Pharmaceutical Sciences (subidivision Bioanalytical Chemistry, dr. Jeroen Kool), where he studied the possibilities of using extremely high temperature LC-MS using only water to separate and detected (candidate) drugs and their metabolites. He wrote his literature thesis about the analytical challenges in metabolic profiling under supervision of dr. Henk Lingeman. In 2011, Andre started his PhD at the Centre for Medical Biomics (prof. dr. Roel Vonk) of the University Medical Center Groningen and the Animal Nutrition Group (prof. dr. Wouter Hendriks) of Wageningen University. In his PhD, he studied the effects of several nutritional adjustments/feeding strategies on (the development of) insulin resistance and glucose homeostasis in veal calves. Furthermore, he explored the applicability of metabolic profiling techniques in the identification of insulin resistance veal calves. The PhD was performed within the framework of the Carbohydrate Compentence Center in collaboration with University Medical Center Groningen, Wageningen University, VanDrie Group and Tereos Syral. Since June 2016 he is working as Study Director LC-MS at the bioanalytical laboratories of PRA Health Sciences.

Acknowledgements

129

Acknowledgements

And then, I managed to finish my PhD project. It was an interesting, funny and challenging period and I have learned a lot. Throughout this period I have received much support from my colleagues and several experts from different disciplines. It is difficult to find the right words to express my gratitude and appreciation to everyone who has helped me throughout this period, but I will do my utmost best.

First of all I want to thank my promoter Roel, who has given me the opportunity to perform my PhD. Before even starting my PhD program I was very interested in your work and the work of the Centre for Medical Biomics in general. During the years I have learned a lot from you. Thanks to you I have not only improved my management and research skills, but I have also developed as a person in general. I will never forget the many healthy discussions we had about the focus of the project. Sometimes a bit frustrating, but we always managed to reach a consensus. Throughout the years we have worked in harmony and it was a great pleasure. I wish you the best with your retirement, and I hope you really enjoy it. Success with your retirement, and I hope you have many wonderful moments with your family and loved ones. Please remember to take enough time to rest and relax.

Dear Han and Marion. Despite the fact you were only physically present as my daily supervisor during the first 1.5-2 years of the project, I could always count on your valuable input and encouragements. I really appreciate this. Marion, I have learned a lot from you and I admire the way you critically analyze your/our work. Its been more than 3-4 years since we don’t see each other on a daily basis anymore, but I hope that you are enjoying your life to the fullest back in Germany. Han, we almost clicked immediately, as we are both (bio)- analytical chemist. We had many discussions on metabolic profiling and LC-MS analysis in general, and were constantly exchanging publications, increasing our knowledge on this field. Due to retirement of Roel our group was dissolved too early, in my opinion, and we had to go separate ways. This is a pity because I had the feeling we could have learned a lot more from each other. Nevertheless, you always remained available for discussions regarding our work. I hope you are enjoying your new career as a teacher and hope we meet again soon.

Dear Marcel. Thanks for all your practical support in the lab. I have always admired your skills as a lab analyst and your knowledge of LC-MS. We had some nice discussions not only on metabolic profiling and LC-MS, but also about research and PhD in general. I will soon pass by again so we can have some coffee and tea and talk about the good times we had working in the lab. How can I forget Gerlof. Thanks for all the assistance in analyzing the 1000’s insulin and glucose samples. I could always count on your help, especially during the labor-intensive periods. I was happy to hear that you finally decided to do a PhD as well. Know that you can always count on my help if needed. Coby and Kees you guys were my “predecessors” and I always looked up to

Acknowledgements

130

you. I could always reach out for tips and tricks. Kees, thanks for the “pep talks” during the last phase of my PhD, it helped a lot. Coby, you were my office mate and we clicked from day one. You always helped me out when needed and you experienced my good and bad days. We should definitely keep our tradition of going out (dancing, drinking, etc.) with the group.

Dear Walter, Joost and Myrthe. As collaborators on this project from the University of Wageningen we had some great but also tough times. Walter, I always admired your skills as the project leader. You always kept us on the right track and made sure we kept our eyes on the main focus of the project and the corresponding timelines. Thanks for all your help, your availability and valuable input. Joost, despite the fact that you were sometimes limitedly available, I have learned a lot from you. Your knowledge on calves and insulin resistance is incredible and I could always count on your in-depth and critical reviews. Myrthe, my “partner in crime”, thanks for putting up with me during this PhD project. I will never forget those early days and long hours at the experimental facilities. I really appreciate the fact that you made me feel at home during the experimental days at Wageningen. I hope you are enjoying your post-doc work. If you are ever up north (in Groningen) please don’t hesitate to give me a call.

Eelke van der Wal (VanDrie Group), Wiebe Mulder (VanDrie Group), Emmanuelle Apper (Tereos Syral) and Frédérique Respondek (Tereos Syral), thank you for your contribution both financially and scientifically in this project. You guys were always present during the meetings and had great ideas about the experiments and experimental conduct.

Also, I want to thank all the animal caretakers who helped me during the 3 large experiments, namely Tonnie, Klaas, Ries, André, Rinie and Bert. You took excellent care of the calves and always came with wonderful suggestions.

Last but not least I will like to thank my parents Jacinto Pantophlet and Joke King. Despite the fact that we live an ocean apart you always encouraged me to make the most out of life. When I turned 18 you allowed me to leave the small Island of Curaçao and move all the way to another continent to realize my dream to be an (bio)- analytical chemist. I cannot wait to see both of you again and hand you this PhD-thesis personally.

Andre.

List of Abbreviations

131


BW Body weightCON ControlDM Dry MatterFOS Fructo-oligosaccharides FRU FructoseFSIGTT Frequently sampled intravenous glucose tolerance test GI Glycemic index GIR Glucose infusion rate GLU GlucoseGLUT Glucose transporter GLY GlycerolHDL High-density lipoproteinHF High-fatHILIC Hydrophilic interactionHL High-lactose iAUC Incremental area under the curve IR Insulin resistant ISi Insulin sensitivity index derived from an intravenous glucose tolerance test ISminmod Insulin sensitivity index derived from the frequently sampled intravenous glucose tolerance test IVGTT Intravenous glucose tolerance test Kg Rate of glucose disappearance LC-MS Liquid chromatography–mass spectrometry LDL Low-density lipoproteinM-value Glucose disposal M/I-value Glucose disposal divided by the average plasma insulin level at steady state ME Metabolizable energyMIS Moderately insulin sensitive MR Milk replacer MS Mass spectrometryMTT Milk tolerance test NEFA Non-esterified fatty acids NMR Nuclear magnetic resonance spectroscopy OPLS-DA Orthogonal projection to latent structures discriminant analysis PDV Portal-drained viscera QC Quality control


132

QUICKI Quantitative insulin sensitivity check index RP Reversed phase scFOS short-chain Fructo-oligosaccharides SEM Standard error of the meanSF Solid feedTmax Time to maximum concentrationTmin Time to minimum concentrationVIP Variable importance in the projection ΔCmax Maximum concentration minus fasting concentrationΔCmin Minimum concentration minus fasting concentration

Documents

University of Groningen Glucose Homeostasis and Insulin ... · Glucose Homeostasis and Insulin Resistance in veal calves Studies on the effects of age, nutritional modulations and