A. A . ..
A,w..,1fi., A.
,mihfiémfi 1 ,2;
atti‘ZLA . Pair: . .5381 "an.
.1

_
“a? A.

a
.

,. . he, ”4...; A

. .r .a $.51. w ..

A .1}. i, .6, A . .
A . 4p ?

In! .31 ,
L m . L’i!f~-C...¢.. .. ..
. A: 5.1. 2..
5...? V ‘BIIK:1N.DAIL

.

t «I... I.) .
Barr». .wafik

, i A, A an
is? ..;.- A , a:
,3... nuhxwwammfimrznm A haw

L: . 9 .
. a . 5...:
A an“

EL}.
(.5.
F x

.4
.3!

. . .a 3.x [3;

. v-2.t A . :1 «33“..

. thiai. . .1312. .A in
Q»??? Arvin hum?

(u. f... .n .. Liliana

‘3‘
2. I‘m
. ‘t\

r .5 . :3 .. v: z!!!
136.. nutshruotlil.
1.223%? -

. L2! .

tr. 2.1:. $131!-
. 01}... .(
.Sra-r‘

G... n :c

 

 

This is to certify that the
dissertation entitled

PROPIONATE REGULATION OF FEED INTAKE

presented by

BARRY JOSEPH BRADFORD

has been accepted towards fulfillment
of the requirements for the

Doctoral degree in Animal Science

 

 

WJWL

' ' Map/Professor’s Signature
fill/06

Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

 

.UBRARY l
M'Chigan State
University

 

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

M

m2

 

 

 

 

 

 

 

 

 

 

 

 

 

2/05 p:/CIRC/DateDue.indd—p.1

 

 

 

 

PROPIONATE REGULATION OF FEED INTAKE
By

Barry Joseph Bradford

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Animal Science

2006

ABSTRACT
PROPIONATE REGULATION OF FEED INTAKE
By

Barry Joseph Bradford

Ruminant diets that include large amounts of readily-fermentable starch are
known to limit feed intake. Experiments conducted over the past 20 years have
indicated that physiological changes in rumen osmolality, rumen pH, or plasma
insulin concentration do not consistently explain hypophagic responses to highly
fermentable diets. Meanwhile, evidence has accumulated in a number of species
that hepatic oxidation of fuels stimulates satiety, and the rapid increase in
propionate flux to the liver during meals likely results in oxidation of some
propionate. Six studies were conducted to evaluate the hypothesis that hepatic
oxidation of propionate limits feed intake in lactating dairy cattle. In a crossover
experiment including 32 cows with a wide range of production levels, increased
dietary starch fermentability depressed dry matter intake (DMI) 8% but did not
alter digestible DMl. Although consistent with the hepatic oxidation hypothesis,
these results could also be explained by other proposed mechanisms for
regulation of intake. Evidence from both ruminant and non-ruminant studies has
suggested that propionate may directly stimulate leptin secretion. Therefore, 2
experiments were conducted utilizing pulse-dose and intermediate-term
propionate infusions to evaluate whether this mechanism is important in lactating

cows. Based on the small, non-linear effects of propionate administration on

leptin secretion that were observed, we concluded that leptin does not likely
mediate intake depression by propionate. We then conducted a set of
experiments using the glucose analog phlorizin to increase glucose demand and
gluconeogenic capacity. By increasing the potential for glucose production from
absorbed propionate, we expected to decrease the proportion of propionate that
was oxidized, and in turn, increase feed intake. Evidence from 3 experiments
suggested that glucose demand was increased by phlorizin treatment, and
measurement of transcript abundance for potentially rate-limiting gluconeogenic
enzymes indicated that hepatic gluconeogenic capacity was increased by
phlorizin treatment. However, phlorizin treatment also consistently stimulated
lipolysis, which likely increased hepatic fatty acid oxidation. This confounding
effect probably played a role in the lack of an effect on feed intake, and
prevented assessment of the hepatic oxidation hypothesis. Finally, we tested the
hypothesis that more rapid infusion of propionate during meals would overwhelm
hepatic gluconeogenic capacity to a greater extent, resulting in quicker satiety
and smaller meals. However, we did not find differences in feeding behavior in
response to infusion of 1.2 mol of Na propionate over the course of 5 or 15
minutes during meals. In conclusion, these experiments have not added to the
existing evidence for the hepatic oxidation hypothesis for ruminants.
Nevertheless, progress has been made toward understanding the mechanism for
propionate regulation of feed intake, because leptin has been ruled out as a
necessary mediator of this response, and because glucose demand has been

shown to have little influence on DMl in the short-term.

ACKNOWLEDGEMENTS

Thanks first to Dr. Michael Allen for his outstanding mentorship, sage advice, and
good nature. I could not have chosen a better advisor for my PhD program, and l
have benefited greatly from his guidance. I am also indebted to the remaining
members of my guidance committee; Dr. MattDoumit, Dr. Tom Herdt, Dr. Dale
Romsos, and Dr. Rob Tempelman have all been generous with their time, and I
feel very fortune to have learned from these experts in such divergent areas of

study.

The expert help and friendship of the Allen Lab made this dissertation possible
and helped me keep my sanity through my graduate training. Thanks to Dave
Main, Dewey Longuski, Jackie Ying, Dr. Steve Mooney, Jenn Linton, Christy
Edwards, and Kevin Harvatine for the late nights of sample collecting, long days
in the lab, and for all of those productive “meetings”. Thanks also to all the crew
at the MSU Dairy Cattle Teaching and Research Center, especially Bob Kreft,

Gordon Galloway, Randy Bontrager, and Rob West.

I am grateful for the friendship and support of the faculty, staff, and graduate
students in the Department of Animal Science. It seems as though nearly
everyone in the department has helped me at some point during my time at
Michigan State, but I will especially remember the conversations with Dr. Herb

Bucholtz, Dr. Dave Beede, Dr. Mike VandeHaar, Dr. Dave Hawkins, Dr. Brian

iv

Nielsen, Dr. Mike Orth, Jim Liesman, Justin Zyskowski, Sue Sipkovsy, Chris
Colvin, Dr. Dave Edwards, Dr. Emilio Ungerfeld, Dr. Brett Etchebarne, Dr. Laurie

Rincker, Dr. Mike Rincker, and Dr. Osman Patel.

Finally, I want to thank my family. Sarah and Hannah have made this time in my
life so much more meaningful, and have always been understanding when
responsibilities got in the way of family time (well, Sarah at least). Our parents
and siblings have always supported us, even though our location made it harder
to get everyone together, and I appreciate that. Most importantly, my family has
instilled in me a faith in God that has made it possible for me to keep my head up
even in the toughest of times. All of you have made me who I am today (for

better or worse)!

TABLE OF CONTENTS

List of Tables ........................................................................................ ix
List of Figures ....................................................................................... xi

CHAPTER 1: Literature review

Starch in ruminant diets .................................................................. 1
Propionate and its role in ruminant metabolism ................................... 4
Hypophagic effects of propionate ...................................................... 7
Proposed mechanisms for depression of feed intake by propionate ......... 9
Osmotic effects .................................................................... 9
Insulin signaling .................................................................. 10
Central nervous system sensors ............................................ 11
Peripheral sensory system ................................................... 13
Potential mechanisms linking hepatic oxidation and feeding behavior.....16
Implications of the hepatic oxidation hypothesis for ruminants ............... 22

CHAPTER 2: Increasing dietary starch fermentability decreases feed intake
of lactating dairy cows

Abstract .................................................................................... 26
Introduction ................................................................................ 28
Materials and methods ................................................................. 29
Data and sample collection ................................................... 30
Sample and statistical analysis .............................................. 33
Results and discussion ................................................................. 37
Plasma metabolites and hormones ......................................... 39
Variability in DMI responses to altered starch fermentability ........ 40
Conclusions ................................................................................ 45

CHAPTER 3: Propionate is not an important regulator of plasma leptin
concentration in dairy cattle

Abstract ..................................................................................... 47
Introduction ................................................................................ 48
Materials and methods ................................................................. 48
Experiment 1 ..................................................................... 48
Experiment 2 ..................................................................... 49
Analytical procedures .......................................................... 49
Results ...................................................................................... 50
Experiment 1 ..................................................................... 50
Experiment 2 ..................................................................... 52
Discussion .................................................................................. 52

vi

CHAPTER 4: Phlorizin administration increases hepatic gluconeogenic
enzyme mRNA abundance but not feed intake in late-
lactation dairy cows

Introduction ................................................................................ 59
Materials and methods ................................................................. 60
Design and treatments ......................................................... 60
Data and sample collection ................................................... 60
Sample analysis ................................................................. 60
Statistical analysis .............................................................. 61
Results ...................................................................................... 61
Glucose loss, milk production, and intake ................................ 61
Plasma metabolites and hormones ......................................... 61
Hepatic mRNA abundance ................................................... 61
Discussion .................................................................................. 61

CHAPTER 5: Phlorizin induces lipolysis and alters meal patterns in both
early and late lactation dairy cows

Abstract .................................................................................... 67
Introduction ................................................................................ 69
Materials and methods ................................................................. 70
Design and treatments ......................................................... 70
Data and sample collection ................................................... 71
Sample analysis ................................................................. 72
Statistical analysis .............................................................. 74
Results and discussion ................................................................. 75
Conclusions ................................................................................ 81

CHAPTER 6: Phlorizin administration did not attenuate hypophagia
induced by intraruminal propionate infusion

Abstract .................................................................................... 82
Introduction ................................................................................ 84
Materials and methods ................................................................. 85
Design and treatments ......................................................... 85
Data and sample collection ................................................... 86
Sample analysis ................................................................. 88
Statistical analysis .............................................................. 89
Results and discussion ................................................................. 90

CHAPTER 7: Rate of propionate infusion within meals did not influence
feeding behavior

Abstract .................................................................................... 97
Introduction ................................................................................ 99
Materials and methods ................................................................. 99
Results and discussion ............................................................... 102

vii

CHAPTER 8: Conclusions .................................................................. 107

BIBLIOGRAPHY ................................................................................. 113

viii

Table 2.1

Table 2.2

Table 2.3

Table 2.4

Table 3.1

Table 4.1

Table 4.2

Table 4.3

Table 4.81

Table 5.1

Table 5.2

Table 5.3

Table 6.1

Table 6.2

Table 6.3

LIST OF TABLES

Ingredients and nutrient composition of experimental diets ............ 29

Effects of corn grain conservation method on intake,
digestibility, and productivity of lactating cows ............................ 38

Effects of corn grain conservation method on blood plasma
metabolites and hormones ..................................................... 40

Pearson correlation coefficients for dry matter intake
depression and potential pre-trial response predictors .................. 41

Nutrient composition of experimental diets (percent of
dietary dry matter) ................................................................. 49

Ingredients and nutrient composition of experimental diet ............. 60

Effects of phlorizin treatment on glucose loss, feed intake,
and milk production in lactating Holstein cows ............................. 61

Effects of phlorizin treatment on blood plasma metabolites
and hormones in lactating Holstein cows .................................... 62

Genes, primers, and homology to reference sequences for
sequences analyzed by qRT-PCR ............................................. 65

Ingredients and nutrient composition of experimental diet ............... 72
Effects of phlorizin treatment and stage of lactation on

glucose excretion, blood plasma metabolites and hormones,

and feeding behavior .............................................................. 76

Effects of phlorizin treatment and stage of lactation on milk
and milk component yield ....................................................... 79

Ingredients and nutrient composition of experimental diet...............87
Effects of phlorizin treatment and SCFA infusion on glucose
excretion, ruminal SCFAs, and blood plasma metabolites and
honnones ............................................................................ 91

Effects of phlorizin treatment and SCFA infusion on feeding
behavior and yield of milk components ...................................... 92

ix

Table 7.1 Ingredients and nutrient composition of experimental diet ............. 101

Table 7.2 Effects of propionate infusion rate on feeding behavior
of lactating cows ................................................................. 103

Figure 1.1

Figure 1.2

Figure 2.1

Figure 2.2

Figure 3.1

Figure 3.2

Figure 3.3

Figure 4.1

Figure 4.81

Figure 5.1

Figure 7.1

Figure 7.2

LIST OF FIGURES

Proposed mechanism for regulation of feed intake by
hepatic oxidation ................................................................ 21

Model by which propionate might affect satiety in ruminants ........ 23
Pre-trial plasma insulin concentration (COV insulin) predicts

DMI depression resulting from an increase in dietary starch
fermentability ..................................................................... 42
Insulin response index from glucose tolerance tests predicts

DMI depression in response to increased dietary starch
fermentability ..................................................................... 42
Responses to intrajular bolus of Na propionate ......................... 51

Pre-infusion leptin concentrations are positively correlated
with body condition score ..................................................... 52

Responses to continuous intraruminal Na propionate infusion ...... 53
Effects of phlorizin administration in lactating Holstein cows

on hepatic mRNA abundance relative to control for genes

involved in gluconeogenesis ................................................. 62

Overview of hepatic metabolic pathways related to
gluconeogenesis ................................................................ 66

Effects of phlorizin administration on feeding behavior ................ 80

Propionate infusion at spontaneous meals depresses feed
intake in a dose-dependent manner ...................................... 104

Distribution of meal lengths on infusion days .......................... 106

xi

CHAPTER 1

LITERATURE REVIEW

Feed intake of ruminants is regulated by an extremely complex system which
integrates disparate signals from various organs (Forbes, 1996). Although
homeorhetic mechanisms can alter the feed intake set point over the long-term,
daily intake is determined only by meal size and meal frequency. From a
physiological perspective, gastrointestinal distention and chemosensory
feedback are the most important determinants of meal patterns. While physical
fill limitations to feed intake have been well characterized (Allen, 2000), the
effects of diet on chemosensory regulation of intake, and the mechanisms
involved, are poorly understood in ruminants. Recent work has focused on
explaining feed intake depression by propionate and on providing a relatively
comprehensive mechanism for feed intake regulation by metabolites. This
review of the literature will highlight the determinants of propionate production, its
role in ruminant metabolism, effects of propionate on feed intake, and potential

mechanisms for these responses.

Starch In ruminant diets

The rates of ruminal starch digestion and passage vary greatly across grains fed
to ruminants and depend upon the type of cereal grain, conservation method,
and processing (National Research Council, 2001 ). Pure starch is readily

degraded by many ruminal microbes; however, starch found in feed grains is

surrounded by a protein matrix. Degradation of this protein matrix is typically the
rate-limiting step in ruminal starch digestion (Kotarski et al., 1992). Differences in
fermentability between types of grains and varieties within a grain type (eg.
vitreous vs. floury corn varieties) are primarily due to differences in the protein
matrices within the grain endosperrn. Likewise, degrading the protein matrix
through ensiling, or increasing the surface area available for enzymatic
degradation by processing, can increase the ruminal fermentability of a grain

source.

Ruminal digestion kinetics determine the site and extent of nutrient digestion,
which can greatly affect the type and temporal pattern of fuels absorbed. While
ruminal starch digestion results in the production of volatile fatty acids (VFAs),
starch that escapes ruminal digestion can be degraded by pancreatic amylase in
the duodenum. Although little glucose appears in the portal circulation in
ruminants, glucose is efficiently absorbed in the small intestine (Huntington and
Reynolds, 1986). Gastrointestinal epithelial cells apparently metabolize most
absorbed glucose to lactate, and a relative high net flux of lactate has been
measured across the splanchnic bed of lactating cows (Reynolds et al., 2003).
Therefore, diets with similar concentrations of starch can provide the animal with
VFAs and lactate in different proportions depending on the physical

characteristics of the starch source.

Increasing the amount of grain in the diet, or the fermentability of grain, results in
increased ruminal VFA production primarily because of greater propionate
production (Bauman et al., 1971; Sutton et al., 2003). Ruminal starch-digesting
microbes have a competitive advantage when high-grain rations are fed because
of their ability to rapidly utilize starch, and these species produce more
propionate than cellulolytic bacteria (Russell, 1998). Furthermore, increased
ruminal fermentation rate is typically associated with a decrease in ruminal pH,
and bacteria that tend to produce propionate are favored in this situation because

of their tolerance for low pH (Russell, 1991 ).

In addition to effects on nutrient absorption and rumen pH, variation in ruminal
starch digestion alters animal productivity and behavior. Compared to feeding
lactating cows forage alone, inclusion of grain can increase dry matter intake
(DMI) by removing physical fill constraints on intake. Because grains also
provide more metabolizable energy than forages, this can result in dramatically
increased energy intake and milk production (Conrad et al., 1964). Not
surprisingly, grain feeding became popular in the mid-1900’s and most diets fed
to lactating cows now contain at least 40% concentrate. However, it is now
known that cereal grains that are highly digestible in the rumen can depress feed
intake of lactating cows. Dry matter intake (DMI) was depressed by nearly 3 kg
DM / d (~13%) when more fermentable grains were substituted in diets of
lactating cows in several studies reported in the literature (Allen, 2000). Oba and

Allen (2003b) demonstrated that a more rapidly fermented starch source reduced

meal size 17%, causing an 8% reduction in food intake despite a 10% decrease
in intermeal interval. The more fermentable treatment nearly doubled the
fractional rate of starch digestion in the rumen, increasing the contribution of
VFAs, especially propionate, as fuels at the expense of lactate. The evidence
that propionate mediates feed intake depression by highly fermentable starch

sources is discussed below.

Propionate and Its role In ruminant metabolism

Propionate is a 3-carbon VFA that plays a central role in ruminant metabolism.
Propionate is produced by microbial degradation of carbohydrates, and to a
lesser extent, amino acids (Bergman, 1990). Work in the late 1800’s led to the
assumption that VFA were an important source of energy for ruminants, but
propionate was first identified as a glucose precursor by Ringer (1912), who
showed that injecting phlorizin-treated dogs with propionate resulted in increased
glucose excretion. A number of investigators, using similar protocols, later
showed that infusion of either propionate or butyrate in ruminants led to
increased plasma glucose concentrations (Kronfeld, 1957; Potter, 1952).
However, Leng and Annison (1963) demonstrated that propionate, but not
butyrate, contributed labeled carbon to glucose during incubations with sheep
liver slices. Ash and colleagues (1964) also showed that infused butyrate did not
contribute labeled carbon to plasma glucose, and did not increase glucose
production by isolated hepatocytes. Propionate is now recognized as the primary

glucogenic precursor In ruminants, and estimates of its contribution to glucose

requirements are typically in the range of 50 - 75% (Bergman, 1990). One
reason that propionate is such an important glucose precursor is that hepatic
uptake exceeds 70% of supply, and more than 90% of absorbed propionate is
taken up by the liver (Reynolds, et al., 2003). Propionate can be used quite
efficiently for glucose production (Steinhour and Bauman, 1988), but direct
assessment of propionate oxidation demonstrated that approximately 13% of
propionate taken up by the liver is oxidized (Black et al., 1966). Propionate not
taken up by the liver can be used by peripheral tissue for oxidative metabolism or
by adipose and mammary tissues for de novo lipogenesis, leading to synthesis of

odd-chain fatty acids.

Propionate production rates vary greatly across diets fed to lactating cows.
Mean calculated production rates were reported in the range of 13.4 to 16.9
mol/d in lactating Holstein cows fed diets containing 30 - 33% starch on a BM
basis (Oba and Allen, 2003b; Sutton, et al., 2003). However, calculated
production was 9.3 mon from a diet containing 21% starch (Oba and Allen,
2003b), and Sutton et al. (2003) measured a mean production rate of 36.2 mol/d
while feeding a 48% starch diet. Therefore, within similar cows, dietary
differences can alter propionate production rates by at least 4-fold. Diet-induced
differences in propionate production are further amplified if one considers
temporal variation in propionate production. Starch that is more completely
digested in the rumen is also more rapidly fermented, resulting in greater

differences in propionate production rates during and immediately after meals.

Propionate absorption may be more limiting than production during times of rapid
fermentation. However, propionate and other VFAs act within the rumen to
increase blood flow to ruminal papillae. Small amounts of VFA applied to the
ruminal epithelium can stimulate an increase in local blood flow (Thorlacius,
1972), and portal blood flow increased with increasing rate of intraruminal
propionate infusion (Weeks and Webster, 1975). Absorption of VFAs is thought
to be primarily a passive process; therefore, greater blood flow to the ruminal
epithelium clears nutrients from intracellular pools more quickly, resulting in a
larger cross-membrane concentration gradients and increased absorption rates.
The combination of increased blood flow and rapid production of VFAs during
meals results in a rapid increase in propionate delivery to the liver during the
course of a meal. Benson et al. (2002) reported that net appearance of
propionate from the portal-drained viscera can increase by more than 40% within

an hour of feeding.

Propionate’s central role in ruminant metabolism suggests that regulation of feed
intake by propionate is feasible from an evolutionary perspective. Although most
wild ruminants consume little grain, propionate is also produced during ruminal
digestion of forages. The production of adequate propionate is especially
important in lactating ruminants consuming only forages, because little lactate is
absorbed and amino acids are ideally conserved for milk protein synthesis.

Therefore, within the constraints imposed by the volume of the gastrointestinal

tract, feed intake should increase until sufficient propionate is available for
glucose production. The mechanism proposed below provides the additional
benefit of responding to changes in glucose demand rather than simply

responding to propionate concentrations per se.

Hypophagic effects of propionate

Early in the study of feed intake regulation, ruminants were recognized as a
valuable model for investigating feed intake depression by metabolites.
Following the suggestion by Mayer (1953) that feed intake is regulated by
changes in blood glucose concentration, a number of investigators began
assessing the possibility that acetate, propionate, or butyrate might play a role in
regulating feed intake of ruminants. Dowden and Jacobsen (1960) first reported
that 8-h intravenous infusions of propionic acid or acetic acid significantly
depressed 24-h feed intake of dairy calves, but that infusions of butyric acid,
lactic acid, and glucose did not influence feeding behavior. Later work from
Baumgardt’s group confirmed that VFA, but not glucose, were capable of
decreasing feed intake of mature dairy cows (Montgomery et al., 1963; Simkins
et al., 1965). Baile and Pfander (1966) reported that pulse-dose intraruminal
infusion of acetate (0.25 mol) prior to feeding depressed feed intake of sheep by
40%, but also showed that infusion of isomolar NaCl decreased feed intake by
25%, suggesting that ruminal osmolality is involved in intake modulation by
VFAs. This led Baile and Mayer (1968) to investigate the relative effects of

intraruminal and intravenous infusions of acetate administered during

spontaneous meals. The hypophagic response to intraruminal acetate infusion
was not observed when acetate was infused intravenously (Baile and Mayer,
1968), supporting the hypothesis that intraruminal sensors are involved in
regulation of feed intake by acetate. At that point in time, most investigators had
focused on acetate as a potential feedback signal because its concentration in
peripheral blood is higher than that of other VFAs. However, using intraruminal
infusion of VFAs at spontaneous meals, Baile and Mayer (1969) demonstrated
that propionate is at least as effective as acetate and more effective than

butyrate at eliciting feed intake depression in sheep.

In the last 30 years, hypophagic effects of propionate infusions have been
documented extensively for ruminants (Anil and Forbes, 1980; Elliot et al., 1985;
Famingham and Whyte, 1993; Hurtaud et al., 1993; Leuvenink et al., 1997;
Mbanya et al., 1993; Shepard and Combs, 1998). Propionate was more
hypophagic than acetate or butyrate when infused into the portal vein of sheep
(Anil and Forbes, 1980), and infusion of propionate into the mesenteric vein of
steers reduced feed intake while acetate infused at similar rates did not (Elliot et
al., 1985). Although propionate might be expected to decrease food intake
compared to acetate because it has higher energy content, propionate linearly
decreased metabolizable energy intake compared to acetate in lactating cows
when infused intra-ruminally as iso-osmotic mixtures (Oba and Allen, 2003d). As
the proportion of propionate increased, the reduction in ME intake from the diet

exceeded that supplied from the infusate. Food intake was reduced primarily

from a linear reduction in meal size from 2.5 to 1.5 kg DM as propionate
increased from 0 to 100% of infusate. Meal frequency also tended to decrease
linearly (P = 0.08) from 7.4 to 6.1 meals during the 12 h monitoring period as
propionate was increased (Oba and Allen, 2003d). Therefore, propionate
decreased energy intake compared to acetate by increasing satiety and possibly
by decreasing hunger. Propionate also decreased food intake of dairy cows
compared to iso—energetic infusions of VFA mixtures (Hurtaud, et al., 1993) or
acetate (Shepard and Combs, 1998). These studies suggest that hypophagic
effects of propionate cannot be explained simply by the additional energy
supplied as propionate. It is unlikely that animals consume feed to meet their
energy requirements per se but rather have fuel-specific mechanisms regulating -

satiety and hunger.

Proposed mechanisms for depression of feed intake by propionate
The mechanism by which propionate regulates satiety is not fully understood. A
wide array of potential mechanisms has been proposed, reflecting the myriad

metabolic and physiological responses to this single metabolite.

Osmotic effects. One possibility is that increased ruminal VFA concentration
causes satiety by increasing the osmolality of rumen fluid during meals.
Osmolality may stimulate satiety through osmoreceptors in the rumen wall
(Carter and Grovum, 1990; Leek and Harding, 1975) or by stimulating release of

vasopressin, a hormone that can affect satiety (Langhans et al., 1991), in

response to water efflux from the blood to the rumen (Allen, 2000). While
mechanisms associated with osmolality likely contribute to satiety, changes in
osmolality per se do not affect food intake. Choi and Allen (1999) compared iso-
osmotic infusions of NaCl, Na acetate, and Na propionate administered
intraruminally at spontaneous meals for 12 h. Although NaCl decreased meal
size 27%, it did not affect daily food intake because intermeal interval decreased
31% compared to sham infusion. In that experiment, both Na acetate and Na
propionate decreased food intake compared to NaCI, primarily because of
greater intermeal interval, indicating delayed hunger (Choi and Allen, 1999). This
makes the important point that regulatory mechanisms must be investigated for
time courses beyond the first meal after treatment, because feed intake is
determined by both satiety and hunger. Feed intake depression by propionate

infusion cannot be explained by osmotic effects alone.

Insulin signaling. One of propionate's physiological effects is to increase insulin
secretion. Peripheral infusion of pr0pionate dose-dependently increases plasma
insulin concentrations, and the peak insulin response precedes the peak in
plasma glucose (Bradford et al., 2006a). This is not simply a pharmacological
effect, because portal infusion of propionate at physiological rates also Increased
plasma insulin concentration (Leuvenink, et al., 1997). Propionate’s role as an
insulin secretagogue led Grovum (1995) to suggest that hypophagic effects of
propionate are mediated by insulin (Allen, 2000). Grovum's hypothesis is

supported by the broad consensus that insulin is an anorexic agent in

10

monogastrics (Benoit et al., 2004; Porte et al., 2005), and by evidence that
intraventricular administration of insulin depressed feed intake of sheep (Foster
et al., 1991). However, hypophagic effects of propionate infusions have been
observed without an increase in plasma insulin concentration (Famingham and
Whyte, 1993; Frobish and Davis, 1977), and hyperinsulinemic—euglycemic
clamps do not depress energy intake of dairy cattle when energy provided by
infused glucose is considered (Griinari et al., 1997; Mackle et al., 1999; McGuire
et al., 1995). Insulin can likely influence feed intake through its effects on
metabolism, nutrient partitioning, and possibly by influencing hypothalamic
satiety centers; however, there is little evidence that it is directly responsible for

the depression of feed intake by propionate (Allen, 2000).

Central nervous system sensors. According to the glucostatic hypothesis of
Mayer (1953), neurons in the central nervous system (CNS) alter feeding
behavior in response to changes in plasma glucose concentration. Indeed,
neurons in the hypothalamus have long been known to alter their firing rates in
response to hyperglycemia and hypoglycemia (Anand et al., 1964). While few
investigators today believe that glucosensors play an important role in normal
energy homeostasis, glucosensors can likely stimulate eating in cases of
glucoprivation (Routh, 2002). Likewise, a sensory system in the CNS must be
considered as a potential mechanism for regulation of feed intake by propionate.
First-pass extraction of propionate by the liver exceeds 90% (Reynolds, et al.,

2003), resulting in much lower concentrations of propionate reaching the brain

11

compared to that reaching the liver. However, increasing rate of intraruminal
propionate infusion linearly increased peripheral plasma propionate
concentrations (Oba and Allen, 2003a), suggesting that the CNS can potentially
monitor propionate concentrations directly. Several important findings make a

CNS sensory system unlikely, however.

Over three decades ago, Baile (1971) found that propionate injections into the
ruminal vein during spontaneous meals decreased feed intake of sheep, yet
injections of larger amounts into the jugular vein did not alter feed intake. Anil
and Forbes (1980) confirmed this finding, showing that infusion of propionate into
the portal vein of sheep depressed feed intake while jugular infusion of the same
dose of propionate had no effect. If the CNS were involved in regulation of feed
intake by propionate, the opposite responses would be expected, because
infusion of propionate into the mminal or portal veins results in efficient clearance
by the liver and limited delivery of propionate to the brain. In a follow-up
experiment, sheep underwent surgery to denervate the liver prior to infusion of
propionate, and in sheep which were successfully denervated, subsequent
propionate infusions did not depress feed intake (Anil and Forbes, 1980). Later
work showed that intake responses to portal infusion of propionate were
eliminated by sectioning the hepatic branch of the vagus nerve or by
anaesthetizing the splanchnic nerve (Anil and Forbes, 1988). An important
shortcoming of these approaches is that interrupting efferent communication from

the CNS to the liver and pancreas can have multiple physiological consequences

12

(Berthoud, 2004), and the elimination of hepatic afferent signaling alone has not
been accomplished in ruminants. Nevertheless, in combination with the infusion
experiments discussed above, these results strongly suggest that signaling from
a peripheral sensory system to the CNS mediates propionate's effects on feed

intake, rather direct sensing within the CNS.

Perlpheral sensory system. In light of his findings regarding the effects of
propionate infusion into different veins, Baile (1971) proposed that propionate
receptors in the ruminal region of sheep might regulate food intake. His
hypothesis was supported in part by evidence that intraruminal administration of
VFA decreased gastric motility (Ash, 1959), presumably by activating
chemosensors in the rumen wall or in the veins draining the rumen. However,
later work showed strong feed intake responses to portal infusion of propionate
(Anil and Forbes, 1980; Famingham and Whyte, 1993), shifting the focus from
the rumen or ruminal vein to the region of the portal vein and liver. The
“propionate sensor" hypothesis continues to be supported by some investigators

(Leuvenink, et al., 1997).

An alternative way to explain hepatic involvement in feed intake depression by
propionate has arisen from the monogastric literature. The liver is in a unique
position to monitor changes in fuel metabolism to control eating behavior
because of its central role in energy metabolism of animals (Friedman and

Stricker, 1976). Langhans and Scharrer (1987) suggested that hepatic oxidation

13

of a variety of fuels contributes to regulation of food intake by altering signals to
brain satiety centers via the hepatic vagus, and a large body of evidence has
accumulated to support this hypothesis. Portal glucose infusion decreased
discharge rates of hepatic vagal afferents in guinea pigs (Niijima, 1983) and
portal infusion of 2-deoxy-D-glucose, which reduced glucose utilization,
increased the firing rate of hepatic vagal afferents and increased feed intake in
rabbits (Novin et al., 1973). Administration of the fructose analog 2,5-anhydro-D-
mannitol (2.5-AM) decreased hepatic ATP concentration by trapping inorganic
phosphate and elicited an eating response in rats (Koch et al., 1998; Rawson et
al., 1994; Tordoff et al., 1988). Furthermore, phosphate loading prevented both
the decrease in hepatic ATP and the stimulation of feeding by 2.5-AM (Rawson
and Friedman, 1994). The hepatic oxidation hypothesis, then, proposes that
meals can be terminated by a signal carried from the liver to the brain via
afferents in the vagus nerve that are affected by hepatic oxidation of fuels and

generation of ATP (Friedman, 1995; Langhans and Scharrer, 1992).

The mechanism by which propionate may alter the firing rate of the hepatic
vagus in ruminants is unknown. However, regulation of food intake by
propionate in ruminants is consistent with the hepatic oxidation hypothesis;
propionate might decrease food intake of ruminants by stimulating oxidative
metabolism in the liver (Allen, 2000). Of fuels metabolized by the ruminant liver,
propionate is likely a primary satiety signal because its flux to the liver increases

greatly during meals (Benson, et al., 2002). While propionate is extensively

14

metabolized by the ruminant liver, there is little net metabolism of acetate
(Reynolds, 1995) because ruminant liver has high activity of propionyl CoA
synthetase but not acetyl CoA synthetase (Demigne et al., 1986; Ricks and
Cook, 1981 ), thus explaining differences in hypophagic effects of infusions of

propionate and acetate in ruminants.

Evidence currently available in the ruminant literature cannot delineate whether
feed intake responses to propionate infusion are mediated by portal propionate
sensors or by hepatic oxidation of propionate. For the purpose of debating the
likelihood of each proposed mechanism, comparisons with the (proposed)
analogous portal glucosensors are useful. Much of the data supporting the
presence of glucosensors in the portal vein of monogastrics relies on the use of
2-deoxy-D-glucose (Novin, et al., 1973; Rezek and Kroeger, 1976; Rossi et al.,
1996), which has its effects by blocking metabolism of glucose. Therefore, both
proposed mechanisms rely on glucose metabolism to initiate a signaling
cascade. Although many ruminant tissues use little glucose, ruminant neural
tissue apparently utilizes glucose as extensively as that of monogastrics (Lindsay
and Setchell, 1976). If one makes the common assumption that mechanisms
regulating feed intake are well conserved across species, it follows that portal
glucosensors should be present in ruminants as well as in monogastrics. While
glucose can clearly alter feeding behavior of monogastrics (Forbes, 1988), it has
been shown to have no effect on feed intake in ruminants (Baile and Forbes,

1974; Dowden and Jacobson, 1960; Frobish and Davis, 1977; Simkins, et al.,

15

1965). To rely on receptors to explain the observed effects of nutrient infusions
on feed intake of ruminants, one must propose that the ability to metabolize
glucose in sensory receptors was lost, while the ability to metabolize propionate
was gained during the evolutionary divergence of ruminants. This seems
especially unlikely given that ruminant neural tissue is known to utilize glucose
(Lindsay and Setchell, 1976). Rather, differences in hypophagic effects of
glucose infusion observed between ruminants and monogastrics are likely
because of differences in hepatic oxidation of glucose; liver hexokinase activity is
low in ruminants compared to nonruminants (Ballard, 1965), and in mature
ruminants, hepatic removal of glucose appears to be negligible (Stangassinger
and Giesecke, 1986). Although results of propionate infusion experiments could
be explained by the presence of propionate sensors, the hepatic oxidation
hypothesis is more attractive because of its “broad explanatory power”

(Friedman, 1995) across different fuels and species.

Potential mechanisms linking hepatic oxidation and feeding behavior
Hepatic ATP is clearly involved in the hypophagic response to fuels in rats;
however, the mechanism by which intracellular ATP concentration affects the
firing rate of the hepatic vagus has not been determined. All mechanisms that
have been proposed involve depolarization of the plasma membrane of the
hepatocyte (Boutellier et al., 1999). Hepatocytes, like nerve cells, contain
voltage-gated ion channels that can rapidly alter membrane potential, and this

depolarization can trigger a number of physiological responses.

16

Perhaps the best-described mechanism for regulating membrane potential by
cellular ATP concentration is through ATP-sensitive K” (KATp) channels (Miki and
Seino, 2005). Because of the large K" ion gradient across the plasma
membrane, K” ions move out of the call when channels are open, driving the
negative resting potential even lower (hyperpolan'zation). However, when ATP
concentrations increase, ATP allosterically alters the conformation of Kim:
channels to stop ion flow. This results in accumulation of K”, potentially leading
to membrane depolarization. This mechanism is involved in insulin secretion by
pancreatic beta cells (Nichols and Koster, 2002), and KATP channels mediate the
increased firing rate of some glucose-sensitive neurons in the brain (Miki et al.,
2001). Although Km: channels are probably found in hepatocytes (Malhi et al.,
2000), they do not serve as an important component of the regulation of feeding
behavior by hepatic oxidation. All evidence regarding this hypothesis indicates
that transmembrane voltage becomes more positive (depolarizes) as hepatic
ATP concentrations decrease (Boutellier, et al., 1999; Friedman et al., 1999); the
opposite relationship would be expected if Kim: channels played a primary role in

linking cellular metabolism to membrane potential in hepatocytes.

Another link between cellular energy status and membrane potential is the
Na*/K+ ATPase pump (Langhans, 1996). This pump drives a net movement of 1
cation out of the cell, and is the primary mechanism responsible for maintaining
the negative resting membrane potential of most cells. Like other catalytic

proteins, activity of Na*/K* ATPase is affected by substrate concentration, and

17

Niijima (1983) proposed that changes in cytosolic ATP concentration could
influence activity of this pump sufficiently to alter resting membrane potential.
Strong evidence supports the conclusion that glucose utilization in the liver of
guinea pigs reduces the firing rate of hepatic vagal afferents by activation of
Na”/K* ATPase; portal infusion of ouabain, which blocks activation of Na‘lK”
ATPase, prevented effects of glucose on the firing rate of vagal hepatic afferents
(Niijima, 1983). More recently, it was reported that 2.5-AM not only decreased
ATP concentration in hepatocytes, but also increased intracellular Na+
concentration, consistent with decreased Na‘lK+ ATPase activity and a less
negative membrane potential (Friedman et al., 2003). Therefore, information
regarding hepatic ATP concentration may be encoded as membrane potential

through substrate-level effects of ATP on Na"/K+ ATPase activity.

Another emerging candidate is AMP-activated protein kinase (AMPK). AMPK is
activated in response to high AMPzATP ratios, and direct activation of AMPK in
the hypothalamus increased food intake of rats (Andersson et al., 2004). The
ability of AMPK to phosphorylate ion channels and other proteins involved in
signal transduction pathways provides a broad range of possible mechanisms for
communicating energy status to sensory neurons. For example, AMPK is known
to alter the function of voltage-gated sodium channels, increasing the length of
time that the channels are in open state following depolarization (Light at al.,
2003). AMPK also affects the function of certain CI‘ (Hallows et al., 2003) and

Na” (Carattino et al., 2005) channels; changing the flow of these ions through the

18

membrane could cause membrane depolarization. In addition, AMPK could
interact with any number of intracellular signaling cascades to amplify secretory
responses to membrane depolarization. Continued characterization of the
downstream targets of AMPK phosphorylation should shed further light on its

potential role in signaling hepatic energy status to the nervous system.

The most tenuous component of the proposed mechanism for hepatic oxidation
control of feeding behavior is the communication between hepatocytes and
neighboring vagal afferent nerve endings (Langhans, 1996). Although little work
has been done to investigate potential modes of communication, several
possibilities are feasible. Hepatocytes contain gap junctions that allow
intercellular movement of ions, including Ca2+ (Gaspers and Thomas, 2005),
which can cause membrane depolarization. Rawson et al. (2003) reported that
2.5-AM caused a phopholipase C —mediated release of intracellular calcium
stores in hepatocytes in a time frame that coincided with decreases in cellular
ATP concentration. However, gap junctions between hepatocytes and afferent
receptor nerves have not been found (Berthoud, 2004), making it unlikely that
direct transport of ions between hepatocytes and neurons occurs. It may be
more important that increased intracellular Ca2+ can stimulate exocytosis of
signaling molecules. Hepatocytes are known to utilize ATP as an
autocrine/paracrine signaling molecule (Wang et al., 1996), and some afferent
nerves are activated by ATP (Kirkup et al., 1999). Depolarization of a hepatocyte

could trigger release of intracellular Ca2+ stores (Weiss and Burgoyne, 2002),

19

leading to exocytosis of ATP. ATP could then move through the extracellular
space to nearby parasympathetic nerves, binding purinergic receptors and

leading to the generation of action potentials.

However communication between hepatocytes and afferent vagal nerves takes
place, the response is an alteration in the number of action potentials that are
generated in the sensory neurons. Action potentials are propagated along the
vagus nerve, and these vagal afferents terminate in the dorsal vagal complex of
the hindbrain; more specifically, they innervate the nucleus of the solitary tract, or
NTS (Berthoud, 2004). Information received at the NTS can initiate the classic
autonomic reflex responses, which include control over digestive processes
(Luckman and Lawrence, 2003). Projections from the NTS also innervate the
hypothalamus, including the paraventricular and arcuate nuclei (Ricardo and
Koh, 1978), which are important sites for regulation of feeding behavior.
Inhibitors of hepatic oxidative metabolism have been shown to activate neurons
throughout this pathway, clearly indicating that vagal-derived signals can

influence behaviors that are under the control of the forebrain (Horn et al., 1999).

Although certain components of the hepatic oxidation hypothesis remain
tentative, sufficient evidence has accumulated to indicate that these proposed
mechanisms are feasible (Figure 1.1). Redundancy has been a common theme

in the study of feed intake regulation (lnui, 2000), and multiple, overlapping

20

mechanisms likely play a role in communicating the energy status of the liver to

the hypothalamus.

 

Hepatic fuel oxidation Potential mediators

l

T Hepatocyte [ATP]
Na“/K+ ATPase activity

{ I AMPK activity
Hepatocyte membrane
I voltage (hyperpolarization)
I lon flow via gap junctions
{ I Release of signaling
molecules (eg. ATP)

I Firing rate of the
hepatic vagus nerve

1 Activation of the NTS

l

Inhibition of hypothalamic
feeding centers

Figure 1.1 Proposed mechanism for regulation of feed Intake by
hepatlc oxldatlon. Pathways that are involved signal transduction from

the liver to the forebrain are shown. Potential mediators are listed for

 

 

components of the mechanism that remain unresolved.

 

21

implications of the hepatic oxidation hypothesis for ruminants

Inconsistent effects of propionate infusion on food intake have led to doubts that
VFA per se are signals of satiety for ruminants (Grovum, 1995). However, the
partitioning of fuels among different tissues and between metabolic pathways
affects food intake substantially (Friedman, 1998). Propionate can be used for
gluconeogenesis, which consumes ATP, or oxidized in the tricarboxylic acid
(TCA) cycle as acetyl CoA, generating ATP. Therefore, the temporal pattern of
oxidative metabolism in the liver is greatly altered by both propionate flux to the

liver and fate of propionate within the liver (Allen, et al., 2005; Figure 1.2).

Insulin and glucagon are important for directing fuel partitioning, especially during
meals. Decreased rate of gluconeogenesis when plasma glucose and insulin are
high is expected to speed oxidation of propionate in the liver and cause satiety
sooner. In support of this, the extent of hypophagia caused by propionate
infusion increased linearly with plasma glucose concentration in dairy cows (Oba
and Allen, 20030). On the other hand, increased plasma insulin concentrations
may also contribute to increased hunger between meals by speeding clearance
of fuels from the blood, especially following relatively smaller meals. This might
explain inconsistent intake effects of peripheral insulin administration reported in
the literature (Hayirli et al., 2002); while increased hepatic oxidation of fuels likely
causes satiety, increased clearance of fuels from the blood is expected to cause

hunger (Oba and Allen, 2000).

22

gluconeogenesis

    

 

 

(+)
groptior'late increased glucose
ux 0 war demand
A
(+) oxidation
(+)
increased diet (+)
fermentability
. I-)
feed Intake ‘ satiety center

 

Figure 1.2 Model by which propionate might affect satiety In ruminants.
Propionate flux to the liver is affected by feed intake and diet fermentability
and increases greatly during meals. Propionate can be used for
gluconeogenesis, which consumes ATP, or oxidized in the tricarboxylic acid
cycle as acetyl CoA, generating ATP. Supply of propionate to the liver
relative to glucose demand should affect the temporal pattern of ATP
concentration in the liver and is expected to affect satiety. When glucose
demand is high, gluconeogenesis increases and less propionate is oxidized,
resulting in greater meal size. When glucose demand is low, a greater
fraction of propionate is oxidized, resulting in satiety and smaller meal size

(Allen et al., 2005).

23

Formulation of diets to shift starch digestion to the small intestine not only
decreases propionate production, but also increases lactate absorption. This
shift often results in greater DMI, which is inconsistent with the hypothesis that
feed intake is regulated to meet energy requirements (NRC, 2001). In contrast,
the hepatic oxidation hypothesis predicts increased DMI in response to shifting
the site of starch digestion. Stimulation of hepatic oxidation by lactate is much
less than propionate, especially during meals (Allen, et al., 2005), because of the
greater lag before lactate absorption, and because hepatic extraction of lactate is
much less than propionate (Reynolds, et al., 2003). Hepatic extraction of lactate
is probably lower because metabolism of lactate to pyruvate is
thermodynamically unfavorable when cellular NAD/NADH is low (Forman, 1988).
Therefore, because of differences in metabolism, the hepatic oxidation
hypothesis predicts that more energy can be absorbed in the form of lactate than

propionate.

Feed intake regulation by hepatic oxidation may also help explain other
phenomena observed in ruminants. Administration of somatotropin to lactating
dairy cows consistently leads to greater DMI, yet this increase lags behind the
increase in milk production by several weeks (Dohoo et al., 2003). This lag may
be caused by a short-term increase in fatty acid mobilization and oxidation in the
liver (Baldwin and Knapp, 1993). After cows deplete their most readily-available
energy stores, a gradual deficit of hepatic fuels may result in lower mean ATP

concentrations in the liver and increased DMI. Another unresolved issue is the

24

elusive mechanism for depressed feed intake in the periparturient period
(lngvartsen and Andersen, 2000). Plasma non-esterified fatty acid (NEFA)
concentrations can be up to tenfold higher in early lactation than during gestation
(lngvartsen and Andersen, 2000) as a result of decreased plasma insulin
concentration, combined with lower insulin sensitivity of adipose tissue during the
transition to lactation (Doepel et al., 2002). This hyperlipidemia likely contributes
to depressed feed intake rather than the reverse because plasma NEFA
concentrations increase preceding periparturient hypophagia (Vazquez-Anon et
al., 1994). Uptake of NEFA by the liver increases with increasing plasma NEFA
concentration, resulting in greater hepatic fatty acid oxidation (Drackley and
Andersen, 2006). Hepatic fatty acid oxidation likely contributes to satiety by
generating ATP (Ji et al., 2000), and negative energy balance is further
exacerbated because intake depression limits insulin secretion and promotes

continued lipolysis.

The hepatic oxidation hypothesis, therefore, provides a unifying mechanism for
explaining behavioral responses to changes in both nutrient digestion and
metabolism. However, despite the strength of the evidence supporting the
hepatic oxidation hypothesis, no mechanistic studies have been conducted to
evaluate the hypothesis in ruminants. To assess its potential explanatory power
for regulation of feed intake by lactating dairy cows, we designed and conducted
5 experiments; the results of these experiments are reported and discussed in

the remaining chapters of this dissertation.

25

CHAPTER 2
INCREASING DIETARY STARCH FERMENTABILITY DECREASES FEED

INTAKE OF LACTATING DAIRY COWS

ABSTRACT
The effects of dietary starch fermentability on feed intake, nutrient digestibility,
milk production, and plasma metabolites and hormones were evaluated in a
crossover study. Thirty-two multiparous Holstein cows [121 :l: 48 d in milk, 41 :l: 9
kg/d 3.5% fat-corrected milk (FCM); mean 1 SD] were randomly assigned to
treatment sequence and were fed a diet intermediate to the treatments during an
initial 21-d period. Treatments were dry ground corn grain (DG) and high
moisture corn (HM) harvested from the same field. Treatment periods were 14 d,
with the final 4 (I used for data and sample collection. Diets included corn silage
and alfalfa haylage at a 2:1 ratio and were ~26% neutral detergent fiber, 17%
crude protein, 32% starch, and 3.5% fatty acids. HM increased plasma glucose,
non-esterified fatty acid, and triglyceride concentrations, but treatment had no
effect on yield of milk or FCM. HM decreased dry matter intake (DMI) by 8%, but
did not significantly alter digestible DMI. Individual DMI responses were highly
variable, and variables from pre-trial propionate challenge tests, glucose
tolerance tests, and hepatic mRNA analysis were assessed as potential
predictors of DMI depression from increased dietary starch fermentability. Of the
covariates tested, only pre-trial plasma insulin concentration and insulin response

to glucose tolerance test were significant predictors of DMI depression. High

26

pre-trial plasma insulin concentration was correlated with greater depression in
DMI with increased fermentability; conversely, greater insulin secretion in
response to glucose infusion was associated with minimal depression in DMI.
Consistent with past results, increased dietary starch fermentability decreased
DMI. Significant correlations between insulin variables and individual DMI

response may warrant further investigation.

27

INTRODUCTION
High-producing dairy cows require high levels of dietary energy to meet
requirements for maintenance, milk synthesis, and reproduction. In the United
States, these energy requirements are often met by adding corn grain to the diet.
While corn grain in general is a highly digestible energy source, its ruminal
fermentability may vary greatly due to differences in preservation method.
Several studies have shown increases in ruminal starch digestibility between
19% and 24% for high moisture corn diets (HM) relative to ground dry corn diets
(DG) (Knowlton et al., 1998; Oba and Allen, 2003e; Ying et al., 1998). These
large differences in starch digestibility are expected to result in very different
rates of propionate production and absorption. While propionate is the primary
glucose precursor in dairy cattle, it may also serve as the primary metabolic
limitation to feed intake (Allen, 2000). Ruminal infusions of propionate decreased
NE intake in lactating dairy cows relative to iso-osmotic infusions of acetate (Oba

and Allen, 2003d).

Oba and Allen (2003b) compared a HM diet to a DG diet, both including
approximately 32% starch. The HM diet caused a 1.7 kg/d depression in DMI
relative to the DG diet without affecting mean or variance of ruminal pH.
However, responses to the change in starch fermentability were highly variable
across cows. The objective of this experiment was to determine if intake and
production responses to a change in diet fermentability could be predicted by

production level or by measures of gluconeogenic capacity.

28

MATERIALS AND METHODS
Thirty-two multiparous Holstein cows (121 :t 48 DIM; mean 1 SD) from the
Michigan State University Dairy Cattle Teaching and Research Center were
assigned randomly to sequence in a crossover experiment. At the beginning of
the experiment, BW of cows was 675 :l: 69 kg and FCM yield was 41.3 t 9.6 kg/d
(mean 1 SD). Prior to the initial treatment period, all cows were fed a single diet
intermediate in composition to the two treatment diets for 28 d. The purpose of
this covariate period (COV) was to obtain baseline values for DMI, milk yield, and
hepatic transcript abundance, and to conduct glucose and propionate challenge

tests, all independent of dietary treatments.

Treatments were conservation method of com grain (HM vs. DG). Treatment
periods were 14 d, with the final 4 d used to collect samples and data. One corn
hybrid (Great Lakes 4526; Great Lakes Hybrids, Ovid, MI) was grown in 2002,
and half of the field was harvested as HM at 69% DM, ground, and ensiled in a
2.4 x 30.0-m silage bag (Ag Bagger, Ag Bag Corp., Blair, NE). The remaining
half of the field was harvested as DG at 83% DM, dried to 86% DM by a high-
temperature forced-air dryer, and finely ground. Experimental diets contained
either HM or DG, corn silage (50% of forage DM), alfalfa silage (50% of forage
DM), a premix of protein supplements, a premix of minerals and vitamins, and

liquid protein supplement (Table 2.1). A diet intermediate to the experimental

29

diets was fed during the COV period. All diets were formulated for 18% dietary

CP concentration and fed as a TMR.

Table 2.1 Ingredients and nutrient composition of

experimental diets1.

 

 

Item COV DG HM
Diet ingredients
DG 14.7 30.3 -
HM 15.8 - 32.1
Alfalfa haylage 15.4 14.9 14.5
Corn silage 30.7 30.8 30.0
Protein mix2 15.2 15.6 15.2
Mineral and vitamin mix3 5.3 5.4 5.3
Liquid protein
supplement4 2.9 3.0 2.9
Nutrient composition
DM 46.1 49.1 47.8
OM 93.3 93.5 93.6
Starch 31.5 31.7 32.8
NDF 25.1 26.2 25.3
Indigestible Nt>1=5 7.0 6.8 6.7
CF 15.3 16.5 16.3

 

jCOV: intermediate covariate diet, 06: dry ground corn diet, HM:
high moisture corn diet. Values other than DM are expressed as a

percent of dietary DM.

2Protein mix contained 75% soybean meal, 20% bloodmeal, and
5% SoyPlus (West Central Soy, Ralston. IA) on a DM basis.

3Mineral and vitamin mix contained 58.5% DG, 17.0% limestone,
10.5% dicalcium phosphate, 7.9% sodium bicarbonate, 2.1%
magnesium oxide, 1.7% trace mineral premix. 1.6% trace mineral

salt, and 0.8% vitamin A, D, and E premix.

4QLF Dairy TMR 20.

5Estimated after 120 h of in vitro ruminal fermentation.

Data and Sample Collection

Throughout the experiment, cows were housed in tie stalls and fed once daily

(1000 h) at 115% of expected intake. One cow was removed from the experiment

prior to its completion clue to health problems unrelated to the experimental

30

treatments. Cows were blocked from feed between 900 h and 1000 h daily, and
the amount of feed offered and orts were weighed for each cow during collection
periods. Samples of all dietary ingredients (0.5 kg) and orts (12.5%) were
collected daily and composited into one sample per cow, per period. Fecal
samples were collected 3 times at 8 h intervals on d 14 of each experimental
period and were composited into one sample per cow period. Cows were milked
twice daily in a milking parlor, and milk yield was measured daily during the
collection period and was averaged over the collection period. Milk was sampled
at every milking on d 11-14 of each period and 3.5% fat-corrected milk (FCM)
and energy-corrected milk (ECM; Tyrrell and Reid, 1965) yield were calculated.
Body weight and BCS were measured on d 1 of each period and at the
conclusion of the experiment. Body condition was scored by three trained
investigators on five-point scale where 1 = thin and 5 = fat, as described by

Wildman et al. (1982).

Blood samples were collected every 9 hours during (I 12-14 of each period to
represent every 3 h of a 24-hour period. Blood was sampled from coccygeal
vessels and collected into 2 evacuated tubes, one containing potassium EDTA
and the other containing potassium oxalate with sodium fluoride as a glycolytic
inhibitor. Both were centrifuged at 2000 x g for 15 min immediately after sample
collection, and plasma was harvested and frozen at -20°C until analysis.

Samples containing K3EDTA were preserved with benzamidine (0.05 M final

31

concentration), a proteolytic inhibitor that prevents glucagon degradation

(Ensinck et al., 1972; Matsunaga et al., 1997).

Glucose tolerance and proplonate challenge tests. Beginning on d 19 of the
COV period, glucose tolerance tests (G‘l'l's) and propionate challenge tests
(PCTs) were conducted. Infusion tests were administered to blocks of 16 cows
on each day; blocks corresponded to treatment sequence in the crossover
portion of the experiment. All cows were fitted with a single jugular catheter 2 d
prior to the GTT. lndwelling polypropylene catheters (0.24 cm o.d. x 0.17 cm i.d.
tubing, MRE 095, Braintree Scientific, Braintree, MA) were inserted through a 10-
gauge needle until approximately 30 cm of tubing was inside the jugular vein.
Catheter patency was checked daily throughout the experiment. On the day of
the GTT. cows were blocked from feed at 800 h and were not allowed access
until GTTs were completed. Sterile solutions of 50% dextrose (wt/vol) were
administered by intrajugular bolus at a dose of 1.67 mmol glucose/kg BW over
the course of 5 - 10 min. Plasma samples were collected from the jugular vein
10 min prior to infusion, immediately before infusion, and every 10 min through
120 min post-infusion. All time points denoted are relative to the beginning of
infusion. Catheters were flushed with 5 mL of sterile 4.2% Na citrate after
infusion and after each blood sample collection. Propionate challenge tests were
conducted after 1 or 2 d of rest. Sodium propionate (USP grade, Spectrum
Chemical, New Brunswick, NJ) was dissolved in distilled, deionized water at a

concentration of 4.5 mol/L, adjusted to pH 7.4 with NaOH, and filtered (#4 filter,

32

Whatman lntemational, Maidstone, UK). On test days, Na propionate was
infused over a period of 5 - 10 min at a dose of 1.04 mmol/kg BW. Infusion,
sampling, and catheter maintenance were completed as for the G'I‘l', and blood
samples were processed and stored as described for the tailbleeding protocol.
One catheter was lost during both the GTT and PCT, so 31 response curves

were determined for each infusion test.

Liver biopsies. On cl 15 (block 1) and 16 (block 2) of the COV period, liver
biopsies were collected for measurement of transcript abundance for several
proteins involved in gluconeogenesis. After local anesthetization with 2%
Iidocaine hydrochloride, biopsy instruments (14-gauge Vet-Core biopsy needles,
Global Veterinary Products, New Buffalo, Ml) were inserted between the eleventh
and twelfth ribs on a line between the olecranon and the tuber coxae on the right
side. Ten samples of approximately 20 mg were collected, immediately placed in
RNAIater (Ambion Inc., Austin, TX) to inhibit endogenous RNAse activity, and

stored at -80°C until further processing.

Sample and Statistical Analysis

Diet ingredients and orts were dried in a 55°C forced-air oven for 72 h and
analyzed for DM concentration. All samples were ground with a Wiley mill (1-mm
screen; Arthur H. Thomas, Philadelphia, PA). Samples were analyzed for ash,
NDF, indigestible NDF, CP, and starch. Ash concentration was determined after

5 h of oxidation at 500°C in a muffle furnace. Concentrations of NDF were

33

determined (Van Soest et al., 1991, method A). Crude protein was analyzed
according to Hach et al. (1987). Starch was measured by an enzymatic method
(Karkalas, 1985) after samples were gelatinized with sodium hydroxide; glucose
concentration was measured with a glucose oxidase method

(Sigma Chemical Co., St. Louis, MO). lndigestible NDF was estimated as NDF
residue after 120 h of in vitro fermentation (Goering and Van Soest, 1970).
Ruminal fluid for the in vitro incubations was collected from a nonpregnant dry
cow fed alfalfa hay only. lndigestible NDF was used as an internal marker to
calculate apparent digestibilities of DM, OM, starch, and NDF (Cochran et al.,
1986). Concentrations of all nutrients except for DM were expressed as
percentages of DM determined by drying at 105°C in a forced-air oven for more

than 8 h.

Milk samples were analyzed for fat, true protein, and lactose with infrared
spectroscopy by Michigan DHIA (East Lansing, MI). Plasma samples were
analyzed using commercial kits to determine the plasma concentrations of
glucose (Glucose kit #510; Sigma Chemical Co., St. Louis, MO), non-esterified
fatty acids (NEFA C-kit; Wako Chemicals USA, Richmond, VA), triglycerides (kit
TR0100, Sigma Chemical Co.), beta-hydroxybutyrate (BHBA; procedure #2440,
Stanbio Laboratory, Boerne, TX), insulin (Coat-A-Count, Diagnostic Products
Corporation, Los Angeles, CA), and glucagon (Glucagon kit #GL-32K, Linco

Research Inc., St. Charles, MO).

Transcript abundance. RNA was isolated from liver samples according to a
method modified from Chomczynski and Sacchi (1987). Approximately 40 mg of
tissue was homogenized in 1 mL Trizol reagent (lnvitrogen, Carlsbad, CA), 200
uL of chloroform was added to the homogenate, and the mixture'was centrifuged.
The aqueous layer was collected and 500 uL of isopropanol was used to
precipitate RNA. RNA isolates were treated with DNAse (Ambion) to remove any
DNA contamination, and quality was verified by analysis with an Agilent 2100

Bioanalyzer (Agilent Technologies, Palo Alto, CA).

Messenger RNA abundance for phosphoenolpyruvate carboxykinase (cytosolic
form: PCK1), glucose-6-phosphatase (catalytic subunit: G6PC), pyruvate
carboxylase (PC), and pyruvate dehydrogenase kinase 4 (PDK4) was analyzed
by quantitative real-time reverse transcriptase polymerase chain reaction (qRT-
PCR) using a commercial kit (Superscript Ill Platinum Two-Step qRT-PCR Kit,
lnvitrogen). Reverse transcription was conducted using oligo-dT primers with 1
ug total RNA added as template. Following RNase H treatment, the cDNA
product was quantified by spectrophotometer. Real-time PCR was carried out in
duplicate using 1 pg cDNA and was monitored using the ABI Prism 7000
Sequence Detection System (Applied Biosystems, Foster City, CA). Labeled LUX
primers (FAM label) and complementary unlabeled primers (lnvitrogen) were
designed as previously described (Bradford and Allen, 2005). Melting point
analysis confirmed that only the transcripts of interest were amplified during

PCR. Copy numbers for each gene were measured with 2 separate qRT-PCR

35

analyses using standard curves to allow for absolute quantification (Whelan et
al., 2003). Clones used for standards were described previously (Bradford and
Allen, 2005). Messenger RNA abundance was normalized using the geometric
mean of copy numbers of cyclophilin, B-actin, and phosphoglycerate kinase 1 in
each sample (Vandesompele et al., 2002). These control genes were selected
from genes used for normalization in the literature, based on consistent

expression patterns in a preliminary study (unpublished data).

Statistical analysis. Area under the curve (AUC) was calculated by the
trapezoidal rule for infusion test variables by using the mean of the -10 and 0 min
time points as the baseline value for each test. Insulin sensitivity was calculated
from GTT responses using the ratio of maximum insulin : maximum glucose.
Main effects of treatment were analyzed using the fit model procedure of JMP

(version 5.0, SAS Institute, Cary, NC) according to the following model:

Yijkl = II + Pi + Ti 1' Sk + CI(Sk) 4' eijkl
Where Yijk] = dependent variable, p = overall mean, Pi = fixed effect of period (i =
1 to 2), Tj = fixed effect of treatment (j = 1 to 2), 8., = fixed effect of treatment
sequence (k = 1 to 2), C.(Sk) = random effect of cow (l = 1 to 16) within

sequence, 30'” = residual. Residual distributions were visually checked for

normality. Treatment effects were declared significant at P < 0.05, and

tendencies for treatment effects were declared at P < 0.10.

36

Individual differences in response to treatments were of interest in this study, and
regression analyses were used to assess potential predictors of individual
response. To quantify the hypophagic effects of increasing dietary starch
fermentability, the following formula was used: DMI depression = DMI during HM
treatment - DMI during DG treatment. There was a significant effect of period on
DMI, so the parameter estimate for the period term was added to the DMI
depression variable for all cows in one sequence and subtracted from all cows in
the other sequence. For regression analysis, the distribution of Cook’s D
statistics was visually checked and outliers were removed from the analysis. No
more than 3 data points were removed from any regression, and quadratic
regressions were tested for each relationship. To control experiment-wise Type I
error rate, significance was declared at P < 0.01 and tendencies were declared at

P < 0.02 for regression analyses.

RESULTS AND DISCUSSION
The HM diet depressed DMI relative to the DG diet (P < 0.001, Table 2.2). The
depression in DMI of 2.0 kg/d for the more highly fermentable diet was similar to
the 1.7 kg/d depression in DMI found by Oba and Allen (2003b). However, total-
tract starch and DM digestibilities were decreased by DG (P < 0.001 ), and as a
result, digestible DMI was not significantly altered by treatment (P = 0.35).
Consistent with the lack of effect on digestible DMI, neither ECM yield nor BCS

change differed by treatment (Table 2.2).

37

Table 2.2 Effects of corn grain conservation method on intake, digestibility,
and productivity of lactating cows.

 

 

Item DG HM SEM P value
DMI (kg/d) 25.6 23.6 0.6 < 0.001
Total-tract digestibility (% of intake)
DM 74.7 77.4 0.5 < 0.001
OM 75.9 78.8 0.4 < 0.001
Starch 94.0 98.0 0.3 < 0.001
NDF 53.6 52.4 0.9 0.24
DM apparently digested (kg/d) 18.3 17.9 0.4 0.35
Yield (kg/d)
Milk 38.5 38.3 1.5 0.57
FCM (3.5%) 38.3 36.5 1.7 0.17
ECM 38.2 36.6 1.6 0.20
Milk fat 1.34 1.24 0.08 0.15
Milk protein 1.19 1.20 0.05 0.94
Milk lactose 1.80 1.82 0.08 0.78
BW change (kg/d) 0.34 0.40 0.19 0.81
BCS change (I14 (I) 0.05 0.09 0.04 0.52

 

A number of studies have shown increased total-tract starch digestibility with HM
relative to 06 (Huntington, 1997; Knowlton, et al., 1998; Oba and Allen, 2003a).
Earlier harvest of HM can result in decreased vitreousness of starch granules
and increased digestibility relative to DG (Correa et al., 2002). In addition,
ensiling leads to microbial degradation of the protein matrix surrounding starch
granules, increasing fermentability of the starch (Kotarski et al., 1992).
Differences in ruminal fermentability of HM and DG are typically greater than
differences in whole-tract digestibility (Huntington, 1997; Oba and Allen, 2003e),
and HM likely caused substantially greater propionate production than the DG

treatment in this experiment.

38

Several mechanisms have been proposed to explain depression of feed intake
by inclusion of highly digestible starch sources in ruminant diets. Propionate has
been implicated in feed intake regulation in ruminants because intraruminal
propionate infusion depresses energy intake (Oba and Allen, 2003d). and portal
infusion decreases DMI (Anil and Forbes, 1980). The expected increase in
propionate production by HM in this experiment may have caused the observed
decrease in DMI. However, other hypotheses have been advanced which point
to reactive oxygen species (lllius et al., 2002) or implicitly identify total energy
requirements (NRC, 2001) as factors limiting feed intake. These hypotheses
have a number of shortcomings; there is currently no proposed mechanism by
which whole-body oxidative stress can be relayed to the brain to influence satiety
or hunger, and models which propose that animals eat to their energy
requirements ignore the fact that diet can alter energy requirements (Harvatine
and Allen, 2006). Nevertheless, these hypotheses, like the propionate oxidation
hypothesis, predict a decrease in DMI with HM. Therefore, these results do not

provide strong support for any proposed mechanism of feed intake regulation.

Plasma Metabolites and Hormones

The HM treatment significantly increased plasma glucose concentration (P <
0.01, Table 2.3). Plasma NEFA (P < 0.001) and triglyceride (P = 0.02)
concentrations were also increased by HM, despite the fact that plasma insulin

concentration was not significantly altered by treatment (P = 0.39). The reason

39

for this unusual result is not obvious. Plasma BHBA (P = 0.58) and glucagon (P

= 0.55) concentrations were not different between treatments.

Table 2.3 Effects of corn grain conservation method on blood
plasma metabolites and hormones.

 

 

Item DG HM SEM P value
Glucose (mg/dL) 58.7 60.6 0.6 < 0.01
NEFA (qu/L) 45.5 54.1 2.1 < 0.001
Triglyceride (mg/dL) 2.9 3.3 0.1 0.02
BHBA (mg/dL) 7.6 7.8 0.3 0.58
Insulin (ulU/mL) 9.5 10.5 0.8 0.39
Glucagon (pg/mL) 189 185 9 0.55

 

Variability in DMI Responses to Altered Starch Fermentabiilty

In addition to assessing the effects of dietary starch fermentability across
animals, this experiment was designed to find animal characteristics that predict
individual DMI response to increased starch fermentability. Individual variability
in DMI response to starch fermentability was great, ranging from an increase of
0.9 mm to a decrease of 6.0 kg/d. DMI depression variables were compared to
independent pre-trial (COV) variables, including plasma hormone and metabolite
concentrations, production characteristics, infusion test responses, and hepatic
transcript abundance. Most COV variables were not related to DMI depression,
however, significant relationships with 2 insulin variables were observed (Table
2.4). Plasma insulin concentration during the COV period was inversely
correlated with DMI depression (r = -0.53, P < 0.01), indicating that cows with
high plasma insulin concentrations prior to treatment initiation experienced
greater depression of DMI when starch fermentability was increased (Figure

2.1). Insulin response index following glucose infusion was positively related to

40

DMI depression (r = 0.57, P < 0.01), suggesting that cows with greater insulin

secretory response to glucose were better able to maintain feed intake on highly

fermentable diets (Figure 2.2).

Table 2.4 Pearson correlation coefficients for dry matter

intake depression and potential pre-trial response

 

 

predictors.1
Variable r
Milk production (3.5% FCM) 0.24 0.19
Days in milk 025 0.17
Body condition score 001 0.95
Plasma glucose concentration -0.04 0.82
Plasma NEFA concentration 0.05 0.78
Plasma BHBA concentration -0.17 0.37
Plasma insulin concentration -0.53 < 0.01
Insulin : glucagon ratio -0.24 0.21
Plasma leptin concentration -0.10 0.62
GTT glucose AUC 0.01 0.95
GTT maximum glucose -0.36 0.06
GTT insulin AUC 0.45 0.03
GTT maximum insulin 0.37 0.07
GTT insulin response index 0.57 < 0.01
PCT glucose AUC -0.13 0.49
PCT maximum glucose -0.25 0.19
PCT insulin AUC 0.12 0.54
PCT maximum insulin 0.14 0.47
Relative PCK1 mRNA abundance 0.08 0.69
Relative PC mRNA abundance 0.18 0.35
Relative G6PC mRNA abundance 0.10 0.58
Relative PDK4 mRNA abundance -0.10 0.60

 

IDMI depression data is corrected for differences in treatment

sequence.

41

 

 

 

DMI depression (kg/d)
‘i’

 

 

'7 I I I

o 5 1i) 15 20
COV insulin (ulU/mL)

Figure 2.1 Pre-trial plasma insulin concentration (COV insulin) predicts DMI
depression resulting from an increase in dietary starch fermentability. DMI
depression is calculated as (HM - DG) DMI, corrected for period effect. Two

outliers were removed prior to regression analysis. R2 = 0.28, P < 0.01, n =
29.

 

 

 

 

 

1 - Q

o. . . ‘
'P I
\ O
a ' '
V 2d 0
C
.9
1” O
In 3-
9 o o T
8 o
D '4'
2
o .5-

.6-1

'7 *0 I I I I I

10 20 30 40 50 60 70

Insulin response Index (ulU insulin / mg glucose)

Figure 2.2 Insulin response index from glucose tolerance tests predicts DMI
depression in response to increased dietary starch fermentability. DMI
depression is calculated as (HM — DG) DMI, corrected for period effect.
Insulin response is quantified as the ratio of maximum insulin : maximum
glucose. Two outliers were removed prior to regression analysis; other

individuals are not included because of missing G'l'l' time points. R2 = 0.33,
P< 0.01, n = 24.

 

42

 

 

High plasma insulin concentrations in the COV period may be indicative of
adequate nutritional status and may provide negative feedback on hepatic
gluconeogenesis. This relationship is consistent with our hypothesis that
decreased use of propionate for glucose production leads to greater propionate
oxidation and decreased DMI. Individual cows with an adequate supply of
glucogenic precursors may respond to a further increase in supply by decreasing

DMI.

However, the positive relationship between GTT insulin response index and DMI
depression seems to refute this proposed explanation. Cows that responded to
increased plasma glucose concentration with greater insulin secretion (at 10 - 20
min post-infusion) were better able to maintain DMI on the more fermentable diet.
If negative feedback on gluconeogenesis by insulin were involved in intake
depression by HM, an inverse correlation would be expected. However, the
multiple, complicated effects of insulin on nutrient homeostasis make this
conclusion less clear. It is possible that cows with strong responses to increased
plasma glucose concentration are able to clear nutrients from the bloodstream
more quickly after meals, potentially decreasing intermeal interval (Oba and
Allen, 2000). Although mean daily plasma insulin concentration and insulin
response to glucose infusion are clearly unique measures (correlation: r = 0.06, P
= 0.78), the reason for their respective relationships with DMI response is less

obvious.

43

Despite indications that measures related to insulin predicted responses to diet
fermentability, the status of insulin as a long-term regulator of feed intake in
lactating ruminants is questionable. Although strong relationships between
adiposity and insulin concentrations have been reported in the biomedical
literature (Bagdade et al., 1967; Polonsky et al., 1988), we have found no
evidence of a similar relationship in lactating dairy cattle. Body condition score
measured during the COV period was highly correlated with plasma leptin

concentration (Bradford et al., 2006b), but was unrelated to plasma insulin

concentration (insulin [pmol/L] = 47.6 + 11.0 - BCS, R2 = 0.03, P = 0.31). These

and other findings (Meikle et al., 2004) suggest that in mature cows, insulin does
not serve as an adiposity signal for long-term control of feed intake.
Nevertheless, Insulin may impact feed intake in the short-term through its

influences on hepatic metabolism and nutrient partitioning.

There are likely a number of reasons that so few relationships with DMI
depression were found. One drawback to the use of dietary treatments is that
diet adaptation periods are required; as a result, COV variables were determined
35 - 42 d prior to period 2 data collection. Variables such as COV plasma NEFA
concentration or hepatic transcript abundance may have had no relationship with
metabolic state 42 (I later. The length of the experiment also prevented the use
of early lactation cows, because their responses tend to change dramatically by
period; this may have contributed to the lack of a relationship with days in milk.

More surprising, perhaps, is the fact that measures of glucose and insulin

responses from the propionate challenge test were not related to DMI
depression; Oba and Allen (2003c) found a clear relationship between plasma
glucose concentration and DMI response to intraruminal propionate infusion.
However, we recently reported that jugular administration of propionate
stimulates glycogenolysis, indicating that changes in plasma glucose
concentration during the PCT do not reflect increased gluconeogenesis
exclusively (Bradford et al., 2006a). In addition, insulin responds to changes in
both propionate and glucose concentrations during the propionate challenge test,
adding noise that may mask differences that were observed during the GTT.
Nevertheless, the fact that simple measures such as body condition score,
production level, and plasma glucose concentration did not predict DMI response
suggests that the mechanisms causing DMI depression on highly fermentable

diets may involve interactions of metabolic factors.

CONCLUSIONS
Increasing the fermentability of starch in a high-concentrate dairy ration
depressed DMI but did not have significant effects on digestible DMI or yield of
milk or milk components. Pre-trial plasma insulin concentration was correlated

with the depression of DMI induced by increasing dietary starch fermentability.

45

CHAPTER 3

Bradford, B. J., M. Oba, R. A. Ehrhardt, Y. R. Boisclair, and M. S. Allen. 2006.
Propionate is not an important regulator of plasma leptin concentration in dairy
cattle. Domest Anim Endocrinol. 30(2):65-75.

46

 

Available onlIne at www.5ciencedirect.com

 

DOMESTIC
sci-non nun-c1:-
@ ANIMAL
-- - ~ ENDOCRINOLOGY
ELSEVIER Domestic Animal Endocrinology 30 (2006) 65-75 ——-—-—

 

www.joumals.elsevierhealth.com/periodicals/dae

Propionate is not an important regulator of plasma
leptin concentration in dairy cattle‘i

Barry J. Bradford“, Masahito Oba“, Richard A. Ehrhardtb,
Yves R. Boisclairb, Michael S. Allen 31*

' Department of Animal Science. Michigan State University, 22656 Anthony Hall,
East Lansing. MI 48824, USA
b Department of Animal Science. Cornell University, Ithaca. NX USA

Received 11 April 2005; received in revised form 25 May 2005; accepted 30 May 2005

 

Abstract

Propionate was recently shown to increase leptin synthesis in rodents. To determine if a similar
effect occurs in ruminants, propionate was administered to lactating dairy cows. In experiment 1,
31 cows were given an intrajugular Na propionate bolus (1040 umol/kg body weight), increasing
plasma propionate from 160 to 5680 p.M and plasma insulin from 6.8 to 77.8 uJU/mL. Plasma leptin
concentration decreased from 2.11 ng/mL before bolus to 1.99 ng/mL after dosing (P<0.05) with
no differences in leptin concentrations at 20, 50, and 100 min post-bolus (P>0.10). In experiment
2, 12 cows were used in a duplicated 6 x 6 Latin square experiment to assess the dose—response
effect of ruminal propionate infusion on plasma leptin concentration. Sodium propionate was infused
at rates of 0, 260, 520, 780, 1040, or 1300 mmol/h, while total short-chain fatty acid infusion rate
was held constant at 1300mmol/h by addition of Na acetate to the infusate. Coccygeal blood was
sampled following 18h of infusion. Increasing the rate of propionate infusion linearly increased
plasma propionate concentration from 180 to 330 ItM (P<0.001) and plasma insulin concentration
from 6.7 to 9.1 uIU/mL (P<0.05). There was a quadratic response in plasma leptin concentration
(P=0.04) with a maximum at 780 mmol/h propionate, but leptin concentrations increased by no
more than 8% relative to the 0mmth propionate infusion. Leptin concentrations were correlated

{5' Presented in part at Experimental Biology 2005. San Diego, 2-6 April 2005 [Bradford Bl. Allen MS, Oba
M, Ehrhardt RA, Boisclair YR. Propionate does not stimulate leptin secretion in lactating dairy cows. FASEB J
2005;19: Abstract #6229].

‘ Corresponding author. Tel.: +1 517 432 1386; fax: +1 517 432 0147.
E-mail address: allenm@msu.edu (M.S. Allen).

0739-7240/5 - see front matter © 2005 Elsevier Inc. All rights reserved.
doi: 10. 1016/ j.domaniend.2005.05.007

47

66 3.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65—75

with insulin concentrations but not with propionate concentrations in plasma. Propionate is not a
physiological regulator of leptin secretion in lactating dairy cows.
© 2005 Elsevier Inc. All rights reserved.

Keywords: Leptin; Propionate; Dairy cow; Adipose tissue

 

1. Introduction

Leptin, a peptide hormone produced primarily by adipocytes, plays a central role
in the regulation of food intake. Leptin was originally proposed to be the long-sought
feedback signal from adipose tissue to hypothalamic nuclei regulating long-term energy
intake [1]. Subsequent research has largely validated this hypothesis, providing strong
evidence that leptin can alter energy intake by modulating the hypothalamic production of
orexigenic (NPY and AGRP) and anorexigenic (POMC and CART) peptides [2]. However,
leptin’s proposed roles have now expanded to the regulation of metabolism, growth, and
reproduction [3]. Indeed, recent evidence has called into question leptin's role as a purely
lipostatic signal. The ability to modulate leptin secretion through short-term nutritional [4]
and environmental changes [5] suggests that leptin can also convey metabolic information
to the central nervous system [6]. .

Short-chain fatty acids (SCFA) are derived primarily from microbial fermentation and
provide roughly 70% of the energy requirements in ruminants [7]. While SCFA represent
only 6-10% of caloric intake in humans, there is evidence that SCFA may be responsible
for some of the health benefits of high-fiber diets [8], possibly through modification of
gene expression patterns [9]. Recently, several groups have suggested that SCFA may
enhance leptin expression and secretion in anterior pituitary cells [10] and adipocytes [l 1].
Furthermore, jugular infusion of propionate increased leptin mRNA abundance in ovine
adipose tissue [12] and oral administration of propionate in mice increased plasma leptin
concentration [ 1 3].

The objective of this study was to determine if an intrajugular sodium propionate bolus or
continuous intraruminal infusion of Na propionate would alter plasma leptin concentrations
in lactating dairy cattle.

2. Materials and methods

Experimental procedures were approved by the All-University Committee on Animal
Use and Care at Michigan State University and cows were selected from the Michigan State
University Dairy Teaching and Research Center herd.
2.1. Experiment 1

Animals, experimental design, and diets for experiment 1 have been described [14].

Thirty-two multiparous lactating Holstein cows (121 :I:48d in lactation, 41.5 :I: 7.8 kg/d
milk yield, 675 :t 69 kg body weight; mean :1: SD.) were offered a single diet (Table I)

48

8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65-75 67

 

 

Table 1
Nutrient composition of experimental diets (percent of dietary dry matter)
Experiment 1 Experiment 2
Organic matter 93.3 93.9
Starch 31.5 27.4
Neutral detergent fiber 25.1 30.0
Crude protein 15.3 18.1

 

for 22 d, and propionate challenge tests (PCT) were conducted over 2 (I. At least 3 d prior
to the PCT, indwelling polyurethane catheters (MRE 095, Braintree Scientific, Braintree,
MA) were inserted in a single jugular vein of each animal. On each infusion day, feed was
removed at 08:00 h, and PCI‘ were initiated for 16 cows over 2h. Catheter patency was
lost for one cow prior to the PCT, so a total of 31 PCI‘ were conducted. Sodium propionate
(USP grade, Spectrum Chemical, New Brunswick, NJ) was dissolved in distilled, deionized
water at a concentration of 4.5 mol/L, adjusted to pH 7.4 with NaOH, and filtered (#4 filter,
Whatman lntemational, Maidstone, UK). Sodium propionate was infused into the jugular
vein (1040 umol/kg body weight) over the course of approximately 8 min. Jugular blood
samples were collected 10min prior to infusion, immediately before infusion, and every
10 min after the initiation of infusion until 120 min post-infusion. Samples were collected
by an automated blood collection system [15], and catheters were flushed with a solution
of 4.2% Na citrate following infusion and after each sampling.

2.2. Experiment 2

Experiment 2 has been described previously [16]. Briefly, 12 multiparous lactating Hol-
stein cows (36.4:l: 6.5 kg/d milk yield, 6403: 63 kg body weight) with ruminal cannulas
were assigned to early lactation (9 :1: 6d in lactation) and mid lactation (192 :I: 17 d in
lactation) blocks in a duplicated 6 x 6 Latin square design. Treatments were intraruminal
infusions of 0, 260, 520, 780, 1040, or 1300 mmol Na propionate/h beginning 6h before
feeding and continuing for 18 h. Rate of total SCFA administration was held constant at
1300 mmol/h by adding required amounts of Na acetate. One cow in each block was removed
from the experiment due to adverse reactions to the treatments, resulting in a total of 66
infusions.

For each stage of lactation, cows were adapted to the experimental diet (Table I) for 14 d.
Blood samples were then collected from coccygeal vessels every 9 h for 3 d and composited
for determination of plasma leptin concentrations prior to treatment initiation. Treatments
were administered on alternate days for the final 1 1 d of the experiment. On treatment days,
cows were blocked from feed for 4 h prior to feeding, but had access to feed during the rest
of the infusion. Blood samples were collected from coccygeal vessels at the conclusion of
each 18-h infusion.

2.3. Analytical procedures

Blood samples were placed on ice immediately after collection and plasma was
harvested and frozen at —20 °C until analysis. Plasma leptin concentrations were

49

68 8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65—75

determined by a radioimmunoassay specific to bovine leptin [17]. The intra-assay
coefficient of variation was 4.1% and the inter-assay CV was 2.6% for leptin analysis.
Insulin concentrations were determined by commercial radioimmunoassay (Coat-A-Count,
Diagnostic Products Corporation, Los Angeles, CA). Intra- and inter-assay CV's for
insulin analysis were 7.8 and 10.1%, respectively. Plasma propionate and acetate concen-
trations were determined by HPLC according to the method described for ruminal fluid
[18].

For experiment 1, plasma propionate and insulin values were log-transformed to
normalize the distribution of the data and were subsequently analyzed by the mixed model
procedure of SAS (version 8.0, SAS Institute, Cary, NC). The mixed model included
fixed effects of infusion day and sample time and the random effect of cow nested within
day. An autroregressive (AR [1]) covariance structure was assigned to the time variable.
Orthogonal contrasts were used to determine whether differences from pre-infusion values
were significant, and propionate and insulin values were back-transformed following
analysis.

Leptin data from experiment I and all data from experiment 2 were analyzed by the fit
model procedure of JMP (version 5.0, SAS Institute, Cary, NC) using the REML method.
The mixed model for experiment 1 included fixed effects of infusion day and sample time
and the random effect of cow within day; differences between sample times were tested by
Tukey’s HSD. The mixed model for experiment 2 included the fixed effects of treatment
and period and the random effect of cow. Linear and quadratic effects of propionate infusion
rate were tested for each dependent variable. Stage of lactation was included in the original
model, but was dropped after establishing the absence of a significant main effect and
interaction effects. In addition to these analyses, correlations among several dependent
variables were investigated. For these correlations, the distribution of Cook’s D statistic
was plotted and as many as three outliers were removed from the analysis before the final
correlation was calculated.

3. Results
3.1. Experiment 1

Intrajugular bolus of Na propionate significantly increased plasma propionate concen-
tration 10 and 20 min after the initiation of infusion (both P < 0.001), but not at subsequent
sampling times. Jugular propionate concentrations reached a mean of 5680 11M at 10 min
post-infusion compared to 160 p.M prior to infusion (Fig. 1). Propionate administration also
stimulated insulin secretion, increasing venous insulin concentration from 6.8 p.1U/mL pre-
infusion to a peak of 77.8 uIU/mL at 10 min post-infusion. Plasma insulin concentrations
remained higher than pre-infusion values through the 40 min sampling time (all P <0.001,
Fig. 1). Plasma leptin concentrations averaged 2.11ng/mL immediately prior to infusion,
decreased to 1.99 ng/mL at 20 min post-infusion, and remained lower than baseline concen-
trations at 50 and 100 min post-infusion (SEM. = 0.09, P < 0.05, Fig. 1). Pre-infusion leptin
concentrations were positively correlated with body condition score (r=0.6l, P<0.001,
Fig. 2), as reported previously [17].

50

3.]. Bradford 2! al. / Domestic Animal Endocrinology 30 (2006) 65—75 69

6000- ***

3000-

Plasma propionate 04M)

2000-

I000‘

**

Insulin (plU/mL)

Leptin (nglmL)
g

 

0 2'0 AI) 60 Iii) 100 I20
Time (min)
Fig. 1. Responses to intrajugular bolus of Na propionate. Propionate was infused at the rate of 1040 “mol/kg body

weight over approximately 8min beginning at 0min. 'P<0.05 compared to pro-infusion values; "'P<0.001
compared to pre-infusion values.

51

70 8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65—75

3.5 -

 

Plasma leptin (ng/mL)

 

 

I T

1.5 i 2.3 3 3.5 4
Body condition score

.4

Fig. 2. Preoinfusion leptin concentrations are positively correlated with body condition score (r=0.61, P< 0.001,
experiment I).

3.2. Experiment 2

Plasma acetate concentration decreased quadratically with increasing propionate infu-
sion rate (and decreasing acetate infusion rate) from 6.3 to 0.9 mM in early lactation cows
and from 2.1 to 0.8 mM in mid-lactation cows. Therefore, it was necessary to verify that
Na acetate was an appropriate control for this experiment. Pre-treatment leptin concentra—
tions were compared to leptin concentrations following the 0 mmol/h propionate treatment
(1300 mmol/h acetate), and no differences were found (P=0.99). We conclude that Na
acetate infusion does not influence leptin production.

Plasma propionate concentrations increased linearly with the amount of propionate
infused (P<0.001), increasing from a baseline of 180 uM to a maximum of 330 pM at
the highest propionate infusion rate (Fig. 3). Plasma insulin concentrations also increased
linearly with propionate infusion rate (P< 0.05) from 6.7 to 9.1 pIU/mL. While the rate of
propionate infusion had no linear effect on plasma leptin concentrations (P >0.15), there
was a significant quadratic effect (P =0.04, Fig. 3; Y: 2.95 + 9.55 x Io-5x - 3.86 x 10-7
(X — 650)2). Plasma leptin concentration was not related to plasma propionate concentra-
tion (P>0.15), but was positively correlated with plasma insulin concentration (r=0.45,
P<0.001). To determine if this relationship simply represented long-term regulation of
leptin by insulin, the changes in plasma insulin, leptin, and propionate concentrations from
baseline values were calculated for each cow period. The change in plasma leptin concen-
tration was positively correlated with the change in plasma insulin concentration across
infusions (r =0.30, P = 0.02), but not with change in plasma propionate concentration.

4. Discussion

Recent evidence has suggested that propionate can alter gene expression in adipose tis-
sue. Lee and Hossner [12] administered 1920 umol propionate/kg body weight in sheep

52

8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65—75 71

350-

N a w
8 o 8
r l 1

Plasma propionate 04M)
9'

6-1

Insulin (ulU/mL)

 

Leptin (ng/mL)

 

 

I I l T f I
200 400 600 800 1000 1200
Propionate infusion rate (mmol/h)

0.1

Fig. 3. Responses to continuous intraruminal Na propionate infusion. Sodium propionate was infused at rates of
0, 260, 520, 780, 1040, or 1300 mmol/h, with total SCFA infusion rate held constant at 1300 mmol/h by addition
of Na acetate to the infusate. Coccygeal blood was sampled following 18h of infusion. Increasing the rate of
propionate infusion linearly increased plasma concentrations of propionate (P< 0.001) and insulin (P < 0.05) and
quadratically altered leptin concentrations (P<0.05).

53

72 8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65-75

during a 30 min intrajugular infusion and measured increased mRNA abundance for several
genes in subcutaneous adipose tissue, including leptin. In mice, 500 pmol of propionate
administered via gavage increased plasma propionate concentrations from 25 to 250 11M
and resulted in an 80% increase in plasma leptin concentration [13]. Several groups have
identified pathways mediating the effects of propionate in adipose tissue [11,13]. The G-
protein coupled receptor GPR41 binds SCFA, particularly propionate, and is expressed
primarily in adipose tissue in humans [1 1]. Recent sequencing of the bovine genome has
identified a sequence (413563485, http://www.ncbi.nih.gov/Tracesl) with a translated pri-
mary structure that shares 74% homology with a portion of human GPR41 protein (81%
coverage). This finding is consistent with the possibility that propionate could regulate
leptin gene expression via the bovine GPR41 ortholog.

The data obtained in these two experiments suggests that propionate plays a minor
role, if any, in regulating plasma leptin concentrations in lactating dairy cows. In our
first experiment, venous propionate concentrations peaked at more than 30 times baseline
concentrations, although they decreased rapidly following this peak. Because leptin mRNA
abundance was greatly increased just 2h after initiation of propionate infusion in the
sheep [12], it is likely that any stimulation of plasma leptin concentration by propionate
would have been observed by 100 min post-infusion. However, it is possible that the
very short-term effect on peripheral propionate concentrations prevented significant
stimulation of leptin expression. Our second experiment addressed this possibility by
administration of SCFA during 18 h intraruminal infusions. Although there was a significant
quadratic response of plasma leptin to increasing rate of propionate infusion, plasma
leptin concentration was increased by no more than 8% and did not produce a linear
response.

It is possible that stimulation of leptin secretion requires peripheral propionate concen-
trations to be elevated to a greater extent than achieved with intraruminal infusions and for
longer periods than in the intrajugular infusion experiment. If this were the case, however,
propionate would not be expected to stimulate leptin secretion in any physiological situ-
ation. Hepatic uptake of propionate is extremely efficient in most species, with extraction
efficiencies generally reported in the range of 90—100% in dairy cows [19]. As a result,
despite the absorption of large amounts of propionate in dairy cows, venous propionate
concentrations are relatively stable and are typically in the range of 70—150 11M. Propionate
concentrations exceeding our treatment mean of 330 [AM (1300mmth Na propionate,
experiment 2) would not occur under normal feeding conditions, and would likely indicate
impaired liver function.

Another possibility is that peripheral concentrations of propionate in well-fed cows are
adequate to constitutively stimulate GPR41 receptors. Baseline venous propionate concen-
trations of 160 uM in the cows in experiment 1 are approximately 50-fold higher than those
in humans; healthy individuals in three studies had a mean propionate concenu’ation of
3 uM [20—22]. Oral gavage of propionate stimulated leptin secretion in mice by increasing
plasma propionate concentrations from 25 to 250 1.1M [13], suggesting that a near-maximal
response might be achieved at 160 uM propionate. The reported EC50 for propionate stimu-
lation of leptin secretion is 200 11M in primary cultures of murine adipocytes [13], however,
differences in the primary structure of murine and bovine GPR41 may alter binding kinetics
[11].

8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65-75 73

One weakness of in vivo experiments implicating propionate in regulation of leptin
secretion [12] is that propionate is a potent insulin secretagogue in ruminants [23]. Insulin’s
effects on leptin secretion have been demonstrated in dairy cows [24], and in experiment 2,
plasma leptin concentrations were significantly related to insulin, but not propionate con-
centrations. Insulin may alter transcription of the leptin gene by stimulating nutrient uptake
and metabolism in adipocytes [25], as evidenced by the stimulatory effect of malonyl-CoA
on leptin production in murine adipocytes [26]. However, increased leptin secretion follow-
ing propionate infusion in rodents is probably not mediated by insulin because propionate
does not stimulate insulin release in non-ruminants [27].

Several studies have considered the effects of fasting and refeeding on plasma leptin
concentrations in cattle [28,29]. While propionate infusion might be expected to have a
greater impact on leptin secretion in fasted animals, confounding effects on insulin, growth
hormone, epinephrine, and other hormones would be an even greater problem in fasted
animals. Furthermore, evidence that propionate can stimulate leptin secretion in rats fed ad
libitum [ I 3] led us to hypothesize that a similar response might occur in lactating cows. There
was no evidence that leptin response to propionate infusion was greater for early lactation
cows in experiment 2 (interaction P = 0.91), despite the fact that early lactation cows were in
negative energy balance. Therefore, we expect that the direct effect of propionate on leptin
secretion (independent of insulin, etc.) would be minor even in fasted animals.

Xiong et al. [13] suggested that hypophagia induced by infusion or dietary supplemen-
tation of propionate might be mediated by increased leptin production. Oba and Allen [16]
reported that the infusions described here in experiment 2 linearly decreased metaboliz-
able energy intake during the final 12 h of infusion (including energy provided by infused
SCFA) as propibnate infusion rate increased. The lack of a corresponding increase in leptin
concentration refutes the hypothesis that leptin mediates the hypophagic effects of propi-
onate in lactating dairy cattle. Potential alternative mechanisms for this response have been
discussed elsewhere [30].

In conclusion, our findings do not support the hypothesis that propionate is an important
regulator of plasma leptin concentration in lactating dairy cattle. Although propionate may
have some influence on leptin production through its role as an insulin secretagogue, it is
unlikely that hypophagic effects of propionate are mediated primarily through leptin.

Acknowledgements

This project was supported in part by National Research Initiative Competitive Grant no.
2004-35206- 14167 (to MSA) and 2000-35206-9352 (to YRB) from the USDA Cooperative
State Research, Education, and Extension Service, and by a National Science Foundation
Graduate Research Fellowship (to BIB).

References

[1] Frederich RC, Lollmann B, Hamann A, Napolitano-Rosen A, Kahn BB, Lowell BB, et al. Expression of ob
mRNA and its encoded protein in rodents. Impact of nutrition and obesity. 1 Clin Invest 1995;96:1658—63.

55

74 8.]. Bradford et al. / Domestic Animal Endocrinology 30 (2006) 65-75

[2] Baskin DG, Hahn TM, Schwartz MW. Leptin sensitive neurons in the hypothalamus. Horm Metab Res
1999;31:345-50.

[3] Harris RB. Leptin—much more than a satiety signal. Annu Rev Nutr 2000;20:45—75.

[4] lngvartsen KL, Boisclair YR. Leptin and the regulation of food intake, energy homeostasis and immunity
with special focus on periparturient ruminants. Domest Anim Endocrinol 2001;21:215-50.

[5] Puerta M, Abelenda M, Rocha M, Trayhum P. Effect of acute cold exposure on the expression of
the adiponectin, resistin and leptin genes in rat white and brown adipose tissues. Horrn Metab Res
2002;34:629-34.

[6] Block SS, Butler WR, Ehrhardt RA, Bell AW, Van Amburgh ME, Boisclair YR. Decreased concentra-
tion of plasma leptin in periparturient dairy cows is caused by negative energy balance. J Endocrinol
2001;171:339—48.

[7] Bergman EN. Energy contributions of volatile fatty acids from the gastrointestinal tract in various species.
Physiol Rev 1990;70:567-90.

[8] Bingham SA. Mechanisms and experimental and epidemiological evidence relating dietary fibre (non-starch
polysaccharides) and starch to protection against large bowel cancer. Proc Nutr Soc 1990;49:153-71.

[9] \Vrlliarns EA, Coxhead lM, Mathers lC. Anti-cancer effects of butyrate: use of micro-array technology to
investigate mechanisms. Proc Nutr Soc 2003;62:107-15.

[10] Yonekura S, Senoo T, Kobayashi Y. Yonezawa T, Katoh K, Obara Y, et al. Effects of acetate and butyrate on
the expression of leptin and short-form leptin receptor in bovine and rat anterior pituitary cells. Gen Comp
Endocrinol 2003;133:165—72.

[l I] Brown Al, Goldsworthy SM, Barnes AA. Eilert MM. Tcheang L, Daniels D, et al. The Orphan G protein-
coupled receptors GPR41 and GPR43 are activated by propionate and other short chain carboxylic acids. J
Biol Chem 2003;278:11312—9.

[12] Lee SH, Hossner KL. Coordinate regulation of ovine adipose tissue gene expression by propionate. J Anim
Sci 2002;80:2840—9.

[l3] Xiong Y, Miyamoto N, Shibata K, Valasek MA, Motoike T, Kedzierski RM, et al. Short-chain fatty acids
stimulate leptin production in adipocytes through the G protein-coupled receptor GPR41. Proc Natl Acad Sci
USA 2004;101:1045-50.

[14] Bradford Bl, Allen MS. Milk fat responses to a change in diet fermentability vary by production level in
dairy cattle. J Dairy Sci 2004;87:3800—7.

[15] Allen MS, Oba M, Mooney CS. Automated system for collection of ruminal fluid and blood of ruminants. J
Dairy Sci 2000;83(Suppl l):288.

[l6] Oba M, Allen MS. Dose-response effects of intrauminal infusion of propionate on feeding behavior of
lactating cows in early or midlactation. J Dairy Sci 2003;86:2922—31.

[l7] Ehrhardt RA, Slepetis RM, Siegal-Willott J, Van Amburgh ME, Bell AW, Boisclair YR. Development of
a specific radioimmunoassay to measure physiological changes of circulating leptin in cattle and sheep. J
Endocrinol 2000;166:519—28.

[18] Oba M, Allen MS. Effects of corn grain conservation method on feeding behavior and productivity of lactating
dairy cows at two dietary starch concentrations. 1 Dairy Sci 2003;86:174—83.

[19] Benson JA, Reynolds CK, Aikman PC. Lupoli B, Beever DE. Effects of abomasal vegetable oil infusion on
splanchnic nutrient metabolism in lactating dairy cows. 1 Dairy Sci 2002;85:1804—14.

[20] Cummings JH, Pomare EW, Branch W1, Naylor CP, Macfarlane GT. Short chain fatty acids in human large
intestine, portal, hepatic and venous blood. Gut 1987;28:1221-7.

[21] Scheppach W, Richter F, Joeres R, Richter E, Kasper H. Systemic availability of propionate and acetate in
liver cirrhosis. Am J Gastroenterol 1988;83:850-3.

[22] Dankert l, Zijlstra JB, Wolthers BG. Volatile fatty acids in human peripheral and portal blood: quantitative
determination vacuum distillation and gas chromatography. Clin Chim Acta 1981;110:301-7.

[23] Leuvenink HG, Bleurner El, Bongers LJ, van Bruchem .1, van der Heide D. Effect of short-term propionate
infusion on feed intake and blood parameters in sheep. Am J Physiol 1997;272zE997-1001.

[24] Leury Bl, Baumgard LH, Block SS, Segoale N, Ehrhardt RA, Rhoads RP, et al. Effect of insulin and growth
hormone on plasma leptin in periparturient dairy cows. Am J Physiol 2003;2852R1107-15.

[25] Moreno-Aliaga Ml, Stanhope KL, Havel Pl. Transcriptional regulation of the leptin promoter by insulin-
stimulated glucose metabolism in 313-11 adipocytes. Biochem Biophys Res Commun 2001;283:544—8.

56

8.]. Bradford er al. / Domestic Animal Endocrinology 30 (2006) 65-75 75

[26] Shirai Y. Yaku S, Suzuki M. Metabolic regulation of leptin production in adipocytes: a role of fatty acid
synthesis intermediates. J Nutr Biochem 2004;15:651-6.

[27] Horino M, Machlin U, Hertelendy F, Kipnis DM. Effect of short-chain fatty acids on plasma insulin in
ruminant and nonruminant species. Endocrinology 1968;83:118-28.

[28] Amstalden M, Garcia MR, Williams SW, Stanko RL, Nizielski SE, Morrison CD, et al. Leptin gene expres—
sion, circulating leptin. and luteinizing hormone pulsatility are acutely responsive to short-term fasting
in prepubertal heifers: relationships to circulating insulin and insulin-like growth factor 1(1). Biol Reprod
2000;63:127—335.

[29] Chelikani PK, Ambrose JD, Keisler DH, Kennelly .11. Effect of short-term fasting on plasma concentrations
of leptin and other hormones and metabolites in dairy cattle. Domest Anim Endocrinol 2004;26:33-48.

[30] Allen MS, Bradford 81. Harvatine K]. The cow as a model to study food intake regulation. Annu Rev Nutr
2005;25:523-47.

57

CHAPTER 4

Bradford, B. J. and M. S. Allen. 2005. Phlorizin administration increases hepatic
gluconeogenic enzyme mRNA abundance but not feed intake in late-lactation
dairy cows. J Nutr. 135(9):2206-2211.

58

 

 

Nutrient Metabolism

 

 

Phlorizin Administration Increases Hepatic Gluconeogenic Enzyme mRNA
Abundance but Not Feed Intake in Late-Lactation Dairy Cows”

Barry J. Bradford and Michael S. Allen“

Department of Animal Science, Michigan State University, East Lansing, MI 48824

ABSTRACT Gluconeogenic capacity may be an important factor regulating dry matter intake (DMI) in lactating
dairy cows. To determine whether increased glucose demand affects feed intake and hepatic gene expression.
lactating Holstein cows were treated with phlorizin or vehicle (propylene glycol) for 7 d. Multiparous cows (n = 12;
269 t 65 d in milk, mean I SD) were randomly assigned to treatment sequence in a crossover design and were
adapted to a common diet for 7 d before the beginning of the experiment. Phlorizin injected s.c. at 4 g/d caused
glucose excretion in urine at the rate of 474 g/d. Although phlorizin decreased lactose synthesis and milk
production (both P < 0.01). DMI and 3.5% fat-corrected milk production were not altered by treatment. A net deficit
of 383 g glucose/d in milk and urine for phlorizin (relative to control) was likely replaced partially through increased
gluconeogenesis. The molar insulin:glucagon ratio was decreased 17% by phlorizin (P < 0.001) and hepatic
phosphoenolpyruvate carboxykinase. glucose-G-phosphatase. and pyruvate carboxylase mRNA abundance in-
creased (all P < 0.05). Late-lactation dairy cows adapted quickly to an increase in peripheral glucose demand;
adaptation mechanisms likely included enhanced gluconeogenic capacity, whereas DMI was not altered. J. Nutr.

135: 2206-2211, 2005.

KEY WORDS: . dairy cows . phlorizin . glucose demand . gluconeogenesis . pyruvate carboxylase

Dry matter intake (DMI)S is regulated by a wide range of
signals related to the intake of specific nutrients, gut disten-
tion, adiposity, cephalic stimulation. and other factors (1).
Forbes (2) suggested that these signals are integrated. driving
feeding behaviors that minimize discomfort to the animal.
Although this signal integration likely occurs in the hypothal-
amus, there is a large body of evidence indicating that some
signals that regulate DMI originate in the liver and are deliw
ered via the hepatic vagus (3.4). These signals are related to
the energy status of the liver: increased hepatic ATP concen-
trations coincide with the ends of meals in rats (5).

Propionate is an important fuel for ruminants because it is
the primary substrate for glumneogenesis in lactating cows fed
highly fennentable diets. Its importance in intake regulation

was demonstrated by Oba and Allen (6). who showed that

‘ Prewnted in part at the 10th International Symposium on Ruminant Phys-

. Aug. ISO-Sept. 4. 2004. Copenhagen. Denmark [Brale 8. J. 8 Allen.

M. S. (2004) Increasing glucose demand increases hepatic pyruvate caboxv

ylase mRNA concentration but not feed intake ‘n late—lactation dairy cows J.
Ann. Feed Sci. 13 (suppl. 1): 377].

3 Supported in part by National Research initiative Competitive Grant no.
2004 45206-14167 from the US. Department of Agriculture Cooperative State
Research, Education. and Extensnon Service. and by a National Sconce Foun-
dation Graduate Research Fellowship,

3 Supplemental Table 1 and Supplemental Figure 1 are available as Online
Supporting Material with the onltne posting of "“5 paper at www.mtntionorg.

‘ To whom correspondence should be addressed. E-rnatl: allenms‘thuedu.

5Abbreviations used: BHBA. B-hydroxybulyrale; DMI. dry matter intake:
GGPC. glucose-G-pnosphatase catalytic subunit: PC. pyruvate carbOxylase:
PCK1. cytosolic phosphoenolpyruvate caboxykinase; PDK4. pyruvate dehydro-
genase kinase 4: qRT-PCR. quantitative RT-PCR; SL037A4. glucose-s-phospha-
tase translocase.

0022-3166/05 $8.00 ‘0 2005 American Society for Nutritional Sciences.

intraruminal infusions of propionate linearly decreased metab—
olizable energy intake (including both feed and infusions)
Compared with isomolar infusions of acetate. The liver was
likely responsible for these effects because hepatic vagotomy
eliminated hypophagic responses to portal infusions of propi—
onate (7). These treatments have physiological importance for
the regulation of meal size because propionate flux to the liver
increases rapidly during meals (8).

if meal size and intermeal interval are dependent on the
energy status of the liver. the intake effects of propionate
should depend on the extent to which it stimulates hepatic
oxrdation. Therefore, increasing the flux of propionate to
glucose tnight decrease its hypophagic effects while simulta-
neously providing more glucose for the mammary gland and
other peripheral tissues. This mechanism, mediated by changes
in insulin and glucagon secretion, may provide the link be«
tween increases in peripheral energy demand and increased
DMI. Although this relation has been established in a variety
of circumstances (9—11), well-controlled mechanistic studies
in ruminants have not been conducted.

To test the effects of increased glucose demand on meta-
bolic regulation and DMl. we used phlorizin to cause irrevers'
ible glucose loss in urine (12). We hypothesized that increased
peripheral glucose demand would cause upregulation of glu-
coneogenic pathways, and that increitsed gluconeogenic ca«
pacity would increase DMl by increasing meal size. Therefore.
the objective of this study was to measure responses in hepatic
mRNA abundance, milk yield. DMI, feeding behavior, and
plasma honnones and metabolites during phlort:in treatment.
Although several studies have measured the effect 0fplll0l’121n

Manuscript received 4 March 2005. Initial review completed 2.8 April 2005. Revision accepted 27 May 2005.

59

GLUCOSE DEMAND AND NUTRIENT METABOLISM 2201

on lactating cows (I 3.14). only one has considered its effecr on
DMI. and that study used only 4 cows with treatment periods
of Z d (14). To our knowledge. there have been no reports
evaluating the effects of phlorizin on expression of metabolic
genes in bovine liver.

MATERIALS AND METHODS

Experimental procedures were approved by the All-University
Committee on Animal Use and Care at Michigan State University.

Design and treatments. Multiparous Holstein cows (ii = 12; 269
I 65 d in milk; 30.7 t 5,0 kg/d iiiilk yield, 2.6 t 0.7 lactations, mean
I 51)) were selected from the Michigan State Universtty Dairy
Cattle Teaching and Research Center and were randomly assigned to
treatment sequence in a crossover design. l’hlorizin, an inhibitor of
renal glucose reabsorption (12). was administered via s.c. injection at
the rate of 4 g/d. with propylene glycol .is vehicle and control.
Treatment periods were 7 d, and injections were given every 6 h
during these periods. The cows were adapted to a single diet for a 741
period before the first treatment period. and a 7-d rest period was
included between the 2 treatment periods. The experimental diet
(41‘) g dry matter/kg diet) contained dry, finely ground corn grain.
corn Silage. .ilfalfa silage. a premix of protein supplements. and a
premix of minerals and vitamins (Table I). The diet was formulated
for 250 g/kg neutral detergent fiber and 160 g/kg crude protein.

Data and sample collection. Throughout the experiment, cows
were housed in tic-stalls and fed once daily (1110 h) at 115% of
expected daily intake. Cows were not allowed access to feed from
1000 to 1110 h. during which time orb and the amount of feed
offered were weighed for each Cow daily. During the final 4 d of each
treatment period. feeding behavior was monitored by a computerized
data acquisition system (15) throughout the day. Data on chewing
activities. feed disappearance. and water consumption were recorded
for each cow every 5 s. and mean daily values for number of meal
bouts. interval between meals. meal size, and time spent eating were
calculated. One cow was removed froin feeding behavior analysis
because equipment malfunction prevented collection of satisfactory
chctving behavtor data. On each of these 4 (1. samples of all dietary
ingrulients (0 5 kg) and orts (12.5%) were collected. and on d 7 of
treatment. fecal samples were collected 1 times at 8-h intervals. Cows
were milked twice daily in the milking parlor during rest periods and
in the tie stalls during treatment periods. Milk yield was recorded and
samples were taken at each milking during the 4 collection days.

 

 

TABLE 1
Ingredients and nutrient composition of experimental diet
Component Content
g/kg dry matter
Diet ingredient
Dry ground com 382
Corn silage 264
Alfalfa silage 238
Protein mix1 94
Mineral and vitamin mix’ 22
Nutrient composition
Organic matter 939
Starch 346
Neutral detergent fiber 278
lndigestible neutral detergent fiber“ 113
Crude protein 138

 

‘ Contents (g/kg dry matter): soybean meal. 750; blood meal. 200;
SoyPlus (West Central Soy). 50.

2 Contents (per kg dry manor): Na. 81 9: Cl, 41 9: Ca. 31 g: P. 26 9:
Mg. 23 9: Fe. 3.2 9; Mn, 2.4 9; Zn. 2.3 9; K, 2 g; S. 2 9; Cu. 574 mg; l.
34 mg; Se. 16 mg; Co, 10 mg; vitamin A. 14 mg; vitamin D, 330 [.192
vitamin E. 113 mg.

3 Measured after 240 h of In vitro ruminal fermentation.

60

Urine was collected for a 24-h period starting on d 4 (1100 h) of
each experimental period. Urinary catheters (Bardcx Lubricath Folcy
MFR. Baird Medical) were inscrtul and urine was collected into a
container with 0.2) mol HLIl to prevent glycolysts. Volumes were
mcztsurcd at IZ'h intervals and samples were taken and frozen until
analysis. According to the data of Amanilvl’hillips ct al. (14). phlo-
rizin injected every 6 h causes glucose excretion .it a constant rate.
indicating that a Single 24-h collection provided a valid estimate of
daily glucose excretion. During the same 24-h period, blood samples
were collected hourly front indwelling jugular catheters. Collected
blood was immediately emptied into 2 tubes. one containing potasv
slum EDTA and the other containing potassium oxalate wrth sodium
fluoride as a glycolytic inhibitor (Vacutaincr. Becton Dickinson).
Both were centrifuged at 2000 X g for 15 min immediately after
sample collection. and plasma was harvested and frozen at ~20°C
until analysis. One plasma sample from each tube containing
K.EI)TA was preserved With benzamidine (0.05 mol/l- final concen-
tration). a protcolytic inhibitor that prevents glucagon degradation
(16.17).

Liver biopsies were collected from phlori:in-treatcd Cows imme-
diately before the first injection on d l of each treatment period
(control sitriiple) and shortly after the final injection .it the end of d
7 (treated sample). After local anesthetization with 2% Iidocamc
hydrochloride. biopsy instruments (l4-gauge ch'Core biopsy nee-
dles, Global Veterinary Products) were inserted between the l lth and
12th ribs on a line between the olecranon and the tuber coxae on the
right “le; 10 samples of ~20 mg were Collected and immediately (<5
s) frozen in liqutd nitrogen; the 3.00 mg sample wme stored at —80°C
until further processing.

Sample analysis. Diet ingredients. orts, and fecal samples were
dried in a 55°C forcedvair oven for 72 h and analyzed for dry matter
concentration. All samples were ground with a Wiley mill (l—mm
screen; Author H. Thomas). Samples were analyzed for ash, neutral
detergent hber. indigestible neutral detergent fiber. crude protein. and
starch. Ash concrntmtion was determined after 5 h of oxidation at
500°C in a muffle furnace. Ncutnil detergent fiber was analyzed
according to Van Soest ct .il. [(18). method A] lndigestible neutral
detergent fiber was IIlL‘Z'NIrL‘t'I as neutral detergent fiber residue after
240 h of in Vitro fermentation (19). Ruminal fluid for the in vitro
incubations was collected from a nonpregnant dn' Cow fed alfalfa hay
only. Crude protein was analyzed according to Hach ct al. (20).
Starch was measured by an enzymatic method (21 ) after samples were
gelatinized with sodium hydroxide: glucose concentration was mea-
sured using the glucose. otridase method. Concentrations of all nutriv
cuts were expressed as percentages of dry matter determined from
drying at 105°C in a forced-air oven. lndigestible neutral detergent
fiber was used as an internal market to calculate total—tract digest-
ibility of other nutrients.

Milk samples were analyzed for fat. true protein, and lactose with
infrared spectroscopy by the Michigan Hairy Herd Improvement
Association. Urinary nitrogen was quantifiul in diiplicatc by Dumas
combustion using it commercial analyzer (Lli(,fO I'll-2000). Urine and
plasma samples were analyzed in duplicate for glucose content by the
glucose oxtdasc method. I‘Lisma samples were analyzed in duplicate
using commercial kits to determine Concentrations of FFA [NEFA
C-kit; Wako Chemicals. as modified (22)]. fi-hydrovautynite
(BHBA; procedure 32440. Stanbio Laboratory), insulin (Coat-A-
(Iount. Diagnostic Products). and glucagon (Glucagon kit 'GLJZK.
Linco Research). Plimna Lrlactate content was quantified with a
clinical analyzer (YSI 1500. 1'81 Life Sacnccs).

Total RNA was isolated from liver tissue using a commercml kit
(Ribol‘urc. Ambion) and samples were trcatul with DNasc to remove
any DNA coiitairiination. “)1: quality of .ill RNA isolates was verified
by analysis with an Agilent 3100 Bioanalyu'r (Agilent Technologies).
mRNA abundance for PhiISPINX'HUIPVI'UVANC carboxykintue (cytosolic
form: PCK1. EC 4 1.1.32). gliicosc—b-phosphatase (EC 5.1.1.9, cata—
lytic subunit: (ibl‘C. regulatory subunit: SL(.')7A4). pyruvate car—
boxylase (PC. EC 6.4.1.1). and pyruvate dehydrogenase kinase 4
(PDK4) was analyzed by quantitative real-time RT-l‘klR (qRT-I’CR)
using a coinincrcuil kit (Superscript lll l‘latinum TwovStcp qRT-I‘CR
Kit, lnvitrogen). Reverse transcription was conducted using oligovdT
primers With 1 pg total RNA added as a template. After RNase H

m BRADFORD AND ALLEN

treatment. the cDNA product was quantified by spectrophotometer.
ReaLtime PCR was carried out in duplicate usnng 1 pg cDNA and
was monitored using the ABI Prism 7000 Sequence Detection System
(Applied Biosystems). Labeled LUX primers (FAM label) and com-
plementary unlabeled primers (lni itrogen) were LleSignL-d with online
LUX Designer software (23) and included at 200 nmol/L in the PCR
mix (Supplemental Table l) Melting point analysis conhrmed that
only the transcripts of intetrLst were amplified during PCR Copy
numbers for each genL we re measured with 2 sepmrite LiRTv PCR
analyses using standard curves to allow for absolute quantification
(34) Clones used for standards were as follows. PC Kl. bPEPC K-
C 2000 [(25). a gift from Dr. Shawn Donkin. Purdue University. West
Lafayette. IN]: GOPC. GenBAinlt accession (13465348 (a gift from
Dr. Tim Smith. USDA Meat Animal Research Center. Clay Center.
NE); SLC37A4. accession IBFO75927; PC. accession J‘BFbO-iMO:
PlilK‘l. accession IBE589I94 (all from the Center for Animal Func-
tional Genomics. Michigan State University. East Lansing, Ml).
Clones were cultured. plasmids were harvested (Wizard Plus Mini-
preps. Promega). and copy numbers were quantified by SPL‘CUOPhO'
rometer for serial dilution. mRNA abundance was normalized using
the geometric mean of copy numbers of cyclophilin. Bvactin, and
phosphoglycerate kinase 1 in each sample (26) Control genes were
selected from genes used for normalization in the literature. based on
consistent expression patterns in a preliminary study (unpublished
data) Quantification of copy numbers for control was was carried
out in the manner LlLscribed above, with the follow 'ing clones and for
standards. cyclophilin, accession '80690429; B—actin. accessmn
#3068903}: phosphoglycerate liinase l. accession 3B06L‘Wi981 (all
from the Center for Animal Functional Genomics. Michigan State
University, East Lansing. Ml).

Statistical analysis. Data were analyzed by the fit model proce—
dure of JMP (vemon 5.0. SAS Institute) using the REML method

according to the following model:
Yutzu‘l'P,+T’+(:E+PT"+(’”‘

where Y a. is a dependent van iblL u is the overall mean. P is the
fixed effect of period (i = l to 2), T is the fixed effect of m A-itant
(} = l to 2). C. is the random effect ofcow (k -— l to [2). PT is the
interaction of period and treatment. and e“ is the residual error.
Plasma analyses were conducted using the above model. but also
included fixed effects of sample time. Sample time was a significant
factor for all plasma variables except glue Migon mRNA abundance
was analy:ed with a model that included fixed effects of treatment
and qRT-PCR run and random effect of cow. Values for PCKI.
GoPC. PC. and PDK4 mRNA abundimce were log-tninsformed for
analysis. and values reported here are bile'IRIDSftlrmt‘tl. Although
statistical analysis of mRNA .ibiindanCe was carried out using values
for copy numbers/pg cDNA. abundance is reported relative to control
values for ease of interpretation. For all main eftects. significance was
declared at P < 005. and tendencies were declared at P < 0.10.
There was a tendency for an interaction between period and treatv
merit for plasma glucagon concentration (P = 0.14). but not for other
variables of interest (all P > 0.15).

RESULTS
Glucose loss, milk production, and intake. Phlorizin

treatment caused urinary excretion of 474 g glucose/d (Table
2). equivalent to 40% of daily lactose secretion in this group
of cows. Milk yield was depressed by treatment (P = 0.001)
primarily because of an 8% decrease in milk lactose production
with phlorizin (P < 0.01). Milk fat and 3.5% fatvcorrected
milk yield were not altered by treatment. although milk pro—
tein production tended to decrease with phlorizin (P = 0.08).
Although the response in milk lactose yield was significant.
the decrease of 90 ng accounted for a relatively small propor'
tion of the 474 g/d of glucose lost in urine from phlorizin
treatment. Contrary to our hypothesis. DMI was not altered by
treatment (Table 2). nor was digestible DMl. Plilori:in did not

61

TABLE 2

Effects of phlorizin treatment on glucose loss. feed intake,
and milk production in lactating Holstein cows'

 

 

Item Control Phlorizin SEM P<
Urinary glucose excretion. g/d 1 474 25 0.001
Dry matter intake. kg/d 20.7 20.2 0.8 0.37
Digestible dry matter intake. kg/d 13.6 13.2 0.5 0.34
Yield, kg/d
Milk 24.9 22.8 1.4 0.001
3.5% lat-corrected milk 27 0 26.2 1.5 0.18
Swirls-corrected milk 24.7 23.7 2.0 0.05
Milk lat 1.01 1.01 0.06 0.99
Milk protein 0.82 0.75 0.06 0.08
Milk lactose 1.18 1.09 0.07 0.01

 

‘Valuos are least-squares means : SEM. n = 12.

alter feeding behavior as quantified by meal size and the
number of meals/d (data not shown).

Plasma metabolites and hormones. Plasma glucose con-
centration was decreased by phlorizin (P < 0.001) and insulin
concentration tended to decrease (P = 0.09). whereas lacrate
concentration was greater during phlorizin treatment (P
< 0.001. Table 3). Concentratiom of FFA and BHBA were
also increased by phlorizin (P < 0.001). comment with a
decrease in fat deposition and increased hepatic oxidation of
fatty acids. Similarly. the increase in plasma glucagon concen-
tration (P = 0.02) and the. decreased insulintglucagon ratio (P
< 0.001) during phlorizin treatment suggest that gluconeo—
genesis was enhanced by phlorizin. Phlorizin also tended to
merease urinary N excretion from 135.7 to 151.6 g/d (P < 0.10).

Hepatic mRNA abundance. Expression patterns for genes
involved in regulation of hepatic gluconeogenesis were altered
by phlorizin (Fig. l). Phlorizin increased PC mRNA abun-
dance by 66‘36 (P < 0.001) and abundance of PCK1 mRNA by
29% (P < 0.05); in addition. GbPC mRNA abundance was
increased by 42% during phlorizin treatment (P < 0.001).
SLC37A4 or PDK4 mRNA abundance was not affected.

DISCUSSION

The evidence available in this study indicates that glu—
coneogenic flux was likely greater in phlorizin-treated cows.
The enzymes encoded by PCK1 and GoPC catalyze reactions
that are considered to be potentially rate-limiting steps in
gluconeogenesis (27.28); thus. the increased mRNA abun-
dance of these genes likely led to enhanced gluconeogenic
capacity. Given the effects on hepatic transcript abundance.
hormone profile. and the relatively small decrease in milk
lactose production. phlorizin treatment likely increased he—
patic glucose export. However. the net increase in glucose
production cannot be calculated because th‘ile—lxxly glucose
oxidation was not measured. Phlorizin treatment decreased the
proportion of plasma glucose oxidized to plasma CO; by 34%
in steers (29). and some decrease in glucose oxidation probably
resulted from phlorizin treatment in this study as well.

Regulation of transcription and/or mRNA stability is an
important component of PC regulation. because Greenfield et
al. (30) demonstrated that mRNA abundance was highly
related to PC activity in bovine liver blopSlcs. The expression
of PC is regulated by both insulin and glucagon (31). and
short-term infusions of glucagon increased PC transcript abunv
dance ()2) and enzyme acliVIty (3}) in ruminants. Therefore.
in this study. increased glucose demand probably increased

GLUCOSE DEMAND AND NUTRIENT METABOLISM 2209

TABLE 3

Effects of phlorizin on blood plasma metabolites and
hormones in lactating Holstein cows‘

 

 

item Control Phlorizin SEM P <
Glucose. mmol/L 3.62 3.54 007 0001
Lactate. mmol/L 0.37 0.44 0.02 0.001
FFA “mol/L 70.0 1001 11.8 0.001
BHBA. “mol/L 624 807 38 0.001
Insulin. pmol/L 94.7 87.5 9.3 0.09

Glucagon. pmot/L 36.9 37.9 2.3 002

Insulin:glucagon ratio. mol/mol 2.48 2.06 0.38 0001

 

‘ Values are least-souaros means : SEM. n = 12.

transcription of PC by causing a decrease in plasma glucose
Concentration and a corresponding decrease in the insulin:
glucagon ratio (Table 3). The Significant increase in PC
mRNA abundance in response to increased glucose demand by
phlorizin also agrees With recent reports in transition dairy
cows with incremed glucose demand (30.34) and suggests that
the liver responds to increased glucose demand in part by
limiting OXIdallOn of intracellular pyruvate.

Pyruvate carboxylase is generally recognized as an impop
rant enzyme for the pnxluction of glucose from lactate and
quCogenic amino aCIds (50.35). However. its importance in
regulating the flux of propionate to glucose is not generally
discussed. Wlth notable exceptions (36.37). For net oxidation
of propionate to occur. it must be converted to pyruvate
(through oxaloacetate and phi)sphoenolpyruvate). oxidized by
pyruvate dehydrogenase. and directed toward the tricarboxylic
ac id cycle as acetyl-CoA (see Supplemental Fig. 1). In one of
the few studies in which this pathway was studied in lactating
cows. Black et al. ()8) calculated that “-13% of pmpionate
was Converted to acetyl-(IoA. Although this seems to be a
relatively insignificant proportion of the propionate entering
the liver. it represents the only net loss of glucogenic carbon
front propionate in the bovine liver; all other propionate that
is not utilized in gluconeogenesis is converted to lactate or
glucogenic amino aCids. Enhancing PC acuvity. therefore.
should not only increase gluconeogenesis from lactate and
amino acids. but also increase the proportion of propionate
that can be utilized for glucose production. This may be
especially true because of the moment nature of propionate
absorption (8); during meals. propionate is rapidly taken up by
the liver. and propionate influx may exceed the liver's capaCity
for gluconeogenesis. If this build—up causes an increase in
pyruvate concentration in hepatocytes. the relative activities
of PC and pyruvate dehydrogenase are important in determin-
ing the proportion of glucogenic substrates that are preserved.

Past research using phlorizin indirectly supports these as—
sertions. Veenhuizen et al. (29) Used phlorizin to increase
irreversible loss of glucose by 36% and measured the end-
products of propionate metabolism. Phlorizin significantly in-
creased the proportion of proplonate that was utilized for
gluconeogenesis. both With and Without supplemental dietary
propionate. Similarly. the relative proportion of propionate
Used for gIUCose production vs. C03 production was thft’ibt‘d
in hepatocytes harvested from phlorizin'treated wethers (39).
Although no enzyme activity or gene expression data are
available from these studies. it is possible that enhanced PC
activity was responsible for increasing the efficiency of glu’
Coneogenesis from propionate.

Although phlorizin treatment increased the demand for
glucose specifically. adaptations to the treatment included

62

altered metabolism of fatty acids as well as glucose and its
precursors. Phlorizin increased plasma FFA concentration.
which likely increased hepatic FFA uptake because the liver
takes up FFA in proportion to its concentration in plasma
(40). The increased uptake of FFA likely resulted in greater
hepatic fatty acid oxidation because plasma BHBA Concen-
tration was increased by phlorizin. Although lipolysis contribl
Utes only small amounts of carbon for gluconeogenesis (glyc-
erol). hepatic fatty acid oxidation plays an important role in
supporting gluconeogenesis (41). Acetyl-CoA and NADH
produced by oxidation of fatty aCIds provide energy and re
ducing equivalents to drive gluconeogenesis and allosterically
inhibit pyruvate dehydrogenase acrivuy. which increases the
proportion of pyruvate retained for glucose production. There—
fore. pyruvate dehydrogenase activity was likely decreased by
phlorizm. despite the fact that mRNA abundance of the
regulatory protein PDK4 (42) was not altered by treatment.
Work by Chow and Jesse (43) indicated that gluconeogenesis
from propionate was decreased when fatty and oxidation. and
therefore acetyl-CoA and NADH accumulation. was limited
by tetradecylglycidic acid. The authors were surprised by this
result because acetyl-ChA does not activate any of the en.
zymes in the direct gluconeogenic pathway from propionate
through phiisphoenolpyruvate. However. carbon cycling
among oxaloacetate. phosphoenolpynivate. and pyruvate (see
Supplemental Fig. 1) could explain this surprismg result be—
cause of the allosteric effects of acetyl—CoA and NADH on PC
and pyruvate dehydrogenase (41).

Plasma lactate concentrations were increased by phlorizin
in this study; however. this does not indicate whether glucose
production from lactate was altered in phlorizin-treated cows.
Baird et al. (44) found that lactate tumover through the Cori
cycle was decreased in lactating cows that were food deprived
for 4 d. even though plasma lactate concentrations nearly
doubled. Although net hepatic uptake of lacrate increases with
increasmg demand for glucose production in ruminants (45).
the majority of this lactate is apparently not used for glucose
production because it accounted for only 7% of total glucose
production in early—lactation cows (46). Fatty acid oXidation
inhibits the conversion of lactate to pyruvate by decreasing the

— E

O
(a

Relative mRNA Abundance

 

 

FIGURE 1 Effects of phlorizin administration In lactating Holstein
cows on hepatic mRNA abundance relative to control tar genes In-
volvod in gluconeogenesis. Values are means : SEM. n = 12. 'Ditterent
from control. P < 0.05; mdifferent from control. P < 0.001.

2210 BRADFORD AND ALLEN

NADNADH ratio. and nonessential amino acids are probably
more important sources of gluconeogenic substrate in cases of
increased glucose demand (39). in this study. the tendency for
an increase in urinary N excretion during phlorizin treatment
suggests that increased PC activity may have increased glucose
production from nonessential amino acrds.

Despite the indirect evrdence that gluconeogenic capacity
was increased (and the possibility that propionate oxidation
was decreased) by phlorizin treatment. voluntary DMI did not
increase. Although this result is contrary to our hypothesis. the
treatment effect on hepatic fatty acid oxidation makes it
impossible to reject the hypothesis. A decrease in propionate
oxrdation was expected to increase DMI only if it delayed
prandial increases in hepatic energy status. In this study,
however. enhzuiced hepatic fatty acrd oxrdation could have
replaced an energy deiicrt created by a decrease in propionate
oxrdation.

The decision to use late-lactation cows in this study was
based on the assumption that DMI is limited primarily by
metabolic factors rather than distention of the reticulorumen
when cows are producing at less than the peak of lactation
(47). In Cows whose intake is limited primarily by rumen
distention. the energy status of the liver likely has less effect
on DMI. However. considering the results of the present study,
other animal or experimental models may be more appropriate
for testing the effects of increased glucose demand on DMI.
Early—lactation Cows generally mobilize large amounts of fat
from adipose tissue. making it possrble that a decrease in the
tnsulin:glucagon ratio would have little additional effect on
lipolysis. Maintaining a constant rate of lipolysis across treat—
ments would likely prevent hepatic fatty acid oxrdation from
confounding the effects of glucose demand on DMI. An alter
native model would be to use late-lactation cows and to
administer tetradecylglyCidic acid. an inhibitor of camitine
palmitoyltransferase. during both phlorizin and control injec—
tions. The drawback to this model is that tetradecylglycrdic
acid would limit fatty acrd oxidation in muscle and adipose
tissue in addition to the liver. Nevertheless, the strength of the
evrdence relat ing meal patterns to the energy status of the liver
in rodents (48) encourages continued investigation of this
potential mechanism of DMI regulation in cattle.

ACKNOWLEDGMENTS

The authors thank R. E. Kreft. T. H. Herdt, R. A. Longuski, l). G.
Math. Y. Ying. C. S. Mooney. J. A. Voelker. R. J. Tempelman. and
S. P. Suchyta for technical assistance; SS. Drink-in. S. Koser. Tl‘. L.
Smith. and K. S. Tennill for help in acquiring cDNA clones; and
West Central Soy for donation of SoyPlus protein supplement.

LITERATURE CITED

1. Allen. M. 8.. Bradford. 8. J. a Harvatine. K. J. (2005) The cow as a
model to study food intake regulation Annu. Rev. Nutr. 25: 523—547.

2. Forbes. J. M. (1996) Integration of regulatory signals controlling forage
intake in rurninants. J. Anim. So. 74. 3029-3035.

3. Langhans. W. a Scharrer. E. (1987) Evidence for a vagalfy mediated
satiety signal derived from hepatic fatty acid Oxidation. J. Auton. Nerv. Syst. 18:
13-18.

4. Scharrer. E. (1999) Control of food intake by fatty acid Oxidation and
ketoger‘iesrs. Nutrition 15: 704 -714.

5. Koch. J. 5.. J1, H.. Osbaklien. M. D. 8 Friedman. M l. (1998) Temporal
relationships between eating behavtor and liver adenne nucleotides in rats
treated With 2.5-AM. Am. J. Physrol. 274: R610—9617.

6. Oba. M. 5 Allen. M. S. (2003) lntrarumnal infusion of propionate alters
feeding behawor aid decreases energy intake of lactating dairy cows. J. Nutr.
133: 1094—1099.

7. Anil. M. H 8 Forbes. J. M. (1988) The roles of hepatic nerves in the
reduction of food intake as a consequence of intraportal sodium propionate
administration in sheep. O. J. Exp. Physrol. 73: 539-546.

8. Benson. J. A. Reynolds, C. K.. Aikman. P. C.. Lupoll. 8. A Beaver. D. E.

63

(2002) Effects of abomasal vegetable oil infusion on splanchnic nutrient metab-
olism in lactating dairy cows. J. Dairy SCI. 85: 1804-1814.

9. Blundell. J. E. Stubbs. R. J., Hughes. 0. A. Whybrow. S. a King. N. A
(2003) Cross talk between physrcal actrvrty and appetite control: does physacal
activity stimulate appetite? Pioc. Nutr. Soc. 62: 651-661.

10. Dohoo, I. R. Leslie. K.. DesColeaux. t... Fredeen. A. Bowling, P . Weston.
A at Shewfelt. W. (2003) A meta-analyse review of the effects of recombinant
bowne somatotropin. 1. Methodology and effects on production. Cm. J. Vet. Rec.
67; 241—251.

11. Johnson. M. 8.. Thomson. 3 C. & Speakmm. J. R. (2001) L'mrts to
sustained energy intake. I. Lactation in the laboratory mouse Mus musculus. J.
Exp. Biol. 204: 1925—1935.

12. Young. J. w- Schmidt. 8. P.. Akowuah. E. 3.. H855. G. S. & McGilliard.
A D. (1974) Effects of phlorizin on glucose Imetics 'n the bowne. J. Dairy 30.
57: 689—694.

13. Rutta. L. M 8 Manns, J. G. (1988) Follicular phase gonadotroon
secretion in cyclic postpartum bee1 cows wrth phlonzm-induced hypoglycemia. J.
Anim. SCI 66:1194-1200.

14. Amara-Phillips. D M..McGilliard.A 0..Lindberg. G. L..Veenhurzen. J. J.
8 Young. J. W. (1993) Effects of decreased availability of glucose for day
cows. J. Dairy Set. 76: 752—761.

15. Dado. R. G. 8 Allen. M. S. (1993) Continums computer acquisition of
feed and water intake. chewng reticular motility. and mm pH of cattle. J. Davy
Scr. 76: 1589—1600.

16. Ensmck. J. W.. Shepard. C.. Dudl. R. J. 8 Williams. R. H. (1972) Use
of benzamidine as a proteotytic inhibitor in the radiommunoassay of glucagon in
plasma. J. Clin. Endocrinol. Metab. 35: 463-467.

17 Matsunaga. N.. Goka. T., Nam. K. T.. Oda. S.. Ohneda. A. a Sasalri. Y.
(1997) inhibition of GH releasing 1actor (ORB-induced 0H secretion by intraru-
mnal mus-on of volatile fatty aCrds (VFA) in sheep. Endocr. J. 44; 133-140.

18. Van Soest. P. J. Robertson. J. B. a Lewis. B. A (1991) Methods for
dietary fiber. neutral detergent fiber. and nonstarch potysaccharides in relation to
animal nutrition. J. Dairy SCI. 74: 3583—3597.

19. Goering. H. K. 5 Van Soest, P. J. (1970) Forage Fiber Analysis (Ap-
paratus. Reagents. Procedures. and Some Applications). Agric. Handbook No.
379. APE/USDA Washington. DC.

20. Hach. C. C.. Bowden. B. K.. Lopolove. A B. s Brayton. S. V. (1987)
More powerful perm-do Kieldahl digestion method. J. Assoc. 011. Mal. Chan. 70.
783—787.

21. Karkalas. J. (1985) An mprovod enzymatic method for the determi-
nation of native and modified starch. J. Scr. Food Agric. 36: 1019-1027.

22. McCutcheon. S. N. it Bauman, D. E. (1986) Effect of chronic growth
hormone treatment on responses to epinephrine and thyrotropn-releasing hor-
mone in lactating cows. J. Dairy Sci. 69: 44 —51.

23. LUX Fluorogenic Primers. www.mvrtrogenconmux. [Last accessed Jm.
6. 2005].

24. Whelan. J. A. Russell. N. B. & Whelan. M. A (2003) A method for the
absolute quantification of cDNA using real-time PCR. J. lmmunol. Methods 278;
261—269.

25 Agca. C.. Greenfield. R. 8.. Hartwell. J. R. a Donlon. S. S. (2002)
Cloning and characterization of bovne cytosolic and mitochondrial PEPCK doing
transition to lactation. Physrol. Genormcs 11: 53—63.

26. Vandesompele. J.. De Prefer. K.. Pattyn. P. Poppa. 8.. Van Roy, N.. 09
Paepe. A 8 Speleman. F. (2002) Accurate normalization of real-time quanti-
tatrve RT-PCR data by geometric averaging of multiple internal control genes.
Genome Biol. 3: RESEARCHOOCM.

27. Pilkis. S. J. 6 Granner. D. K. (1992) Molecular physiology of the
regulation of hepatic gluconeogenesis and glycotySis Annu. Rev. Physot 54;
885 909

28. Nordl'ie. R. C..Foster. J. D.&Lange.AJ. (1999) Regilationofgbcose
production by the lNef. Annu. Rev. Nutr. 19: 379—406.

29 Veenhiiizen. J. J.. Russell, R. W. at YWfig, J. W. (1988) Knetics of
metabolism of glucose. propionate and CO; in steers as affected by meeting
phlorizin and feeding propionate. J. Nutr. 118: 1366—1375.

30. Greenfield. R. 8.. Cecava M. J. a. Doniun. S. S. (2000) Oiariges in
mRNA expressron for gluconeogenic enzymes in irva of dairy cattle during the
transmon to lactation. J. Dairy SCI. 83: 1228-1236.

31. Jitrapakdee. S. a Wallace. J. C. (1999) Structure. functm UK! rogu-
latlon of pyruvate carboxylase. Biochern. J. 340 (Pt. 1): 1—16.

32. She,P..LindDerg.G. L. Hippen.A R..Beitz.D.C.&Young.J.W. (1999)
Regulation of messenger ribonucleic ac-d expresson for gluconeogenic enzymes
during glucagon irifusrons into lactating cows. J. Dairy SCI. 82. 1153—1163.

33. Brockman. R. P. 8 Manns, J. G. (1974) Effects of glucagon on xtiv-
"rites of hepatic enzynm in sheep. Comell Vol. 64: 217-224

34. Hartwell. J. Ft. Cecava. M. J. & Donirin. S. S. (2001) Roman undo
gradable protein. men-protected choline and mRNA expression for enzymes in
gluconeogeneSrs and ueagenesis in periparturient dairy cows. J. Dairy Scr. 84:
490-497.

35 Looney. M. C. Baldwn. R. L. &Calvart, C. C. (1987) Guconeogenesls
n isolated lamb hepatocytes. J. Anim. Sei. 64: 283—294.

36. Arrrientano. L. E. (1992) Runinant hepatlc metabolism of volatile fatty
acrds. lactate and pyruvate. J. Nutr. 122; 838-842.

37. Sternhou. W. D. 8 Barman. D. E. (1988) Propionate metabolism; 3
new interpretation. in: Aspects of Digestive Phys-ology in Run-nants (Dobson. A
a Dobson. M. J.. eds). pp. 238—256 Comstoctc Ithaca. NY.

38. Black. A L.. Luck. J., Mollar. F. 6 Anand, R. S. (1966) Pyruvate and

GLUCOSE DEMAND AND NUTRIENT METABOLISM 2211

propionate metabolism in lactating cows. Effect of butyrate on pymvate metab-
olism. J. Biol. Chem, 241: 5233—5237.

39. Ovaton T. R. Dracldey. J. K.. OttemanAbWte. C. J. Beam A 0..
ErmtatLS.&Cla‘KJ.H. (19%) Substratemilmtmtorhepanc ‘
Bdtaedbymeasedglmsedemmd hmtsJ. mm Sci. 77: 1940-1951.

40. Emery. Ft. 8. & Herdt. T. H. (1991) Lipid nutntion. Vet. Clin. Noah Am.
Food Anim. Pract. 7: 341-352.

41. GonzalezManchon. C.. Ma'tln-Rmuero. A. Ayuso. M. S. a Parrilla. R.
(1992) Hole 01 endogenous fatty acnds in the control oi hepatic gluconeogenesis.
Arch. Biochem. Biophys. 292: 95—101.

42. Huang. 8.. Wu. P.. Bowker-Kinley. M. M. 8 Harris. R. A. (2002) Reg-
ulation oi pyruvate dehydrogenase kinase expression by peioxlsome proliferator-
activated receptor-alpha ligands. glucocorl'coids. and insub'n Diabetes 51: 276 4&3.

43. Glow.J.C&Jeme.B.w. (1992) WWQW
mmmthMWJDa-‘ysa 75:2142—2149

44. Baird. G. 0.. meter Wait. J. G. a Bergman. E. N. (1983) Wholevbody
metabolism oi glucose and lactate in productive sheep ma cows. Br. J. Nutr. 50'.
249-265.

45. Baird. G. 0.. Lornax. M. A. Symonds. H. w. 8 Shaw. 8. R. (1980)
Net hepatic and splanchnic metabolism 0! lactate. pyruvate and propionate in
dairy cows in vivo m relation to lactation and nutrient supply. Biochem. J. 186:
47-57.

46. van der Walt. J. G.. Baird. G. D. a Bergman. E. N. (1983) Tissue
glucose and lactate metabolism and interconvevsms in pregnmt and actating
sheep. Br. J. Nutr. 50: 267-280.

47. Allen. M, s. (2000) Effects 01 diet on short-term regulation 01 lead
htake by lactating dairy cattle. J. Davy Sci. 83: 1598—1624.

48. Horn. C. 0.. J1. H. & Friedman. M. I. (2004) Etomoxir. a fatty acid
Oxidation hhioitor. ‘ncreases tood intake and reduces hepatlc energy status in
rats. Physiol. Behav. 81: 157-162.

SUPPLEMENTAL TABLE 1

Genes, primers. and homology to reference sequences for sequences analyzed by qRT-PC R

 

 

Gene of Forward primer (5' - 3‘) Reference Sequence Alignment

interest' Reverse primer (5' - 3‘) Sequence Homology Region

PCK1 GTCCTGGCCCTGAAGCAGA AY145503.1 100% 1972 - 2060
TCC'I‘GCTCCTGGTGCG'ITG'

G6PC GTCCTC'ITCCCCA'I'ITGGTTC' NM000151.1 92% 212 - 282
TCCFCT‘GGGTAGCTGTGATFGG (H. sapiens)

SLC37A4 CACCTCCCATGCCATTG' NM001467.3 96% 1762 - 1839
TITGGCAATGGTGCTGAAGG (H. saplens)

PC CGTCTI'I‘GCCCACTTCAAGG NM177946 100% 3113 - 3177
GGCGCGTA'I‘TGAGGCI‘G'

PDK4 CAGGCCAAOCAATTCATA'ITG' NM002612.2 89% 904 — 985
GGCCCI‘CATI‘GCA’I'I‘CFI‘GA (H. saplens)

PP] H TGGCAAATI‘CAAGCCCTG' NM028677.2 90% 78 - 142
AAC'I'I‘ CCTGGCCGCCA AT (H. sapiens)

ACT B TGGGCCAGAAGGACI‘CG' AY141970.1 100% 179 — 248
GGGTACTI‘GAGGGTCAGGATGC

PGK] CAGTGGAGCCAAGTCAGTTG' NM000291.2 90% 228 - 314
GCAACI‘GGCTGCAAGGAGTA (H. sapiens)

 

'PCKI. cytosolic phosphoenolpyruvate carboxykinase; G6PC. catalytic gluoosob—phosphatasc; SLC37A4.
negulatory subunit glucose-o-phosphatasc; PC. pyruvate carboxylase; PDK4. pyruvate dehydrogenase kinase 4;
PPIH, cyclophilin; ACT B, Bactin; PGK I . phosphoglywnte kinase 1.

TAM-labeled LUX primers.

65

 

Glucose ' Glycogen

GGPase

 

Pyruvate .
a, _ m... .......

PCK1 0K0”!

9 m
PCK2 PC PDH

u
Beta-oxidation
m -

Mohandas

 

Sue: Propionate

SUPPLEMENTAL FIGURE 1 Overview of hepatic metabolic pathways related to
gluconeogenesis. Pathways and enzymes discussed in the text are labeled (G6Pase,
glucose-6-phosphatase; PCK1, PCKZ, cytosolic and mitochondrial phosphoenolpyruvate
carboxykinase; PC, pyruvate carboxylase; PDI-I, pyruvate dehydrogenase).

66

CHAPTER 5
PHLORIZIN INDUCES LIPOLYSIS AND ALTERS MEAL PATTERNS IN BOTH

EARLY AND LATE LACTATION DAIRY COWS

ABSTRACT
Increased glucose demand likely alters hepatic propionate metabolism, but it also
stimulated lipolysis in past studies in ruminants. To assess whether increased
glucose demand selectively increases dry matter intake (DMI) for cows in
negative energy balance, phlorizin was administered to blocks of early and late
lactation cows. Six Holstein cows in early lactation (19 1: 6 DIM, 50.0 1 1.8 kg/d
milk, mean :1: SD) and six Holstein cows in late lactation (228 :t 18 DIM, 30.6 t
1.9 kg/d milk) were randomly assigned to treatment sequence in a crossover
design. Periods were 14 d with 7 d adaptation periods and 7 d of treatment.
Phlorizin (4 g/d) and propylene glycol (carrier and control) were administered
subcutaneously every 6 h throughout the treatment periods. Feeding behavior
and DMI data were collected for the final 4 d of each treatment period, and blood
samples and total urine output were collected on d 4 of each treatment period.
Phlorizin caused urinary loss of glucose at 333 g/d in early lactation and 532 g/d
in late lactation cows. Phlorizin increased plasma non-esterified fatty acid
concentration similarly in early and late lactation cows, but did not significantly
alter plasma insulin concentrations. Phlorizin treatment tended to decrease meal

size, but also decreased intermeal interval, resulting in no effect on DMI. We

67

conclude that phlorizin's effects on lipolysis, feeding behavior, and DMI are not

dependent on relative energy balance.

68

INTRODUCTION
Physiological regulation of feed intake must respond to a variety of environmental
and endogenous cues to allow growth or maintain body weight, support milk
production, and supply nutrients for fetal growth. The mechanisms that direct
changes in feeding behavior often interact, making it difficult to identify the
signals that are most responsible for increased dry matter intake (DMI) at
parturition or following the initiation of bST treatment, for example.
Pharmacological treatments that have more specific impacts on physiological
processes can help identify specific mechanisms that may contribute to the

regulation of feed intake.

Phlorizin inhibits renal reabsorption of glucose (Ehrenkranz et al., 2005),
resulting in urinary excretion of glucose and a significant increase in whole-body
glucose demand. Increasing glucose demand in lactating cows led to an
increase in transcript abundance for potentially rate-limiting gluconeogenic
enzymes (Bradford and Allen, 2005). We hypothesized that greater
gluconeogenic capacity would increase utilization of propionate for glucose
production and decrease its oxidation, altering hepatic energy status. Decreased
hepatic ATP production may result in delayed satiety during meals, because
preventing ATP production by trapping inorganic phosphate increased feed
intake in rats (Rawson et al., 1994), as did inhibition of fatty acid oxidation (Horn
et al., 2004).. Therefore, we predicted that phlorizin would increase meal size

and DMI in lactating cows by limiting propionate oxidation (Allen et al., 2005).

69

However, in our previous work, we showed that late-lactation cows were able to
adapt to phlorizin administration for 7 d without increasing DMI (Bradford and
Allen, 2005). Phlorizin's effects on glucose metabolism were confounded by a
concomitant increase in fatty acid (FA) delivery to the liver, which likely supplied
substrate to replace propionate removed from oxidative pathways. We
speculated that early-lactation cows, already in negative energy balance, would
be less flexible in their use of nutrients to drive gluconeogenesis, and that
phlorizin treatment would result in increased DMI in this model. This experiment

was designed to test our revised hypothesis.

MATERIALS AND METHODS
Experimental procedures were approved by the All-University Committee on

Animal Use and Care at Michigan State University.

Design and treatments

Multiparous Holstein cows were selected from the Michigan State University
Dairy Cattle Teaching and Research Center and assigned to blocks of early
lactation (n = 6; 19 i 6 DIM; 2.3 :1: 0.5 lactations; 657 :t: 68 kg BW; mean :1: SD)
and late lactation (n = 6; 228 1 18 DIM; 2.3 1: 0.8 lactations; 689 :t 77 kg BW)
cows. Stalls were assigned to block and treatment sequence to assure balance
within blocks, and cows were randomly assigned to stalls within block. Phlorizin
(Sigma Chemical Co., St. Louis, MO) was administered via subcutaneous

injection at the rate of 4 g/day, with propylene glycol as vehicle and control.

70

Treatment periods were 7 d, and injections were given every 6 h during these
periods. Animals were adapted to a single diet for a 7-d period prior to the first
treatment period, and a 7-d rest period was included between the two treatment
periods. One of the 28 injections during period 2 was missed, however, no data
was collected for 36 h after the missed injection, and glucose excretion stabilizes
relatively quickly following injection of phlorizin (Amaral-Phillips et al., 1993).
Because there were no significant period x treatment interactions for DMI,
feeding behavior, or plasma analyte concentrations in subsequent statistical
analyses (all P > 0.15), we conclude that the missed injection had little or no

effect on measured responses to treatment.

Data and sample collection

Throughout the experiment, cows were housed in tie-stalls and fed a single
experimental diet (Table 5.1) as a TMR once daily (1130 h) at 115% of expected
daily intake. Cows were not allowed access to feed from 1000 to 1130 h, during
which arts and the amount of feed offered were weighed for each cow daily.
During the final 4 d of each treatment period, feeding behavior was monitored by
a computerized data acquisition system (Dado and Allen, 1993) throughout the
clay. Data on chewing activities, feed disappearance and water consumption
were recorded for each cow every 5 s, and mean daily values for number of meal
bouts, interval between meals, and meal size were calculated. On each of these
4 (1, samples of all dietary ingredients (0.5 kg) and arts (12.5%) were collected

and frozen for later analysis. Starting on d 4 (1000 h) of each experimental

71

period, urine was collected for a 24-h period and blood was sampled hourly from
an indwelling jugular catheter as previously described (Bradford and Allen, 2005).
Cows were milked twice daily in the milking parlor during rest periods and in the
tie stalls during treatment periods. Milk yield was recorded and samples were

taken at each milking during the 4 collection days.

Table 5.1 Ingredients and nutrient composition of
experimental diet1.

 

 

Item

Diet ingredients
Corn silage 31.8
High moisture corn grain 29.1
Alfalfa haylage 11.5
Soybean meal 11.8
Modified expeller soybean meal2 9.0
Mineral and vitamin mix3 6.9

Nutrient composition
DM, % as-fed 43.4
Organic matter 94.0
Starch 28.2
NDF 27.1
CF 17.8

 

1Values other than DM are expressed as % of dietary DM.
2SoyPlus (West Central Soy, Ralston, IA).

Mineral and vitamin mix contained 74.7% dry ground corn,
10.9% limestone, 5.5% salt, 5.3% dicalcium phosphate,
1.8% magnesium oxide, 1.4% trace mineral premix, and
0.4% vitamin ADE premix.

Sample analysis

Diet ingredients, arts, and fecal samples were dried in a 55°C forced-air oven for
72 h and analyzed for dry matter concentration. Feed samples were ground with
a Wiley mill (1-mm screen; Arthur H. Thomas, Philadelphia, PA) and analyzed for

ash, neutral detergent fiber, crude protein, and starch concentrations as

72

previously described (Bradford and Allen, 2005). Concentrations of all nutrients
were expressed as percentages of dry matter determined from drying at 105°C in

a forced-air oven.

Milk samples were analyzed for fat, true protein, and lactose with infrared
spectroscopy (AOAC, 1990) by Michigan DHIA (East Lansing, MI). In addition,
milk samples from d 7 of each treatment period were analyzed for fatty acid
profile by gas chromatography as previously described (Bradford and Allen,
2004). Plasma samples were composited into a single sample for each cow
period for all analyses except non-esterified fatty acid (NEFA) and growth
hormone (GH) concentration. Urine and plasma samples were analyzed in
duplicate for glucose concentration by the glucose oxidase method (Raabo and
Terkildsen, 1960). Plasma samples were analyzed in duplicate using
commercial kits to determine concentrations of NEFA [NEFA C-kit; Wako
Chemicals USA, Richmond, VA, as modified (McCutcheon and Bauman, 1986)],
B-hydroxybutyrate (procedure #2440, Stanbio Laboratory, Boerne, TX), insulin
(Coat-A-Count, Diagnostic Products Corporation, Los Angeles, CA), and
glucagon (Glucagon kit #GL-32K, Linco Research Inc., St. Charles, MO). Plasma
GH concentrations were quantified with a double-antibody radioimmunoassay

(Gaynor et al., 1995).

73

Statistical analysis

One cow in the late lactation group decreased DMI by approximately 50% at the
beginning of each data collection period, likely because of sensitivity to the chew
halter apparatus. Because the cow was an outlier for DMI, and plasma analyses
were indicative of an unusual catabolic state, this cow was removed from all

analyses.

Data were analyzed according to the following mixed-effects model:

Yijk = lJ "' Pi + S) + Tk + CI(Sj) "‘ STjk 1' PTik + eijkl
in which Yuk. is a dependent variable, p is the overall mean, Pi is the fixed effect
of period (i = 1 to 2), S] is the fixed effect of block (j = 1 to 2) T., is the fixed effect
of treatment (k = 1 to 2), C. is the random effect of cow nested within stage (I = 1

to 6), STjk is the interaction of block and treatment, PT". is the interaction of

period and treatment, and eijkl is the residual error. Analysis of plasma NEFA
data also included fixed effects of sample time, time by treatment interaction, and
time by stage interaction. Urinary glucose and plasma BHBA values were not
normally distributed and were log-transformed for analysis; reported means are
back-transformed. For all main effects, significance was declared at P < 0.05,
and tendencies were declared at P < 0.10. For interactions, significance was

declared at P < 0.10, and tendencies were declared at P < 0.15.

74

RESULTS AND DISCUSSION
As expected, urinary glucose excretion increased dramatically with phlorizin
administration, from 0.4 g/d to 333 g/d in early lactation and 532 g/d in late
lactation (P < 0.001, Table 5.2). The significant treatment by stage of lactation
interaction (P < 0.01) is likely related to the significantly higher plasma glucose
concentrations in late lactation. Phlorizin-treated cows in late lactation had a
mean glucose concentration of 68.2 mg/dL compared to 59.9 mg/dL for treated
early-lactation cows. Because phlorizin inhibits renal glucose reabsorption,
greater delivery of glucose to the kidneys in late-lactation cows would be

expected to result in greater glucose excretion.

As reported previously (Amaral-Phillips, et al., 1993; Bradford and Allen, 2005;
Overton et al., 1998), phlorizin treatment increased plasma NEFA concentration
(P < 0.03), which is indicative of the rate of lipolysis in ruminants (Dunshea et al.,
1989). Although Amaral-Phillips and coworkers (1993) reported increased
lipolysis in response to phlorizin treatment in cows at 6 weeks postpartum, the
mean NEFA concentration for the control treatment (181 qu/L) suggested that
the cows were not in an severe catabolic state. Nevertheless, our results in cows
with higher plasma NEFA concentrations (278 qu/L) confirm these findings, and
further demonstrate that early and late lactation cows respond to phlorizin with a
comparable increase in lipolysis. We and others have suggested that phlorizin-
induced lipolysis is mediated by decreased plasma insulin concentrations

(Bradford and Allen, 2005; Vranic et al., 1984), however, in this experiment, we

75

Eugen: new x003 be 530835 u E. ”Eugene .6 Soto EmE u E. ”€3.38. .0 8va x003 .0 «onto 59: u omfimw

.5598. 2m. .8 m u : use 5388. Emu .2 m u : .zwm a memo... Susan ammo. 9n moan?

 

 

 

Rd 26 85 No a m. R No a a. R No a QR no a Nam A29; 9.95 .89: >5
«no Ed :55 v 8.0 a 8m and a «Z and a one moo a 8e 3555 :o 2.85
208295
86 and Sod v 9.0 a 24 who a one 8.0 a :3 2.0 a and
59329533.
3569
one 5.0 86 :32 :32 :n s: a; RF
comnoam mEmmE
mod 85 Sod v mod H was mod n as Ed a new ad H 5N 3.523 5.3.5 «585
mod 3.0 36 we a we 3 n 4.9 we a 4.2 we a m. : 3.2.3.5 <mzm 2.85
33 3o Sod v on a 8 on a a. R « Rm R n EN 3533 sz «58:
5d and :3 v an a «.8 on a QR em a m8. 3 a N. 5 32mg 88:5 «Emma
Nod Sod v :3 v on 4 2a. 86 n 93 2 n mum mod n 93 E9 88% 8::
NE. at... momma c_~_._o_.._n_ .2200 ENtoEn. 6:50 Em:
a gag 12583 cam
125:8 atom“.

new .3888; can amazon—36E memE 0005 63906 $836 :0 Eamon. Lo 693m new ENtoEa Co maomtm a...” win...

76

found that phlorizindid not significantly alter plasma insulin concentration or

insulinzglucagon ratio (Table 5.2).

We then assessed whether GH was responsible for this Iipolytic response,
because plasma GH concentrations increase in response to hypoglycemia (De
Feo et al., 1989) and GH acts to suppress lipogenesis in ruminant adipose tissue
(Liesman et al., 1995). Again, we found no effect of phlorizin on plasma
concentration of GH (Table 5.2). Finally, we considered the work of Brockman
(1984), who demonstrated that hypoglycemia directly stimulated lipolysis in an
elegant study of alloxan diabetic, adrenalvdenervated sheep. However, phlorizin
treatment did not cause hypoglycemia in this study; in fact, phlorizin did not
significantly alter plasma glucose concentration (Table 5.2). While we are unable
to address the possibility that phlorizin treatment caused a stress response in
these animals, Brockman’s observation that phlorizin causes lipolysis in animals
without active adrenal glands (1984) demonstrates that a stress response is not
a required component of phlorizin-induced lipolysis. Therefore, the cause of
phlorizin-induced lipolysis in lactating dairy cows remains unclear. It is possible
that small, statistically undetectable changes in plasma insulin and glucose
concentration were adequate to increase plasma NEFA concentration to the
extent that was observed in this experiment. Also, despite the validation of
plasma NEFA concentration as an index of lipolysis (Dunshea, et al., 1989), it is

important to note that we did not directly measure Iipolytic rate in this experiment.

77

We found a significant stage of lactation by treatment interaction for plasma

BHBA concentration (P < 0.03), but the reason for this response is unclear.

Phlorizin did not alter yield of milk, milk lactose, or milk fat (Table 5.3). However,
there was a tendency for an interaction between stage of lactation and treatment
for milk protein yield (P < 0.15). We also used milk FA profile to help assess
metabolic changes in response to phlorizin injection. Consistent with observed
effects on plasma NEFA concentration, long-chain milk FA yield tended to
increase with phlorizin treatment (P = 0.07, Table 5.3). Long-chain FA in milk are
derived from circulating FA rather than de novo synthesis in the mammary gland
(Barber et al., 1997), so a tendency for increased secretion of long-chain FA
suggests that more circulating FA were available for uptake by the mammary
gland. We also found that early lactation cows had a significantly greater
proportion of unsaturated FA in milk fat than late lactation cows (39.6 vs. 31.3
g/100 9 FA, P < 0.001). Greater delta-9 desaturase activity in early lactation
cows (delta-9 desaturase index: 0.40 vs. 0.30, P < 0.01) accounted for the
majority of the difference in milk FA saturation, in agreement with the findings of

Kay et al. (2005).

Contrary to our primary hypothesis, there was no interaction of treatment and
stage of lactation for DMI in this study (P = 0.37). This is perhaps not surprising
given that indicators of lipolysis (plasma NEFA concentration and long-chain FA

secretion) responded to treatment in a similar manner in both early and late

78

.2? .E 5299.5 .90 .. 80 .E 59.0.5585 do i 3 .E canoeofi ”555.0. 8 85.8 8.8 band
5:958... can v.83 .6 5.6825 n E. ucopammb .0 Soto 5mE u t... ӣ55905 .0 mafia. 5025 do Soto 59: u $sz
.5538. 2m. .2 m u E can ceased. Eco 8.. o u : .zww « mcmmE madam ammo. Em moan?

 

 

 

8d Ed Edd v 8 a Rm. 8 n 5. mm a 32 B n 98 .29 ”55> E 520-93
odd Ed Edd v 8 a N3. 8 H «mm on an So on a N8 .29 55> E .80
mod ddd 5d v t a new 2 a 8m 9 a 3m 5: H Rd .29 n55> E 520-5582
Ed Ed Edd v 3 a .8 3 a 8 S n «3 S a E .23 .”55> E 520.50%
5d 2d Edd v 2d a 85 2d « >5 add 5 Ed 8d 5 Ed Edd 55> a... 552
3d V dd Ed V 3d an 8d 3d a 8.5 8d a 85 8d a 8.5 .29.. 55> 559.. 552
2d odd Edd v 3d n 3.. 3d H of «d a 8d fid a 85 .29.. 55> 828. 5.:
mod dmd Sod v ad a mdm ad a >. 5 SN a. dds. E « ddm .29.. 55> 552
NE. «E. momma 55.6...n. .obcoo 55.0.5.5 .2580 Em:
n. 8.558.. 53 3

 

.555» 295958 o.__E 5cm 5...: co 5.5.905 .6ng can 55.5.55 .6 muootw «6 0.58.

79

lactation cows. However, the results of this study do contradict our original
hypothesis that phlorizin would increase DMI by increasing meal size. Phlorizin
treatment tended to decrease mean meal size (P = 0.07, Figure 5.1), which
resulted in a compensatory decrease in intermeal interval (P = 0.02) and no

effect on DMI (P = 0.15, Table 5.2).

      
   

 

 

 

 

 

2.5- T * — 120
V
I — 100
2..
—80
1.5-
-60
11
-40
0.5- A _20
Control
0_ A ////A L0

Meal size (kg DM) Intermeal interval (min)

Figure 5.1 Effects of phlorizin administration on feeding behavior.

Values are least square means :t: SEM, n = 11. 1' P < 0.10; " P < 0.05.

Although we have previously shown that phlorizin administration results in

increased gluconeogenic capacity, we also observed an increase in NEFA

available for hepatic oxidation (Bradford and Allen, 2005). In this study we have

80

again shown that phlorizin increased circulating NEFA, and the total amount of
substrate available for oxidation in the liver may have increased with phlorizin
treatment. Furthermore, we have previously suggested that propionate and
acetyl-00A may interact to shorten meal size in cows mobilizing body fat (Oba
and Allen, 2003a). Cows in negative energy balance often combine high rates of
acetyl-CoA production (from FA oxidation) with a relative shortage of glucogenic
substrate. This situation results in a deficit of TCA cycle intermediates and high
NADH:NAD+, preventing oxidation of acetyl-00A (Lopes-Cardozo et al., 1975).
Propionate can stimulate acetyl-CoA oxidation both by providing TCA
intermediates and by utilizing NADH during its conversion to glucose. Such an
interaction would lead to rapid ATP production during meals, and could have
played a role in the tendency for a decrease in meal size for the phlorizin

treatment.

CONCLUSIONS
In agreement with past results, phlorizin not only increased glucose demand but
also stimulated lipolysis in both early and late lactation cows. Phlorizin did not
significantly alter plasma concentrations of glucose, insulin, glucagon, or CH.
Finally, phlorizin tended to decrease meal size but also decreased intermeal
interval, resulting in no change in DMI in either early or late lactation cows.
Phlorizin provides a model for increased glucose demand, but confounding
effects on lipid metabolism complicate interpretation of its effects on hepatic

energy metabolism in lactating cows.

81

CHAPTER 6
PHLORIZIN ADMINISTRATION DID NOT ATTENUATE HYPOPHAGIA

INDUCED BY INTRARUMINAL PROPIONATE INFUSION

ABSTRACT
Infusion data from ruminants has shown that propionate stimulates satiety and
decreases meal size, possibly because of increased propionate oxidation in the
liver. In this experiment, phlorizin was used to increase glucose demand, which
was expected to decrease propionate oxidation and attenuate the decrease in
dry matter intake (DMI) caused by propionate infusion. Twelve multiparous,
ruminalIy-cannulated Holstein cows (49 :t 33 DIM, 40 :t: 7 kg/d milk; mean :t: SD)
were randomly assigned to square and treatment sequence in a replicated 4x4
Latin square experiment with a 2x2 factorial arrangement of treatments.
Treatments were subcutaneous injection of phlorizin or propylene glycol in
combination with intraruminal infusion of either Na acetate or Na propionate.
Following a 7 d adaptation period, phlorizin (4 g/d) and control injections were
administered every 6 h for 7 (1. During the final 2 d of injections, Na acetate or
Na propionate solutions (1 mol/L, pH 6.0) were infused continuously at the rate of
0.80 Uh. Feeding behavior data were collected during the final 2 d of treatment.
Phlorizin caused urinary excretion of 400 g glucose/d across infusion treatments.
Phlorizin tended to increase plasma non-esterified fatty acid and beta-
hydroxybutyrate concentrations to a greater extent with Na acetate compared to

Na propionate infusion. Phlorizin decreased and Na propionate increased

82

plasma insulin and glucose concentrations. Infusion of Na propionate decreased
DMI (18.4 vs. 21.1 kg/d) through an increase in intermeal interval (89.2 vs. 77.3
min), resulting in fewer meals/d (11.6 vs. 13.7). Phlorizin did not alter DMI or
measures of feeding behavior, nor were there interactions with infusion type.
Increasing glucose demand does not limit the extent to which propionate

decreases DMI in lactating dairy cows.

83

INTRODUCTION
Propionate, a product of ruminal carbohydrate metabolism, depresses feed
intake of ruminants. Although the mechanism for this hypophagic response is
unknown, a strong body of evidence indicates that the liver is involved.
Propionate infusion depresses feed intake to a greater extent when infused into
the portal circulation compared to peripheral infusion (Anil and Forbes, 1980;
Baile, 1971 ), and blocking autonomic nervous system pathways between the
brain and liver eliminated hypophagic responses to propionate infusion (Anil and
Forbes, 1980; Anil and Forbes, 1988). Forbes has pointed out that propionate
and many other substrates oxidized by the liver have demonstrable effects on
food intake, while the products of these oxidation reactions do not (Forbes,
1988). Increased feed intake was induced in nonruminants by blocking hepatic
oxidation of fatty acids (Horn et al., 2004; Langhans and Scharrer, 1987) and by
preventing ATP production through phosphate trapping (Rawson et al., 1994),
providing more direct evidence for the involvement of hepatic oxidation in feed

intake regulation.

Evidence regarding the feed intake effects of hepatic oxidation by ruminant liver
is currently lacking. Although at least some propionate is oxidized by ruminant
liver (Black et al., 1966), the majority of propionate entering the liver is utilized for
gluconeogenesis. Propionate infusion alters feed intake primarily through
decreased meal size (Oba and Allen, 2003a), and we hypothesized that the

increased flux of propionate into the liver during meals (Benson et al., 2002)

results in excess glucogenic substrate availability and greater oxidation of
propionate, stimulating satiety. To test this hypothesis, we used phlorizin to
cause urinary loss of glucose; we previously provided evidence of an adaptive
increase in gluconeogenic capacity in response to phlorizin administration in
lactating cows (Bradford and Allen, 2005). We expected phlorizin to direct more
propionate toward glucose production and to limit the hypophagic response to

propionate infusion.

MATERIALS AND METHODS
Experimental procedures were approved by the All-University Committee on

Animal Use and Care at Michigan State University.

Design and treatments

Twelve multiparous lactating Holstein cows (49 :l: 33 days in milk, 40 :l: 7 kg/d
milk; mean :I: SD) with ruminal cannulas were selected from the Michigan State
University Dairy Cattle Teaching and Research Center and randomly assigned to
square and treatment sequence in a replicated 4x4 Latin square design balanced
for carry-over effects. A 2x2 factorial arrangement of treatments was used to
assess effects of phlorizin in combination with infusion of Na acetate or Na
propionate. During treatment periods, phlorizin (Sigma Chemical, St. Louis, MO)
was administered via s.c. injection every 6 h at the rate of 4 g/d, with propylene
glycol as vehicle and control. Treatment periods were 7 d, and during the final 2

d of treatment, solutions of Na acetate or Na propionate (1.0 mol/L, pH 6.0) were

85

continuously infused into the rumen at the rate of 0.80 :l: 0.07 Uh. Solutions were
infused using 4-channel peristaltic pumps (#78016—30, Cole-Farmer Instrument,
Vernon Hills, IL) and Tygon tubing (7.5 m x1.6 mm id; Fisher Scientific 00.,
Pittsburgh, PA). The cows were adapted to a single diet for a 7-d period before
the first treatment period, and 7 d of rest were included between treatment
periods. Throughout the experiment, cows were housed indoors in tie stalls and
fed a total mixed ration once daily (1130 h) at 110% of expected intake. The diet
(455 g dry matter / kg diet, Table 6.1) was formulated to meet nutrient
requirements (NRC, 2001) and to promote endogenous propionate production.
One cow was removed from treatment in period 1 for clinical ketosis and a
second cow was removed from the experiment in period 4 for mastitis, leaving 11

observations for each of the acetate treatments.

Data and sample collection

Access to feed was prevented for 90 min each day (1000 to 1130 h) while arts
and the amount of feed offered were weighed for each cow. During the final 2 d
of each treatment period, feeding behavior was monitored by a computerized
data acquisition system (Dado and Allen, 1993) throughout the day. Data on
feed disappearance and water consumption were recorded for each cow every 5
s, and mean daily values for number of meal bouts, interval between meals, and
meal size were calculated. On each data collection day, samples of all dietary
ingredients (0.5 kg) and arts (12.5%) were collected. Cows were milked in the tie

stalls during data collection and in the milking parlor during the rest of the

86

experiment. Milk yield was recorded and samples were taken at each milking

during collection days.

Table 6.1 Ingredients and nutrient composition of the
experimental diet.

 

 

Component Content
g/kg dry matter
Diet ingredient
Corn silage 297
High moisture com 291
Alfalfa silage 165
Modified expeller soybean meal1 81
Soybean meal 76
Mineral and vitamin mix2 39
Nutrient composition
Organic matter 922
Starch 316
Neutral detergent fiber 275
Crude protein 153

 

1 SoyPlus (donation from West Central Soy, Ralston, IA).
2 Contained (lkg dry matter): Ca, 44 9; Na, 38 9; Cl, 36 9; Mg, 14

g; P, 13 9; K, 2.8 9; Fe, 1.4 g; S, 1.4 9; Mn, 467 mg; Zn, 453

mg; Cu, 113 mg; l, 6.7 mg; Se, 3.4 mg; Co, 1.7 mg; B-carotene,

24 mg; cholecalciferol, 369 pg; 2—ambo-a-tocopherol, 148 mg.
Urine was collected for a 24-h period on d 7 (1100 h) of each treatment period.
Urinary catheters (Bardex Lubricath Foley 24FR, Bard Medical) were inserted
and urine was collected into a container with 0.23 mol HCI to prevent glycolysis.
Volumes were measured at 6-h intervals and samples were taken and frozen
until analysis. During the same 24—h period, blood samples were collected at 3-h
intervals from indwelling jugular catheters. Collected blood was processed and
stored as described (Bradford and Allen, 2005). At the end of each data
collection period, ruminal contents were sampled from 5 sites throughout rumen

and squeezed through a nylon screen (1 mm pore size) to collect the liquid

phase. Ruminal fluid pH was measured using a portable pH meter (model 230A,

87

ATl Orion, Boston, MA), and samples were frozen at -20° C until analysis.

Sample analysis

Diet ingredients, orts, and fecal samples were dried in a 55°C forced-air oven for
72 h and analyzed for dry matter concentration. Ingredient samples were ground
with a Wiley mill (1-mm screen; Authur H. Thomas) and analyzed for ash, neutral
detergent fiber, crude protein, and starch content as previously described
(Bradford and Allen, 2005). Concentrations of all nutrients are expressed as
percentages of dry matter determined from drying at 105°C in a forced-air oven.
Milk samples were analyzed for fat, true protein, and lactose with infrared
spectroscopy by the Michigan Dairy Herd Improvement Association (Lansing,
MI). In addition, milk samples from d 7 of each treatment period were analyzed
for fatty acid profile by gas chromatography as previously described (Bradford
and Allen, 2004). Urine and plasma samples were analyzed in duplicate for
glucose content by the glucose oxidase method (Raabo and Terkildsen, 1960).
Plasma samples were analyzed using commercial kits to determine
concentrations of F FA, fi-hydroxybutyrate (BHBA), insulin, and glucagon as
previously described (Bradford and Allen, 2005). Concentrations of short-chain
fatty acids (SCFAs) in ruminal fluid were determined as previously described

(Oba and Allen, 2003b).

88

Statistical Analysis
Data were analyzed by the fit model procedure of JMP (version 5.0, SAS

Institute, Cary, NC) using the REML method according to the following model:
Yijkl = lJ + Pi 1' S] + CkISj)+ TI "’ PTii 1' V 1' VTI 'l' eijkl

where Yuk. is a dependent variable, [.1 is the overall mean, Pi is the fixed effect of

period (i = 1 to 4), Si is the fixed effect of square 0 = 1 to 3), Ck is the random

effect of cow within square (k = 1 to 4), T. is the fixed effect of treatment (I = 1 to

4), PT" is the interaction of period and treatment, V is the effect of infusion

volume, VT. is the interaction of infusion volume and treatment, and em. is the

residual error. Infusion volume was included in the model to account for potential
bias caused by variation in infusion rate across cows and periods. Period by
treatment interactions were not significant for any variable except milk lactose
yield; modeling potential carryover effects for this variable did not resolve the
issue, so treatment effects could not be assessed. Plasma data were analyzed
by a model including the above terms as well as the effects of sample time and
time by treatment interaction; repeated measures overtime were modeled with a
heterogeneous autoregressive [ARH(1)] covariance structure using the repeated
statement of SAS (version 9.0, SAS Institute, Cary, NC), and denominator
degrees of freedom were estimated using the Kenward Rogers method. For all
main effects, significance was declared at P < 0.05, and tendencies were
declared at P < 0.10. Interactions were declared significant at P < 0.10 and

tendencies for interactions were declared at P < 0.15.

89

RESULTS AND DISCUSSION
Phlorizin administration caused urinary loss of glucose at the rate of 400 g/d
across infusion treatments (Table 6.2), consistent with previous results in
lactating cows (Bradford and Allen, 2005). Also as expected, infusion of acetate
or propionate increased ruminal concentrations of the respective SCFAs (P <
0.001). More surprisingly, phlorizin treatment tended to decrease ruminal
propionate concentration (P < 0.10), and propionate infusion decreased ruminal
butyrate concentration relative to acetate (P < 0.01). Plasma glucose
concentration was altered both by substrate availability and by peripheral
demand; glucose concentrations increased in response to propionate infusion (P
< 0.001) and decreased with phlorizin treatment (P < 0.01). Changes in plasma
insulin and glucagon concentrations were expected responses to treatment

effects on plasma glucose concentration (Table 6.2).

In past experiments, we have consistently found that both increased dietary
starch fermentability and propionate infusion depress feed intake primarily
through decreased meal size (Oba and Allen, 2003a; Oba and Allen, 2003b; Oba
and Allen, 2003d). However, in this experiment, propionate infusion increased
intermeal interval (P < 0.04, Table 6.3) with no effect on meal size.
Management, ration formulation, and infusion protocols were similar in this
experiment compared to our past work, making it difficult to determine why the

feeding behavior response was different in this case.

90

.3553: Lo 5.8825 n E. EBEEEEE 55:2...“ Co Soto u 5.5.5 65825 22552.. .o 89.3 n SEN

9:95am: Encoaoa Lo“. NF u c new 35.58: 288m .8 S u c Sum « memos. 23am “mom. can moan?

 

 

8d mdd Edd Rd NE o.d.. m. 5 4.8 5.qu deduced 9555
Ed Ed :dd 2 do 2 t. dd .5323 .5555 «58:
Rd 8d Rd mmd 3d «ed 3d mod SSE 82825.. «585
2d Ed add 2: 8... d; 85 8d .285. .<m_._m «585
3d mud Ed on 2 mm m? 3 .308: Ed 2.85
Ed add Ed 2d and 9d 3d 8d SSE ._98>_d memes.
Ed 5d Sod ddd Rd add add. mod .SoEE .8836 555a
Ed mdd 5d Rd Pd 3. ed? 52 :25: .5533 3553".
8d Ed :dd 3 >. 5 mdm E; 4.2 .3225 decoded 3553".
Ed ddd Edd md 5% o. E at dds .5655 .288 E553.
Ed Edd «dd 9. 8... 5 mod N Ed .8885... 88% >555
NE. «SEQ wooed 2mm 55.5.5. .2550 558.5. .9200 Em:
v n. 3 wolmfloflz

 

9555 500... new .w<u.om .m5EE £55995 coco—.6 cc .5625 <n.om 5cm 55.6.5... .0 mdomtm Nd m._m<._.

..

wmcogoc UCG mOEODmaOE

91

55.5825 505.55... >5 5255 58557. n 5 3583 593035 5: 2:8 38:0 555.52...

.9598 5.. 55 55E 5.3 3.3 bun. we danced 55 no.8 35.. 555.90..

v
n

9.55555 5 5.5825 n E. .co.§5.5Eun 555.5 5 585 u 5.5.”. 25525 555555 5 59cc .1. nocdm
55.5.5525 55.5505 5.. up u c new «55.58: 288m .8 3 u c .zwm « wcmoE 55:3 fine. 55 3:5?

 

 

dad ad add dd ddm odd «.8 in 58 >55 520%th
Ed d«d Rd 8 mom dad a: one 58 >55 cacowmwofimm
Ed 3d 5d 8d 8.5 8.5 of 8.5 .09. 55> a. 5...).
odd d«d 3d mod 85 «3 8.5 m: 59. 55> 5505 552
3d d«.« >«.« ««.« o«.« Ede. .m55> 8205 5.52
8d 5d 3d dd «.8 5.«d QR dd» 55 .5255 8555
«ad Ed Edd ed d. : d. : 4.2 «.3 Q .585 co 285:2
odd Ed 3d mod and 5.5 m3 m3 885 >5 9. d8 :82
5d and add 3 E: «.2 dd« m. 5« Re. .8555 adds >5
«5. «coin. «moan. 2mm 55.5....n. .0550 555.55 .2500 :5:
v Q 3 waglmlz

 

..58:an8 5:5 .0 55> use .5228 968. 8 8525 Eow Em 55.5.55 58% n... 5.55

92

In previous experiments, phlorizin induced a catabolic state resulting in increased
availability of FFA at the liver (Bradford and Allen, 2005). Phlorizin caused
increases of similar magnitude in plasma FFA and BHBA concentrations with
acetate infusion in this experiment (Table 6.2). However, propionate infusion
significantly decreased FFA and BHBA concentrations; furthermore, tendencies
for treatment interactions for both FFA and BHBA suggest that marginal
responses to phlorizin were decreased by propionate infusion (Table 6.2).
Plasma FFA analyses provide a reasonable assessment of Iipolytic activity in
ruminants (Dunshea et al., 1989). On the other hand, plasma glycerol is also
typically indicative of lipolysis (Brockman, 1984), and phlorizin increased plasma
glycerol concentrations across infusion treatments (P < 0.02, Table 6.2).
Analysis of milk components offers the advantage of providing a composite view
of nutrient availability over the course of the day. De novo lipogenesis in the
ruminant mammary gland produces fatty acids with chain lengths up to 16
carbons, but long-chain fatty acids (LCFA) exceeding 16 carbons are plasma-
derived (Barber et al., 1997). Because phlorizin treatment had no influence on
dietary lipid intake, any treatment effects on LCFA can be attributed to
differences in Iipolytic activity. Although there was no significant effect of
phlorizin on LCFA yield (P < 0.29, Table 6.3), numerical increases in LCFA yield
accounted for 32% and 88%, respectively, of the significant increase in milk fat
yield for phlorizin treatment with acetate and propionate infusions. No other

class of fatty acids could account for the increase in milk fat yield with phlorizin.

93

Taking into account treatment effects on plasma NEFA, plasma glycerol, and
milk LCFA yield, it seems likely that lipolysis, and delivery of fatty acids to the

liver, was increased by phlorizin across infusion treatments.

As expected, propionate infusion decreased DMI relative to acetate infusion
(18.7 vs. 21.5 kg/d, P <0.001, Table 6.3); according to our working hypothesis,
this was because propionate infusion stimulated hepatic oxidation of propionate
(Allen et al., 2005). Phlorizin is expected to increase utilization of propionate for
gluconeogenesis (Veenhuizen et al., 1988), and we predicted that this would limit
the hypophagic effects of propionate infusion. However, we found no overall
effect of phlorizin on DMI (P < 0.39, Table 6.3), nor was an interaction with
infusion treatment evident (P < 0.91 ). These data do not support our hypothesis;
however, treatment effects on fatty acid availability for hepatic oxidation confound

the targeted effects on propionate and glucose metabolism.

A number of alternative hypotheses have been suggested to explain the
hypophagic effects of propionate in ruminants. Propionate is an insulin
secretagogue, and both propionate infusions and rations with large amounts of
highly digestible starch typically increase plasma insulin concentrations (Harmon,
1992). Insulin is a potent satiety factor when administered via
intracerebroventricular infusion in sheep (Foster et al., 1991), and it can access
the central nervous system through the blood-brain barrier (Plum et al., 2005).

However, propionate infusions have decreased DMI without increasing plasma

94

insulin concentrations (Famingham and Whyte, 1993), and in one study,
propionate infusion failed to decrease DMI despite nearly tripling jugular insulin
concentrations (Leuvenink et al., 1997). Within propionate infusions in this
experiment, phlorizin treatment tended to decrease insulin concentration (68 vs.
79 pmol/L, P < 0.06) while numerically decreasing DMI; at least in this case,

insulin did not mediate propionate’s hypophagic effects.

Nutrient receptors in the veins draining the splanchnic bed have been proposed
for propionate in ruminants (Baile, 1971; Leuvenink, et al., 1997) and glucose in
nonruminants (Novin et al., 1973). Glucose is hypophagic in a variety of
nonruminants (Forbes, 1995), but intestinal and intravenous glucose infusions
have not decreased energy intake of ruminants (Allen, 2000). The absence of
effects of portal glucose infusion on feed intake by sheep (Baile and Forbes,
1974) casts doubt on the hypothesis that sensory neurons in the portal vein
mediate the hypophagic effects of glucose, given that mechanisms regulating
feed intake are well conserved across divergent species (Chiang and
MacDougald, 2003). There is broad consensus now that portal glucose infusions
modulate feed intake of nonruminants through oxidation of glucose (Mobbs et al.,
2005), but there is disagreement as to whether this occurs in sensory neurons or
hepatocytes. Because ruminant neural tissue metabolizes glucose (Lindsay and
Setchell, 1976), it seems unlikely that hypophagia is induced by glucose
oxidation in sensory receptors. Rather, differences in hypophagic effects of

glucose infusion observed between ruminants and nonruminants are likely

95

because of differences in hepatic oxidation of glucose; liver hexokinase activity is
low in ruminants compared with nonruminants (Ballard, 1965), and in mature
ruminants, hepatic removal of glucose appears to be negligible (Stangassinger
and Giesecke, 1986). Likewise, the responses attributed to the proposed
propionate receptor can be explained equally well by the hepatic oxidation
hypothesis, without the need to propose dramatic evolutionary divergence in

mechanisms regulating feed intake.

Evidence regarding the mechanism for propionate's hypophagic effects in
ruminants remains elusive. These data and other recent findings suggest that
short-term (7 d) increases in glucose demand do not alter feed intake response
to propionate infusion or highly fermentable diets (Bradford and Allen, 2005).
Adaptive physiological responses such as increased lipolysis and protein
breakdown may provide oxidative fuel and glucogenic substrate for enhanced
gluconeogenesis during phlorizin administration, preventing the increase in feed

intake that was expected in response to urinary loss of ~5% of net energy intake.

96

CHAPTER 7
RATE OF PROPIONATE INFUSION WITHIN MEALS DID NOT INFLUENCE

FEEDING BEHAVIOR

ABSTRACT
Propionate has been shown to depress feed intake of ruminants, but it is
unknown whether the rate of propionate infusion influences this response. To
test this possibility, rate of propionate infused within meals was altered while total
amount of propionate infused was held constant. Eight multiparous Holstein
cows (51 :t 19 DIM, 44.0 i 4.8 kg/d milk; mean :1: SD) were randomly assigned to
treatment sequence in a crossover experiment with a 10-d diet adaptation period
and 3 d between treatments. Treatments were intraruminal infusion of 1.26 mol
Na propionate (2.33 :l: 0.06 L, 0.54 M, pH 6.0) over the course of either 5 min
(FAST) or 15 min (SLOW) at each spontaneous meal. The experimental diet
included high moisture corn and was formulated for 27% NDF, 36% starch, and
17.5% CF. Feeding behavior was monitored by a computerized data acquisition
system that triggered infusion pumps at the initiation of meals, and consecutive
infusions began at least 15 min apart under both treatment protocols. Feeding
behavior data were analyzed to quantify meal size, meal length, number of
meals, and intermeal interval. Propionate infusions depressed feed intake by
20% and 23%, respectively, for SLOW and FAST. However, rate of propionate
infusion did not significantly alter dry matter intake, meals/d, meal size, meal

length, or intermeal interval. We found no evidence that rate of infusion, within

97

the range of typical meal lengths, determines the extent of hypophagia from

propionate.

98

INTRODUCTION
Propionate plays an important role as a regulator of feed intake in ruminants fed
high starch diets. Continuous infusion of propionate depressed feed intake of
lactating cows in a dose-dependent manner, primarily by decreasing meal size
(Oba and Allen, 2003a; Oba and Allen, 2003d). We have also shown that
intraruminal infusion of Na propionate at spontaneous meals decreases feed
intake relative to isomolar infusions of Na acetate and NaCl (Choi and Allen,
1999). These observations, in combination with data suggesting that propionate
delivery to the liver can increase significantly within the course of a meal (Benson
et al., 2002), lead us to hypothesize that propionate depresses feed intake by
decreasing the time required to stimulate satiety within a meal (Allen et al.,
2005). However, if hepatic oxidation of propionate is required for its satiating
effects, then the fate of propionate within the liver will determine the extent to
which feed intake is depressed. If propionate flux into the liver outpaces capacity
for glucoeneogenesis, excess propionate will likely be oxidized. Therefore, rate
of propionate absorption may be as important as total propionate uptake as a
limitation to feed intake. To test this hypothesis, we assessed the effects of
propionate infusion rate, independent of infusion amount, on feeding behavior

and dry matter intake (DMI) of lactating dairy cows.

MATERIALS AND METHODS

Eight multiparous Holstein cows (51 :t 19 DIM, 44.0 :I: 4.8 kg/d milk; mean 1: SD)

with ruminal cannulas were selected from the Michigan State University Dairy

99

Cattle Teaching and Research Center and randomly assigned to treatment
sequence in a crossover experiment with a 10-d diet adaptation period and 3 d
between treatments. Treatments were intraruminal infusion of 1.26 mol Na
propionate (2.33 :l: 0.06 L, 0.54 M, pH 6.0) over the course of either 5 min (FAST)
or 15 min (SLOW) at each meal for 24 h. The experimental diet is shown in
Table 7.1; the diet was formulated to promote endogenous propionate production
so that propionate would provide the primary limitation to feed intake during the
experiment. Throughout the experiment, cows were housed indoors in tie stalls
and fed a total mixed ration once daily (1130 h) at 110% of expected intake.
Cows were not allowed access to feed between 1000 h and 1130 h each day
while orts and the amount of feed offered were weighed for each cow. Diet
ingredients and orts samples were dried in a 55°C forced-air oven for 72 h and
analyzed for dry matter concentration. Ingredient samples were ground with a
Wiley mill (1-mm screen; Authur H. Thomas) and analyzed for ash, neutral
detergent fiber, crude protein, and starch content as previously described
(Bradford and Allen, 2004). At the end of each infusion period, ruminal contents
were sampled from 5 sites throughout rumen and squeezed through a nylon
screen (1 mm pore size) to collect the liquid phase. Ruminal fluid pH was
measured using a portable pH meter (model 230A, ATI Orion, Boston, MA), and
samples were frozen at -20° C until analysis. Concentrations of volatile fatty
acids in ruminal fluid were determined as previously described (Oba and Allen,

2003b)

100

Table 7.1 Ingredients and nutrient composition of
experimental diet.1

 

 

Item

Diet ingredients
Corn silage 31.0
High moisture corn grain 28.4
Alfalfa haylage 14.8
Modified expeller soybean meal2 8.9
Mineral and vitamin mix3 8.6
Soybean meal 8.4

Nutrient composition
DM 46.8
Organic matter 92.8
Starch 30.1
NDF 27.3
CP 16.1

 

1Values other than DM are expressed as % of dietary DM.
ZSoyPlus (donation from West Central Soy, Ralston, IA).

Mineral and vitamin mix contained 67.9% dry ground corn,
13.1% limestone, 8.5% magnesium sulfate, 6.4% salt,
2.9% dicalcium phosphate, 0.9% trace mineral premix, and
0.3% vitamin ADE premix.

On treatment days, feeding behavior was monitored by a computerized data
acquisition system (Dado and Allen, 1993) that triggered peristaltic infusion
pumps at the initiation of each new meal. Weight of the feed manger was
monitored every 5 s, and when the running SD reached a threshold, a
computerized “eating flag” was triggered. Because of differences in behavior, the
SD threshold was set for each cow individually based on 3 d of observation prior
to the initial treatment (range = 1.25 to 2.75). Infusions were initiated only when
the eating flag was triggered 5 times in a 100 s period. Mid-meal infusions were
prevented by requiring that the eating flag was triggered less than 13 times in the

preceding 5 min. Consecutive infusions began at least 15 min apart under both

101

treatment protocols to prevent potential treatment bias. Feeding behavior data
were analyzed to quantify meal size, meal length, number of meals, and
intermeal interval, and statistical analyses were carried out using a mixed model
with fixed effects of period, treatment, and period by treatment interaction, and

the random effect of cow.

RESULTS AND DISCUSSION

Triggering infusions based on real-time monitoring of feeding behavior resulted in
reasonably good relationships between infusion events and meals identified by
post-hoe analysis. There were a mean of 2.9 false positives (infusions without
meals) out of a total of 15.6 infusions per cow/d, and 0.1 false negatives (meals
without infusions) per cow/d. A numerically greater volume of Na propionate was
infused for the SLOW treatment compared to FAST (38.6 vs. 34.1 Ud, P = 0.21),
due in part to a numerically greater number of false positive infusions (3.63 vs.

2.25 per d, P = 0.19).

Compared to the days prior to infusion, propionate infusion at meals depressed
DMI by 20% and 23%, respectively, for SLOW and FAST. However, rate of
infusion had no detectable influence on DMI or on measures of feeding behavior
(all P > 0.10, Table 7.2). In addition, to assess whether the numerically greater
amount of propionate infused for SLOW biased these results, energy intake from
both infusate and feed consumed was calculated; treatment had no effect on

calculated ME intake (data not shown). Propionate infusion rate did not alter

102

ruminal pH (P = 0.38, mean = 6.38) or ruminal propionate concentration (P =

0.74) at the end of the infusion periods.

Table 7.2 Effects of propionate infusion rate on feeding behavior of lactating

 

 

cows1.

Item FAST SLOW SEM P
Dry matter intake (kg/d) 18.6 19.1 1.3 0.61
Meal size (kg DM) 1.48 1.49 0.17 0.93
Meals/d 12.88 12.75 0.70 0.87
Meal length (min) 22.6 24.3 1.1 0.20
lntermeal interval (min) 82.8 82.6 5.7 0.98

 

1
Values are least square means, n = 8.

Data from 2 preliminary studies were included in an expanded dataset to
examine whether DMI responses to infusion rate are dose-dependent. Protocols
from these 2 studies were identical to those described above, except that
solutions of Na propionate were of different concentrations (0.27 M and 1.0 M,
respectively). The preliminary experiments were randomized complete block
designs, so DMI depression is expressed as a percent of DMI on the preceding 2
d to account for individual variation in baseline DMI. lntraruminal infusion of
propionate at the beginning of spontaneous meals decreased DMI in a dose-
dependent manner (Figure 7.1A), similar to what has been observed previously
with continuous infusion in cows (Oba and Allen, 2003d) and meal-initiated
infusions in goats (Baile and Mayer, 1969). Depression of DMI was not
influenced by FAST or SLOW across experiments. However, amount of
propionate infused per meal better predicted intake responses than did total

amount of propionate infused (Figure 7.1 B).

103

.3"

100-

40-

20-

04

Dry matter intake depression (%)

 

 

-20

 

I I l I

5 13 1'5 20 2s 30 35
Total propionate infused (mol)

100 -
80 ~
60 -
4o -

20.

 

Dry matter intake depression (%)

 

’20 I I I l
0.5 1 1.5 2 2.5 3

 

Propionate per infusion (mol)

Figure 7.1 Propionate infusion at spontaneous meals depresses feed intake in a
dose-dependent manner. FAST: O, SLOW: A. (A) Total amount of propionate
infused over 24 h, but not infusion rate, determined degree of dry matter intake
depression by infused propionate. R2 = 0.33, P < 0.001, n = 38. (B) Amount of
propionate infused at each meal was a better predictor of dry matter intake

depression than total amount infused. R2 = 0.83, P < 0.001, n = 38.

104

The lack of effect of infusion rate on feeding behavior and dry matter intake may
be because our infusion rates were not sufficiently different in this experiment.
Meal lengths across treatments averaged 23.1 :t 14.2 min (SD), and 73% of
meals were longer than 15 min (Figure 7.2), suggesting that the SLOW
treatment could have been infused over a longer period. Another possibility is
that rate of propionate activation by propionyl-CoA synthetase may be limiting,
resulting in little change in hepatic uptake of propionate across treatments.
Propionyl-CoA synthetase has a reported Km for propionate of 1.3 mM (Ricks
and Cook, 1981), and enzyme activity increases as the concentration of
propionate reaching the liver increases. However, portal propionate
concentrations in cows nearing peak lactation averaged 1.86 mM over the course
of the day (calculated from Reynolds et al., 2003). Here we observed similar
DMI with higher starch and NDF concentrations in the diet (compared to that of
Reynolds et al., 2003), which was expected to promote greater endogenous
propionate production. Combining propionate infusion with rapid propionate
production during meals in this experiment may have increased propionate
concentrations at the liver to the extent that hepatic propionyl-CoA synthetase
activity was maximized. Although hepatic extraction of propionate is typically
quite efficient, we have reported that peripheral propionate concentrations
increased linearly with rate of intraruminal propionate infusion (Bradford et al.,
2006b). Therefore, it is plausible that liver capacity for activation of propionate

was saturated during meals in this experiment. Given these potential limitations,

105

the possibility remains that the post-absorptive fate of propionate influences DMI
in lactating dairy cows.

60-

Count

 

 

0 1 0 20 30 4O

50 60 7o 80 90
Meal length (min)

Figure 7.2 Distribution of meal lengths on infusion days. The histogram

represents 210 meal bouts from 2 d (n = 8 cows).

106

CHAPTER 8

CONCLUSIONS

The aim of these investigations was to provide insight into the mechanism by
which propionate affects feeding behavior of lactating dairy cows. We initially
used dietary treatments to alter propionate flux to the liver and measured a large
number of potentially related variables in 32 cows. Few relationships between
feed intake response and metabolic parameters, production variables, or hepatic
gene expression patterns were discovered. However, significant relationships
between feed intake response and 2 measures of insulin dynamics do suggest

that metabolic state influences individual response to propionate.

In our second study, we tested the hypothesis that propionate depresses feed
intake by stimulating leptin secretion. Plasma leptin responses to both short-term
jugular infusions and longer-term ruminal infusions of propionate were minimal
and did not explain feed intake responses to ruminal infusion. We concluded that

propionate's effects on feed intake are not mediated by leptin.

In a series of 3 experiments, we investigated the effect of increased glucose
demand on feed intake response to highly fermentable diets and propionate
infusion. In each of the experiments, phlorizin not only increased glucose
demand but also stimulated lipolysis in cows in both positive and negative energy

balance. This secondary effect of phlorizin most likely increased hepatic uptake

107

of fatty acids, which serve as important fuels for the ruminant liver. Because of
this confounding effect on fatty acid metabolism, the lack of an effect of phlorizin

on feed intake is not inconsistent with the hepatic oxidation hypothesis.

Finally, we investigated whether rate of propionate infusion administered at
voluntary meals would influence feed intake. We expected that a more rapid
infusion rate would depress feed intake to a greater extent because more
propionate would be forced into oxidative pathways, as opposed to
gluconeogenesis. However, we found no effect of infusion rate on feeding
behavior. Despite the fact that our hypothesis was not supported, a number of
factors could explain the lack of treatment effect, including the possibility that our
treatments did not change flux of propionate into the liver. Therefore, in this

experiment, the hepatic oxidation hypothesis could not be critically evaluated.

The complexities of the hepatic oxidation hypothesis make it extremely difficult to
design experiments that could reasonably lead to its rejection. According to the
hypothesis, satiety is driven by the accumulation of hepatic ATP; this single
concentration is dependent on the balance between both catabolic and anabolic
processes in the liver, and is undoubtedly variable over time. While it is attractive
to attempt to measure changes in hepatic energy charge associated with
treatments, we have not had success in measuring significant changes over the
course of a meal (unpublished data, Longuski, Bradford, and Allen).

Furthermore, in vivo assessment of fluxes through oxidative and anabolic

108

processes in ruminants is technically challenging, and because multiple fuels
must be considered (eg. fatty acids and propionate), extremely expensive.
Despite the difficulty of “ruling in or ruling out” the hepatic oxidation hypothesis,
the evidence in its favor is compelling enough to warrant further investigation into

its potential role in regulation of feed intake of ruminants.

One difficulty in studying feed intake regulation in general is that multiple
mechanisms are always at play in determining feed intake and feeding behavior,
and it can be difficult to ensure that the mechanism in question is the dominant
one during the experiment. A potential solution to this problem is to focus on
intermediates in the proposed pathway. Although this approach is somewhat
less satisfying in that feeding behavior is the variable of greatest interest,
biochemical measurements are less likely to be masked by confounding factors

than are measures of animal behavior.

If the hepatic oxidation hypothesis is valid, then some means of communication
between hepatocytes and vagal afferents is required. This may be a fruitful point
in the pathway for assessing the effects of propionate infusions on potential
mediators. One clear candidate for investigation is AM PK; phosphorylation state
of AMPK can be maintained by careful preservation of liver samples, and specific
antibodies for phospho-AMPK are commercially available for use in
immunoblotting procedures. Infusion of Na propionate vs. Na acetate over the

course of several hours should decrease the relative phosphorylation state of

109

hepatic AMPK if it is involved in the signaling pathway mediating propionate's
effects. The other proposed mediator at that step, Na‘lK+ ATPase, is more
difficult to investigate, because ATP is thought to alter its activity primarily at the
substrate level. However, phosphorylation of various ion channels could be
quantified in the same manner as for AMPK. Because of the large number of ion
channels that could be involved, a proteomics approach may be ideal for this
investigation. Plasma membranes could be isolated from liver samples, and
mass spectroscopy could be used to identify peptide fragments derived from ion
channels with and without phosphate groups attached. In this way, the relative
phosphorylation state of multiple ion channels could be measured

simultaneously.

In addition to a greater focus on pathway intermediates, different approaches to
applied studies may also yield more conclusive data. The results of the phlorizin
experiments described in chapters 4 —6 suggest that our focus at the outset was
too narrowly focused on metabolism of propionate. One of the strengths of the
hepatic oxidation hypothesis is its ability to integrate metabolism of carbohydrate,
fats, and proteins into a single mechanism, and future research needs to more
explicitly consider potential interactions of nutrient classes. For example,
increasing fatty acid oxidation (e.g. in early lactation or with phlorizin treatment)
increases the amount of acetyl-CoA generated through B-oxidation. When
glucose demand is high relative to availability of gluconeogenic substrates,

oxidation of acetyl-CoA can be limited by accumulation of reducing equivalents

110

(high NADHzNAD“) and by availability of TCA cycle intermediates. When
propionate enters the hepatocyte, it can be quickly converted to oxaloacetate,
providing substrate for the citrate synthase reaction, and during conversion to
glucose, propionate utilizes reducing equivalents. Therefore, propionate can
stimulate ATP production by removing the primary constraints to oxidation of

acetyl-CoA, even if it is not oxidized itself.

In support of this view, Oba and Allen (2003a) showed in a dose-response
experiment that low rates of propionate infusion decreased feed intake in early
lactation cows, but not in mid lactation cows. Infusion of propionate at lower
doses increased plasma glucose concentration in both stages of lactation,
suggesting that this rate of infusion did not overwhelm gluconeogenic capacity
and likely did not greatly increase propionate oxidation. However, propionate
likely stimulated oxidation of acetyl-CoA, increasing ATP production in early
lactation cows despite the use of propionate for glucose production. This
hypothesis is supported by the observed decrease in plasma BHBA
concentration in early lactation cows at lower infusion rates while NEFA
concentration remained elevated (Oba and Allen, 2003a). Differences in the
concentration of acetyl-CoA in the liver might help explain differences in feeding
behavior responses between early and mid-lactation cows as well as differences
among individual cows. The experiment described in chapter 1 is an ideal design
to assess whether hepatic concentration of acetyl-CoA explains individual

differences in response to highly fermentable diets; unfortunately, samples were

111

not stored for preservation of acetyl-CoA. Future work regarding the hepatic
oxidation hypothesis for ruminants will undoubtedly need to consider such

potential interactions.

Despite our inability to add support for the hepatic oxidation hypothesis in these
experiments, we have made progress on the mechanism for propionate
regulation of feed intake. In the work described in chapter 2, we presented
strong evidence that leptin secretion is not an important component of feed
intake regulation by propionate in lactating cows fed ad libitum. The experiment
reported in chapter 6 added to the evidence that insulin is not a required
component of the mechanism for propionate's effects on feed intake. Finally, the
series of phlorizin studies provide strong evidence that increased glucose
demand and energy requirements do not result in compensatory increases in
feed intake within 7 d. These results narrow the scope of possible mechanisms
by which propionate may act to alter feeding behavior. Hopefully, future research
will be able to more clearly define the mechanism for propionate regulation of
feed intake, allowing for better predictions of individual responses to dietary

formulations and improving the management of lactating dairy cows.

112

BIBLIOGRAPHY

Agca, C., R. B. Greenfield, J. R. Hartwell, and S. S. Donkin. 2002. Cloning and
characterization of bovine cytosolic and mitochondrial PEPCK during transition to
lactation. Physiol. Genomics. 11(2):53-63.

Allen, M. S. 2000. Effects of diet on short-term regulation of feed intake by
lactating dairy cattle. J Dairy Sci. 83(7):1598-1624.

Allen, M. S., B. J. Bradford, and K. J. Harvatine. 2005. The cow as a model to
study food intake regulation. Annu Rev Nutr. 25:523-547.

Allen, M. S., M. Oba, and C. S. Mooney. 2000. Automated system for collection
of ruminal fluid and blood of ruminants. J Dairy Sci. 83(Suppl. 1):288.

Amaral-Phillips, D. M., A. D. McGilIiard, G. L. Lindberg, J. J. Veenhuizen, and J.
W. Young. 1993. Effects of decreased availability of glucose for dairy cows. J
Dairy Sci. 76(3):752-761.

Amstalden, M., M. R. Garcia, S. W. Williams, R. L. Stanko, S. E. Nizielski, C. D.
Morrison, D. H. Keisler, and G. L. Williams. 2000. Leptin gene expression,
circulating leptin, and luteinizing hormone pulsatility are acutely responsive to
short-term fasting in prepubertal heifers: relationships to circulating insulin and
insulin-like growth factor l(1). Biol Reprod. 63(1):127-133.

Anand, B. K., G. S. Chhina, K. N. Sharma, S. Dua, and B. Singh. 1964. Activity of
single neurons in the hypothalamic feeding centers: effect of glucose. Am J
Physiol. 207(5):1 146-1 154.

Andersson, U., K. Filipsson, C. R. Abbott, A. Woods, K. Smith, S. R. Bloom, D.
Carling, and C. J. Small. 2004. AMP-activated protein kinase plays a role in the
control of food intake. J Biol Chem. 279(13):12005-12008.

Anil, M. H. and J. M. Forbes. 1980. Feeding in sheep during intraportal infusions
of short-chain fatty acids and the effect of liver denervation. J Physiol. 298:407-
414.

Anil, M. H. and J. M. Forbes. 1988. The roles of hepatic nerves in the reduction
of food intake as a consequence of intraportal sodium propionate administration
in sheep. Q J Exp Physiol. 73(4):539-546.

AOAC. 1990. Official Methods of Analysis. Vol. 2. 15th ed. AOAC Intl., Arlington,
VA.

Arrnentano, L. E. 1992. Ruminant hepatic metabolism of volatile fatty acids,
lactate and pyruvate. J Nutr. 122(3 Suppl):838-842.

113

Ash, R. W. 1959. Inhibition and excitation of reticqu-rumen contractions following
the introduction of acids into the rumen and abomasum. J Physiol. 147(1):58—73.

Ash, R. W., R. J. Pennington, and R. S. Reid. 1964. The effect of short-chain
fatty acids on blood glucose concentration in sheep. Biochem J. 90(2):353-360.

Bagdade, J. D., E. L. Blerman, and D. Porte, Jr. 1967. The significance of basal
insulin levels in the evaluation of the insulin response to glucose in diabetic and
nondiabetic subjects. J Clin Invest. 46(10):1549-1557.

Baile, C. A. 1971. Metabolites as feedbacks for control of feed intake and
receptor sites in goats and sheep. Physiol Behav. 7(6):819-826.

Baile, C. A. and J. M. Forbes. 1974. Control of feed intake and regulation of
energy balance in ruminants. Physiol Rev. 54(1):160—214.

Baile, C. A. and J. Mayer. 1968. Effects of intravenous versus intraruminal
injections of acetate on feed intake of goats. J Dairy Sci. 51 :1490-1494.

Baile, C. A. and J. Mayer. 1969. Depression of feed intake of goats by
metabolites injected during meals. Am J Physiol. 217(6):1830—1836.

Baile, C. A. and W. H. Pfander. 1966. A possible chemosensitive regulatory
mechanism of ovine feed intake. Am J Physiol. 210(6):1243—1250.

Baird, G. D., M. A. Lomax, H. W. Symonds, and S. R. Shaw. 1980. Net hepatic
and splanchnic metabolism of lactate, pyruvate and propionate in dairy cows in
vivo in relation to lactation and nutrient supply. Biochem J. 186(1):47-57.

Baird, G. D., J. G. van der Walt, and E. N. Bergman. 1983. Whole-body
metabolism of glucose and lactate in productive sheep and cows. Br J Nutr.
50(2):249-265.

Baldwin, R. L. and J. R. Knapp. 1993. Recombinant bovine somatotropin's
effects on patterns of nutrient utilization in lactating dairy cows. Am J Clin Nutr.
58(2 Suppl):282S-286$.

Ballard, F. J. 1965. Glucose utilization in mammalian liver. Comp Biochem
Physiol. 14:437-443.

Barber, M. C., R. A. Clegg, M. T. Travers, and R. G. Vernon. 1997. Lipid
metabolism in the lactating mammary gland. Biochim Biophys Acta. 1347(2-
3):101-126.

Baskin, D. G., T. M. Hahn, and M. W. Schwartz. 1999. Leptin sensitive neurons
in the hypothalamus. Horm Metab Res. 31(5):345-350.

114

Bauman, D. E., C. L. Davis, and H. F. Bucholtz. 1971. Propionate production in
the rumen of cows fed either a control or high-grain, low-fiber diet. J Dairy Sci.
54(9): 1 282-1 287.

Benoit, S. C., D. J. Clegg, R. J. Seeley, and S. C. Woods. 2004. Insulin and
leptin as adiposity signals. Recent Prog Horm Res. 59:267-285.

Benson, J. A., C. K. Reynolds, P. C. Aikman, B. Lupoli, and D. E. Beever. 2002.
Effects of abomasal vegetable oil infusion on splanchnic nutrient metabolism in
lactating dairy cows. J Dairy Sci. 85(7):1804-1814.

Bergman, E. N. 1990. Energy contributions of volatile fatty acids from the
gastrointestinal tract in various species. Physiol Rev. 70(2):567-590.

Berthoud, H. R. 2004. Anatomy and function of sensory hepatic nerves. Anat Rec
A Discov Mol Cell Evol Biol. 280A(1):827-835.

Bingham, S. A. 1990. Mechanisms and experimental and epidemiological
evidence relating dietaryfibre (non-starch polysaccharides) and starch to
protection against large bowel cancer. Proc Nutr Soc. 49(2):153-171.

Black, A. L., J. Luick, F. Moller, and R. S. Anand. 1966. Pyruvate and propionate
metabolism in lactating cows. Effect of butyrate on pyruvate metabolism. J Biol
Chem. 241 (22):5233-5237.

Block, S. S., W. R. Butler, R. A. Ehrhardt, A. W. Bell, M. E. Van Amburgh, and Y.
R. Boisclair. 2001. Decreased concentration of plasma leptin in periparturient
dairy cows is caused by negative energy balance. J Endocrinol. 171 (2):339-348.

Blundell, J. E., R. J. Stubbs, D. A. Hughes, S. Whybrow, and N. A. King. 2003.
Cross talk between physical activity and appetite control: does physical activity
stimulate appetite? Proc Nutr Soc. 62(3):651-661.

Boutellier, S., T. A. Lutz, M. Volkert, and E. Scharrer. 1999. 2-Mercaptoacetate,
an inhibitor of fatty acid oxidation, decreases the membrane potential in rat liver
in vivo. Am J Physiol. 277(1):R301-305.

Bradford, B. J. and M. S. Allen. 2004. Milk fat responses to a change in diet
fermentability vary by production level in dairy cattle. J. Dairy Sci. 87(11):3800-
3807.

Bradford, B. J. and M. S. Allen. 2005. Phlorizin administration increases hepatic
gluconeogenic enzyme mRNA abundance but not feed intake in late-lactation
dairy cows. J Nutr. 135(9):2206-2211.

Bradford, B. J., A. D. Gour, A. S. Nash, and M. S. Allen. 2006a. Propionate

challenge tests have limited value for investigating bovine metabolism. J Nutr.
136(7):1915-1920.

115

Bradford, B. J., M. Oba, R. A. Ehrhardt, Y. R. Boisclair, and M. S. Allen. 2006b.
Propionate is not an important regulator of plasma leptin concentration in dairy
cattle. Dom Anim Endocrinol. 30(2):65-75.

Brockman, R. P. 1984. Effect of glycemic changes on lipolysis in sheep in vivo.
Metabolism. 33(4):329-331.

Brockman, R. P. and J. G. Manns. 1974. Effects of glucagon on activities of
hepatic enzymes in sheep. Cornell Vet. 64(2):217-224.

Brown, A. J., S. M. Goldsworthy, A. A. Barnes, M. M. Eilert, L. Tcheang, D.
Daniels, A. l. Muir, M. J. Wigglesworth, l. Kinghom, N. J. Fraser, N. B. Pike, J. C.
Strum, K. M. Steplewski, P. R. Murdock, J. C. Holder, F. H. Marshall, P. G.
Szekeres, S. Wilson, D. M. lgnar, S. M. Foord, A. Wise, and S. J. Dowell. 2003.
The Orphan G protein-coupled receptors GPR41 and GPR43 are activated by
propionate and other short chain carboxylic acids. J Biol Chem. 278(13):11312-
11319.

Carattino, M. D., R. S. Edinger, H. J. Grieser, R. Wise, D. Neumann, U.
Schlattner, J. P. Johnson, T. R. Kleyman, and K. R. Hallows. 2005. Epithelial
sodium channel inhibition by AMP-activated protein kinase in oocytes and
polarized renal epithelial cells. J Biol Chem. 280(18):17608-17616.

Carter, R. R. and W. L. Grovum. 1990. Factors affecting the voluntary intake of
food by sheep. 5. The inhibitory effect of hypertoniclty in the rumen. Br J Nutr.
64(1 ):285-299.

Chelikani, P. K., J. D. Ambrose, D. H. Keisler, and J. J. Kennelly. 2004. Effect of
short-term fasting on plasma concentrations of leptin and other hormones and
metabolites in dairy cattle. Domest Anim Endocrinol. 26(1):33-48.

Chiang, S.-H. and O. A. MacDougald. 2003. Will fatty worms help cure human
obesity? Trends Genet. 19(10):523-525.

Choi, B. R. and M. S. Allen. 1999. Intake regulation by volatile fatty acids and
physical fill. S. Afr. J. Anim. Sci. 29(lSRP):40-41 (Abstr.).

Chomczynski, P. and N. Sacchi. 1987. Single-step method of RNA isolation by
acid guanidinium thiocyanate-phenol-chloroforrn extraction. Anal Biochem.
162(1):156-159.

Chow, J. C. and B. W. Jesse. 1992. Interactions between gluconeogenesis and
fatty acid oxidation in isolated sheep hepatocytes. J Dairy Sci. 75(8):2142-2148.

Cochran, R. C., D. C. Adams, J. D. Wallace, and M. L. Galyean. 1986. Predicting
digestibility of different diets with internal markers: evaluation of four potential
markers. J Anim Sci. 63:1476-1483. .

116

Conrad, H. R., A. D. Pratt, and J. W. Hibbs. 1964. Regulation of feed intake in
dairy cows. I. Change in importance of physical and physiological factors with
increasing digestibility. J Dairy Sci. 47:54-62.

Correa, C. E., R. D. Shaver, M. N. Pereira, J. G. Lauer, and K. Kohn. 2002.
Relationship between corn vitreousness and ruminal in situ starch degradability.
J Dairy Sci. 85(11):3008-3012.

Cummings, J. H., E. W. Pomare, W. J. Branch, C. P. Naylor, and G. T.
Macfarlane. 1987. Short chain fatty acids in human large intestine, portal, hepatic
and venous blood. Gut. 28(10):1221-1227.

Dado, R. G. and M. S. Allen. 1993. Continuous computer acquisition of feed and
water intake, chewing reticular motility, and ruminal pH of cattle. J Dairy Sci.
76(6):1589-1600.

Dankert, J., J. B. Zijlstra, and B. G. Wolthers. 1981. Volatile fatty acids in human
peripheral and portal blood: quantitative determination vacuum distillation and
gas chromatography. Clin Chim Acta. 110(2-3):301-307.

De Feo, P., G. Perriello, E. Torlone, M. M. Ventura, F. Santeusanio, P. Brunetti,
J. E. Gerich, and G. B. Bolli. 1989. Demonstration of a role for growth hormone in
glucose counterregulation. Am J Physiol Endocrinol Metab. 256(6):E835-843.

Demigne, C., C. Yacoub, C. Remesy, and P. Fafoumoux. 1986. Propionate and
butyrate metabolism in rat or sheep hepatocytes. Biochim Biophys Acta.
875(3):535-542.

Doepel, L., H. Lapierre, and J. J. Kennelly. 2002. Peripartum Performance and
Metabolism of Dairy Cows in Response to Prepartum Energy and Protein Intake.
J. Dairy Sci. 85(9):2315-2334.

Dohoo, l. R., K. Leslie, L. DesCoteaux, A. Fredeen, P. Dowling, A. Preston, and
W. Shewfelt. 2003. A meta-analysis review of the effects of recombinant bovine
somatotropin. 1. Methodology and effects on production. Can J Vet Res.
67(4):241-251.

Dowden, D. R. and D. R. Jacobson. 1960. Inhibition of appetite in dairy cattle by
certain intermediate metabolities. Nature. 188:148-149.

Drackley, J. K. and J. B. Andersen. 2006. Splanchnic metabolism of long-chain
fatty acids in ruminants. Pages 199-224 in Ruminant physiology: Digestion,
metabolism and impact of nutrition on gene expression, immunology and stress.
K. Serjsen, H. T., and M. O. Nielsen, eds. Wageningen Academic Publishers,
Wageningen, The Netherlands.

117

Dunshea, F. R., A. W. Bell, and T. E. Trigg. 1989. Relations between plasma
non-esterified fatty acid metabolism and body fat mobilization in primiparous
lactating goats. British J Nutr. 62(1):51-61.

Ehrenkranz, J. R., N. G. Lewis, C. R. Kahn, and J. Roth. 2005. Phlorizin: a
review. Diabetes/Metabolism Research and Reviews. 21(1):31-38.

Ehrhardt, R. A., R. M. Slepetis, J. Siegal-Willott, M. E. Van Amburgh, A. W. Bell,
and Y. R. Boisclair. 2000. Development of a specific radioimmunoassay to
measure physiological changes of circulating leptin in cattle and sheep. J
Endocrinol. 166(3):519-528.

Elliot, J. M., H. W. Symonds, and B. Pike. 1985. Effect on feed intake of infusing
sodium propionate or sodium acetate into a mesenteric vein of cattle. J Dairy Sci.
68(5):1165-1170.

Emery, R. S. and T. H. Herdt. 1991. Lipid nutrition. Vet Clin North Am Food Anim
Pract. 7(2):341-352.

Ensinck, J. W., C. Shepard, R. J. Dudl, and R. H. Williams. 1972. Use of
benzamidine as a proteolytic inhibitor in the radioimmunoassay of glucagon in
plasma. J Clin Endocrinol Metab. 35(3):463-467.

Famingham, D. A. and C. C. Whyte. 1993. The role of propionate and acetate in
the control of food intake in sheep. Br J Nutr. 70(1):37-46.

Forbes, J. M. 1988. Metabolic aspects of the regulation of voluntary food intake
and appetite. Nutrition Research Reviews. 1:145-168.

Forbes, J. M. 1995. Voluntary Food Intake and Diet Selection in Farm Animals.
CAB International, Oxon, UK.

Forbes, J. M. 1996. Integration of regulatory signals controlling forage intake in
ruminants. J Anim Sci. 74(12):3029-3035.

Forman, D. T. 1988. The effect of ethanol and its metabolites on carbohydrate,
protein, and lipid metabolism. Annals of Clinical and Laboratory Science.
18(3):181-189.

Foster, L. A., N. K. Ames, and R. S. Emery. 1991. Food intake and serum insulin
responses to intraventricular infusions of insulin and lGF-l. Physiol Behav.
50(4):745-749.

Frederich, R. C., B. Lollmann, A. Hamann, A. Napolitano-Rosen, B. B. Kahn, B.

B. Lowell, and J. S. Flier. 1995. Expression of ob mRNA and its encoded protein
in rodents. Impact of nutrition and obesity. J Clin Invest. 96(3):1658-1663.

118

Friedman, M. l. 1995. Control of energy intake by energy metabolism. Am J Clin
Nutr. 62(5 Suppl):10968-1 100$.

Friedman, M. l. 1998. Fuel partitioning and food intake. Am J Clin Nutr. 67(3
Suppl):513S-5188.

Friedman, M. l., G. Graczyk-Millbrandt, H. Ji, N. E. Rawson, and M. D.
Osbakken. 2003. 2,5-Anhydro--mannitol increases hepatocyte sodium:
transduction of a hepatic hunger stimulus? Biochim Biophys Acta (BBA) -
Molecular Cell Research. 1642(1-2):53-58.

Friedman, M. l., R. B. Harris, H. Ji, I. Ramirez, and M. G. Tordoff. 1999. Fatty
acid oxidation affects food intake by altering hepatic energy status. Am J Physiol.
276(4):R1046-1053.

Friedman, M. l. and E. M. Stricker. 1976. The physiological psychology of
hunger: a physiological perspective. Psychological Review. 83(6):409-431.

Frobish, R. A. and C. L. Davis. 1977. Effects of abomasal infusions of glucose
and propionate on milk yield and composition. J Dairy Sci. 60:204-209.

Gaspers, L. D. and A. P. Thomas. 2005. Calcium signaling in liver. Cell Calcium.
38(3-4):329-342.

Gaynor, P. J., K. J. Lookingland, and H. A. Tucker. 1995. 5-
Hydroxytryptaminergic receptor-stimulated growth hormone secretion occurs
independently of changes in peripheral somatostatin concentration. Proceedings
of the Society for Experimental Biology and Medicine. 209(1):79-85.

Goering, H. K. and P. J. Van Soest. 1970. Forage Fiber Analysis (Apparatus,
Reagents, Procedures, and Some Applications). Agric. Handbook No. 379.
ARS—USDA, Washington, DC.

Gonzalez-Manchon, C., A. Martin-Requero, M. S. Ayuso, and R. Parrilla. 1992.
Role of endogenous fatty acids in the control of hepatic gluconeogenesis. Arch
Biochem Biophys. 292(1):95-101 .

Greenfield, R. B., M. J. Cecava, and S. S. Donkin. 2000. Changes in mRNA
expression for gluconeogenic enzymes in liver of dairy cattle during the transition
to lactation. J Dairy Sci. 83(6):1228-1236.

Griinari, J. M., M. A. McGuire, D. A. Dwyer, D. E. Bauman, and D. L. Palmquist.
1997. Role of insulin in the regulation of milk fat synthesis in dairy cows. J Dairy
Sci. 80(6):1076-1084.

119

 

Grovum, W. L. 1995. Mechanisms explaining the effects of short chain fatty acids
on feed intake in ruminants - osmotic pressure, insulin and glucagon. Pages 173-
197 in Ruminant Physiology: Digestion, Metabolism, Growth and Reproduction.
W. v. Englehardt, S. Leonhard-Marek, G. Breves, and D. Geisecke, eds.
Ferdinand Enke Verlag, Stuttgart, Germany.

Hach, C. C., B. K. Bowden, A. B. Lopelove, and S. V. Brayton. 1987. More
powerful peroxide Kjeldahl digestion method. J AOAC. 70:783—787.

Hallows, K. R., G. P. Kobinger, J. M. Wilson, L. A. Witters, and J. K. Foskett.
2003. Physiological modulation of CFTR activity by AMP-activated protein kinase
in polarized T84 cells. Am J Physiol Cell Physiol. 284(5):C1297-1308.

Harmon, D. L. 1992. Impact of nutrition on pancreatic exocrine and endocrine
secretion in ruminants: a review. J Anim Sci. 70(4):1290-1301.

Harris, R. B. 2000. Leptin—much more than a satiety signal. Annu Rev Nutr.
20:45-75.

Hartwell, J. R., M. J. Cecava, and S. S. Donkin. 2001. Rumen undegradable
protein, rumen-protected choline and mRNA expression for enzymes in

gluconeogenesis and ureagenesis in periparturient dairy cows. J Dairy Sci.
84(2):490-497.

Harvatine, K. J. and M. S. Allen. 2006. Effects of Fatty Acid Supplements on Milk
Yield and Energy Balance of Lactating Dairy Cows. J Dairy Sci. 89(3):1081-1091.

Hayirli, A., S. J. Bertics, and R. R. Grummer. 2002. Effects of slow-release insulin
on production, liver triglyceride, and metabolic profiles of Holsteins in early
lactation. J Dairy Sci. 85(9):2180-2191.

Horino, M., L. J. Machlin, F. Hertelendy, and D. M. Kipnis. 1968. Effect of short-
chain fatty acids on plasma insulin in ruminant and nonruminant species.
Endocrinol. 83(1):1 18-128.

Horn, C. C., A. Addis, and M. l. Friedman. 1999. Neural substrate for an
integrated metabolic control of feeding behavior. Am J Physiol. 276(1):R113-119.

Horn, C. C., H. Ji, and M. I. Friedman. 2004. Etomoxir, a fatty acid oxidation
inhibitor, increases food intake and reduces hepatic energy status in rats. Physiol
Behav. 81 (1 ):157-162.

Huang, B., P. Wu, M. M. Bowker-Kinley, and R. A. Harris. 2002. Regulation of
pyruvate dehydrogenase kinase expression by peroxisome proliferator-activated
receptor-alpha ligands, glucocorticoids, and insulin. Diabetes. 51(2):276-283.

Huntington, G. B. 1997. Starch utilization by ruminants: from basics to the bunk.
J Anim Sci. 75(3):852-867.

120

Huntington, G. B. and P. J. Reynolds. 1986. Net absorption of glucose, L-Iactate,
volatile fatty acids, and nitrogenous compounds by bovine given abomasal
infusions of starch or glucose. J Dairy Sci. 69(9):2428-2436.

Hurtaud, C., H. Rulquin, and R. Verite. 1993. Effect of infused volatile fatty acids
and caseinate on milk composition and coagulation in dairy cows. J Dairy Sci.
76(10):301 1-3020.

lllius, A. W., B. J. Tolkamp, and J. Yearsley. 2002. The evolution of the control of
food intake. Proc Nutr Soc. 61 (4):465-472.

lngvartsen, K. L. and J. B. Andersen. 2000. Integration of metabolism and intake
regulation: a review focusing on periparturient animals. J Dairy Sci. 83(7):1573-
1 597.

lngvartsen, K. L. and Y. R. Boisclair. 2001. Leptin and the regulation of food
intake, energy homeostasis and immunity with special focus on periparturient
ruminants. Dom Anim Endocrinol. 21(4):215-250.

lnui, A. 2000. Transgenic approach to the study of body weight regulation.
Pharrnacol Rev. 52(1):35-62.

Ji, H., G. Graczyk-Milbrandt, and M. l. Friedman. 2000. Metabolic inhibitors
synergistically decrease hepatic energy status and increase food intake. Am J
Physiol. 278(6):R1579-1582.

Jitrapakdee, S. and J. C. Wallace. 1999. Structure, function and regulation of
pyruvate carboxylase. Biochem J. 340 ( Pt 1):1-16.

Johnson, M. S., S. C. Thomson, and J. R. Speakman. 2001. Limits to sustained
energy intake. I. Lactation in the laboratory mouse Mus musculus. J Exp Biol.
204(Pt 11):1925-1935.

Karkalas, J. 1985. An improved enzymatic method for the determination of native
and modified starch. J Sci Food Agric. 36:1019-1027.

Kay, J. K., W. J. Weber, C. E. Moore, D. E. Bauman, L. B. Hansen, H. Chester-
Jones, B. A. Crooker, and L. H. Baumgard. 2005. Effects of week of lactation and

genetic selection for milk yield on milk fatty acid composition in holstein cows. J
Dairy Sci. 88(11):3886-3893.

Kirkup, A. J., C. E. Booth, I. P. Chessell, P. P. A. Humphrey, and D. Grundy.
1999. Excitatory effect of P2X receptor activation on mesenteric afferent nerves
in the anaesthetised rat. J Physiol. 520(2):551-563.

Knowlton, K. F., B. P. Glenn, and R. A. Erdman. 1998. Performance, ruminal

fermentation, and site of starch digestion in early lactation cows fed com grain
harvested and processed differently. J Dairy Sci. 81(7):1972-1984.

121

Koch, J. E., H. Ji, M. D. Osbakken, and M. l. Friedman. 1998. Temporal
relationships between eating behavior and liver adenine nucleotides in rats
treated with 2,5-AM. Am J Physiol. 274(3 Pt 2):R610-617.

Kotarski, S. F ., R. D. Waniska, and K. K. Thum. 1992. Starch hydrolysis by the
ruminal microflora. J Nutr. 122(1):178-190.

Kronfeld, D. S. 1957. The effects on blood sugar and ketone bodies of butyrate,
acetate and beta-OH butyrate infused into sheep. Aust J Exp Biol Med Sci.
35(3):257-265.

Langhans, W. 1996. Role of the liver in the metabolic control of eating: What we
know—and what we do not know. Neurosci Biobehav Rev. 20(1):145—153.

Langhans, W., E. Delprete, and E. Scharrer. 1991. Mechanisms of vasopressin's
anorectic effect. Physiol Behav. 49(1 ):169-176.

Langhans, W. and E. Scharrer. 1987. Evidence for a vagally mediated satiety
signal derived from hepatic fatty acid oxidation. J Auton Nerv Syst. 18(1):13-18.

Langhans, W. and E. Scharrer. 1992. Metabolic control of eating. World Rev Nutr
Diet. 7021-67.

Lee, S. H. and K. L. Hossner. 2002. Coordinate regulation of ovine adipose
tissue gene expression by propionate. J Anim Sci. 80(11):2840-2849.

Leek, B. F. and R. H. Harding. 1975. Sensory nervous receptors in the ruminant
stomach and the reflex control of reticulo-ruminal motility. Pages 60-76 in
Digestion and Metabolism in the Ruminant. l. W. McDonald and A. C. I. Warner,
eds. University of New England Publishing Unit, Armidale, Australia.

Leng, R. A. and E. F. Annison. 1963. Metabolism of acetate, propionate and
butyrate by sheep-liver slices. Biochem J. 86:319-327.

Leury, B. J., L. H. Baumgard, S. 8. Block, N. Segoale, R. A. Ehrhardt, R. P.
Rhoads, D. E. Bauman, A. W. Bell, and Y. R. Boisclair. 2003. Effect of insulin
and growth hormone on plasma leptin in periparturient dairy cows. Am J Physiol
Regul Integr Comp Physiol. 285(5):R1107-1115.

Leuvenink, H. G., E. J. Bleumer, L. J. Bongers, J. van Bruchem, and D. van der
Heide. 1997. Effect of short-term propionate infusion on feed intake and blood
parameters in sheep. Am J Physiol. 272(6 Pt 1):E997-1001.

Liesman, J. S., J. P. McNamara, A. V. Capuco, M. Binelli, W. K. Vanderkooi, R.
S. Emery, H. A. Tucker, and W. M. Moseley. 1995. Comparison of growth
horrnone-releasing factor and somatotropin: lipid and glucose metabolism in
dairy cows. J Dairy Sci. 78(10):2159-2166.

122

Light, P. E., C. H. R. Wallace, and J. R. B. Dyck. 2003. Constitutively active
adenosine monophosphate-activated protein kinase regulates voltage-gated
sodium channels in ventricular myocytes. Circulation. 107(15):1962-1965.

Lindsay, D. B. and B. P. Setchell. 1976. The oxidation of glucose, ketone bodies

and acetate by the brain of normal and ketonaemic sheep. J Physiol. 259(3):801-
823.

Looney, M. C., R. L. Baldwin, and C. C. Calvert. 1987. Gluconeogenesis in
isolated lamb hepatocytes. J Anim Sci. 64(1):283-294.

Lopes-Cardozo, M., l. Mulder, F. van Vugt, P. G. Hermans, S. G. van den Bergh,
W. Klazinga, and E. de Vries-Akkerrnan. 1975. Aspects of ketogenesis: control
and mechanism of ketone-body formation in isolated rat-liver mitochondria. Mol
Cell Biochem. 9(3):155-173.

Luckman, S. M. and C. B. Lawrence. 2003. Anorectic brainstem peptides: more
pieces to the puzzle. Trends Endocrinol Metab. 14(2):60-65.

Mackle, T. R., D. A. Dwyer, K. L. lngvartsen, P. Y. Chouinard, J. M. Lynch, D. M.
Barbano, and D. E. Bauman. 1999. Effects of insulin and amino acids on milk
protein concentration and yield from dairy cows. J Dairy Sci. 82(7):1512-1524.

Malhi, H., A. N. lrani, P. Rajvanshi, S. O. Suadicani, D. C. Spray, T. V.
McDonald, and S. Gupta. 2000. KATP channels regulate mitogenically induced
proliferation in primary rat hepatocytes and human liver cell lines. Implications for
liver growth control and potential therapeutic targeting. J Biol Chem.
275(34):26050-26057.

Matsunaga, N., T. Goka, K. T. Nam, S. Oda, A. Ohneda, and Y. Sasaki. 1997.
Inhibition of GH releasing factor (GRF)-induced GH secretion by intraruminal
infusion of volatile fatty acids (VFA) in sheep. Endocr J. 44(1):133-140.

Mayer, J. 1953. Glucostatic mechanism of regulation of food intake. N Engl J
Med. 249:13-16.

Mbanya, J. N., M. H. Anil, and J. M. Forbes. 1993. The voluntary intake of hay
and silage by lactating cows in response to ruminal infusion of acetate or
propionate, or both, with or without distension of the rumen by a balloon. Br J
Nutr. 69(3):713-720.

McCutcheon, S. N. and D. E. Bauman. 1986. Effect of chronic growth hormone
treatment on responses to epinephrine and thyrotropin-releasing hormone in
lactating cows. J Dairy Sci. 69(1):44-51.

McGuire, M. A., J. M. Griinari, D. A. Dwyer, and D. E. Bauman. 1995. Role of

insulin in the regulation of mammary synthesis of fat and protein. J Dairy Sci.
78(4):816-824. -

123

Meikle, A., M. Kulcsar, Y. Chilliard, H. Febel, C. Delavaud, D. Cavestany, and P.
Chilibroste. 2004. Effects of parity and body condition at parturition on endocrine
and reproductive parameters of the cow. Reproduction. 127(6):727-737.

Miki, T., B. Liss, K. Minami, T. Shiuchi, A. Saraya, Y. Kashima, M. Horiuchi, F.
Ashcroft, Y. Minokoshi, J. Roeper, and S. Seino. 2001. ATP-sensitive K+
channels in the hypothalamus are essential for the maintenance of glucose
homeostasis. Nat Neurosci. 4(5):507-512.

Miki, T. and S. Seino. 2005. Roles of KATP channels as metabolic sensors in
acute metabolic changes. J Mol Cell Cardiol. 38(6):917-925.

Mobbs, C. V., F. lsoda, H. Makimura, J. Mastaitis, T. Mizuno, I. W. Shu, K. Yen,
and X. J. Yang. 2005. Impaired glucose signaling as a cause of obesity and the
metabolic syndrome: The glucoadipostatic hypothesis. Physiol Behav. 85(1):3-
23.

Montgomery, M. J., L. H. Schultz, and B. R. Baumgardt. 1963. Effect of
intraruminal infusion of volatile fatty acids and lactic acid on voluntary hay intake.
J Dairy Sci. 46:1380-1384.

Moreno-Aliaga, M. J., K. L. Stanhope, and P. J. Havel. 2001. Transcriptional
regulation of the leptin promoter by insulin-stimulated glucose metabolism in 3t3-
l1 adipocytes. Biochem Biophys Res Commun. 283(3):544-548.

Nichols, C. G. and J. C. Koster. 2002. Diabetes and insulin secretion: whither
KATP? Am J Physiol Endocrinol Metab. 283(3):E403-412.

Niijima, A. 1983. Glucose-sensitive afferent nerve fibers in the liver and their role
in food intake and blood glucose regulation. J Auton Nerv Syst. 9(1):207-220.

Nordlie, R. C., J. D. Foster, and A. J. Lange. 1999. Regulation of glucose
production by the liver. Annu Rev Nutr. 19:379-406.

Novin, D., D. A. VanderWeele, and M. Rezek. 1973. Infusion of 2-deoxy-D-
glucose into the hepatic-portal system causes eating: evidence for peripheral
glucoreceptors. Science. 181 (102):858-860.

NRC. 2001. Nutritional Requirements of Dairy Cattle. 7th rev. ed. Natl. Acad.
Sci., Washington, DC.

Oba, M. and M. S. Allen. 2000. Effects of brown midrib 3 mutation in corn silage
on productivity of dairy cows fed two concentrations of dietary neutral detergent
fiber: 1. Feeding behavior and nutrient utilization. J Dairy Sci. 83(6):1333-1341.

Oba, M. and M. S. Allen. 2003a. Dose-response effects of intrauminal infusion of

propionate on feeding behavior of lactating cows in early or midlactation. J Dairy
Sci. 86(9):2922-2931.

124

Oba, M. and M. S. Allen. 2003b. Effects of corn grain conservation method on
feeding behavior and productivity of lactating dairy cows at two dietary starch
concentrations. J Dairy Sci. 86(1):174-183.

Oba, M. and M. S. Allen. 2003c. Extent of hypophagia caused by propionate
infusion is related to plasma glucose concentration in lactating dairy cows. J Nutr.
133(4):1105-1112.

Oba, M. and M. S. Allen. 2003d. lntraruminal infusion of propionate alters feeding
behavior and decreases energy intake of lactating dairy cows. J Nutr.
133(4): 1 094-1 099.

Oba, M. and M. S. Allen. 2003e. Effects of corn grain conservation method on
ruminal digestion kinetics for lactating dairy cows at two dietary starch
concentrations. J Dairy Sci. 86(1):184-194.

Overton, T. R., J. K. Drackley, C. J. Ottemann-Abbamonte, A. D. Beaulieu, and J.
H. Clark. 1998. Metabolic adaptation to experimentally increased glucose
demand in ruminants. J Anim Sci. 76(11):2938-2946.

Overton, T. R., J. K. Drackley, C. J. Ottemann-Abbamonte, A. D. Beaulieu, L. S.
Emmert, and J. H. Clark. 1999. Substrate utilization for hepatic gluconeogenesis
is altered by increased glucose demand in ruminants. J Anim Sci. 77(7):1940-
1951.

Pilkis, S. J. and D. K. Granner. 1992. Molecular physiology of the regulation of
hepatic gluconeogenesis and glycolysis. Annu Rev Physiol. 54:885-909.

Plum, L., M. Schubert, and J. C. Bruning. 2005. The role of insulin receptor
signaling in the brain. Trends Endocrinol Metab. 16(2):59-65.

Polonsky, K. S., B. D. Given, L. Hirsch, E. T. Shapiro, H. TilliI, C. Beebe, J. A.
Galloway, B. H. Frank, T. Karrison, and E. Van Cauter. 1988. Quantitative study
of insulin secretion and clearance in normal and obese subjects. J Clin Invest.
81 (2):435-441.

Porte, 0., Jr., D. G. Baskin, and M. W. Schwartz. 2005. Insulin signaling in the
central nervous system: a critical role in metabolic homeostasis and disease from
C. elegans to humans. Diabetes. 54(5):1264-1276.

Potter, B. J. 1952. Relief of hypoglycaemic convulsions with butyric acid. Nature.
170(4326):541-541.

Puerta, M., M. Abelenda, M. Rocha, and P. Trayhum. 2002. Effect of acute cold
exposure on the expression of the adiponectin, resistin and leptin genes in rat
white and brown adipose tissues. Horm Metab Res. 34(11-12):629-634.

125

Raabo, E. and T. C. Terkildsen. 1960. On the enzymatic determination of blood
glucose. Scand J Clin Lab Invest. 12:402-407.

Rawson, N. E., H. Blum, M. D. Osbakken, and M. l. Friedman. 1994. Hepatic
phosphate trapping, decreased ATP, and increased feeding after 2,5-anhydro-D-
mannitol. Am J Physiol. 266(1):R112-117.

Rawson, N. E. and M. I. Friedman. 1994. Phosphate loading prevents the
decrease in ATP and increase in food intake produced by 2,5-anhydro-D-
mannitol. Am J Physiol. 266(6):R1792-1796.

Rawson, N. E., H. Ji, and M. l. Friedman. 2003. 2,5-Anhydro--mannitol increases
hepatocyte calcium: implications for a hepatic hunger stimulus. Biochim Biophys
Acta. 1642(1-2):59-66.

Reynolds, C. K. 1995. Quantitative aspects of liver metabolism in ruminants.
Pages 351-372 in Ruminant Physiology: Digestion, Metabolism, Growth and
Reproduction. W. v. Englehardt, S. Leonhard-Marek, G. Breves, and D.
Geisecke, eds. Ferdinand Enke Verlag, Stuttgart, Germany.

Reynolds, C. K., P. C. Aikman, B. Lupoli, D. J. Humphries, and D. E. Beever.
2003. Splanchnic metabolism of dairy cows during the transition from late
gestation through early lactation. J Dairy Sci. 86(4):1201-1217.

Rezek, M. and E. A. Kroeger. 1976. Glucose antimetabolites and hunger
(theoretical article). J Nutr. 106(2):143-157.

Ricardo, J. A. and E. T. Koh. 1978. Anatomical evidence of direct projections
from the nucleus of the solitary tract to the hypothalamus, amygdala, and other
forebrain structures in the rat. Brain Res. 153(1):1-26.

Ricks, C. A. and R. M. Cook. 1981. Regulation of volatile fatty acid uptake by
mitochondrial acyl CoA synthetases of bovine liver. J Dairy Sci. 64(12):2324-
2335.

Ringer, A. I. 1912. The chemistry of gluconeogenesis. I. The quantitative
conversion of propionic acid into glucose. J Biol Chem. 12(3):511-515.

Rossi, R., A. Meissner, E. Del Prete, and E. Scharrer. 1996. Hepatic branch
vagotomy enhances feeding in response to centrally elicited glucose deprivation
in rats. Exp Physiol. 81(1):119-129.

Routh, V. H. 2002. Glucose-sensing neurons: Are they physiologically relevant?
Physiol Behav. 76(3):403-413.

Russell, J. B. 1991. Intracellular pH of acid-tolerant ruminal bacteria. Appl
Environ Microbiol. 57(11):3383-3384.

126

Russell, J. B. 1998. The importance of pH in the regulation of ruminal acetate to
propionate ratio and methane production in vitro. J Dairy Sci. 81(12):3222-3230.

Rutter, L. M. and J. G. Manns. 1988. Follicular phase gonadotropin secretion in
cyclic postpartum beef cows with phlorizin-induced hypoglycemia. J Anim Sci.
66(5):1194-1200.

Scharrer, E. 1999. Control of food intake by fatty acid oxidation and ketogenesis.
Nutrition. 15(9):704-714.

Scheppach, W., F. Richter, R. Joeres, E. Richter, and H. Kasper. 1988. Systemic
availability of propionate and acetate in liver cirrhosis. Am J Gastroenterol.
83(8):850-853.

She, P., G. L. Lindberg, A. R. Hippen, D. C. Beitz, and J. W. Young. 1999.
Regulation of messenger ribonucleic acid expression for gluconeogenic enzymes
during glucagon infusions into lactating cows. J Dairy Sci. 82(6):1153-1163.

Sheperd, A. C. and D. K. Combs. 1998. Long-term effects of acetate and
propionate on voluntary feed intake by midlactation cows. J Dairy Sci.
81 (8):2240-2250.

Shirai, Y., S. Yaku, and M. Suzuki. 2004. Metabolic regulation of leptin
production in adipocytes: A role of fatty acid synthesis intermediates. J Nutr
Biochem. 15(1 1 ):651-656.

Simkins, J., K. L., J. W. Suttie, and B. R. Baumgardt. 1965. Regulation of food
intake in ruminants. 4. Effect of acetate, propionate, butyrate, and glucose on
voluntary food intake in dairy cattle. J Dairy Sci. 48:1635-1642.

Stangassinger, M. and D. Giesecke. 1986. Splanchnic metabolism of glucose
and related energy substrates. Pages 347-366 in Control of Digestion and
Metabolism in Ruminants. L. P. Milligan, W. L. Grovum, and A. Dobson, eds.
Prentice Hall, Englewood Cliffs, NJ.

Steinhour, W. D. and D. E. Bauman. 1988. Propionate metabolism: a new
interpretation. Pages 238-256 in Aspects of digestive physiology in ruminants. A.
Dobson and M. J. Dobson, eds. Comstock Pub. Associates, Ithaca, NY.

Sutton, J. D., M. S. Dhanoa, S. V. Morant, J. France, D. J. Napper, and E.
Schuller. 2003. Rates of production of acetate, propionate, and butyrate in the
rumen of lactating dairy cows given normal and low-roughage diets. J Dairy Sci.
86(1 1):3620-3633.

Thorlacius, S. O. 1972. Effect of steam-volatile fatty acids and carbon dioxide on
blood content of rumen papillae of the cow. Am J Vet Res. 33(2):427-430.

127

Tordoff, M. G., R. Raflta, M. J. DiNovi, and M. I. Friedman. 1988. 2,5-anhydro-D-
mannitol: a fructose analogue that increases food intake in rats. Am J Physiol.
254(1):R150-153.

Tyrrell, H. F. and J. T. Reid. 1965. Prediction of the energy value of cow's milk. J
Dairy Sci. 48(9):1215-1223.

van der Walt, J. G., G. D. Baird, and E. N. Bergman. 1983. Tissue glucose and
lactate metabolism and interconversions in pregnant and lactating sheep. Br J
Nutr. 50(2):267-280.

Van Soest, P. J., J. B. Robertson, and B. A. Lewis. 1991. Methods for dietary
fiber, neutral detergent fiber, and nonstarch polysaccharides in relation to animal
nutrition. J Dairy Sci. 74(10):3583-3597.

Vandesompele, J., K. De Preter, F. Pattyn, B. Poppe, N. Van Roy, A. De Paepe,
and F. Speleman. 2002. Accurate normalization of real-time quantitative RT-PCR
data by geometric averaging of multiple intemal control genes. Genome Biol.
3(7):RESEARCH0034.

Vazquez-Anon, M., S. Bertics, M. Luck, R. R. Grummer, and J. Pinheiro. 1994.
Peripartum Liver Triglyceride and Plasma Metabolites In Dairy Cows. J. Dairy
Sci. 77(6):1521-1528.

Veenhuizen, J. J., R. W. Russell, and J. W. Young. 1988. Kinetics of metabolism
of glucose, propionate and 002 in steers as affected by injecting phlorizin and
feeding propionate. J Nutr. 118(11):1366-1375.

Vranic, M., C. Gauthier, D. Bilinski, D. Wasserrnan, K. El Tayeb, G. Hetenyi, Jr.,
and H. L. Lickley. 1984. Catecholamine responses and their interactions with
other glucoregulatory hormones. Am J Physiol Endocrinol Metab. 247(2):E145-
156.

Wang, Y., R. Roman, S. D. Lidofsky, and J. G. Fitz. 1996. Autocrine signaling
through ATP release represents a novel mechanism for cell volume regulation.
PNAS. 93(21):12020-12025.

Weeks, T. E. and A. J. Webster. 1975. Metabolism of propionate in the tissues of
the sheep gut. Br J Nutr. 33(3):425-438.

Weiss, J. L. and R. D. Burgoyne. 2002. Sense and sensibility in the regulation of
voltage-gated Ca(2+) channels. Trends Neurosci. 25(10):489-491.

Whelan, J. A., N. B. Russell, and M. A. Whelan. 2003. A method for the absolute
quantification of cDNA using real-time PCR. J Immunol Methods. 278(1-2):261-
269.

128

Wildman, E. E., G. M. Jones, P. E. Wagner, R. L. Boman, J. Troutt, H. F., and T.
N. Lesch. 1982. A dairy cow body condition scoring system and its relationship to
selected production characteristics. J Dairy Sci. 65:495-501.

Williams, E. A., J. M. Coxhead, and J. C. Mathers. 2003. Anti-cancer effects of
butyrate: use of micro-array technology to investigate mechanisms. Proc Nutr
Soc. 62(1):107-115.

Xiong, Y., N. Miyamoto, K. Shibata, M. A. Valasek, T. Motoike, R. M. Kedzierski,
and M. Yanagisawa. 2004. Short-chain fatty acids stimulate leptin production in
adipocytes through the G protein-coupled receptor GPR41. PNAS. 101(4):1045-
1050.

Ying, Y., M. S. Allen, M. J. VandeHaar, and N. K. Ames. 1998. Effects of
fineness of grinding and conservation method of corn grain on ruminal and whole
tract digestibility and ruminal microbial protein production of Holstein cows in
early lactation. J Dairy Sci. 81(Suppl. 1):339.

Yonekura, S., T. Senoo, Y. Kobayashi, T. Yonezawa, K. Katoh, and Y. Obara.
2003. Effects of acetate and butyrate on the expression of leptin and short-forrn
leptin receptor in bovine and rat anterior pituitary cells. Gen Comp Endocrinol.
133(2):165-172.

Young, J. W., S. P. Schmidt, E. S. Akowuah, G. S. Hess, and A. D. McGilIiard.
1974. Effects of phlorizin on glucose kinetics in the bovine. J Dairy Sci.
57(6):689-694.

129