EM

 

LIBRARY
Michigan State
’ University

 

 

This is to certify that the

thesis entitled

THE EFFECT OF INSULIN-LIKE GROWTH FACTOR 1 ON
CHANGES IN PROLIFERATION-RELATED GENE EXPRESSION
IN BOVINE MAMMARY EPITHELIAL CELLS

presented by

MICHAEL ALAN JACOBSEN

has been accepted towards fulfillment
of the requirements for the

A

/

degree In

 

Animal Science

 

 

LIA

 

ijor rofessor’s SBnature

ASE/7’— 07

Date

MSU is an affirmative-action, equal-opportunity employer

 

 

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

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

6/07 p:/CIRC/DateDue.indd-p.1

 

THE EFFECT OF INSULIN-LIKE GROWTH FACTOR 1 ON CHANGES IN
PROLIFERATION-RELATED GENE EXPRESSION IN BOVINE MAMMARY
EPITHELIAL CELLS
By

MICHAEL ALAN JACOBSEN

A THESIS

Submitted to
Michigan State University
In partial fulfillment of the requirements
For the degree of

MASTER OF SCIENCE
DEPARTMENT OF ANIMAL SCIENCE

2007

THE EFFECT OF INSULIN-LIKE GROWTH FACTOR 1 ON CHANGES IN
PROLIFERATION-RELATED GENE EXPRESSION IN BOVINE MAMMARY
EPITHELIAL CELLS
By

MICHAEL ALAN JACOBSEN

ABSTRACT

Insulin-like growth factor 1 (IGF-l) is a potent mitogen for mammary epithelial
cells. My objective was to determine if IGF-l treatment alters the expression of genes
in the MAC-T bovine mammary epithelial cell line in a manner consistent with increased
proliferation. Cells were treated with O or 100 ng/mL of IGF-I for 8 or 24 hours. Gene
transcript abundance was measured with a bovine metabolism microarray of 2360 genes.
IGF-l increased cell confluency by 40% after 24 hr of treatment (P < 0.05). IGF-l
altered the expression (P < 0.05) of 89 genes after 8 hours (70 increased, 18 decreased)
and 184 genes afier 24 hours (139 increased, 45 decreased). IGF-l altered the expression
of several regulatory genes that might increase cell proliferation and several metabolic
genes to support increased proliferation. The fold-changes of 9 of 10 genes as measured
with RT-PCR were similar to those with microarray analysis, although the statistical
significance of the change was the same for only 6 of the genes. In conclusion, IGF-l
alters the expression of proliferative and metabolic genes in a manner consistent with

increased cell proliferation.

For my beautiful wife Ruth

iii

ACKNOWLEDGEMENTS

First, I would like to thank my advisor, Dr. Michael VandeHaar, without whos
guidance, support, and grace this degree would never have been possible. I would also
like to extend my gratitude to my committee, Dr. Smith, Dr. Weber-Neilsen, and Dr.
Sordillo, for their input and support. Second, I would like to thank Jim Liesman for his
. invaluable statistical and technical support. Next, I give my appreciation to Chris Colvin,
Sue Sipkofsky, Xiaoning Rei, Patty Weber and Janet Ireland for the use of their lab
equipment. I would also like to thank my fellow and former lab mates, Myunggi Baik,

J injong Bong, Laurie Davis-Rinker, Brett Etchebarne, and Jianguo Li for their
encouragement. Finally, I would like to give my deepest gratitude to my family, John,
Susan, and Katie, and my wife Ruth who have kept me focused on completing my MS.

degree.

iv

TABLE OF CONTENTS

List of Tables ........................................................................................... vi
List of Figures .................................................................................................................... vii
List of Abbreviations ........................................................................................................ viii
IntrOduction ......................................................................................................................... 1
Literature Review ................................................................................................................ 4
Overview of mammary development ...................................................................... 4
Effect of diet on mammary development in prepubertal heifers ............................. 6
The somatomedin hypothesis .................................................................................. 8
IGF-1 and regulation of the cell cycle ................................................................... 11
IGF-1 in the mammary tissue... . .. ........................................................................ 12
IGF binding proteins .............................................................................................. 13
Interaction of IGF-1 and other hormones on mammary development .................. 15
Intracellular signaling pathways activated by IGF-1 ............................................. 16
The gene microarray .............................................................................................. 18
The MAC-T cell line as a model of mammary epithelial cells .............................. 20
Summary ................................................................................................................ 22
Materials and Methods ...................................................................................................... 23
Cell culture ............................................................................................................ 23
RNA extraction ..................................................................................................... 24
Microarray hybridization ....................................................................................... 26
Quantitative RT-PCR ............................................................................................ 28
Data analyses ......................................................................................................... 32
Results ................................................................................................................................ 34
Cell proliferation .................................................................................................... 34
Microarray results .................................................................................................. 34
Quantitative RT-PCR results ................................................................................. 42
Discussion of Results ......................................................................................................... 44
Discussion of Approach ..................................................................................................... 49
Conclusion ......................................................................................................................... 53
Appendices ........................................................................................................................ 54
Bibliography ...................................................................................................................... 76

LIST OF TABLES

Table 1. Quantitative RT-PCR primer sets ........................................................................ 31
Table 2. Number of genes altered by IGF-l at P < 0.05 ................................................... 35

Table 3. Emotional categories of genes whose expression was altered by IGF-1
treatment for 8 hr and 24 hr ............................................................................................... 37

Table 4. Proliferation and survival-related genes whose expression was regulated by
IGF-1 at 8 hr or 24 hr ......................................................................................................... 39

Table 5. Regression and correlation values of all genes expressed at a P-value less than
0.1 between the 8 hr arrays ......................................................................... 41

Table 6. F old-change values from microarray analysis and qRT-PCR for the validated
genes .................................................................................................. 43

vi

LIST OF FIGURES

Figure l. High-energy diets fed to prepubertal heifers increase body growth and IGF-1

production in prepubertal dairy heifers ............................................................................... 8
Figure 2. The current form of the somatomedin hypothesis ............................................. 10
Figure 3. Proliferative effect of 100 ng/mL IGF-1 on MAC-T cells after 18 hr of IGF-1

treatment ............................................................................................................................ 24
Figure 4. Average cell confluencies at 24 hours ............................................................... 34
Figure 5. P-value distribution histogram of 8 hr 24 hr microarray data ........................... 35
Figure 6. Quantitative RT—PCR verification of genes altered by IGF-1 ........................... 42
Figure 7. Infusion design for the udder quarters of each heifer ........................................ 56

Figure 8: Mean index of BrdU-labeled mammary epithelial cells sampled from two
heifers infused once daily with 0, 23, or 46 pg leptin over a seven-day period ................ 60

vii

LIST OF ABBREVIATIONS

ADG .............................................................................. Average daily gain
BSA ........................................................................... Bovine serum albumin
cDNA ...................................................... Complementary deoxyribonucleic acid
DEPC ....................................................................... Diethylpropyl carbonate
DMSO ............................................................................ Dimethyl sulfoxide
DNA ........................................................................... Deoxyribonucleic acid
dNTP ........................................................... Deoxyribonucleotide triphosphates
DTT. . ................................................................................... Dithiothreitol
E ................................................................................................ Estrogen
ECM .............................................................................. Extracellular matrix
EtOH ............................................................................................ Ethanol
FBS ................................................................................ Fetal bovine serum
GH ................................................................................... Growth hormone
hr .................................................................................................................................. hours

IGF-1 .................................................................... Insulin-like growth factor 1
IGF-2 .................................................................... Insulin-like growth factor 2
IGFBP .................................................. Insulin-like growth factor binding protein
IRS ........................................................................... Insulin related substrate
min ............................................................................................................................ minutes

TEB ................................................................................ Terminal end buds

viii

J AK ....................................................................................... Janus kinase

MAPK ............................................................ Mitogen-activated protein kinase
MDGI ......................................................... Mammary-derived growth inhibitor
mRN A .................................................................. Messenger ribonucleic acid
PBS ........................................................................ Phosphate buffered saline
PI3K ................................................................. Phosphotidylinositol -3 kinase
qRT-PCR ........................ Quantitative reverse-transcriptase polymerase chain reaction
STAT ............................................ Signal transducer and activator of transcription

ix

INTRODUCTION

Prepubertal mammary development in the dairy heifer sets the foundation for
future epithelial cell growth and activity. Sinha and Tucker (1969) found that between 3
months and 9 months of age, the mammary parenchyma grows at a 3-fold greater rate
than body growth. Because dairy heifer rearing costs account for 20% of farm costs,
dairy scientists have studied the effect of feeding heifers for an average daily gain (ADG)
of greater than 1.0 kg/d to decrease rearing times. In an experiment studying the effect of
ADG greater than 1 kg/d on body growth and mammary development in dairy heifers
between the ages of 11 and 23 weeks, Davis-Rinker (2005) found that heifers fed for an
ADG of 1.1 kg/d possessed more total mammary gland mass but less parenchyma per
unit of body mass than heifers fed for 0.7 kg/d. Davis-Rinker then stained the mammary
parenchymas of heifers from both treatments for Ki-67, a proliferation-related cell
marker, and counted the number of stained cells. The mammary parenchymas of heifers
fed for an ADG of 1.1 kg/d had 30% less Ki-67-labeled cells than the parenchymas of
heifers fed for 0.7 kg/d, demonstrating a decrease in cell proliferation. The high-gain diet
increased circulating concentrations of insulin-like growth factor 1 (IGF-1) by 77%.
IGF-1 is considered to be a major regulator of proliferation. For example, after infusing
the udders of prepubertal heifers with 10 pg IGF-l for 7 days, Silva (2002) discovered
that IGF -1-infused quarters contained 52% more proliferating epithelial cells than saline-
infused quarters. Therefore, a paradox exists in which heifers fed for high rates of gain
have higher circulating levels of IGF-1 but decreased mammary parenchyma

development compared to heifers fed for more moderate rates of gain. The current state

of knowledge on mammary development, effects of diet, and IGF-1 are discussed in the
literature review.

IGF-1 can affect proliferation via a number of different possible mechanisms.
IGF-1 binding to mammary epithelial cells could be altered by changes in the presence
and concentration of IGF binding proteins (IGFBP) in the mammary parenchyma. Weber
et al. (2000) found that IGFBP2 expression was increased and IGFBP] expression was
decreased in the parenchymas of heifers fed for high rates of gain. IGF-1 can also
influence or be influenced by other hormones. Leptin impairs the proliferative effect of
IGF-1 in mammary epithelial cells as compared with cells treated with only IGF-l (Silva,
2002). However, there are other possible mechanisms, such as gene expression changes,
protein synthesis and modification changes, IGF-1 receptor number changes, and
phosphorylation changes. To further explore the proliferative effect of IGF-1 on bovine
mammary epithelial cells, I examined the changes in gene expression due to IGF-1
stimulation. To our knowledge, no one had previously explored the effect of IGF-1 on
gene expression in bovine mammary epithelial cells. F urtherrnore, our lab has the means
(the BMET microarray; Etchebarne, 2005) to study gene expression changes. The BMET
microarray contains all of the known genes associated with metabolism and proliferation
in the cow. Four spots per gene on the BMET array provides the technical replication for
reducing possible effects of probe spot and spot position on the final results. By finding
which genes are directly altered by IGF-1, we now have a foundation to explore which
genes are altered by IGF—1 in conjunction with other factors (feeding levels, hormones,

puberty, etc.).

To localize changes in gene expression to mammary epithelial cells, we used the
MAC-T cell line. The MAC-T cell line is an immortalized bovine mammary epithelial
cell line that was developed by transfecting the SV40 T-antigen into epithelial cells taken
from a lactating Holstein cow (Huynh et al., 1991). To test my hypothesis, it is critical
that the cells proliferate in response to IGF-1. Primary cells can undergo senescence, in
which they stop proliferating after being passaged too many times (Matitashvili et a1,
1997). The MAC-T cell line is a pure population of bovine mammary epithelial cells,
albeit with modifications, and, most importantly, MAC-T cells consistently increase
proliferation in response to IGF-1. In fact, in my preliminary results, MAC-T cells
treated with 100 ng/mL IGF-1 synthesized DNA at greater than 3 times the rate of control
cells. Therefore, much of the signaling pathways are likely still intact in the MAC-T
cells.

Therefore, my hypothesis is that IGF-1 alters the expression of genes in a manner
consistent with increased proliferation in bovine mammary epithelial cells. To test this
hypothesis, my objective was to determine if IGF-1 treatment for 8 and 24 hours alters
the transcript abundance of genes in the MAC-T bovine mammary epithelial cell line in a
manner consistent with increased proliferation. I used the BMET microarray, an
oligonucleotide array constructed entirely of genes related to proliferation and

metabolism in the cow.

Literature Review
Overview of mammary development

Mammary development has been described in detail by Williams and Daniel
(1983) and Akers (2000). As these authors explain, mammary gland development in the
bovine fetus is initiated at 30 days of gestation. Ectodermal cells combine together to
form the mammary streak. AS the fetus ages, the rudimentary gland progresses through
the crest, hillock, bud, and sprout stages. The primary sprout appears as a solid mass of
cells but it canalizes into a hollow structure containing an epithelial cell border two to
three layers thick that will become the future milk cistern. Secondary sprouts, which will
become the large ducts emptying into the milk cistern, extend from the primary sprout at
around day 90 in the fetus. The mammary fat pad appears at the same time as the
primary and secondary sprouts and the teat begins to form soon after the development of
the sprouts. At birth, the streak canal, milk cistern and a few ducts budding fi'om the milk
cistern are present. As the heifer ages, the ducts extend into the fat pad, and each duct
gives rise to subtending ducts.

The parenchyma develops at the same rate as the body until 3 months of age.
From 3 months to about the third or fourth estrous cycle, the parenchyma grows at
roughly three times the rate of the body (Sinha and Tucker, 1969). In rodents, most of
the growth is ductal in nature, with few alveolar-like structures branching off the ducts.
Ductal extension occurs via the rapid invasion of the terminal end buds (TEB) into the
mammary fat pad with very few branches until the edge of the fat pad is reached
(Williams and Daniel, 1983). However, prepubertal mammary development in heifers

involves less extensive growth into the mammary fat pad from the nipple and a higher

degree of branching from the ducts into a loose sheath of connective tissue, giving rise to
a broccoli-like appearance, as shown by Ellis and Capuco (2002) using computerized
tomography. Extensive branching was apparent in regions of actively proliferating
epithelial cells and occurred in conjunction with ductal elongation. Furthermore,
branching occurred along the ducts within these regions of ‘terminal ductal units’ as they
were labeled by the researchers. After puberty, the parenchyma] growth rate again
matches that of the body until pregnancy (Sinha and Tucker, 1969). The vast majority of
mammary parenchyma growth occurs during pregnancy.

The amount of DNA in a tissue is considered to be proportional to the number of
cells in the tissue. Thus, an increase in the DNA content indicates an increase in cell
numbers in the mammary parenchyma, and one way to study cell proliferation is to
measure the DNA content of a tissue. Another method to determine cell proliferation is
to measure radioisotope-labeled thymidine incorporation, which is considered to be a
“snapshot” view of cell proliferation. Proliferating cells take up nucleotides for DNA
replication during the S-phase of the cell cycle; therefore, this assay provides an estimate
of cells undergoing DNA replication during a particular period of time. Weber et al.
(2000) measured tritiated thymidine incorporation in primary bovine mammary epithelial
cells that were treated with mammary extracts prepared from prepubertal heifers fed for a
high or a low rate of gain. They discovered that thymidine incorporation was increased
by about 40% in cells treated with extracts from the low rate-of—gain heifers compared to
cells treated with extracts from heifers fed for high rates of gain. Therefore, measuring

DNA content provides a basis for measuring mammary development.

Effect of diet on mammary development in prepubertal dairy heifers

Feeding for higher rates of gain decreases rates of prepubertal mammary
development relative to body grth (for review, see Sejrsen and Purup, 1997; Sejrsen et
al., 2000; and Akers, 2002). Prepubertal heifers reared at an average daily gain (ADG)
above 1.0 kg/d have less parenchymal mass at puberty than heifers reared at 0.7 kg/d
(Sejrsen et al., 1982). In another study, Harrison et al. (1983) showed that heifers reared
at 1.1 kg/d contained 68% less secretory tissue than in heifers reared at 0.7 kg/d. Other
studies have confirmed this effect (Little and Kay, 1979, Petitclerc et al., 1984 and
Stelwagen and Grieve, 1990), although it has not been universally proven (Van Amburgh
et al., 1998). Meyers et al. (2006) measured mammary gland weight and DNA content of
mammary gland samples from heifers fed a diet for either restricted gain (0.65 kg/d) or
high gain (0.95 kg/d). Heifers were slaughtered at 50-kg increments from 100 kg to 350
kg of body weight. Parenchymal weight and DNA content was decreased in heifers fed
the high-gain diet versus heifers fed the restricted diet. When age at tissue collection was
added as a covariate to the model, the diet effects disappeared. They argued that the
observations of adverse effect of diet on mammary development were actually due to age
of the heifer at tissue harvest. Davis-Rinker (2005) discovered that heifers fed for 1.1
kg/d between 11 weeks and 23 weeks of age had 23% less grams parenchyma per unit
body weight, and a 30% reduction of Ki-67-labeled epithelial cells (indicating decreased
cell proliferation). Therefore, the full effects of high rates of gain on mammary
development are still being explored.

Sejrsen et a1. (1983) examined the effects of feeding for high or moderate rates of

gain on serum growth hormone (GH) as a basis for understanding mammary growth.

Heifers fed for restricted rates of gain (0.7 kg/d) had higher concentrations of GH in
serum than those fed for rapid growth. Serum GH concentrations were positively
correlated with parenchyma mass and negatively correlated with extraparenchymal
adipose mass. Feeding for rapid growth led to a reduction in hepatic GH mRNA
abundance (Smith et al., 2002). VandeHaar et al. (1995) measured the effect of negative
energy balance on hepatic and luteal IGF-1 expression in post pubertal heifers. Heifers in
negative energy balance had increased serum GH concentrations and decreased serum
IGF -1 levels. They also found that hepatic IGF-1 mRNA levels were also decreased in
heifers in negative energy balance. Radcliff et al., (2004) discovered that prepubertal
heifers fed for high rates of gain had increased levels of serum IGF-1. However, this
difference disappeared after the heifers had attained puberty. High rates of gain also
increased liver IGF-1 mRNA abundance, and rate of gain was positively correlated with
serum IGF-1 concentrations (r = 0.60, P < 0.01). Davis-Rinker (2005) discovered that
heifers fed for an ADG of 1.1 kg/d had a 73% greater circulating IGF-1 concentration
than heifers fed for 0.7 kg/d during the same time period. Therefore, as shown in figure
1, high-energy diets promote high rates of body growth and increase serum IGF-1
concentrations in prepubertal heifers; however, high rates of gain lead to diminished
mammary parenchymal development. The reasons for this paradox are not clear.
Perhaps other hormones, such as leptin, are involved (Silva et al., 2005). I hope to find

gene pathways that serve as targets to understand this paradox.

Figure 1. Hi gh-energy diets fed to prepubertal heifers increase body growth and IGF-1
production in prepubertal dairy heifers. IGF-1 also stimulates mammary development.
However, feeding hi gh-energy diets to prepubertal heifers diminishes mammary
parenchymal development.

   

High-energy
diets

 
   
 

 

 

     

Body growth Mammary
development

 

The Somatomedin Hypothesis

Growth hormone (GH) increases body tissues growth; yet the theory explaining
the mechanism of GH action changed numerous times over the past sixty years. The
initial theory that GH stimulates growth through an intermediary factor was proposed in
the 1950’s when it was demonstrated that GH treatment on costal cartilage slices only
minimally affected cartilage growth (Daughaday and Reeder, 1966). Given that
hypophysectomy reduces bone growth and GH administration re-establishes growth in
hypophysectomized animals (Denko and Bergenstal, 1955), it was suggested that GH
affects growth via another signal, termed “sulfation factor”. This “sulfation factor” was
partial purified from the serum of acromegalic patients and could mimic the mitogenic

effects of GH. Thus, it was renamed “somatomedin” because it mediated the action of

OH on growing tissues (Daughaday et al., 1972). Six years later, IGF-1 and IGF-2 were
purified and found to be the “sulfation factor” that affected growth in rats. Furthermore,
IGF-1 levels were found to be affected by GH administration, thereby cementing its
identity as the proposed “sulfation factor” (Klapper et al., 1983). Thus, the original
hypothesis stated that GH was released from the pituitary gland and traveled to the liver,
where it stimulated the release of IGF-1. IGF-1 in turn provided negative feedback on
GH production in the pituitary gland.

However, this theory was questioned when it was found that IGF-1 was produced
by several fetal tissues (D’Ercole et al., 1980). Furthermore, IGF-1 was found to be
expressed in numerous tissues other than the liver. This prompted the idea that IGF-1
could be an autocrine/paracrine factor and that GH could stimulate localized production
of IGF-1. Even this view may not fully explain IGF-l production since GH-dependent
IGF-1 synthesis in the mammary gland has never been explicitly demonstrated (Glimm et
al., 1992). Therefore, according to the most recent proposal of the somatomedin theory,
GH travels from the pituitary gland to the liver where it induces IGF-1 synthesis and
release. Furthermore, GH can bind to GH receptors on other tissues and perform various
functions. The liver synthesized IGF-1 is then bound to IGFBPS and travels to the target
tissues, where it initiates primarily proliferative and survival Signaling pathways. Finally,
IGF-1 is produced by local tissues and acts upon the target tissue, as shown in Figure 2
on page 10. For many tissues, serum IGF-1 is probably less important than local IGF-1
(Le Roith, 2001). However, because the bovine mammary gland lacks GH receptors

(Glimm et al., 1992), serum IGF-1 may exert a greater effect on proliferation.

Figure 2. The current form of the somatomedin hypothesis. Taken from Akers, 2006.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Pituitary
GH
I ilGFBP
Direct GH ,
effects L'Ve’

v Local tissue

IGF-1 A IGF-1 synthesis

/ \ "I,
Growth Metabolism
Proliferation Survival

 

 

 

 

 

 

The IGF system is a complex hormone system in the body. It consists of three
ligands (IGF-l, IGF-2 and insulin), three receptors (IGF-1 receptor, IGF-2/mannose-6-
phosphate receptor and insulin receptor), and six known IGF-binding proteins (IGFBP-1-
6). IGF-1 and IGF-2 exert their mitogenic activities via binding to the IGF-l receptor.
The IGF-2/mannose-6-phosphate receptor does not seem to have any effect on IGF-l
signaling but is thought to sequester and remove circulating IGF-2 during fetal
development (Kiess et al., 1987; Baker et al., 1993). All of the binding proteins bind to
both IGF-1 and IGF-2. Insulin and IGF-1 can weakly bind to the other’s receptor.

The IGF-1 and insulin receptors are heterodimeric proteins that possess about
60% overall homology. They both contain an extracellular a-subunit and a membrane-
spanning B-subunit that transmits the signal to the intracellular signaling pathways. The

a-subunit is made up of two ligand binding sites that are separated by a cysteine-rich

10

domain and two fibronectin HI binding domains (F n0 and Fnl) towards the N-terminus.
The extracellular domain of the [I-subunit is made up of two fibronectin III-binding
domains (F n1 and Fn2). The intracellular domain contains a juxtarnembrane domain
close to the plasma membrane, a tyrosine kinase domain that acts as an anchor for
intracellular signaling molecules and a C-terminal domain that also anchors signaling
molecules. A disulfide bridge between the Fnl domain on the (It-subunit and the Fn2
domain on the B-subunit connects the two proteins. Assembled holoreceptors are
connected by disulfide bridges between the FnO and Fnl domains on the a-subunit.
While each dimer is capable of binding to the ligand, the holoreceptor forms a binding
pocket that increases the affinity of the receptor for the ligand (DeMeyts et al., 2004).
IGF-1 and regulation of the cell cycle

The cell cycle refers to the period in the cell’s life when it undergoes cell division.
The cell cycle is separated into four different phases: the M (mitosis) phase, the G (gap) 1
phase, the S (DNA synthesis) phase, and the G2 phase. In each phase, the cell performs
certain tasks that prepare it for mitosis. Therefore, the tasks that are performed in each
phase must be regulated to prevent errors in the creation and transmission of parental
DNA to the new cells. Cyclins, cyclin dependent kinases (cdk), and cyclin dependent
kinase inhibitors (cdki) act as cell cycle machinery within the cells to promote the
accurate progression of cells through the cell cycle. Furthermore, mitogens can direct
their Signals to regulate the cell cycle machinery, which then regulate progression of the
cell through the cell cycle.

IGF-l exerts its proliferative effects by regulating the cell cycle. In breast cancer

cells, IGF-l promotes passage of the cell through the G1 phase by increasing cyclin D1

11

transcription and translation via the PI3K/Akt signaling pathway (Dufourny et al., 1997;
Muise-Helmericks et al., 1998). Cells from IGF-l-knockout mice display a retarded
progression through the G2 phase, suggesting that IGF-l regulates the passage of cells to
the M phase (Adesanya et al., 1999). IGF-1 increases the expression of cyclin A, cyclin
Bl , and cdkl in human osteosarcoma cells, genes that are known to regulate passage
through the G2 phase (F urlanetto et al., 1994). Furthermore, IGF-1 inhibits expression of
the cdki p27 in rat satellite muscle cells and p27 and p21 in cardiomyocytes (Medema et
al., 2000; von Harsdorf et al., 1999). However, the effect of IGF-1 on cdki may be cell-
specific as IGF-1 increases the mRNA and protein levels of p21 in MCF-7 cells (Lai et
al., 2001).
IGF-l in mammary tissue

IGF-1 increases mammary epithelial cell proliferation in the bovine mammary
gland in both in vivo and in vitro models. Cultured mammary epithelial cells proliferate
when exposed to IGF-l (Collier et al., 1993; Matitashavili et al., 1997). IGF-l increases
DNA content (tritiated thymidine incorporation) compared to untreated cells (Zhao et al.,
1992). Mammary explants treated with different doses of IGF-1 increase mammary
epithelial cell proliferation in a dose dependent manner as measured by tritiated
thymidine incorporation (Baumrucker and Stemberger, 1989). Furthermore,
intramammary infusion of IGF-1 increases DNA content per gland and the number of
cells undergoing mitosis (Collier et al., 1993; Silva et al., 2005).

The mammary gland produces IGF-1. IGF-l mRNA and protein have been
localized to the stromal elements of the bovine mammary gland (Hauser et al., 1990),

ovine mammary gland (Hovey et al., 1998a) and human breast (Yee et al., 1989).

12

Epithelial cells express IGF-1 mRNA but seem incapable of producing the IGF-1 protein
(Campbell et al., 1991), suggesting that IGF-1 is transported fi'om the stromal tissue to
the epithelial cells. IGF-l receptor mRNA was found in the alveolar epithelial cells in
bovine mammary glands (Glimm et al., 1992; Pump et al., 1995). In an immortalized
bovine mammary epithelial cell line, researchers found that the cells expressed very little
IGF-1 (Romagnolo et al., 1994).
IGF binding proteins

While six IGF-l binding proteins (IGFBP) with high affinity for IGF-l are known
to exist, research on the effects of these proteins in the mammary gland have focused
primarily on IGFBP-2, IGFBP-3, and IGFBP-5. IGFBP-3 is a 46-53 kDa protein that
acts as the main carrier of circulating IGFS. It is estimated that around 75% of the
circulating IGF-l is transported in the blood bound to IGFBP-3 that forms a 150-kDa
complex with the acid labile subunit protein. This binding extends the half life of IGF-1
from between 30 and 90 minutes for the freely circulating IGFS to 12 to 15 hours (Zapf et
al., 1986; Guler et al., 1989). IGFPB-3 is synthesized in numerous tissues, including
mammary epithelial cells (Cohick and Turner, 1998; Strange et al., 2002), and acts to
regulate IGF binding to the IGF-1 receptor. In vitro studies utilizing chick fibroblasts
show that IGFBP-3 inhibits IGF-l action when co-cultured with IGF-1 at a 3 to 4-fold
molar excess (Blat et al., 1989). In primary bovine mammary epithelial cells, IGFBP-3
inhibited DNA synthesis at equimolar or greater concentrations relative to IGF-1 (Weber
et al., 1999). Jones and Clemmons (1995) showed that the IGF-inhibitory effect of
IGFBP-3 is due to sequestration of IGF-1 away from its receptor. However, research on

IGFBP-l seems to support an IGF-independent mechanism for inhibition of DNA

13

synthesis by some IGFBP. Proteolysis of a 16-kDa fragment led to inhibition of insulin
action in chick embryo fibroblasts and the mitogenic activity of fibroblast growth factor
in both wild-type and IGF-1 receptor-knockout cells (Zadeh and Binoux, 1997).
Furthermore, endogenous IGFBP-3 from transfected bovine mammary epithelial cells
enhanced the mitogenic activity of IGF -1 by as much as ll-fold as compared to mock-
transfected controls treated with the same amount of IGF-1 (Grill and Cohick, 2000).
IGFBP-2 is synthesized in many tissues in the bovine, including mammary
epithelial cells (Cohick and Turner, 1998; Weber at al., 2000). IGFBP-2 primarily acts as
a competitor with the IGF-1 receptor for IGF-1 and IGF-2. Thus, IGFBP-2 inhibits IGF-
stimulated DNA synthesis by sequestering IGF-1 away from the IGF-1 receptor (Jones
and Clemmons, 1995). IGFBP-2 synthesis from mammary epithelial cells is not altered
by the presence of IGF-1 in vitro (Cohick and Turner, 1998; Weber at al., 2000).
IGFBP-5 inhibits IGF-mediated cell proliferation and is associated with
involution and apoptosis. Treating osteosarcoma cells with a molar excess of IGFBP-5
inhibited IGF-l stimulated DNA synthesis (Conover and Kiefer, 1993). IGFBP-5 is
highly expressed in both the pubertal and the pregnant murine mammary gland (Wood et
al., 2000). In bovine mammary epithelial cells, IGFBP-5 mRNA expression increases
during late lactation and tapers off during the dry period (Plath-Gabler et al., 2001). Mice
overexpressing IGFBP-5 in the mammary gland Show decreased expression of the
antiapoptotic bcl-2 and bcl-x proteins and an increase in the expression of caspase-3

(Tonner et al., 2002), thereby implicating IGFBP-5 as a mediator of apoptosis.

14

Interaction of IGF-1 and other hormones on mammary development

In a series of experiments examining the effects of ovariectomy and GH
administration on the local GH/IGF-l system in the udders of prepubertal heifers, Berry
et al. (2003a) found that ovariectomy reduced IGF-1 mRNA expression in the mammary
gland and reduced IGF-1 binding to mammary parenchyma microsomes. Administration
of estrogen (E) and OH to intact heifers led to an increase in mammary epithelial cell
proliferation. While estrogen administration significantly increased mammary
development, GH administration alone only tended to increase mammary development (P
< 0.10). The researchers noted that there was no significant interaction of GH and E.
Thus, they suggested that the effect of both hormones on cell proliferation is additive.
The effect of ovariectomy on local IGF-l production is unclear. Berry et al. (2003a)
noted that E administration to intact heifers tended to increase IGF-1 mRN A levels in the
mammary gland (P < 0.09). Furthermore, E administration significantly increased IGF-1
protein content in all of the mammary tissues (Berry et al., 2001). This suggests that
estrogen may mediate mammary development through increased synthesis of IGF- l.
Estrogen increases IGF-1 expression via the AP-l enhancer region in the IGF-l promoter
region, thereby supporting the idea that estrogen mediates IGF-l synthesis (Umyahara et
aL,1994)

IGF-1 also interacts with other hormones to affect mammary development. In
rodents, epidermal growth factor (EGF) is needed for IGF-1 to affect mammary epithelial
cell development in the absence of serum (Imagawa et al., 1986). Both EGF and IGF-1
induce early G1 cyclin expression but IGF-1 also induces late GI and G2 cyclin

expression and is needed by the cells to enter the S phase of the cell cycle (Stull et al.,

15

2002). Bovine mammary epithelial cells cultured with IGF-1 in serum-free media do not
require EGF for growth (Shamay et al., 198 8). However, when serum is added to the
media, IGF-1 and EGF show strong synergism, suggesting that other factors present in
serum that are necessary to influence the additive effect of IGF-1 and EGF on bovine
mammary epithelial cell proliferation (Shamay et al., 1988).

The extracellular matrix (ECM) affects IGF-1 actions in the mammary gland.
Hovey et al. (1998b) showed that mun'ne mammary epithelial cells, when cocultured with
a mammary fat pad, showed a 21-fold increase in IGF-l mediated epithelial cell
proliferation as compared to epithelial cells cultured with IGF-l in the absence of a
mammary fat pad. Mammary epithelial cells grown on different ECM proteins Show an
increase in the number of IGF-1 and EGF receptors (Woodward et al., 2000). Thus, the
actions of IGF-1 are influenced by a number of different factors.
Intracellular signaling pathways activated by IGF-1

The binding of IGF-1 to its receptor initiates signal cascades down a number of
pathways. Ligands bind to the a-subunit and induce structural changes in the B-subunit
that leads to autophosphorylation of specific tyrosine residues in the tyrosine kinase
domain of the B-subunit. Upon autophosphorylation, the ligand-bound receptor is
internalized via endocytosis which enhances intracellular signaling by IGF-1 (F urlanetto,
1988; Lin et al., 1988). A number of different signaling molecules can then bind to the
phosphorylated receptor. Most research has focused on the mitogen-activated protein
kinase (MAPK) pathway and the phosphotidylinositol-3 kinase (P13 K) pathways. These
pathways are initiated by the insulin related substrates (IRS 1-4) and She.

Autophosphorylation of the tyrosine residues in the juxtamembrane region of the B-

16

subunit provides an anchor for IRS and She to bind. When bound, IRS then can transmit
the signal through different pathways via the signaling molecules Fyn, Syp, ch, and
p85. Binding of p85 to IRS leads to activation of phosphoinositol-B’ kinase and then the
serine/threonine kinase Akt. Akt phosphorylates the proapoptotic molecule Bad, which
allows 14-3-3C to bind to and inactivate Bad, thereby preventing apoptosis (Butler et al.,
1998). She binding to the phosphorylated receptor activates the MAPK pathway and, in
turn, the Ras-Raf signaling molecules. This pathway leads to the transcription of genes
that stimulates cell proliferation. In support of this idea, IGF-1 stimulates MAPK activity
in nonmalignant mouse mammary cells (Merlo et al., 1996).

The pathways do not operate independently but interact with each other.
Interactions among the pathways allow for signaling to occur if components of one
pathway are not available. For example, 14-3-38 interacts with the mitochondrial version
of Raf and inactivates Bad via phosphorylation. IGF-1 may also initiate transcription
through the Janus kinase/signal transducer and activator of transcription (J AK/STAT)
pathway. Zong et al. (2002) demonstrated that STAT3 is activated by the IGF-1 receptor
and that JAKl or JAK2 is required for IGF-l-induced STAT activation. The STAT
family of proteins plays an important role in cellular proliferation and transformation, and
STAT3 has been shown to be important in EGF-regulated cell proliferation (Grandis et
aL,1988)

The end result of IGF-1 transmitting its signal through numerous pathways is that
different cellular mechanisms are influenced so that the cell may proliferate. One such
mechanism is the regulation of gene expression. Because IGF-1 transmits its signal

through different pathways, it can alter the expression of many genes at a given time. To

17

best capture the full extent of changes in gene expression in genes related to proliferation
and cell survival, a method of examining the changes in expression of a large number of
genes in a tissue is needed. Microarrays provide the means for examining global gene
expression changes.

The gene microarray

The concept of microarray design could be seen in dot blot experiments. Dot blot
experiments allowed for Simultaneous analysis of multiple recombinant DNA libraries.
In dot blot experiments, nucleic acids collected from samples (the targets) are spotted
onto a porous support, such as nitrocellulose. Next, nucleic acids of known sequences
(the probes) are labeled with fluorescent or radioactive markers and are hybridized to the
targets on the porous support. A deviation on this procedure, the reverse dot blot, was
created by Saiki et al. (1989) in which the probes were attached to the support. The
introduction of impermeable supports, such as glass and polypropylene, allowed for
consistent and defined spotting of nucleic acids onto the support and, more importantly,
the in situ synthesis of probes directly onto the support. The adaptation of ink-jet printing
and flow channel technologies provided for economically viable large-scale design and
creation of microarrays.

The actual procedure of performing microarray experiments is relatively
straightforward and consists of several steps. First, researchers collect messenger
ribonucleic acid (mRNA) from the experimental tissues. The mRNA is then amplified
into complimentary DNA (cDNA) using one of several commercially available reverse
transcriptase kits. The cDNA is then labeled with a fluorescent or radioactive marker and

hybridized to the probes on the array. After hybridization, the array is washed and

18'

scanned by a laser that is attached to a confocal microscope and a digital camera or is
measured for radioactivity levels. Pictures or radiograms are taken of the array and the
spots on the images are aligned using a spot alignment program such as Spotfire or
GenePix Pro. The data is then log transformed and normalized using a normalization
procedure such as LOESS before it can be analyzed for differences in gene expression
between treatments. Microarray analysis allows for rapid and cost-effective data
collection.

The bovine metabolism (BMET) array was designed to analyze the expression of
genes related to metabolism and metabolic regulation, including proliferation, in the dairy
cow (Etchebarne, 2005). A list of genes in metabolic and proliferative pathways was
extracted from online human genome databases such as the Kyoto Encyclopedia of Genes
and Genomes, Swiss-Prot Metabolic Pathway, and the Biocarta website. The human
sequences of the genes in this list were then paired for homology to bovine expressed
sequence tags using the Basic Local Alignment Search Tool. Highly homologous

sequences were found by selecting those tags that had an expectancy value of less than
10'”. A total of 2,360 bovine sequences related to metabolism and proliferation were

selected from these search methods for oligonucleotide design. Oligonucleotides of the
selected bovine sequences were custom made by the Massachusetts General Hospital
Microarray Core Facility and attached to poly-L-lysine coated slides. To reduce the
effects of spot position on array and improve the detection of small changes in gene
expression, each sequence was spotted 4 times on the array. Furthermore, housekeeping

gene sequences and sequences of genes from Arabidopsis thaliana were spotted on the

19

array to act as internal controls. Thus, the BMET array will allow us to accurately
determine changes in gene expression.
The MAC-T cell line as a model of mammary epithelial cells

To analyze gene expression changes, we need a bovine mammary epithelial cell
model that proliferates in response to IGF-1. The MAC-T cell line is an immortalized
bovine mammary epithelial cell line retains some epithelial cell characteristics.
Mammary epithelial cells fiom lactating Holstein cows were transfected with the simian
virus 40 large-T antigen. The transfected cells demonstrated the cobblestone morphology
and a cytokeratin fibril mesh that is characteristic of epithelial cells. Furthermore, upon
differentiation the cells reportedly rearrange themselves into lumen-like organoids and
express casein proteins (Huynh et al., 1991).

Zavizion et a1. (1995) examined the MAC-T cell line as a viable epithelial cell
model. They discovered that the MAC-T line was comprised of three different types of
mammary epithelial cells, each possessing different characteristics. The researchers
subcloned the MAC-T cells into three different clones: CU-l, CU-2, and CU—3. Each
subclone possesses different morphologies. CU-l did not form a “cobblestone” pattern
until it reached confluence and required fetal bovine serum (FBS) for growth. CU-3
contained epithelial-like cells but also had much larger, multinucleated cells.
Furthermore, the CU-3 subclone did not require FBS for growth. Finally, each subclone
exhibited differences in the chromosome arrangements. Zavizion et a1. looked for
evidence that one or more of these subclones may be myoepithelial in nature. The
subclones were devoid of vimentin, u-actinin, and u-smooth muscle actin filaments,

indicating that the cells were epithelial and not myoepithelial in nature. Furthermore, the

20

cells did not contract in the presence of lO'SM of oxytocin. The researchers concluded
that there was some instability in the phenotype of the MAC-T line. While this instability
may call into question the validity of the MAC-T line as an adequate mammary epithelial
cell model for examining differentiation, I believe that the MAC-T cell line is a viable
model to test the effects of IGF-1 on cell proliferation, and to test my hypothesis, it is
critical that the cells proliferate in response to IGF-1. Primary cells can undergo
senescence, in which they stop proliferating after being passaged too many times
(Matitashvili et al, 1997). The MAC-T cell line is a pure population of bovine mammary
epithelial cells, albeit with modifications, and, most importantly, MAC-T cells
consistently increase proliferation in response to IGF-1.

The MAC-T cell line responds in proliferation to IGF-1 treatments in a manner
similar to primary bovine mammary epithelial cells. In primary bovine mammary
epithelial cells, the maximal proliferative response using tritiated thymidine incorporation
occurs at ~25 ng/mL IGF-1 and is ~3.S times basal proliferation (Weber et al., 1999). In
MAC-T cells, the maximal proliferative response using tritiated thymidine incorporation
occurs at 5 to 10 ng/mL IGF-1 and is ~3 times basal proliferation (Silva, 2002; Jacobsen,
unpublished results). Therefore, much of the signaling pathways are likely still intact in

the MAC-T cells, and MAC-T cells should serve as a good model for my study.

21

Summary

Mammary development is impaired in heifers fed for high rates of gain.
However, IGF-1, a potent mammary epithelial cell mitogen, is increased in the serum of
heifers fed for high rates of gain. The binding of IGF-1 to its receptor initiates a
signaling cascade that promotes proliferation and cell survival, yet how IGF -1 affects
mammary development in heifers is not well understood. A better understanding of the
intracellular pathways involved in mediating the mitogenic effects of IGF-1 in bovine
mammary epithelial cells may help us understand how nutrition influence mammary
development.

My hypothesis is that IGF-l alters the expression of genes in a manner consistent
with proliferation in bovine mammary epithelial cells. Microarray technology allows for
rapid determination of changes in overall gene expression. The MAC-T cell line
provides an adequate model for looking at IGF-1 effect on bovine mammary epithelial
cells. Therefore, my objective was to determine if IGF-l treatment for 8 and 24 hours
alters the expression of genes in the MAC-T bovine mammary epithelial cell line in a

manner consistent with increased proliferation.

22

Materials and Methods
Cell Culture

In my thesis, “experiment” refers to the sets of cultures raised for gene expression
analysis at 8 hr and 24 hr on a given day. Experiments were repeated 3 times.

Lyophilized recombinant human IGF-1 (GroPep Pty, Adelaide, Australia) was
reconstituted in 2.5 mL of 100 mM HCl and then mixed with 2.5 mL of 100 mM NaOH.

The IGF-l was separated into 90-ul aliquots and stored at -20°C. Frozen immortalized
mammary epithelial (MAC-T) cells were plated at a density of 5x103 cells/cm2 in 12 T-

75 flasks. The MAC-T cell line was chosen as the biological model because it is a
homogenous cell population in terms of cell type and stage of differentiation.
Homogeneity avoids the phenotypic and genetic variation associated with collecting
primary cells from different animals (Mattitashvili et al., 1997). The cells were treated
with DMEM-F12 media (GIBCO) and 10% fetal bovine serum (FBS; Invitrogen,
Carlesbad, CA) and incubated overnight at 37°C in 5% C02. The media was
supplemented with 2 ng/mL insulin, 2 ng/mL sodium selenite, 10 ug/mL apo-transferrin,
2 ug/mL soybean trypsin inhibitor, and 2 ug/mL glutathione (Sigma). After incubating
for 24 hours, the cells were washed with phosphate buffered saline (PBS) and were
treated with serum-free medium, which consisted of DMEM-F 12 and 750 ug/ml bovine
serum albumin (BSA; Invitrogen) and incubated for another 48 hours.

Next, the cells were washed twice with PBS and treated with either 0 ng/mL or
100 ng/mL IGF-1 for 8 hours. The 100 ng/mL dose was chosen based upon the IGF-l
doses-response data from Silva et al. (2002). Silva showed a maximal proliferative
response of MAC-T cell to 100 ng/mL IGF-l. To verify the proliferative effect of this

dose on cell proliferation, MAC-T cells were treated with either 100 ng/mL or a serum-

23

free medium control for 18 hr. After 18 hr, the cells were exposed to tritiated thymidine
and proliferation was estimated by measuring the amount of tritiated thymidine uptake.
IGF-1 increased tritiated thymidine uptake by over 3-fold, demonstrating that 100 ng/mL
of IGF-1 has a proliferative effect on MAC-T cells. The IGF-l effect for this project was
validated by estimation of confluency by two observers blinded to treatment at 24 hours
after IGF-1 administration.

Figure 3. Proliferative effect of 100 ng/mL IGF-1 on MAC-T cells after 18 hr of IGF-1
treatment.

40000 a
35000 -
30000 ~
25000 a

20000 —

DPM

15000 -
10000 A

5000 i

 

   

0 — .
0 ng/mL 100 ng/mL

 

RNA extraction

The cells were washed with PBS before RNA isolation. RNA was extracted by
the Trizol method (Invitrogen). After washing the cells one time with PBS, cells were
scraped into 2 mL of Trizol per flask and they were incubated at room temperature for 5
min. Next, 200 uL of chloroform per ml of Trizol was added and the lysate was shaken

vigorously for 10 seconds. After incubating at room temperature for 3 min, the lysate

24

was centrifuged at 10,500 rpm for 15 min and the aqueous phase was mixed with 500 ILL
isopropanol. After incubating for 10 min at room temperature, the RNA was centrifuged
at 10,500 rpm for 10 min. The isopropanol was removed and the RNA pellet was washed
with 1 mL of 75% ethanol and centrifuged at 8500 rpm for 5 min. The ethanol was
removed and the pellet was allowed to dry at room temperature for 10 min. The pellet
was resuspended in 14.8 uL of RNase-free water and incubated in a 37°C water bath for
5 min. Next, the RNA from three flasks was combined to form a total volume of 89 pL
(14.8 mL water per mL Trizol from 2 mL Trizol per flask from 3 flasks). Then, 10 uL of
10X RQ DNase buffer and 1 uL of RQl DNase (lU/mL, Ambion) were added to each
RNA sample and the samples were incubated at 37°C for 30 min. The RNA was mixed
with 100 uL of a 25:24:] phenolzchloroformzisoamyl alcohol solution (pH 4.0,
Invitrogen). The RNA was centrifuged at 14,000 rpm for 2 min and the aqueous phase
was mixed with 9 pL of 3M sodium acetate (Ambion) and 250 pL ethanol. The RNA
was precipitated overnight at -20°C. After precipitation, the RNA was centrifuged at
14,000 rpm for 15 min and washed with 500 pL of cold 75% ethanol. The RNA was
centrifuged at 14,000 rpm for 10 min and the ethanol was removed. The pellet was
allowed to dry in a chemical hood for 15 min and then was resuspended in 25 uL of
nuclease-free water. After incubated at 55°C for 10 min, the RNA was immediately
transferred to ice. The concentration and quality were determined in the Center for
Animal Functional Genomics using a NanoDrop ND-1000 Spectrophotometer and an
Agilent 2100 Bioanalyzer. Samples were then stored at -80°C until use for

hybridizations.

25

Microarray hybridization
Complementary DNA (cDNA) was transcribed from mRN A and dye-coupled

using the SuperScript Indirect cDNA Labeling Core Kit and the cDNA Labeling
Purification Module (Invitrogen). First, 10 pg of RNA was mixed with 5 pg of anchored
oligo(dT) primers. DEPC-treated water was added to bring the volume to 18 pL. This
mix was incubated at 70°C for 5 min in a therrnocycler (GeneAmp PCR System 9700).
A “master mix” consisting of 6 pL 5X first-strand buffer, 1.5 pL 10 mM dNTP mix, 1.5
pL DTT, 1 pL RNaseOUT, and 2 pL SuperScript 111 per sample was added to each
sample and the reaction was continued at 46°C for 3 hr. Then, 15 pL of 1N NaOH was
added to each reaction, and the reactions were continued at 70°C for 10 min. The NaOH
was neutralized with an equal amount of 1N HCl at the end of the amplification.

Next, 20 pL of 3 M sodium acetate was mixed into each sample, and 500 pL of
the kit’s loading buffer were added to each sample. The cDNA was pipetted into a
S.N.A.P. columnTM (Invitrogen) and centrifuged at 12,000 x g for 1 min. A loading
buffer and a washing buffer from the Purification Module kit were mixed with 10 mL
isopropanol and 25 mL ethanol, respectively. The cDNA, which was trapped in the
column, was washed twice with 700 pL washing buffer and centrifuged at 12,000 x g for
1 min. The cDNA was eluted with two 50-pL washes of DEPC-treated water, and the
concentration was determined on the NanoDrop ND-1000 Spectrophotometer. The
cDNA was then mixed with 10 pL of 3M sodium acetate and 2 pL of 20 mg/mL
glycogen and precipitated overnight in -20°C.

The next day, the cDNA was centrifuged at 12,000 x g for 20 min and was

washed in 75% ethanol. The cDNA was then spun at 14,000 x g for 5 min and then dried

26

at room temperature for 12 min. The cDNA was resuspended in 5 pL of 2X coupling
buffer. The fluorescent dyes Cy3 and Cy5 (CyTMDye Post-labeling Reactive Dye Pack,
Amersham Bioscienees) were mixed in 11 pL of dirnethyl sulfoxide (DMSO). 5 pL of
the dye/DMSO mix was mixed into each sample. The samples were then covered in foil
and were incubated in the dark at room temperature for 2 hr. The dye-coupling reaction
was halted with 20 pL of 3M sodium acetate and the dye-coupled cDNA was mixed in
500 pL of washing buffer from the kit. The cDNA was loaded onto a S.N.A.P. column
and centrifuged at 14,000 x g for 1 min. The cDNA was washed with 700 pL of washing
buffer and centrifuged at 14,000 x g for 1 min twice per sample. The dye-coupled cDNA
was eluted by centrifuging 50 pL of DEPC-treated water through the column at 14,000 x
g for 1 min. The concentrations of dye-coupled cDNA were measured using the
NanoDrop ND-lOOO spectrophotometer. The Cy3 dye-coupled samples were mixed with
their respective Cy5 dye-coupled samples in one-1.5 mL Eppendorf tube and the
combined samples were centrifuged at 12,000 x g for 12 min through a Microcon filter
(Millipore). Next, 20 pL of SlideHybe #1 (Ambion) was added to the Microcon filter and
the filter placed upside-down into a clean Microcon tube. The combined sample was
centrifuged at 2500 rpm for 3 min. The probe was then brought to a volume of 110 pL.
The samples were then stored in a 68°C water bath until arrays were prepared for the
samples.

The samples were hybridized onto 2 BMET arrays per time. A dye-swap design
was used for the microarray experiment to remove the preferential binding of dyes to
certain transcripts. The arrays were loaded into a Genomic Solutions Hbetation and the

probe was allowed to hybridize to the array for 18 hr. After hybridization, the arrays

27

were washed three times in 0.06X SSC (Ambion) and were scanned on an Axon 40003
Scanner and the Agilent Scanner. Gains were adj ustcd such that the intensities of each
dye were approximately 1:1. Spots were aligned using the GenePix Pro software before
analysis.

Quantitative RT-PCR

Genes of interest were validated by quantitative RT-PCR using the SuperScript II
reverse transcriptase procedure. The genes of interest were selected due to their relation
to proliferation. To accurately validate the microarray data, I selected some genes that
were significantly upregulated, some that were significantly downregulated, and some
that were not significantly altered according to the microarray data. For genes to be
considered significantly altered (either upregulated or downregulated), they had to pass
two criteria: 1) they were expressed at P < 0.05, and 2) they were altered at a ratio of>1.2
or <0.8. This removes all genes that Show small changes in expression, which may be
attributed to random noise.

All products came from Invitrogen unless otherwise noted. First, cDNA was
synthesized from each RNA sample. I mixed 2 pg of RNA per sample with 1 pL (100
pM) of oligo dT12-13 primers, and the reaction was brought to 10 pL with RNase-free
water in a labeled PCR tube. The samples were then loaded into a GeneAmp PCR
System 9700 and were heated to 70°C for 5 min and then 20°C for 5 min. During this
time, a ‘master mix’ was prepared on ice for each reaction in which 4 pL 5X First Strand
Buffer, 2 pL 0.1 M DTT, 1 pL SuperScript II RNase H reverse transcriptase, 2 pL
RNase-free water, and 1 pL 10mM dNTP mix were combined for a total of 10 pL for

each sample. After the 20°C incubation, 10 pL of the master mix was added to each

28

reaction and the samples were heated to 42°C for 60 min and 70°C for 5 min. The
samples were cooled to 37°C, and 0.5 pL of RNase H (lOU/pL) was added to each
sample. The samples were kept heated at 37 °C for 20 min, after which 0.2 pL of 0.5 M
EDTA (pH 8.0) was mixed into each sample. Next, 5 pL of 3 M sodium acetate
(Invitrogen), 25 pL of RNase-free water, and 125 pL of ice-cold EtOH was added to each
sample and the cDNA precipitated overnight at -20°C. After precipitation, the cDNA
was spun at 14,000 x g for 20 min and washed with 250 pL of cold 75% EtOH. The
cDNA was spun at 14,000 x g for 6 min and the EtOH was removed. The pellet air-dried
for 15 min and was reconstituted in 50 pL of RNase-free water. The cDNA
concentration was measured using the NanoDrop ND-1000 spectrophotometer.

Quantitative real-time RT-PCR (qRT-PCR) was carried out using the SYBR
Green Master Mix from Applied Biosystems. Primer sets for approximately 20 genes
were tested at various concentrations for successful amplification of my samples. For
those primer sets that were successfully amplified, their amplification efficiencies were
measured for similarity with glyceraldehyde-3-phosphate dehydrogenase (GAPDH). The
primer sets 1 used are shown in table 1 on page 31. Primers were designed by Primer
Express (v. ) except for RPS9, RPSlS, and UTX, which were taken from Bionaz and
Loor (2007).

The RT-PCR reactions were run on two separate plates with three genes plus
GAPDH on each plate. All 6 treatment samples were performed in duplicate. First, 20
pg (2 pL of 10ng/pL) of sample cDNA, 3 pL of 5 pM primer mix, 7.5 pL RNase-free
water, and 12.5 pL SYBR Green Master Mix (Applied Bioscience) were mixed on ice for

a total reaction volume of 25 pL. The RT-PCR reaction were carried out in a ABI Prism

29

7500 RT-PCR System using the following protocol: 50°C for 2 minutes, 95°C for 10 min,

41 cycles of 95°C for 15 seconds and 60°C for l min, and 95°C for 15 seconds and

cooled to room temperature.

30

Table 1. Quantitative RT-PCR primer sets.

 

Melting Product
Accession # Symbol Primers set Temp (°C) leggth
XM_602780 SLC39A6 F — GCACTI‘ACTGCAGGCT'I‘GTCAT 79 92
R - CGGCTACATCCATGGTCACTAG

XM_581382 IRS 1 F — TGCGGCCACTCAGAGAACTT 85 124
R - CCAGGATTGTCTCGTGCATGT

NM__174000.2 CALR F - CCGTITACTTI‘AAGGAGCAGTTI‘CTG 82 70
R - TTGTGCTTGGATTCGATCCA

NM_174308.1 EDNRIA F -ATGGACACGAACCGATGTGA 77 72
R - GGTTGCCAAGTTAATACCGATGT

NM_174313.2 F ABP3 F - CCACAGCAGATGACAGGAAAGTC 82 68
R - CTGCACGTGGACAAGTTTGC

NM_175801.1 FST F - TCCCTTGTAAAGAAACGTGTGAGA 79 67
R - TCGCCCTCGTCCTTGTCA

BF606842 HSPAS F - AAGATGTTCGGAAGGACAACAGA 82 67
R - GCCCGTTTGGCCI I I ICTAC

AB072368.1 HSPCA F - CACCGGCATTGGGATGA 82 63
R - CCGGACTTGGCGATGGT

NM_174130.2 ODCl F - CGCATTGTTGAGCGCTGTA 80 66
R - CATGTTCTCAAAGAGCATCCAATC

NM_174217.1 VIL2 F - GCAGCI I I I IGATCAGGTGGTT 75 90
R - TCCACATACTGGAGGCCAAAGT

Not 188 F — GAGAAACGGCTACCACATCCA Not Not
Available R - GACACTCAGCTAAGAGCATCGA Available Available
Not B-actin F — CGCCATGGATGATGATAT‘TGC Not Not
Available R - AAGCCGGCCTTGCACAT Available Available
DT860044 RPS9 F — CCTCGACCAAGAGCTGAAG Not 54
R — CCTCCAGACCTCACGTITGTTC Available

XMS 85783 RPSIS F - GCAGCTTATGAGCAAGGTCGT Not 151

R - GCTCATCAGCAGATAGCGCTT Available
BQ676558 UTX F — TGTGGCCCTTGGATATGGTT Not l 10

R - GGYYGYCGCTGAGCTCTGTG Available
Not Available GAPDH F — GCATCGTGGAGGGACTTATGGA Not Not

R - GGGCCATCCACAGTCTTCTG Available Available

 

31

Data Analyses

Differences in cell confluencies between treatments were analyzed using a two-
tailed t-test. Median intensities of microarray spots were log-transforrned (base-2) and
the microarray data was normalized by the LOESS procedure. The microarray data
collected from both arrays was analyzed utilizing two mixed linear models, as outlined by
Wolfinger et a1. (2001). First, differences across all microarrays were standardized using
the following model,

yijk = p + dyei + array j + dye * array”- + block(array)jk + dye * block(array),-1* + gijk

Where dye represented the fixed effect and dye*array, block(array), and dye*bloek(array)
represented the random effects for gene i, dye j, array k, and block 1.

In the second model, residuals for each gene were calculated by subtracting fitted
values obtained from the first model from the observed intensities. Differences in
individual gene intensities were analyzed by the following model,

rag-[mm = p,- + treatmentm + scanner" + cultureo + treatment * culturemo

+ treatment * scanner * culturemno + gijkl

Where treatment and scanner represent the fixed effects and culture, treatrnent*culture,
and treatrnent*scanner* culture represent the random effects of treatment m, scanner n

and culture 0. The data was analyzed using the PROC MIXED procedure of SAS (v.

9.0) using both the Type III Sums of Squares method.

Data for the quantitative RT-PCR experiments was analyzed using the 2M0

method proposed by Livak and Schmittgen (2001). The cycles-to-threshold (Ct) values
for 188 for each replicate in each sample was subtracted from the Ct values for the genes

of interest for each replicate in each sample to obtain the delta Ct values. Independent t-

32

tests were performed on the delta Ct values to determine differences in gene expression
due to IGF-1 treatment. Because I was verifying the fold-change directions, I analyzed
the genes shown to be statistically significant on the microarray data using a one-tailed t-
test. I used a two-tailed t-test on the non-significantly expressed genes. The fold-change
values for the quantitative RT-PCR experiment were calculated using the following
equation:

F 01 61- Ch an ge = 2(-(gene Ct value — 188 Ct value)-(average 0 ng/pl gene Ct value — average 0 ng/pl 18S Ct value))

33

Results
Cell Confluency Results

I verified the biological effect of IGF-1 on bovine mammary epithelial cells by
estimating cell confluencies after treating the cells with IGF-l for 24 hours in separate
flasks. The results are shown in figure 4. IGF-1 treatment for 24 hr increased cell
confluencies by 40% as compared to control cells that were not treated with IGF-1 (P < ’

0.05). Furthermore, IGF-1 tended to increase cell confluencies at 8 hr by 20% (P = 0.06).

Figure 4. Average cell confluencies at 8 and 24 hours. Error bars are expressed as SEM.

907
804
707
60'
50— ’-'§----------‘-+

40' -e-Control

30 -
20 - —o—IGF-I

10 ~
0 I r I
0 8 16 24

Time of Confluency Estimation, hr

 

% Estimated Confluency

 

 

Microarray results

To determine if IGF-1 treatment altered global gene expression, the histograms of
the P-values of treatment comparisons for each gene were examined in both the 8 hr and
the 24 hr microarray data, as shown in figure 5. These histograms should not Show any
differences in bar heights if IGF-1 treatment did not affect gene expression. However,
the large frequency of genes that were altered at average P-value < 0.05 showed that IGF-

] altered the expression of genes in MAC-T cells at 8 hr and 24 hr of treatment.

34

Figure 5. P-value distribution histogram of 8 hr 24 hr microarray data.

800 j
700 ‘

 

 

 

 

0.05 0.15 0.25 0.35 0.45 0.55 0.65 0.75 0.85 0.95
Average P-value of each probablllty declle

Genes that showed fold-changes in expression of either greater than 1.2 or less
than 0.8 due to IGF-1 and had P < 0.05 were deemed significantly altered by IGF-1.
Table 2 shows the number of upregulated and downregulated genes at 8 hr and 24 hr of
treatment.

Table 2. Number of genes altered by IGF-l at P < 0.05.

 

 

8 hr 24 hr
Upregulated > 20% 70 139
Downregulated > 20% 19 45
Total 89 184

These genes were then organized into groups based upon their function as defined
by the KEGG pathway and Gene Ontology function. Table 3 shows the functional
categories of genes significantly altered by IGF-1. IGF-1 treatment for 24 hr altered a
greater percentage of transcribed genes across most categories than treatment for 8 hr.
However, a greater percentage of coagulation factor genes (16.7%) and DNA replication

and repair genes (16.3%) were altered and also upregulated at 8 hr than at 24 hr (12.5%

35

respectively). A greater percentage of genes were upregulated in both of these categories
at 8 hr than at 24 hr. IGF-1 also downregulated the transcription of a higher percentage

of cell signaling (5.1%) genes at 8 hr than at 24 hr (3.1%).

36

Table 3. Functional categories of genes whose expression was altered by IGF-1

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

treatment for 8 hr and 24 hr.
8 hr 24 hr
total % %
number upreg % altered % % %
on of downreg of upreg downreg altered

Function array total of total total of total of total of total
Carbohydrate
Metabolism 113 1.8 0.9 2.7 9.7 6.2 15.9
C°" Gym and 259 0.8 , 0.0 0.8 1.9 0.3 2.7
Death
Cell signaling 98 10.2 5.1 15.3 15.3 3.1 18.4
Cell Structure
and 44 91 91 132 341 205 545
Extracellular ' ' ' ' ' '
Matrix
C°°g“'°“°" 24 16 7 0 0 16 7 8 3 4 2 12 5
Factors ' ' ' ' ' '
DNA
Replication 49 16.3 0.0 16.3 6.1 2.0 8.2
and Repair
Energy
Metabolism 51 21.6 9.8 31.4 21.6 15.7 37.3
Folate
Metabolism 22 0.0 0.0 0.0 31.8 0.0 31.8
Glycan
Metabolism 38 5.3 5.3 10.5 26.3 10.5 36.8
Lipid
Metabolism 198 2.0 0.0 2.0 4.0 2.0 6.1
Protein
Synthesis and 187 4.8 0.0 4.8 10.2 1.6 11.8
Metabolism
Purine and
Pyrimidine 96 0.0 0.0 0.0 4.2 0.0 4.2
Metabolism
5'3““ . 252 2.0 0.4 2.4 4.8 0.4 5.2
Transduction
Transcription 236 0.0 0.0 0.0 1.3 0.0 1.3
Transport 69 8.7 0.0 8.7 8.7 2.9 11.6
Unknown 624 0.5 0.2 1.7 1.3 0.0 1.3

 

 

 

 

 

 

 

 

 

37

 

IGF-1 increases proliferation in mammary epithelial cells, yet there are different
mechanisms by which it promotes cell proliferation. We examined changes in the
expressions of genes related to proliferation and survival to see if IGF-l regulates cell
proliferation through gene expression. The results can be seen in tables 3 and 4. IGF-1
altered the expression of genes related to the cell cycle and polyamine synthesis at both 8
and 24 hours of treatment. IGF-1 also altered the expression of intracellular signaling
genes that are related to proliferation, particularly genes related to the mitogen-activated
protein kinase (MAPK) pathway and the Wnt signaling pathway. Furthermore, evidence
of IGF-1 signaling through the Janus kinase — signal transducer and activator of
transcription (J AK-STAT) pathway could be seen by the downregulation of protein

inhibitor of activated STAT 1 (PIASl) at 8 hours.

38

Table 4. Proliferation and survival-related genes whose expression was regulated by

IGF-1 at 8 hr or 24 hr.

Gene Name

Symbol

8hr

24hr

 

Fold-

change values

p-

FDR

Fold-

change values

P-

FDR

 

Cell cycle

 

CDC6 cell division
cycle 6 homolog (S.
cerevisiae)

CDC6

1.85

0.03

0.30

1.58

0.02

0.11

 

cyclin E1

CCNEl

1.37

0.03

0.31

1.31

0.04

0.14

 

cyclin-dependent
kinase inhibitor 28

CDKN2B

0.89

0.39

0.52

0.77

0.01

0.09

 

Cell signaling

 

epidermal growth
factor

EGF

l .44

<0.01

0.20

1.34

0.10

0.18

 

fatty acid binding
protein 3, muscle and
heart (mammary-
derived grth
inhibitor)

FABP3

2.13

<0.01

0.26

1.95

<0.01

0.08

 

insulin-like growth
factor binding protein
2, 36kDa

IGFBP2

1.00

0.94

0.66

0.78

0.02

0.11

 

insulin-like growth
factor binding protein
3

IGFBP3

1.40

0.03

0.31

1.42

0.01

0.08

 

Stress

 

heat shock 70kDa
protein 5 (glucose-

regulated protein,
78kDa)

HSPAS

1.30

0.09

0.40

1.33

<0.01

0.05

 

heat shock 90kDa
protein 1, alpha

HSPCA

1.33

0.18

0.45

1.36

0.03

0.13

 

Polyamine synthesis

 

omithine
decarboxylase 1

ODCl

1.60

<0.01

0.26

1.47

0.01

0.10

 

spermidine synthase

SRM

1.47

0.10

0.42

1.36

0.02

0.12

 

Signal transduction

 

FOS-like antigen 1

FOSLI

1.38

0.04

0.34

1.41

0.01

0.09

 

 

insulin receptor
substrate 1

 

IRS l

 

0.87

 

0.40

 

0.52

 

0.76

 

0.01

 

0.10

 

39

 

Table 4 (continued).

 

 

 

 

 

 

 

 

 

8 hr 24 hr
Fold- P- Fold- P-

Gene Name Symbol chan e values FDR change values FDR
lyfiififfir‘lgimf" LEFl 0.77 0.03 0.29 0.83 0.00 0.05
firtggfi'ffuvztjd MAPK4 0.95 0.21 0.46 0.74 <0.01 0.08
“figfgfl’gxtgd MAPK6 0.92 0.86 0.64 1.29 0.02 0.11
mitogen—activated
91‘9“"; $32233 MAP3K7IP2 0.83 0.38 0.51 1.22 0.05 0.15

protein 2
mitogen-activated
mid“ km” MAPKAPK3 123 00 036 125 3 012
activated protein ' ' 5 ° ' 0'0 '
kinase 3
1222;336:181??? PIASl 0.73 0.04 0.34 0.95 0.27 0.29
protein tyrosine
phosphatase, receptor PTPRR 1.38 0.01 0.27 0.88 0.81 0.52
type. R

 

 

 

 

 

 

 

 

Samples from two different experiments were used for one set of 8 hr arrays. To
determine if this influenced the results, the regression of the differences in residuals of all
genes that were expressed at P < 0.1 between all three array sets were examined. This
significance level was chosen to remove all genes that showed very little differences in
dye intensities. As Show in Table 5 on page 41, the array sets were strongly and
positively correlated. Experiments 1 and 2 represent those arrays sets that had samples
within an experiment. Experiment 3 had samples from different experiments. The
correlations were highly Significant, thereby suggesting that the samples from two

different experiments did not affect the 8 hr microarray data.

40

 

Table 5. Regression and correlation values of all genes expressed at P < 0.1 between the
8 hr arrays

Comparison R-squared Correlation P-value

 

Exp 2 vs Exp 1 0.60 0.77 <0.001
Exp 2 vs Exp 3 0.60 0.77 <0.001
Exp 1 vs Exp 3 0.59 0.73 <0.001

 

The full lists of differentially expressed genes are found in the appendices.

Appendix B lists the genes altered after 24 hr IGF-l treatment and Appendix C lists the
genes altered after 8 hr IGF-l treatment.

41

Quantitative RT-PCR results

Quantitative RT-PCR was used to validate significant gene expression changes as
reported by the microarray data analysis. A suitable housekeeping gene was needed to
accurately analyze the RT-PCR data using the AACt method. The fold-changes in the
expressions of 6 potential housekeeping genes were measured across all experiments.
The results can be seen in figure 6. UTX, RPSIS, and B-actin were significantly altered
by IGF-1 treatment at 24 hr. Of the 3 candidate genes remaining, GAPDH was chosen
because it was not significantly altered by IGF-1 and others have used it as a
housekeeping gene in MAC-T cells (Smith and Sheffield, 2002).

Figure 6. Quantitative RT-PCR results of candidate housekeeping genes in cells treated
with IGF-1 for 24 hr.

RT—PCR Expression of Housekeeping Genes
2 5 .00

l

20.00 ~

 

 

15.00 —
(’3
I Control
1:1 IGF-l

10.00 3

5.00 *

 

 

 

 

 

 

 

 

 

0.00 ~

——1

GAPDH 18$ UTX RPS] 5 RPS9 B-actin

P = 0.10 0.80 0.03 0.03 0.08 0.04
Gene

The expression levels often genes were analyzed using quantitative real-time RT-
PCR to verify the array results. These genes were selected because they were involved in

cell proliferation and were either upregulated, downregulated, or not regulated (P < 0.05)

42

by IGF-1 treatment. The microarray and qRT-PCR results for each gene are shown in

table 6. Quantitative RT-PCR confirmed the upregulation of only 2 genes (FABP3 and

ODCl) at P < 0.05 out of the 5 that were upregulated in the microarray analysis. The

qRT-PCR fold-change values for each gene were correlated to the microarray fold-

change values for that same gene to determine if the RT-PCR fold-change values were

similar to the microarray fold-change values. For 7 of the 10 genes, the correlations

between the microarray fold-change values and the qRT-PCR fold-change values were

positive. For the other 3 genes, there was no correlation.

Table 6. F old-change values from microarray analysis and qRT-PCR for the validated

 

 

 

 

 

 

 

 

 

 

 

genes.
Microarray RT-PCR

Gene fold- P- fold- P- .
Gene Name Symbol change value change value Correlation
fatty acid binding protein 3,
muscle and heart (mammary- FABP3 1.95 <0.01 2.06 <0.01l 0.74
derived growth inhibitor)
endothelin receptor type A EDNRA 1.47 .0'01 1.58 0.061 -0.08
ornithine decarboxylase 1 ODCl 1.47 0.01 1.24 0.041 0.86
h°°fish°°k 90kDa 9mm“ 1’ HSPCA 1.36 0.03 1.32 0.141 0.76
heat shock 70kDa protein 5 7
(glucose-regulated protein, HSPAS 1.33 <0.01 1.61 0.061 0.71
78kDa)
calreticulin CALR 1.19 0.07 2.00 0.222 0.85
villin 2 (ezrin) VIL2 0.98 0.61 0.92 0582 0.37
follistatin FST 0.88 0.08 0.80 0.06 0.44
insulin receptor substrate 1 IRSl 0.76 0.01 0.93 0.401 -0£04
°°lm° mm“ “my 39 (mm SLC39A6 0.74 <0.01 1.01 0.481 -005

 

transporteerember 6

 

 

 

 

 

 

 

1 = one-tailed t-test
2 = two-tailed t-test

43

 

Discussion of Results

When mammary epithelial cells are exposed to IGF-1, they initiate cell-cycle
progression and activate antiapoptotic mechanisms to promote proliferation. My
hypothesis is that IGF-1 alters the expression of genes in a manner consistent with
increased proliferation in bovine mammary epithelial cells. The results suggest that my
hypothesis is correct. The qRT-PCR results confirmed the expression levels of 9 of our
10 genes of interest, although the statistical significance was the same for only 6 of those
genes. We still need to determine which analysis is biologically accurate.

My microarray data Shows that IGF-1 altered the expression of genes regulating
proliferation and cell cycle. I found that IGF-l increased the expression of ornithine
decarboxylase (ODCl) and spermidine synthase (SRM), enzymes involved in polyamine
synthesis. ODCl converts omithine to putrecine via decarboxylation and is the rate-
limiting step in polyamine synthesis. SRM catalyzes the conversion of putrecine to
spermidine in a similar fashion. Proliferating cells require polyamines to continue DNA
elongation during the S phase (Oredsson, 2003). However, the type of polyamines
required during the different phases of the cell cycle varies. Putrecine levels are doubled
during the S phase and during the S/G2 transition, while Spermine levels double during
the GI phase (Fredlund et al., 1995). Polyamines are regulated in conjunction with the
cell cycle. ODCl synthesis and activity is cell-cycle specific. In Chinese hamster ovary
cells that were made to proliferate synchronously during the cell cycle, ODCl activity
was Shown to increase at the Gl/S transition and again at the S/G2 transition (Oredsson,
2003). Furthermore, ODC mRNA levels were increased during the Gl/S transition but

not the S/GZ transition. Blocking ODCl activity with a-difluromethylomithine led to a

44

cessation of cell proliferation and an increase of cells accumulating at the GI phase
(Oredsson, 2003). Inhibition of SRM activity in chick embryo fibroblasts also blocks
DNA synthesis (Caruso et al., 1992). IGF-1 treatment for 24 hr increased ODCl mRNA
expression in breast cancer cells by 3.5-fold (Huber and Poulin, 1996). Therefore, my
results suggested that IGF-1 promotes progression of proliferating mammary epithelial
cells through the cell cycle by upregulating of ODCl and SRM expression. ODCl and
SRM then increase polyamine synthesis, thereby promoting DNA synthesis.

IGF-1 also increased the expression of the heat-shock proteins HSPCA and
HSPAS in my study. The expression of the chaperone heat-shock protein 90 (HSPCA) is
increased by a number of growth factors, including IGF-1, just before DNA synthesis
occurs (Jerome et al., 1991). However, HSPCA appears to have a buffering function on
IGF-1 signaling. Blocking HSPCA leads to an amplification of Akt activation, increased
p38 activation and an increased duration of ERK1/2 activation (Meares et al., 2004).
Therefore, in our MAC-T model, IGF-1 might have increased the expression of HSPCA
as a safeguard against overactivity of the proliferative signal. Heat-shock 70kDa protein
5 (HSPAS), an endoplasmic reticulum-localized chaperone protein, blocks the
proapoptotic activities of caspase 7 in cells challenged with topoisomerase inhibitors,
thereby promoting cell survival (Reddy et al., 2003).

In my study, IGF-1 increased the expression of cyclin E1 and decreased the
expression of cyclin-dependent kinase inhibitor 2B (p15, also known as INK4B), genes
directly associated with the cell cycle. Cyclin E1 regulates the passage of the cell
through the 01/8 transition and is required for the initiation of DNA replication (Harper

and Brooks, 2005). Mouse mammary explants exposed to IGF-1 showed an increase in

45

cyclin B mRNA levels (Stull et a1. 2002). The cyclin-dependent kinase inhibitor p15 is
part of the INK4 family of inhibitors. These bind specifically to cyclin D/cdk complexes
and inhibit their actions, thereby allowing the cell to begin the S-phase of the cell cycle.
Based upon the microarray data, IGF-1 appears to be promoting passage of the cells
through the Gl/S transition and the S-phase of the cell cycle by increasing the expression
of cyclin E1 and decreasing the expression of pl 5.

The expression of two parts of the endothelin signaling system, endothelin 1
(EDNl) and the endothelin receptor type 1 alpha (EDNRA), were regulated by IGF-1.
Endothelin 1 increase cell proliferation and DNA synthesis in many different systems
(Battistini et al., 1993). Our results Show that EDNl expression is downregulated by
IGF-1. This is supported by the findings that EDNl expression is increased in the aortas
of liver-specific IGF-l-knockout rats (Tivesten et al., 2002). However, IGF-l increased
EDNl expression in cultured chondrocytes (Messai et al., 2000), thereby suggesting that
the effect of IGF-1 on EDNl expression is system specific. Furthermore, our results
Show that EDNRA expression was increased by IGF-l. Similar results have been found
in vascular smooth muscle cells (Kwok et al., 2005). The net result of IGF-1 promoting
the expression of the receptor and not the ligand is that IGF-1 would make the cell more
responsive to endothelin signaling from other cells. The presence of more endothelin
receptors on the cell will increase endothelin signaling to the epithelial cell. Yet, by not
increasing the expression of endothelin, IGF-1 may control proliferation by preventing
the formation of an autocrine positive-feedback loop. This suggestion needs to be more
fully explored as IGF-1 might affect endothelin signaling via posttranscriptional and

translational analysis methods.

46

IGF-1 also increased the expression of fatty-acid binding protein 3 (FABP3),
otherwise known as mammary-derived growth inhibitor (MDGI). MDGI causes
inhibition of cell proliferation in serum-deprived cells. MDGI inhibits MAC-T cell
proliferation in a dose-dependent manner (Zavizion et al., 1993). However, this
inhibition of growth disappears after six days of MDGI treatment. Furthermore, cell
quiescence is required for the actions of MDGI as cells that were not serum-starved for
14 hours showed a minimal inhibition of proliferation (Zavizion et al., 1995). In our cell
model, IGF-1 increased the expression of MDGI by almost 2-fold. One explanation for
this seeming incongruous result is that cell-contact inhibition may be responsible for this
increase. MAC-T cells proliferate in colonies (Huynh et al., 1991); consequently after a
period of proliferation, cells will form large groups of cells. Cells along the fringe of the
colony have room to divide; thus, they are not affected by cell crowding. However, they
surround the cells in the center of the colony, which do not divide due to cell-contact
inhibition. Related to this is evidence Shows that MDGI transcripts are highly expressed
in lactating mammary glands (Kurtz etal., 1990). Given that cell contact is a necessary
condition for differentiation in cultured mammary cell models, the cells within each
colony could be expressing MDGI to prepare the cells for differentiation. In mouse
mammary explants treated with MDGI, lobuloalveolar formation and beta-casein
expression was increased and epithelial cell growth was decreased (Kurtz et al., 1998).
However, this needs to be explored further by estimating gene expression differences
between crowded and not crowded cells.

Another reason that IGF-1 may increase the expression of MDGI is that the

MDGI protein is a mixture of two fatty acid binding proteins with highly homologous

47

sequences. Specht ct al. (1996) demonstrated that the amino-acid sequence of heart-
derived fatty acid binding protein (H-FABP, which is considered to be the MDGI protein)
contains only 7 different amino acids compared with adipocyte-derived fatty acid binding
protein. Furthermore, they discovered that MDGI mRNA from lactating bovine
mammary tissue demonstrated the presence of the adipocyte-derived fatty acid binding
protein along (A-FABP) with H-FABP. Whether A-FABP is the protein that inhibits
proliferation has yet to be determined. However, Specht et al (1996) showed that H-
FABP inhibited mammary cell proliferation. Therefore, the increase in MDGI mRNA in

this experiment could be due to the presence of A-FABP mRNA.

48

Discussion of Approach

This experiment tested my hypothesis that IGF-1 alters the expression of genes in
a manner consistent with increased proliferation in bovine mammary epithelial cells. The
biological model used for this experiment was the MAC-T cell line, a transformed cell
line. Transformation induces the cell to be able to continuously proliferate under the
appropriate signal. Thus, one of the factors for continual proliferation could be a constant
upregulation of proliferation-inducing genes or a constant downregulation of
proapoptotic genes. An alternative model that could have been used is primary mammary
epithelial cells. Primary cells are not transformed to continually proliferate, thereby
avoiding a possible bias towards cell proliferation. However, primary cells can undergo
senescence, in which they stop proliferating after being passaged too many times
(Matitashvili et al, 1997). To test my hypothesis, it is critical that the cells proliferate in
response to IGF-1. As previously discussed, MAC-T cells respond to IGF-1 by
increasing proliferation. In fact, in my preliminary results, MAC-T cells treated with 100
ng/mL IGF-1 synthesized DNA at greater than 3 times the rate of control cells.
Therefore, much of the signaling pathways are likely still intact in the MAC-T cells.
While my study provides a foundation for understanding the effects of IGF-1 on gene
expression in bovine mammary epithelial cells, future studies Should be conducted to
determine if primary mammary epithelial cells respond in a similar manner.

One effect that should be examined is the effect of substratum on gene expression.
The MAC-T cells in this experiment were grown in collagen-coated flasks. However, a
different substratum, such as a collagen gel, could alter gene expression in place of IGF-

]. Huynh et al. (1991) showed that MAC-T cells grown on floating collagen gels

49

produced more B-casein mRN A than cells plated on plastic substratum. When examining
the differences between the clonal and parental MAC-T cells, Zavizion et al. (1995)
noticed that one of the colonies grew in an atypical manner on collagen than the other
colonies and the parental cell line. However, when treated with mammary extracts fi'om
prepubertal dairy heifers, cells plated on plastic substratum grew in a Similar dose-
dependent manner as cells plated on a collagen substratum (Berry et al. 2003).

To verify the biological actions of IGF-1 in this study, I measured cell confluency
in the culture flasks. I wanted to confirm that the IGF-1 used in this genomics study was
biologically active and was likely stimulating proliferation as it had in my preliminary
study. Percentages of confluency were used to qualitatively verify the biological activity
of IGF-1 without compromising my ability to isolate high-quality RNA from the cells.
Previous studies have consistently shown that IGF-1 increases proliferation in MAC-T
cells (Zhao et al., 1992; Woodward et al., 1994; Robinson et al., 2000). Furthermore, 100
ng/mL IGF-1 increased cell proliferation along with confluency in my preliminary study.
Therefore, I was confident that because IGF-l increased confluency in this genomics
study, it likely also increased proliferation. However, I recognize that confluency is not
necessarily proportional to rate of proliferation because confluency can be affected by
changes in cell size.

In my project, I tried to collect mRNA at both 8 and 24 hr of control and IGF-1
treatments for each of three experiments. However, I did not collect quality mRNA from
each sample of each experiment. Thus, I actually conducted four experiments instead of
three. In experiment 1, all of the 24-hr mRN A was lost due to poor hybridizations, so

gene expression from only the 8-hr arrays was measured. Experiment 2 worked as

50

planned. The mRNA fi'om the 24-hr cells was of high quality for both experiments 3 and
4. However, the mRN A from the control cells at 8 hr fiom experiment 3 and the mRNA
from the IGF-1 treated cells at 8 hr from experiment 4 were degraded. Therefore, I used
the mRNA from the 100 ng/ml treatment from experiment 3 and the mRN A from the 0
ng/ml treatment from experiment 4 as a set of 8-hr mRNA for treatment comparisons. I
justify this because the MAC-T cells used in all of the experiments are from the same
passage. Therefore, there should be very little genetic variation between the experiments,
as confirmed by Table 5.

Even though the microarray analysis demonstrated changes in gene expression,
these results were not confirmed for 4 of the 10 genes with qRT-PCR. One possible
explanation is the number of biological replicates used in this experiment was too low to
reduce the random variation. Three replicates of different culture times were used. A
power test was not conducted because no previous data with the BMET array and the
MAC-T cells were available. Three biological replicates were used to establish a
foundation for determining the effects of IGF-1 on gene expression. However, this may
not have been enough to detect significant changes in gene expression as measured with
qRT-PCR.

The small sample number can affect the false discovery rate (FDR). The FDR is a
statistic that calculates the probability of false discoveries that researchers are expecting
to see in microarray data. Given that we had only three biological replicates, we would
expect the FDR to be less significant. Pawitan et al. (2005) stated that to control for FDR
the sample size should be large, for example, 45 arrays per group to get a 10% FDR if the

proportion of non-Significantly expressed genes is 99% and the researchers was to select

51

the top 1% of significantly expressed genes. For smaller sets of arrays, genes must be
highly significant to control for FDR. Given that we used a small set of arrays (6 arrays
per time period), we would expect the FDR to be less significant.

Another effect that may explain the lack of Significant results by qRT-PCR is the
use of SYBR Green. SYBR Green is relatively inexpensive and easy to use. However,
SYBR Green binds to any double-stranded nucleic acid sequence. Therefore, if primer
dimers have formed between the forward and reverse primers for a gene of interest, the
SYBR Green will incorrectly label that cDNA amplification product. Another RT-PCR
procedure, the Taqman method, uses fluorescence tagged probes that bind to the target
strand between where the two probes anneal. The ends of the probe are labeled with
fluor tag and a quencher tag. Because the two tags are in close proximity, the quencher
blocks light emission from the fluor. During RT-PCR, 5’-exonuclease of the Taq
polymerase removes bases from the probe, including the fluor- and quencher-tagged
bases. When these tagged bases are removed, the fluor emits light Since it is not in as
close proximity to the quencher. However, the fluor-tagged base only is cleaved when
the probe anneals to it complementary sequence on the target gene transcript. Therefore,

Taqman increases the sensitivity of RT-PCR.

52

Conclusion

The act of proliferation requires different cellular machinery than the act of
differentiation. My hypothesis was that IGF-1 alters the expression of genes in a manner
consistent with increased proliferation in bovine mammary epithelial cells. My objective
was to determine if IGF-1 treatment for 8 and 24 hours alters the expression of genes in
the MAC-T bovine mammary epithelial cell line in a manner consistent with increased
proliferation.

In summary, IGF -1 increased cell confluency by 40% after 24 hr of treatment (P
< 0.05). IGF-1 altered the expression (P < 0.05) of 89 genes after 8 hours (70 increased,
18 decreased) and 184 genes after 24 hours (139 increased, 45 decreased). IGF-1 altered
the expression of several regulatory genes that might increase cell proliferation, such as
those for polyamine synthesis, cell cycle progression, and stress response, and several
other genes that support increased proliferation, such as metabolism and cell structure
genes. The fold-changes of 9 of 10 genes as measured with RT-PCR were Similar to
those with microarray analysis, although the statistical significance of the change was the
same for only 6 of the genes. In conclusion, IGF-1 alters the expression of proliferative
and metabolic genes in a manner consistent with increased cell proliferation.

Transcriptional regulation is not the only mechanism that IGF-1 can use to
promote proliferation. Proteins can be modified or destroyed, pathways can be sped up
or Slowed down, and physical migration of the cells may be increased or decreased.
These mechanisms have been studied in other cell systems in other animals. Examining
whether the same mechanisms are altered by IGF-1 in its mitogenic effects would assist

in better understanding the biological effects of IGF-1 in the bovine mammary gland.

53

Appendix A
Introduction

This work was a preliminary project that I did before examining IGF-l effects on
gene expression. I include these results as an extension of my graduate work and not as
part of the main thesis.

Leptin is a 16-kDa peptide that is secreted primarily by adipocytes and informs
the brain on the energy status of the body. Feed intake increases circulating leptin levels
in the body (Ahima and Flier, 2000). However, leptin has numerous other functions,
including regulating cell proliferation (Maor et al., 2002). Circulating leptin levels are
increased in dairy heifers fed to gain 1 or more kg/d (Block et al., 2003). Silva et al.
(2002) hypothesized that leptin negatively affected the stimulatory actions of insulin-like
growth factor 1 (IGF-1) on mammary epithelial cell proliferation in prepubertal Holstein
heifers. To accomplish this, they infused four treatments, 0 or 100 pg leptin mixed with
0 or 10 pg IGF-1, into each of the four quarters of the mammary gland in six Holstein
heifers. Based upon previous work, Silva (2002) Showed that each quarter of a bovine
mammary gland was not influenced by hormonal treatments in the other quarters and
could act as its own experimental unit. The heifers were infused with the treatments once
a day for 6 days then twice on day 7, with 14 hours separating the last two infusions. On
day 8, the heifers were infused with bromodeoxyuridine (BrdU) for 2 hours, slaughtered,
and the glands were sampled to measure incorporation of BrdU into the DNA of dividing
cells. Silva found that quarters treated with leptin and IGF-1 showed reduced mammary

development by 52% when compared to quarters treated IGF-1 alone.

54

While the results seem to implicate leptin as a mediator for decreased mammary
development in heavy prepubertal heifers, the dose of leptin used could have been
supraphysiological. Indeed, mammary extracts fiom leptin-treated quarters contained
171 ng leptin per mL extract, while saline treated quarters only contained around 4
ng/mL. Therefore, we hypothesized that smaller doses of leptin would also impair IGF-
l-induced mammary epithelial cell growth, albeit at lower levels. My objective was to
examine differences in mammary epithelial cell growth in quarters treated with three
different doses of leptin. IGF-1 was included in all of the doses.

Materials and Methods

Two Holstein heifers (8 months, average 400 lbs) were obtained from and housed
on the Michigan State University Dairy Cattle Teaching and Research Center. Using
Spartan Dairy v 2.0, diets were designed for an average daily gain of 0.7 kg/d and a crude
protein to metabolizable energy (CP/ME) ratio of approximately 60 g CP/Mcal ME.
Throughout the adaptation and experimental periods, the heifers were housed in the
metabolism unit and were exercised in a small paddock for about an hour per day. The
heifers underwent a 19-day adaptation period in which they were accustomed to the diet
and handling. Five days before the infusion period, the front quarters of each of the
heifers was infused with 12 mL physiological saline. The infusions were administered
using a 12 mL syringe and a modified 200-pL pipette tip in which part of the wide end
was cutoff for better attachment to the syringe. Each tip was covered in Surgilube to
allow easier insertion of the tip into the teat. The heifers received the infusions at 0800

hours (hr) and their udders were palpated for mastitis at 1400 hr and 1700 hr. No

55

hardness or soreness of the udder was detected. This project was approved by the
Michigan State University All-University Committee on Animal Use and Care.

Lyophilized recombinant human IGF-1 (GroPep Pty, Adelaide) was reconstituted
to 1 mg/mL in equal amounts of sterile 100 mM hydrochloric acid and sterile 100 mM
sodium hydroxide and stored at -20°C. Recombinant ovine leptin was kindly provided by
Dr. Ari Gertler, was reconstituted to 1 mg/mL in sterile MilliQ water and stored at -20°C.
All infusions were prepared at 0600 and the animals were infused at 0800. The different
doses of leptin (0, 23, 46 pg) were combined in 10 mL sterile physiological saline per
treatment that also contained 1 pg/mL IGF-1 and 1 mg/mL bovine serum albumin. Each
treatment was separated into 12-mL syringes and 200-pL pipette tips that were modified
as described above and covered in Surgilube. The animals were allowed to exercise in a
grassy paddock for one hour before infusions. Upon infusion, the teats were cleaned with
3% iodine and 70% ethanol. The hormones were infused in each animal according to the
design shown in Figure 7.

Figure 7. Infirsion design for the udder quarters of each heifer. Each quarter received 1
pg/mL IGF-1 and 1 mg/mL BSA along with the treatments daily for 7 days.

 

 

 

 

 

 

 

 

 

 

 

 

4160 4162
Front
Left 0 pg leptin 23 pg leptin 0 pg leptin 46 pg leptin Right
0 pg leptin 46 pg leptin 0 pg leptin 23 pg leptin
Rear

Blood samples from either the tail vein or the jugular vein were obtained before the first

infusion. On day 7, the heifers were infused at 0800 hr and again at 2000 hr.

56

On the day of Slaughter, the animals were weighed and two blood samples were
taken from the jugular vein of each heifer. BrdU (reconstituted to 10 mg/mL and pH set
to 7.38, Sigma) was infused at an amount of 5 mg/kg BW via jugular catheter. Heifers
were slaughtered between 2.5 to 2.75 hours later with 85 mg/kg body weight of sodium
pentobarbital. The mammary gland was incised from each heifer and the sebaceous fluid
was collected from the milk cistern of each quarter. 3 to 4 grams of parenchyma from
each quarter was flash-frozen in liquid nitrogen. Parenchyma samples were taken from
three regions in each quarter: the area proximal to the teat, the area opposite of the teat
and closest to the fat pad and the area in between (labeled proximal, distal, and
intermediate, respectively). Each sample was fixed in 10% formalin and shipped to the
Diagnostic Center for Population and Animal Health for embedding. The University
Animal Laboratory Resources disposed of the carcasses.

The immunohistochemistry protocol was adapted from Silva et al. (2000) and
used the Zymed Histostain—SP kit. Briefly, tissue was sectioned into 6-pm Slices and
transferred to a poly-L-lysine-coated slide (Sigrna-Aldritch). The slides were baked for
30 min at 65°C and then either were stained immediately after or transferred to -20°C for
storage. The tissues were subjected to 3 xylene washes at 3 min per wash and then a
series of decreasing ethanol washes (100%, 90%, and 70%) at 3 min each wash. The
tissues were washed in methanol containing 3% hydrogen peroxide for 10 min and then a
series of three phosphate-buffered saline (PBS) washes for 2 min per wash. The tissues
were then immersed in citrate buffer (10 mM; pH 6.0) that was heated to between 90 and
95°C in a vegetable steamer. The tissues were heated for 20 min and then were cooled to

45°C. After another three washes in PBS for 2 min each wash, the tissue sections were

57

immersed in PBS plus 5% nonirnmune goat serum for 30 min in a humid chamber. The
BrdU antibody (Clone 9318, Roche Applied Science) was diluted 1:50 in water
containing 0.1% bovine serum albumin and 150 pL was applied to each section. The
slides were then incubated at 4°C overnight in a humid chamber. The next day, the slides
were washed in PBS (3 x 2 min) and then covered with a biotinylated secondary antibody
for 10 min in a humid chamber. After another series of PBS washes, the tissues were
covered with a streptavidin-peroxidase conjugate and incubated at room temperature in a
humid chamber for 20 min. The tissues were washed again in PBS and were then
exposed to diaminobenzidine in a humid chamber for 5 min at room temperature. The
tissues were then rinsed in deionized water and covered with hematoxylin for 20 seconds.
After being washed in PBS for 30 seconds, the tissues were washed in deionized water
and subjected to an ascending series of ethanol washes (70%, 90%, and 100%) for 3 min
per wash. The slides were washed in xylenes for 5 min and were mounted using
Histomount and a 24 x 60 mm cover glass (Corning). Approximately 8 pictures were
taken from each slide at using a Leica DFC480 camera attached to a Leica DMl L
microscope (set to 40X) and hooked up to a Hewlett-Packard Pavilion a5 1 On computer.
All epithelial cells were counted directly from the microscope until approximately
100 BrdU-positive cells were counted. Only the cells in the distal parenchymal region of
the quarters were counted because there is evidence that more proliferation occurs in this
region. This preliminary data would determine whether this would be a viable project to
pursue. A serious problem that occurred was that much of the tissue failed to attach to
the slide. The result was that, when viewed under a microscope, portions of the tissue

would come out of focus relative to other portions of the tissues. Furthermore, parts of or

58

the entire baked sectioned tissue would fall off the slide during the staining procedure.
Consequently, it was extremely difficult in determining individual epithelial cells for
counting. This was a continual occurrence with these tissues. It is believed that the
paraffin did not fully invade the tissue Since attempts to attach bovine mammary
parenchymal tissue prepared by another scientist and rat mammary carcinoma tissue
prepared by another outside laboratory proved successful. Alterations of batches of wash
reagents used, changing the antigen presentation method (previously, the tissues were
heated for 5 min, cooled for 5 min, heated again for 5 min and cooled to 45°C in 10 mM
citrate buffer in a 600 MW microwave), using a different batch of poly-L-lysine slides,
cooling the tissues before sectioning, and altering the wash times all proved unsuccessful
in improving attachment of the tissues to the slide. Lack of resources and the increasing
scarcity of tissue prevented the reparrafinization of the tissues. Counts were analyzed
using the GLM procedure in SAS (v8.0) using treatment and heifer as classes.

The least squared means and standard deviations are presented in Figure 8. There
was no difference in the number of BrdU-labeled cells between treatments (P = 0.9121)
or heifers (P = 0.2593). It should be noted that only two heifers were used in this study
and that this was preliminary data to determine whether to continue with the study.
Furthermore, the standard deviation for the 46-pg leptin treatment is very high. This
most likely was due to the difficulty in obtaining cell counts from the tissue; however, no
statistical analyses were made to determine this. There were no differences in BrdU

labeling between the three regions that were sampled.

59

Figure 8: Mean index of BrdU-labeled mammary epithelial cells sampled from two
heifers infused once daily with 0, 23, or 46 pg leptin over a seven-day period.

 

 

 

 

 

 

 

 

s “l
a 5 “1
i 4“
.3 3 ~
g Dneleptin
:- 2 —« .
a Ileptm
8 1 —‘
2
0 w

 

60

Appendix B

Upregulated in MAC-T cells after 24 hr of IGF-1 treatment.
Fold- P-

 

 

 

Accession # Gene Name chapgp values Symbol
Carbohydrate Metabolism
CK965677 aldolase A, fructose-bisphosphate 1.29 0.05 ALDOA
CB433477 ATP citrate lyase 1.55 0.01 ACLY
‘ cadherin 1, type 1, E-cadherin
AY508164. 1 (epithelial) 1.21 0.03 CDHl
dihydrolipoarnide S-
acetyltransferase (E2 component of
CF613505 pyruvate dehydrogenase complex) 1.26 0.04 DLAT
CK949721 GDP-mannose pyrophosphorylase B 1.24 0.02 GMPPB
AF043228.1 glucose phosphate isomerase 1.55 0.04 GPI
NM_174319.1 hexokinase 1 1.22 0.03 HKl
CK952050 hexokinase 2 1.92 0.01 HK2
isocitrate dehydrogenase 3 (NAD+)
AF 090321.] beta 1.22 0.01 IDH3B
NM_174099.2 lactate dehydrogenase A 1.39 <0.01 LDHA
NM_174100.1 lactate dehydrogenase B 1.22 0.01 LDHB
lipase A, lysosomal acid, cholesterol
Bauman2461 esterase (Wolman disease) 1.26 0.03 LIPA
malate dehydrogenase 1, NAD
CK948244 (soluble) 1 .22 0.01 MDHl
CK772109 phosphogluconate dehydrogenase 1.33 0.03 PGD
CK849264 phosphoglycerate kinase 1 1.40 <0.01 PGKl
CK769449 phosphomannomutase 2 1.26 0.01 PMM2
CK775519 protease, serine, 16 (thymus) 1.21 0.01 PRSSl6
CB165376 triosephosphate isomerase l 1.40 0.02 TPIl
NM_174211.2 UDP-glucose dehydrogenase 1.22 <0.01 UGDH
Cell signaling
5-hydroxytryptamine (serotonin)
AJ491865.1 receptor 2C 1.44 0.01 HTR2C
AJ277986.1 angiotensin II receptor, type 2 1.22 0.01 AGTR2
NM_174308.1 endothelin receptor type A 1.47 0.01 EDNRA
insulin-like growth factor binding
NM_174556.1 protein 3 1.42 0.01 IGFBP3
NM_174375.2 KIT ligand 1.31 0.01 KITLG
NM_174753.1 parathyroid hormone-like hormone 1.20 0.05 PTHLH
AW465454 stratifin 1.33 0.01 SFN
CK948130 transferrin receptor (p90, CD71) 1.32 0.04 TF RC
Cell Cycle and Death '

61

NM_174003.2 calpastatin 1.31 <0.01 CAST
CDC6 cell division cycle 6 homolog

 

CK960396 (S. cerevisiae) 1.58 0.02 CDC6
chromatin assembly factor 1, subunit

CK846204 A (p150) 1.22 0.01 CHAF 1A

BM287438 cyclin E1 1.31 0.04 CCNEl
tumor necrosis factor (ligand)

CB43 8089 superfamily, member 10 1.33 0.01 TNFSFI 0

Cell Structure and Extracellular Matrix

CK769227 abl interactor 2 1.23 0.02 ABI2

CB222344 formin-like 2 1.24 <0.01 F MNL2

NM_173934.1 lumican 1.31 0.01 LUM

neural precursor cell expressed,
NM_174764.2 developmentally down-regulated 8 1.23 <0.01 NEDD8

NM_174718.1 pinin, desmosome associated protein 1.61 0.01 PNN

 

Coagulation Factors
coagulation factor IX (plasma
thromboplastic component,

J00007.l Christmas disease, hemophilia B) 1.25 0.04 F9
coagulation factor XIII, A1
CK77783 8 polypeptide 1.27 0.01 F13A1

serine (or cysteine) proteinase
inhibitor, cladc E (nexin,
plasminogen activator inhibitor type
NM_174137.2 1), member 1 1.53 0.02 SERPINE]
tissue factor pathway inhibitor
(lipoprotein-associated coagulation
CK776402 inhibitor) 1.25 <0.01 TFPI

 

DNA Replication and Repair
polymerase (DNA directed), epsilon

 

CK846301 2 (p59 subunit) 1.31 0.01 POLE2
RAN, member RAS oncogene

CB426829 family 1 .35 0.03 RAN

Energy Metabolism
aldehyde dehydrogenase 18 family,

CK941391 member A1 1.25 0.04 ALDH18A1

BF775817 coproporphyrinogen oxidase 1.41 <0.01 CPOX
cytochrome P450, family 26,

CK964867 subfamily B, polypeptide l 1.35 0.05 CYP26B1

glutarnic-oxaloacetic transaminase
1, soluble (aspartate

NM_177502.2 aminotransferase 1) 1.32 0.02 GOTl
CB453756 heme oxygenase (decycling) 2 1.45 <0.01 HMOX2
CK974609 holocytochrome c synthase 1.21 0.03 HCCS

62

(cytochrome c heme-lyase)

 

 

 

CK772343 hydroxymethylbilane synthase 1.29 0.04 HMBS
NADH dehydrogenase (ubiquinone)
NM_175820.2 1 alpha subcomplex, 4, 9kDa 1.27 0.02 NDUFA4
NADH dehydrogenase (ubiquinone)
NM_175791.2 1 alpha subcomplex, 6, l4kDa 1.21 0.01 NDUFA6
NADH dehydrogenase (ubiquinone)
CB468421 1, alpha/beta subcomplex, 1, 8kDa 1.22 0.02 NDUFABl
NADH dehydrogenase (ubiquinone)
NM_174564.2 l, subcomplex unknown, 1, 6kDa 1.27 0.02 NDUFCI
NADH dehydrogenase (ubiquinone)
Fe-S protein 4, 18kDa (NADH-
NM__175800.2 coenzyme Q reductase) 1.23 <0.01 NDUFS4
NM_173968.2 thioredoxin 1.31 0.02 TXN
Folate Metabolism
garnma-glutamyl hydrolase
(conjugase,
BP100358 folylpolygarnmaglutamyl hydrolase) 1.22 0.01 GGH
methylenetetrahydrofolate
dehydrogenase (N ADP+ dependent),
methenyltetrahydrofolate
CK960935 cyclohydrolase 1.41 0.01 MTI-IF D1
sepiapterin reductase (7,8-
dihydrobiopterinzNADP+
CK775 888 oxidoreductase) 1.38 0.02 SPR
Glycan Metabolism
asparagine-linked glycosylation 5
homolog (yeast, dolichyl-phosphate
CK972901 beta-glucosyltransferase) 1.21 0.04 ALGS
asparagine-linked glycosylation 6
homolog (yeast, alpha-1,3-
AW425955 glucosyltransferase) 1.21 <0.01 ALG6
core 1 UDP-galaetose:N-
acetylgalactosarnine-alpha-R beta
CK845990 1,3-galactosyltransferase 1.23 0.02 C 1 GALTl
famesyltransferase, CAAX box,
NM_177498.2 alpha 1.30 0.02 FNTA
fucosyltransferase 8 (alpha (1,6)
NM_177501.1 fucosyltransferase) 1.24 0.03 FUT8
CK769632 glucosidase I 1.24 0.01 GCSl
Rab geranylgeranyltransferase, beta
AW426143 subunit 1.27 0.02 RABGGTB
Lipid Metabolism

63

acetyl-Coenzyme A acyltransferase
2 (mitochondrial 3-oxoacyl-

 

 

CK973155 Coenzyme A thiolase) 1.22 0.01 ACAA2
acyl-CoA synthetase long-chain
CK846911 family member 6 1.22 <0.01 ACSL6
dual-specificity tyrosine-(Y)-
CK778309 phosphorylation regulated kinase 3 1.28 0.04 DYRK3
elongation of very long chain fatty
acids (F ENl/EloZ, SUR4/Elo3,
CB444358 yeast)-like 4 1.30 0.02 ELOVL4
famcsyl diphosphate synthase
(farnesyl pyrophosphate synthetase,
dirnethylallyltranstransferasc,
NM_177497.2 geranyltranstransferase) 1 .26 <0.01 FDPS
CK944276 glycerol kinase 1.28 0.01 GK
CK771258 lysophospholipase I 1.28 0.05 LYPLAl
CK776702 phosphatidylinositol glycan, class B 1.27 0.03 PIGB
Bauman781 phytoceramidase, alkaline 1.34 0.02 PHCA
CK832399 putative acyl-CoA dehydrogenase 1.24 0.05 FL] 12592
Protein Synthesis and Metabolism
CK848612 aminolevulinate, delta-, synthase 1 1.22 <0.01 ALAS]
eukaryotic translation initiation
NM_175813.1 factor 2, subunit 1 alpha, 35kDa 1.40 0.01 EIFZSl
eukaryotic translation initiation
CB534551 factor 4A, isoform 1 1.31 0.02 EIF4A1
glutamate-cysteine ligase, catalytic
CB444175 subunit 1 .64 0.02 GCLC
glutathione peroxidase 2
CK948205 (gastrointestinal) 1 .26 0.01 GPX2
CBI71170 glutathione S-transferase omega 1 1.24 0.02 GSTOl
CK968451 glycyl-tRNA synthetase 1.23 0.01 GARS
AV602991 lysyl-tRNA synthetase 1.20 0.04 KARS
CB451602 mitochondrial ribosomal protein 814 1.22 0.01 MRPS 14
NM_174130.2 omithine decarboxylase l 1.47 0.01 ODCl
phenylalanine-tRNA synthetase-like,
CK953114 beta subunit 1.25 0.05 FARSLB
serine hydroxymethyltransferase 2
CB45 8343 (mitochondrial) 1 .30 <0.01 SHMT2
NM_174175.2 seryl-tRNA synthetase 1.29 <0.01 SARS
BI898927 spermidine synthase 1.36 0.02 SRM
Purine and Pyrimidine Metabolism
CK983189 adenosine kinase 1.22 0.01 ADK
NM_173889.1 adenylate kinase 2 1.42 0.01 AK2
CK772896 dihydroorotate dehydrogenase 1.53 <0.01 DHODH

64

IMP (inosine monophosphate)

 

CB 1 72231 dehydrogenase 2 1.33 0.05 IMPDH2

CK849436 nucleoside phosphorylase 1.34 <0.01 NP
phosphoribosyl pyrophosphate

CK777978 synthetase 1 1.23 0.01 PRPSl
phosphoribosylglycinamide
fonnyltransferase,
phosphoribosylglycinamide
synthetase,
phosphoribosylaminoirnidazole

CK958159 synthetase 1.36 0.01 GART
ribonucleotide reductase M2

CK979761 polypeptide 1 .71 0.01 RRM2

NM_174625.2 thioredoxin reductase 1 1.47 <0.01 TXNRDI

CK970228 UMP-CMP kinase 1.24 0.01 UMP-CMPK
uridine monophosphate synthetase
(orotate phosphoribosyl transferase

NM_177508.1 and orotidine-5'-decarboxylase) 1.39 0.02 UMPS

Signal Transduction

CL513Contig1 chemokine (C-C motif) ligand 4-like 1.25 0.01 CCL4L

CK846020 FOS-like antigen 1 1.41 0.01 FOSLl
heat shock 70kDa protein 5

BF606842 (glucose-regulated protein, 78kDa) 1.33 <0.01 HSPAS

AB072368.1 heat shock 90kDa protein 1, alpha 1.36 0.03 HSPCA

B153 8908 mitogen-activated protein kinase 6 1.29 0.02 MAPK6
mitogen-activated protein kinase

CK838207 kinase kinase 7 interacting protein 2 1.22 0.05 MAP3K7IP2
mitogen-activated protein kinase-

CK77 7104 activated protein kinase 3 1.25 0.03 MAPKAPK3
nuclear factor of activated T-cells,
cytoplasmic, calcineurin-dependent

BP106653 3 1.29 0.04 NFATC3
pleiotropic regulator 1

BM431413 (PRLlhomolog, Arabidopsis) 1. .33 <0.01 PLRGl
protein kinase, AMP-activated,

NM_174586.1 gamma 1 non-catalytic subunit 1.23 0.04 PRKAGl
protein phosphatase 2 (formerly

NM_181031.2 2A), catalytic subunit, alpha isoform 1.21 0.04 PPP2CA

CK770419 RuvB-like 1 (E. coli) 1.32 0.01 RUVBLI

BM364201 suppressor of cytokine signaling 1 1.26 0.05 SOCSl
TGFB inducible early growth

BM435193 response 1.39 0.01 TIEG
tyrosine 3-

CK953368 monooxygenase/tryptophan 5- 1.28 0.01 YWHAQ

 

65

monooxygenase activation protein,
theta polypeptide

 

 

 

Transcription
activated RNA polymerase II
CK728106 transcription cofactor 4 1.30 0.04 PC4
CL3817Contig1 ets variant gene 1 1.39 0.02 ETVl
CK971624 general transcription factor IIB 1.23 <0.01 GTF2B
methyl-CpG binding domain protein
CB461430 2 1.30 0.02 MBD2
CK951297 pleiomorphic adenoma gene-like 2 1.21 0.01 PLAGL2
polymerase (RNA) I polypeptide B,
AF 461 104.1 128kDa 1.26 0.04 POLRl B
CK838008 suppressor of S. cerevisiae gcr2 1.30 0.04 HSGTl
CK955167 T-box 3 (ulnar mammary syndrome) 1.47 0.03 TBX3
transcription elongation factor B
(SIII), polypeptide 3 (110kDa,
CK774454 elongin A) 1.21 0.02 TCEB3
CK769868 transcription factor-like 4 1.29 0.03 TCFL4
BF605641 zinc ribbon domain containing} 1 1.40 0.01 ZNRDl
Transport
ATP-binding cassette, sub-family A
CB462017 (ABCl), member 1 1.24 0.01 ABCAl
ATP-binding cassette, sub-family B
AB006985.1 (MDR/TAP), member 1 1.37 <0.01 ABCBl
fatty acid binding protein 3, muscle
and heart (mammary-derived growth
NM_1743 1 3 .2 inhibitor) 1 .95 <0.01 FABP3
solute carrier family 2 (facilitated
NM_174602.2 glucose transporter), member 1 1.21 <0.01 SLC2A1
solute carrier family 25
(mitochondrial carrier; adenine
NM_174658.1 nucleotide translocator), member 4 1.29 0.01 SLC25A4
solute carrier family 3 (activators of
dibasic and neutral amino acid
CBl65860 transport), member 2 1.33 0.02 SLC3A2
Unknown
factor for adipocyte differentiation
CB533649 158 1.22 0.01 FAD158
CK848911 FK506 binding protein 1A, 12kDa 1.22 0.03 FKBPlA
CB430950 hypothetical protein MGC2744 1.21 <0.01 MGC2744
RAB27A, member RAS oncogene
CK946480 family 1.22 0.04 RAB27A
SH3-domain kinase binding protein
CBS35077 l 1.30 0.02 SH3KBP1

66

AF198054.1 1.70 <0.01
NM_181810.1 1.28 <0.01
NM 1746622 1.32 0.02

67

Appendix C

Downregulated genes in MAC-T cells after 24 hr of IGF-1 treatment.
Fold

 

 

 

 

 

 

 

Accession # Gene Name change P-value Symbol
Carbohydrate Metabolism
CB454232 mam°51°°8°’ alpha: “ass 2A: 0.67 <0.01 MAN2A1
member 1
CK770297 UDP‘Gal‘°°‘°G'°NA° °°t° 1’4? 0.78 0.04 B4GALT5
galactosyltransferase, polypeptrde 5
NM 1742242 23? 1'C°°”zym° A °°I°°xylas° 0.79 0.02 ACACA
_ P a
Cell Cycle and Death
cyclin-dependent kinase inhibitor 2B
CK944043 (p15, inhibits CDK4) 0.77 0.01 CDKN2B
CL8903Contig1 A kinase (PRKA) anchor protein 8 0.78 0.02 AKAP8
Cell signaling
epidermal growth factor receptor
AY486452.1 (erythroblastic leukemia viral (v-erb- 0.75 <0.01 EGF R
b) oncogene homolog, avian)
NM_181010.2 endothelin 1 0.66 <0.01 EDNl
NM_194266.1 adrenergic, beta-1-, receptor 0.78 0.01 ADRBl
NM 17 4 5 5 5 1 insulin-like growth factor binding 0 78 0 02 IGFBP2
- ' protein 2, 36kDa ' '
Cell Structure and Extracellular Matrix
CK941880 lamin B2 0.78 0.02 LMNB2
CK944548 nuclear mitotic apparatus protein 1 0.78 0.02 NUMAl
CB468342 dedicator of cytokinesis 1 0.68 0.01 DOCK]
AB055312.1 °°th°P°m D ay°°°°m°1 ”WWI 0.76 0 01 CTSD
protease) '
CK776003 collagen, type IV, alpha 6 0.74 0.04 COL4A6
BE752701 agrin 0.79 0.03 AGRN
CK975649 gelsolin (amyloidosis, Finnish type) 0.70 0.02 GSN
DNA Replication and Repair
CB420483 MAX interactor 1 0.67 0.01 MXll
Energy Metabolism
NM 1743041 °yt°°hr°m° ”50’ family 17’ 0.80 0.05 CYP17A1
- subfamrly A, polypeptrde l
Lipid Metabolism
CK948274 °l°°hy°° °°hydr°g°nas° 3 family’ 0.78 0.03 ALDH3A2
member A2
sialyltransferase 7 ((alpha-N-
CK971583 °°°tyln°mm‘“yl'2’3'°°t°' 0.75 0.01 SIAT7B

galactosyl- 1 ,3)-N-acetyl
galactosaminide alpha-2,6-

68

sialyltransferase) B
sulfotransferase family, cytosolic,

 

 

 

 

 

NM_177521.2 1 A, phenol-preferring, member 1 0.67 0.04 SULTlAl
CK774100 P°.’°’°S°m°l 1°“g’°h°in °°y1'°°A 0.74 0.04 ZAP128
thIoesterase
Protein Synthesis and Metabolism
CK849902 g1ethionine adenosyltransferase II, 0.79 0.01 M AT2B
eta
alanyl (membrane) aminopeptidase
CK833665 (”Emptidm N: m°f°°°°°°° 0.79 0.02 ANPEP
M, nncrosomal armnopeptIdase,
CD13, p150)
AW65 8968 histidine decarboxylase 0.55 <0.01 HDC
Purine and Pyrimidine Metabolism
CK769403 adenylate cyclase 8(11rain) 0.72 0.01 ADCY8
Signal Transduction
CK945745 mitogen-activated protein kinase 4 0.74 <0.01 MAPK4
inhibitor of DNA binding 1,
CK950713 dominant negative helix-loop-helix 0.67 0.03 IDl
protein
inhibitor of DNA binding 3,
CK770014 dominant negative helix-loop-helix 0.73 0.01 ID3
protein
CK943734 frizzled homolog 4 (Drosophila) 0.72 0.02 FZD4
CB467921 beta-transducin repeat containing 0.76 0.04 BTRC
CB 422127 nuclear factor of activated T-cells 5, 0.77 0 03 NF AT5
tomcrty-responsrve '
CK777672 retinoic acid receptor, beta 0.63 0.02 RARB
AV610239 insulin receptor substrate 1 0.76 0.01 IRSl
CK778883 ral guanine nucleotide dissociation 0.74 <0.01 RALGDS
stimulator
Transcription
CK974450 basic transcription factor 3 0.77 0.03 BTF3
AV591750 zinc finger protein 192 0.77 <0.01 ZNF192
CB531176 general transcription factor 11, i 0.80 <0.01 GTF2I
BM481287 homeo box D10 0.43 0.03 HOXDIO
CL1270Contig1 2:22:31: r°pr°ss°r °fE1A'S°m“l°t°° 0.65 0.02 CREGI
AY398 689 1 microphthalmia-associated 0 71 0 03 MITF
' transcription factor ' '
CK972308 UBX domain containing 2 0.78 <0.01 UBXD2
CK837990 ets variant gene 1 0.76 0.01 ETVl
Transport
BE217451 ”lute “mi“ funny 39 (mm 0.74 <0.01 SLC39A6

transporter), member 6

69

solute carrier family 1 (glial high
CB442833 affinity glutamate transporter), 0.78 <0.01 SLC1A2
member 2

 

70

Appendix D

Upregulated genes in MAC-T cells after 8 hr of IGF-1 treatment.

 

 

Fold-
Accession # Gene Name change P-value Symbol
Carbohydrate Metabolism
AF 054834.1 amylase, alpha 2B; pancreatic 1.25 0.03 AMY2B
CB433477 ATP citrate lyase 1.33 0.03 ACLY
AF461103.1 citrate synthase 1.28 <0.01 CS
NM_174319.1 hexokinase 1 1.26 0.02 HK1
NM_174100.1 lactate dehydrogenase B 1.24 <0.01 LDHB
CK770445 m°1i° °sz‘“° 1’ NADP(+)'°°P°°°°“" 1.44 0.01 MEI
cytosolrc
CK849264 phosphoglycerate kinase 1 1.22 <0.01 PGKl
CK769449 phosphomannomutase 2 1.44 0.05 PMM2

succinate dehydrogenase complex,
NM_175814.2 subunit C, integral membrane protein, 1.23 0.03 SDHC
15kDa

 

DNA Replication and Repair
chromatin assembly factor 1, subunit

 

 

CK846204 A (p150) 1.28 0.05 CHAF 1A
NM_182651.1 IENA (°y‘°°‘°°'5')'m°°’yl°°°°f°’°°° 1.30 0.01 DNMTI
methylenetetrahydrofolate
dehydrogenase (N ADP-l- dependent),
CK960935 methenyltetrahydrofolate 1 .52 0.01 MTHFDI
cyclohydrolase,
forrnyltetrahydrofolate synthetase
CK940683 replication protein A1, 70kDa 1.22 0.02 RPAl
Cell signaling
AJ491865.1 fégggfixzygypmm (S°’°‘°°‘”) 1.28 0.01 HTR2C
AY191360.2 fiéiggigwm f°°t°r (baa' 1.44 <0.01 EGF
NM_1 745 56.1 31:33?“ ngIh f°°t°r “mung 1.40 0.03 IGFBP3
AW465454 stratifin 1.35 0.03 SFN
Signal Transduction
CK846020 FOS-like antigen 1 1.38 0.04 FOSLl
protein kinase, AMP-activated,
NM_174586.1 g a 1 non-catalytic subunit 1.34 0.02 PRKAGl
Baumanl 784 firotern kInase, cGMP-dependent, type 13 0 0.03 PRKG2
CK773728 protein tyrosrne phosphatase, receptor 1. 38 0.01 PTPRR

type, R

71

Energy Metabolism
cytochrome P450, family 26,

 

 

 

 

 

CK964867 subfamily B, polypeptide 1 1.70 0.02 CYP26B1
CK974609 h°l°°yt°°hI°m° ° 8mm“ 1.29 0.02 HCCS
(cytochrome c heme-lyase)
CK772343 hydroxymethylbilane synthase 1.35 0.04 HMBS
NADH dehydrogenase (ubiquinone) 1
NM_175809.1 beta subcomplex, l, 7kDa 1.22 0.05 NDUFBl
NADH dehydrogenase (ubiquinone) 1,
NM_174564.2 subcomplex unknown, 1, 6kDa 1.23 0.03 NDUFCl
Cell Cycle and Death
CK9 603 96 CDC6 celldIVISIon cycle 6 homolog 1.8 5 0.03 CD C 6
(S. cerev1srae)
BM287438 cyclin E1 1.37 0.03 CCNEI
Lipid Metabolism
farnesyl diphosphate synthase
(farnesyl pyrophosphate synthetase,
NM_177497.2 dime thylallyl trans trans ferase, 1 .42 <0.01 FDPS
geranyltranstransferase)
BM251520 insulin induced gene 1 1.21 0.02 INSIGl
Protein Synthesis and Metabolism
acetyl-Coenzyme A acetyltransferase
CK77353° 2 (acetoacetyl Coenzyme A thiolase) 1'32 0'03 ACAD
CB534551 :fai‘s’ycggfinslatm "mm” f°°t°r 1.54 0.02 EIF4A1
NM_177515.2 glutathione S-transferase A1 1.25 0.04 GSTAl
CK848917 histidyl-tRNA synthetase 1.23 0.02 HARS
hypothetical protein F LJ 22649 similar
CK976501 to signal peptidase SPC22/23 1.23 0.01 FLJ22649
NM_174130.2 ornithine decarboxylase l 1.60 <0.01 ODCl
CK771294 phosphoglycerate dehydrogenase 1.37 0.03 PHGDH
CBl68605 ribophorin I 1.24 0.01 RPNI
BM258870 serine dehydratase 1.35 0.05 SDS
CK951402 vanin 1 1.29 <0.01 VNNl
Purine and Pyrimidine Metabolism
CK769403 adenylate cyclase 8 (brain) 1.24 0.01 ADCY8
NM_173 889.1 adenylate kinase 2 1.51 0.02 AK2
CK849570 adenylosuccinate lyase 1.30 0.04 ADSL
CK772896 dihydroorotate dehydrogenase 1.52 0.05 DHODH
CB 1 72231 MP ("”51“ m°°°°h°5°h°I°I 1.70 0.02 IMPDH2
dehydrogenase 2
CK777978 gxmgbfsyl pyr°Ph°S°h°t° 1.38 0.02 PRPSI
CK9797 61 ribonucleotide reductase M2 1 62 0 02 RRM2
polypeptide ' '

72

uridine monophosphate synthetase

 

 

 

NM_177508.1 (orotate phosphoribosyl transferase 1.69 0.01 UMPS
and orotidine-5'-decarboxylase)

Cell Structure and Extracellular Matrix

NM_174307.2 dennatan sulfate proteoglycan 3 1.29 0.03 DSPG3
transient receptor potential cation

AW356495 channel, subfamily C, member 6 1.32 0.01 TRPC6

Transcription
basic helix-loop-helix domain

CB420822 containing, class B, 2 1.22 0.03 BHLHB2

NM_174000.2 calreticulin 1.26 0.01 CALR

c13531724 LAGI l°f$°Vity ”swan“ h°m°l°g 2 1.22 <0.01 LASSZ
(S. cereVISrae)

CK778210 nuclear receptor subfamily 4, group A, 1 47 0 01 NR4 A3
member 3 ' '

AF461104.1 Il’ggy’kgflm (RNA) I p°lyp°p°°° B: 1.33 0.05 POLRIB

a

CK770419 RuvB-like 1 (E. coli) 1.48 0.04 RUVBLl
SWI/SNF related, matrix associated,

CK847247 actin dependent regulator of 1.20 0.02 SMARCA4
chromatin, subfamily a, member 4
TAP 12 RNA polymerase II, TATA

CK976188 box binding protein (TBP)-associated 1.31 0.02 TAF 12
factor, 20kDa
TAF2 RNA polymerase II, TATA box

CK9803 88 binding protein (TBP)-associated 1.24 0.01 TAF2
factor, 150kDa

CK955167 T-box 3 (ulnar mammary syndrome) 1.28 0.03 TBX3
transcription elongation factor B

CK774454 (SIII), polypeptide 3 (l lOkDa, elongin 1.26 0.04 TCEB3
A)

Transport
ATP-binding cassette, sub-family A

CB462017 (ABCl), member 1 1.28 0.04 ABCAl
fatty acid binding protein 3, muscle

NM_174313.2 and heart (mammary-derived growth 2.13 <0.01 FABP3
inhibitor)

CK848911 FK506 binding protein 1A, 12kDa 1.22 0.04 FKBPlA
solute carrier family 25 (mitochondrial

BM030917 carrier; omithine transporter) member 1.20 <0.01 SLC25A15
15

NM_174657.2 S°'“.t° °°°i°r fan‘i'y 25? (mit°°h°“°’i°1 1.21 0.04 SLC25A3
earner; phosphate earner), member 3

- solute carrier family 3 (activators of
CBl65860 dibasic and neutral amino acid 1.32 0.04 SLC3A2

transport), member 2

 

73

Unknown

CK943898 polyamine-modulated factor] 1.54 0.02 PMF 1
AF 1 98054.1 1.28 0.0]
NM 174086.] 1.53 <0.01

 

74

Appendix E

Dowmegulated genes after 8 hr of IGF-1 treatment.

 

Fold-
Accession # Gene Name change P-value Symbol

Cell Structure and Extracellular Matrix
v-erb-b2 erythroblastic leukemia viral
BM258099 oncogene homolog 2, neuro/glioblastoma 0.69 0.01 ERBB2

derived oncogene homolog (avian)

 

 

 

Energy Metabolism
AY265991.1 $70233: 5450: “my 1’ subfmly A’ 0.79 0.01 CYP1A2
Lipid Metabolism
sialyltransferase 7 ((alpha-N-
acetylneurarninyl-2,3-beta-galactosyl-
CK9715°3 1,3)-N-acetyl galactosaminide alpha-2,6- 0'79 0'04 SIAT7B
sialyltransferase) B
AW653 508 diac l l cerol kinase, alpha 80kDa 0.78 0.02 DGKA
I g Y
Protein Synthesis and Metabolism
NM_173939.1 methylmalonyl Coenzyme A mutase 0.77 0.02 MUT
CK959627 pr°pl°°yl C°°‘.‘zym° A °°I°°xylas° 0.79 0.03 PCCA
alpha polypeptrde
CB4 53808 glptlizssphloadenosme 5 -phosphosulfate 0.73 0.01 P APS Sl

CK772118 cysteine sulfinic acid decarboxylase 0.76 0.04 CSAD
dopamine beta-hydroxylase (dopamine

 

NM 1809952 0.77 0.03 DBH
- beta-monooxygenase)

Transcription

CK837990 ets variant gene 1 0.73 0.01 ETV]

CK957547 GRB2-associated binding protein 1 0.72 0.05 GAB]

transcription factor 12 (HTF4, helix-

loop-helix transcription factors 4) 0'71 0'05 TCFIZ

AV 6 64749 2111c finger. proteIn 42 (myeIOId-Specrfic 0. 68 0.01 ZNF42
retrnOIc acId-responsrve)

CK778931 piggyBac transposable element derived 1 0.78 <0.01 PGBDl

CB468243

 

Signal Transduction

 

CK953091 protein inhibitor of activated STAT, 1 0.73 0.04 PIAS]
CB439479 protein kinase C, eta 0.78 0.01 PRKCH
CK943734 frizzled homolog 4 (Drosophila) 0.74 0.05 FZD4
CK847494 lymphoid enhancer-binding factor 1 0.77 0.03 LEFl
Unknown

AB019395.1 0.80 0.04

 

75

Bibliography

Adesanya OO, Zhou J, Sarnathanam C, Powell-Braxton L, Bondy CA. Insulin-like
growth factor ] is required for G2 progression in the estradiol-induced mitotic cycle.
Proc Natl Acad Sci USA 96 (1999): 3287-3291.

Ahima RS, Flier J S. Leptin. Annu Rev Physiol 62 (2000): 413-437.

Akers, R. Michael. L_actation and the mammgy gland. Ames, Iowa : Iowa State
University Press, 2002.

Akers, RM. Major advances in hormone and growth factor regulation of mammary
growth and lactation in dairy cows. J Dg'g Sci 89 (2006): 1222-1234.

Baker J, Liu JP, Robertson EJ, Efstratiadis A. Role of insulin-like growth factors in
embryonic and postnatal growth. C_e_ll 75 (1993): 73-82.

Battistini B, Chailler P, D'Orleans-Juste P, Briere N, Sirois P. Growth regulatory
properties of endothelins. Peptides 14 (1993): 385-399.

Baumrucker CR, Stemberger BH. Insulin and insulin-like growth factor-I stimulate DNA
synthesis in bovine mammary tissue in vitro. J Anim Sci 67 (1989): 3503-3514.

Berry SD, McFadden TB, Pearson RE, Akers RM. A local increase in the mammary
IGF-1: IGFBP-3 ratio mediates the mammogenic effects of estrogen and growth
hormone. Domest Anim Endocrinol 21 (2001): 39-53.

Berry SD, Howard RD, Jobst PM, J iang H, Akers RM. Interactions between the ovary
and the local IGF-I axis modulate mammary deveIOpment in prepubertal heifers. J
Endocrinol 177 (2003a): 295-304.

Berry SD, Weber Nielsen MS, Sejrsen K, Pearson RE, Boyle PL, Akers RM. Use of an
immortalized bovine mammary epithelial cell line (MAC-T) to measure the mitogenic
activity of extracts from heifer mammary tissue: effects of nutrition and ovariectomy.
Domest Anim Endocrinol 25 (2003b): 245-253.

Bionaz M, Loor JJ. Identification of reference genes for quantitative real-time PCR in
the bovine mammary gland during the lactation cycle. Physiol Genomics 29 (2007): 312-
319.

Blat C, Delbe J, Villaudy J, Chatelain G, Golde A, Harel L. Inhibitory diffusible factor

45 bifunctional activity. As a cell growth inhibitor and as an insulin-like growth factor 1-
binding protein. J Biol Chem 264 ( 1989): 12449-12454 .

76

Block SS, Smith JM, Ehrhardt RA, Diac MC, Rhoads RP, Van Amburgh ME, Boisclair
YR. Nutritional and developmenta regulation of plasma leptin in dairy cattle. J Dairy
SQ 86 (2003): 3206-3214.

Butler AA, Shoshana Y, Gewolb 1H, Karas M, Okubo Y, LeRoith D. Insulin-like grth
factor-1 receptor signal transduction: at the interface between physiology and cell
biology. Comp Biochem thsiol B 121 (1998): 19-26.

Campbell PG, Skaar TC, Vega JR, Baumrucker CR. Secretion of insulin-like growth
factor-I (IGF-I) and IGF-binding proteins from bovine mammary tissue in vitro.
J Endocrinol 128 (1991): 219-228.

Caruso A, Pellati A, Bosi P, Arena N, Stabellini G. Effects of spermidine synthase
inhibition in cultured chick embryo fibroblasts. Cell Biol Int Rep 16( 1992): 349-358.

Cohick WS, Turner JD. Regulation of IGF binding protein synthesis by a bovine
mammary epithelial cell line. J Endocrionl 157 (1998): 327-336.

Collier RJ, McGrath MF, Byatt J C & Zurfluh LL. Regulation of bovine mammary
growth by peptide hormones: involvement of receptors, growth factors and binding
proteins. Livestockjrod Sci 35 (1993): 21-33.

Conover CA, Kiefer MC. Regulation and biological effect of endogenous insulin-like
growth factor binding protein-5 in human osteoblastic cells. J Clin Endocrinol Metab 76
(1993):]153-1159.

Daughaday WH, Reeder C. Synchronous activation of DNA synthesis in
hypophysectomized rat cartilage by growth hormone. J Lab Clin Med 68 (1966): 357-68.

Daughaday WH, Hall K, Raben MS, Salmon WD Jr, van den Brande J L, van Wyk JJ.
Somatomedin: proposed designation for sulphation factor. Nature 235 (1972): 107.

 

Davis-Rinker LED. 2005 Nutritional influences on body and mammary growth and
effect of leptin and IGF-1 in prepubertal dairy heifers. Ph.D. Thesis. Michigan State
University, East Lansing.

De Meyts P, Palsgaard J, Sajid W, Theede AM, Aladdin H. Structural biology of insulin
and IGF-1 receptors. Novartis Found Symp 262 (2004): 160-71.

D'Ercole AJ, Applewhite GT, Underwood LE. Evidence that somatomedin is
synthesized by multiple tissues in the fetus. Dev Biol 75 (1980): 315-28.

Denko CW, Bergenstal DM. The effect of hypophysectomy and growth hormone on S35
fixation in cartilage. Endocrinology 57 (1955): 76-86.

77

Dufourny B, Alblas J, van Teeffelen HA, van Schaik FM, van der Burg B, Steenbergh
PH, Sussenbach J S. Mitogenic signaling of insulin-like growth factor I in MCF-7 human
breast cancer cells requires phosphatidylinositol 3-kinase and is independent of mitogen-
activated protein kinase. J Biol Chem 272 (1997): 31163-31171.

Ellis S, Capuco AV. Cell proliferation in bovine mammary epithelium: identification of
the primary proliferative cell population. Tissue Cell 34 (2002): 155-163.

Etchebarne, BE. 2005 IGF-1 and leptin hormonal infusions alter the cell proliferation
and transcriptional profile of the prepubertal bovine mammary gland. Ph.D Thesis.
Michigan State University, East Lansing.

Fredlund, J ., Johansson, M.C., Dahlberg, E. and Oredsson, S.M. Omithine decarboxylase
and S-adenosyhnethionine decarboxylase expression during the cell cycle of Chinese
hamster ovary cells. Exp Cell Res 216 (1995): 86—92.

Furlanetto RW. Receptor-mediated endocytosis and lysosomal processing of insulin-like
growth factor I by mitogenically responsive cells. Endocrinology 122 (1988): 2044-
2053.

Glimm DR, Baracos VE, Kennelly JJ. Northern and in situ hybridization analyses of the
effects of somatotropin on bovine mammary gene expression. J Daig Sci 75
(1992):2687-2705.

Grandis JR, Drenning SD, Chakraborty A, Zhou MY, Zeng Q, Pitt AS, Tweardy DJ.
Requirement of Stat3 but not Stat] activation for epidermal growth factor receptor-
mediated cell growth in vitro. J Clin Invest 102 (1998): 1385-1392.

Grill CJ, Cohick WS. Insulin-like growth factor binding protein-3 mediates IGF-I action
in a bovine mammary epithelial cell line independent of an IGF interaction. J Cell
Physiol 183 (2000): 273-283.

Guler HP, Zapf J, Schmid C, F roesch ER. Insulin-like growth factors I and II in healthy
man. Estimations of half-lives and production rates. Acta Endocrinol 121 (1989): 753-
758.

Harper JV and Brooks G. The mammalian cell cycle: an overview. Methods Mol Biol
296 (2005): 113-153.

Harrison RD, Reynolds IP, Little W. A quantitative analysis of mammary glands of dairy
heifers reared at different rates of live weight gain. J Dairy Res 50 (1983): 405-412.

Hauser SD, McGrath MF, Collier RJ, Krivi GG. Cloning and in vivo expression of
bovine growth hormone receptor mRNA. Mol Cell Endocrinol 72 (1990): 187-200.

78

Hovey RC, Davey HW, Mackenzie DD, McFadden TB. Ontogeny and epithelial-stromal
interactions regulate IGF expression in the ovine mammary gland. Mol Cell Endocrinol
136 (1998a): 139-144.

Hovey RC, DDS Mackenzie and TB McFadden. The proliferation of maouse mammary
epithelial cells in response to specific mitogens is modulated by the mammary fat pad. 111
Vitro Cell Dev Biol Anim 34 (1998b): 385-392

Huber, M and Poulin R. Post-translational cooperativity of ornithine decarboxylase
induction by estrogens and peptide growth factors in human breast cancer cells. Mol Cell
Endocrinol 117 (1996): 211-218.

Huynh HT, Robitaille G, Turner JD. Establishment of bovine mammary epithelial cells
(MAC-T): an in vitro model for bovine lactation. Exp Cell Res 197 (1991): 191-199.

Irnagawa W, Spencer EM, Larson L, Nandi S. Somatomedin-C substitutes for insulin for
the growth of mammary epithelial cells from normal virgin mice in serum-free collagen
gel cell culture. Endocrinology 199 (1986): 2695-2699.

Jerome V, Leger J, Devin J, Baulieu EE, Catelli MG. Growth factors acting via tyrosine
kinase receptors induce HSP90 alpha gene expression. Growth F agors 4 (1991): 317-
327.

Kiess W, Haskell JF, Lee L, Greenstein LA, Miller BE, Aarons AL, Rechler MM,
Nissley SP. An antibody that blocks insulin-like growth factor (IGF) binding to the type
II IGF receptor is neither an agonist nor an inhibitor of IGF-stimulated biologic responses
in L6 myoblasts. J Biol Chem 262 (1987): 12745-12751.

Klapper DG, Svoboda ME, Van Wyk JJ. Sequence analysis of somatomedin-C:
confirmation of identity with insulin-like growth factor 1. Endocrinology 112 (1983):
2215-2217.

Kurtz A, Vogel F, Funa K, Heldin CH, Grosse R. Developmental regulation of
mammary-derived growth inhibitor expression in bovine mammary tissue. J Cell Biol
110 (1990): 1779-1789.

Kurtz A, Spitzer E, Zschiesche W, Wellstein A, Grosse R. Local control of mammary
gland differentiation: mammary-derived growth inhibitor and pleiotrophin. Biochem Soc
Sm 63 (1998): 51-69.

Kwok CF, Juan CC, Shih KC, ku CM, Jap TS, Ho LT. Insulin-like growth factor-1
increases endothelin receptor A levels and action in cultured rat aortic smooth muscle

cells. J Cell Biochem 94 (2005): 1126-1134.

Lai A, Sarcevic B, Prall OW, Sutherland RL. Insulin/insulin-like growth factor-1 and
estrogen cooperate to stimulate cyclin E-Cdk2 activation and cell Cycle progression in

79

MCF-7 breast cancer cells through differential regulation of cyclin B and
p21(WAF1/Cip1). J Biol Chem 276 (200]): 25823-25833.

Le Roith D, Bondy C, Yakar S, Liu J -L, Butler A. The Somatomedin Hypothesis: 2001.
Endocr Rev 22 (2001): 53 - 74.

Lin FT, Daaka Y, Leflcowitz RJ. Beta-arrestins regulate mitogenic signaling and
clathrin-mediated endocytosis of the insulin-like growth factor I receptor. J Biol Chem
273 (1998): 31640-31643.

Little W, Kay RM. The effects of rapid rearing and early calving on the subsequent .
performance of dairy heifers. Anim Prod 29 (1979): 131-142.

Maor G, Rochwerger M, Segev Y, Phillip M. Leptin acts as a growth factor on the
chondrocytes of skeletal growth centers. J Bone Miner Res 17 (2002): 1034-1043.

Matitashvili E, Brarnley AJ, Zavizion B. An in vitro approach to ruminant mammary
gland biology. Biotechnol Adv 15 (1997): 17-41.

McGrath MF, Collier RJ, Clemmons DR, Busby WH, Sweeny CA, Krivi GG. The direct
effect of insulin-like growth factors (IGFS) on normal bovine mammary epithelial cell
proliferation and production of IGF binding proteins. Endocrinology 129(1991): 671-
678.

Meares GP, Zmijewska AA, J ope RS. Heat shock protein-90 dampens and directs
signaling stimulated by insulin-like growth factor-1 and insulin. F EBS Lett 574 (2004):
181-186.

Medema RH, Kops GJ, Bos JL, Burgering BM. AFX-like F orkhead transcription factors
mediate cell-cycle regulation by Ras and PKB through p27kip1. Nature 404 (2000): 782-
787.

Merlo GR, Graus-Porta D, Cella N, Marte BM, Taverna D, Hynes NE. Growth,
differentiation and survival of HCll mammary epithelial cells: diverse effects of receptor
tyrosine kinase-activating peptide growth factors. Eur J Cell Biol 70 (1996): 97-105.

Messai H, Khatib AM, Lebrun G, Aubin P, Florina M, Jean F, Mitrovic DR. Endothelin-
] in monolayer cultures of articular chondrocytes from young and old rats: regulation by
growth factors and cytokines. Mech Ageing Dev 114 (2000): 37-48.

Meyer MJ, Capuco AV, Ross DA, Lintault LM, Van Amburgh ME. Developmental and

nutritional regulation of the prepubertal heifer mammary gland: I. Parenchyma and fat
pad mass and composition. J Daigy Sci 89 (2006): 4289-97.

80

Muise-Helmericks RC, Grimes HL, Bellacosa A, Malstrom SE, Tsichlis PN, Rosen N.
Cyclin D expression is controlled post-transcriptionally via a phosphatidylinositol 3-
kinase/Akt-dependent pathway. J Biol Chem 273 (1998): 29864-29872.

Oredsson SM. Polyamine dependence of normal cell-cycle progression. Biochem Soc
Trans 31 (2003): 366-370.

Pawitan Y, Michiels S, Koscielny S, Gusnanto A, Ploner A. False discovery rate,
sensitivity and sample size for microarray studies. Bioinforrnatics 21 (2005): 3017-3024.

Petitclerc D, Chapin LT, Tucker HA. Carcass composition and mammary development
responses to photoperiod and plane of nutrition in Holstein heifers. J Anim Sci 58
(1984): 913-9.

Plath-Gabler A, Gabler C, Sinowatz F, Berisha B, Schams D. The expression of the IGF
family and GH receptor in the bovine mammary gland. J Endocrinol 168 (2001): 39-48.

Purup S, Sejrsen K, Akers RM. Effect of bovine GH and ovariectomy on mammary
tissue sensitivity to IGF-I in prepubertal heifers. J Endocrinol 144 (1995):153-158.

Radcliff RP, VandeHaar MJ, Kobayashi Y, Sharma BK, Tucker HA, Lucy MC.Effect of
dietary energy and somatotropin on components of the somatotropic axis in Holstein
heifers. J Daig Sci 87 (2004): 1229-35.

Reddy RK, Mao C, Baumeister P, Austin RC, Kaufman RJ, Lee AS. Endoplasmic
reticulum chaperone protein GRP78 protects cells from apoptosis induced by

topoisomerase inhibitors: role of ATP binding site in suppression of caspase-7 activation.
J Biol Chem 278 (2003): 20915-20924.

Robinson RM, Akers RM, Forsten KE. Real-time detection of insulin-like growth factor-
] stimulation of the MAC-T bovine mammary epithelial cell line. Endocrine 13 (2000):
345-352.

Romagnolo D, Akers RM, Byatt J C, Wong EA, Turner JD. IGF-I-Induced IGFBP-3
potentiates the mitogenic actions of IGF-I in mammary epithelial MD-IGF-I cells. Mo_l
Cell Endocrinol 102 (1994): 131-139.

Saiki RK, Walsh PS, Levenson CH, Erlich HA. Genetic analysis of amplified DNA with
irmnobilized sequence-specific oligonucleotide probes. Proc Natl Acad Sci USA 86
(1989): 6230-6234.

Sejrsen K, Huber JT, Tucker HA, Akers RM Influence of nutrition of mammary
development in pre- and postpubertal heifers. J Dairy Sci 65 (1982): 793-800.

Sejrsen K, Huber JT, Tucker HA. Influence of amount fed on hormone concentrations
and their relationship to mammary growth in heifers. J Daig Sci 66 (1983): 845-55.

81

Sejrsen K, Purup S. Influence of prepubertal feeding level on milk yield potential of
dairy heifers: a review. J Anim Sci 75 (1997): 828-35.

Sejrsen K, Purup S, Vestergaard M, Foldager J. High body weight gain and reduced
bovine mammary growth: physiological basis and implications for milk yield potential.
Domest Anim Endocrinol 19 (2000): 93-104.

Shamay A, Cohen N, Niwa M, Gertler A. Effect of insulin-like growth factor I on
deoxyribonucleic acid synthesis and galactopoiesis in bovine undifferentiated and
lactating mammary tissue in vitro. Endocrinology 123 (1988): 804-809.

Silva LF. 2000 Role of leptin in mammary gland development in prepubertal heifers.
Ph.D. Thesis. Michigan State University, East Lansing.

Silva LF, Liesman J S, Etchebarne BE, Nielsen MS, Vandehaar MJ. Short
communication: intramammary infusion of IGF-I increases bromodeoxyuridine labeling
in mammary epithelial cells of prepubertal heifers. J Dairy Sci 88 (2005): 2771-2773.

Sinha YN, Tucker HA. Mammary development and pituitary prolactin level of heifers
from birth through puberty and during the estrous cycle. J Dairy Sci 52 (1969): 507-12.

Specht B, Bartetzko N, Hohoff C, Kuhl H, F rankc R, Borchers T, Spener F. Mammary
derived growth inhibitor is not a distinct protein but a mix of heart-type and adipocyte-
type fatty acid-binding protein. J Biol Chem 271(1996): 19943-19949.

Smith J M, Van Amburgh ME, Diaz MC, Lucy MC, Bauman DE. Effect of nutrient
intake on the development of the somatotropic axis and its responsiveness to GH in
Holstein bull calves. J Anim Sci 80 (2002): 1528-37.

Stelwagen K, Grieve DG. Effect of plane of nutrition on grth and mammary gland
development in Holstein heifers. J Dairy Sci 73 (1990): 2333-41.

Strange KS, Wilkinson D, Emerrnan JT. Mitogenic properties of insulin-like growth
factors I and II, insulin-like growth factor binding protein-3 and epidermal growth factor
on human breast epithelial cells in primary culture. Breast Cancer Res Treat 75 (2002):
203-212.

Stull MA, Richert MM, Loladze AV, Wood TL. Requirement for IGF-I in epidermal
grth factor-mediated cell cycle progression of mammary epithelial cells.
Endocrinology 143 (2002): 1872-1879.

Tivesten A, Bollano E, Andersson I, Fitzgerald S, Caidahl K, Sjogren K, Skott 0, Liu JL,
Mobini R, Isaksson OG, Jansson JO, Ohlsson C, Bergstrom G, Isgaard J. Liver-derived
insulin-like growth factor-I is involved in the regulation of blood pressure in mice.
Endocrinologv 143 (2002): 4235-4242.

82

Tonner E, Barber MC, Allan GJ, Beattie J, Webster J, Whitelaw CB, Flint DJ. Insulin-
like growth factor binding protein-5 (IGFBP-5) induces premature cell death in the
mammary glands of transgenic mice. Development 129 (2002): 4547-57.

Umayahara Y, Kawarnori R, Watada H, Irnano E, Iwama N, Morishima T, Yamasaki Y,
Kaj imoto Y, Kamada T. Estrogen regulation of the insulin-like growth factor I gene
transcription involves an AP—l enhancer. J Biol Chem 269 (1994): 16433-16442.

Van Amburgh ME, Galton DM, Bauman DE, Everett RW, Fox DG, Chase LE, Erb HN.
Effects of three prepubertal body growth rates on performance of Holstein heifers during
first lactation. J Daig Sci 81 (1998): 527-538.

Vandehaar MJ, Shanna BK, F ogwell RL. Effect of dietary energy restriction on the
expression of insulin-like growth factor-I in liver and corpus luteum of heifers. J D '
§9_i 78 (1995): 832-41.

von Harsdorf R, Hauck L, Mehrhof F, Wegenka U, Cardoso MC, Dietz R. E2F-1
overexpression in cardiomyocytes induces downregulation of p21CIP1 and p27KIP1 and
release of active cyclin-dependent kinases in the presence of insulin-like growth factor I.
Circ Res 85 (1999): 128-136.

Weber MS, Purup S, Vestergaard M, Akers RM, Sejrsen K. Regulation of local synthesis
of insulin-like growth factor-1 and binding proteins in mammary tissue. J Dairy Sci 83
(2000): 30-37.

Weber MS, Purup S, Vestergaard M, Ellis SE, Scndergard-Andersen J, Akers RM,
Sejrsen K. Contribution of insulin-like growth factor (IGF)-I and IGF-binding protein-3
to mitogenic activity in bovine mammary extracts and serum. J Endocrinol 161 (1999):
365-373.

Weber MS, Purup S, Vestergaard M, Akers RM, Sejrsen K. Nutritional and somatotropin
regulation of the mitogenic response of mammary cells to mammary tissue extracts.
Domest Anim Endocrinol 18 (2000):]59-64.

Williams JM and Daniel CW. Mammary ductal elongation: differentiation of
myoepithelium and basal lamina during branching morphogenesis. Dev Biol 97 (1983):
274-290.

Wood TL, Richert MM, Stull MA, Allar MA. The insulin-like growth factors (IGFS) and
IGF binding proteins in postnatal development of murine mammary glands. J Mammy
Gland Biol Neoplasia 5 (2000): 31-42.

Woodward TL, Akers RM, Turner JD. Lack of mitogenic response to EGF, pituitary,
and ovarian hormones in bovine mammary epithelial cells. Endocrine 2 (1994): 529-535.

83

Woodward TL, Xie J, Fendrick JL, Haslam SZ. Proliferation of mouse mammary
epithelial cells in vitro: interactions among epidermal growth factor, insulin-like growth
factor I, ovarian hormones, and extracellular matrix proteins. Endocrinology 141 (2000):
3578-3586.

Yee D, Paik S, Lebovic GS, Marcus RR, Favoni RE, Cullen KJ, Lippman ME, Rosen N.
Analysis of insulin-like growth factor I gene expression in malignancy: evidence for a
paracrine role in human breast cancer. Mol Endocrinol 3 (1989): 509-517.

Zadeh SM, Binoux M. The 16-kDa proteolytic fragment of insulin-like growth factor
(IGF) binding protein-3 inhibits the mitogenic action of fibroblast grth factor on
mouse fibroblasts with a targeted disruption of the type 1 IGF receptor gene.
Endocrinology 138 (1997): 3069-3072.

Zapf J, Hauri C, Waldvogel M, F roesch ER. Acute metabolic effects and half-lives of
intravenously administered insulinlike growth factors I and II in normal and
hypophysectomized rats. J Clin Invest 77 (1986): 1768-1775.

Zavizion B, Politis I, Gorewit RC, Turner JD, Spitzer E, Grosse R. Effect of mammary-
derived growth inhibitor on proliferation of MAC-T bovine mammary epithelial cells. J
Daigy Sci 76 (1993):3721-3726.

Zavizion B, Gorewit RC, Politis I. Subcloning the MAC-T bovine mammary epithelial
cell line: morphology, growth properties, and cytogenetic analysis of clonal cells. J Daig
§_c_i 78 (1995): 515-527.

Zhao X, McBride BW, Politis I, Huynh HT, Akers RM, Burton JH, Turner JD. Receptor
binding and growth-promoting activity of insulin-like growth factor-I in a bovine
mammary epithelial cell line (MAC-T3). J Endocrinol 134 (1992): 307-312.

Zong CS, Chan J, Levy DE, Horvath C, Sadowski HB, Wang L. Mechanism of STAT3

activation by insulin-like growth factor 1 receptor. J Biol Chem 275 (2000): 15099-
15105.

84

1“iiiijijijrjjji11:11ng1

   

, .w . -_ x—z i:.*::fi.......‘ :1.“ng‘4