Identification and characterization of proteins secreted by the corpus luteum of the cow during the estrous cycle and pr...

MISSING IMAGE

Material Information

Title:
Identification and characterization of proteins secreted by the corpus luteum of the cow during the estrous cycle and pregnancy
Physical Description:
xvii, 280 leaves : ill. ; 29 cm.
Language:
English
Creator:
Ndikum-Moffor, Florence Maboh
Publication Date:

Subjects

Subjects / Keywords:
Proteins -- Identification   ( lcsh )
Proteins -- Synthesis   ( lcsh )
Corpus luteum   ( lcsh )
Estrus   ( lcsh )
Cows -- Reproduction   ( lcsh )
Animal Science thesis, Ph. D
Dissertations, Academic -- Animal Science -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1995.
Bibliography:
Includes bibliographical references (leaves 165-191).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Florence Maboh Ndikum-Moffor.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002069929
oclc - 34345372
notis - AKQ8190
System ID:
AA00009036:00001


This item is only available as the following downloads:


Full Text










IDENTIFICATION AND CHARACTERIZATION OF PROTEINS SECRETED BY
THE CORPUS LUTEUM OF THE COW DURING THE ESTROUS CYCLE AND
PREGNANCY













By

FLORENCE MABOH NDIKUM-MOFFOR


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA

1995


UNIVERSITY OF FLORIDA UBRARIES
























Dedicated to Mandi, Koga, Kongwenebime, Gaston,

and

my mother and late father.














ACKNOWLEDGMENTS


I would like to thank my major professor, Dr. Michael J. Fields, for his

constant guidance throughout the course of my studies. His patience, positive

attitude, and constant encouragement, gave me the strength to withstand the

difficult and trying moments of this learning process. Dr. Fields did not only

supervise my graduate training and research, but also gave me informal lectures

about the philosophies of life and the importance to strive for excellence. I am

grateful for his leadership and the kindness his family showed towards mine.

I also wish to thank Dr. Rosalia C.M. Simmen for giving me the

opportunity to work and carry out the second part of my research work in her

laboratory. The time I spent in her laboratory was very resourceful and

reaffirmed the importance for thoroughness and detail, which I appreciate.

I wish to thank the other members of my supervisory committee, Dr.

William Buhi, Dr. Peter J. Hansen, and Dr. Lynn Larkin for their readiness to

discuss my research and give helpful suggestions. I wish to specifically thank

Dr. William Buhi for his advice with the first part of my research, and Dr. Peter

Hansen for letting me use his laboratory equipment. Dr. Phillip A. Fields is

acknowledged for the important contributions to the histological component of

this work. Many thanks to Dr. Michael F. Smith and Dr. Harry Nick for providing









some of the cDNA probes used in this work.

I am grateful to Dr. John T. Banser, the Director of the Institute of Animal

and Veterinary Research, IRZ, for providing initial financial support, and the

Animal Science Department, University of Florida for the Graduate

Assistantship. I also wish to acknowledge Dr. Daniel A. Mbah for encouraging

me to pursue graduate training at the University of Florida.

My sincere thanks go to Dr. Shou-Mei Chang for the technical and moral

support she gave me throughout my stay in the laboratory. My thanks go to Ms.

Jill Davidson for her help with flushing of the reproductive tracts of cows used in

this study. I also wish to thank Mr. Frank Michel, Dr. Lokenga Badinga and Dr.

Mike Green for their technical and moral support during the time I spent in their

laboratory. Many thanks to Mrs. Glenda Walton for her moral support and help

with word processing, and to Christa Jenssen for help with preparing slides.

My thanks go to Jack Stokes and Dean Glicco for managing the herd, and

to Larry Eubanks and Leeroy Washington for slaughtering the animals used in

this work. The graduate school was less stressful because of friends like Dr.

Thais Diaz, Mr. Keith Rollyson, Ms. Lannett Edwards, and Mr. Andres Kowalski. I

thank all my family friends, Dr. Esther Smith, Ms. Patience Njofang, Mr. Andrew

Kweh, Mr. Odemari Mbuya and Ms Marie Gaffney, for their friendship. Special

thanks go to my mother, sisters, and brother for their encouragement.









I could not have survived this program without the support, patience, and

love from my husband, Gaston, and my kids, Kongwenebime, Koga and Mandi,

and I wish to thank them for their love and patience.















TABLE OF CONTENTS


ACKNOWLEDGMENTS .........

LIST OF TABLES ..............

LIST OF FIGURES .............

ABSTRACT ...................

CHAPTERS

1 INTRODUCTION .........


2 REVIEW OF LITERATURE ..................
The Corpus Luteum ........................
Histology of the Corpus Luteum .........
Biochemical Differences Between Small and
Large Luteal Cells ..............


........... 3
. . 4

. . 7


Ontogeny of Small and Large Luteal Cells ............... .
Intercellular Communication Among Luteal Cells .................
Cell-Cell Communication in Luteal
Angiogenesis and Development .................. .
Cell-Cell Interaction in Luteal Hormone
S ynthesis ................................ .....
Corpus Luteum of the Estrous Cycle ......................... .
Luteinization and CL Development ..................... .
Luteal Angiogenesis ..................................
Regulation of Luteal Function During The
Estrous C ycle .......................... ............
Luteal Regression ................... ..............
Role of PGF2a in Luteal Regression .............. .
Role of Oxygen Free Radicals in Luteal
R egression ..............................
Corpus Luteum of Pregnancy .............................. .
CL Morphology During Pregnancy ..................... .
Maternal Recognition of Pregnancy .................... .


10
12

13

15
16
16
18

21
23
23

27
28
29
30









Protein Synthesis by the Corpus Luteum ................. .... 36
General Overview of Protein
Synthesis and Release ................... 37
Protein Secretion ............................ 38
Protein Targeting ............................ 38
Translocation Across ER Membrane ............... 39
Processing and Sorting of Proteins in the Golgi ......... ... 39
Apolipoprotein A-1 ................... .. ...... ............ 40
Biochemical Characterization .................... 40
Apo A-1 gene ....................................... 41
Metabolism .............. ........... .............. 41
Role of Apo A-1 .......... ...................... 43
Regulation of Apo A-1 Synthesis by Steroid
Horm ones ................ .... ............. 44
Effects of Nutrition on Apo A-1 Synthesis ............. ... 45
A polipoprotein E ......... .......................... .. 46
Biochemical Characterization .................... 46
Effects of FSH, LH, cAMP, and Phorbol Ester
on Apo E Synthesis ................ ........... 47
Regulation of Apo E Synthesis by Cell
Cholesterol ......... ....................... 49
Regulation of Apo E Secretion by Cytokines ............... 49
Role of Apo E in Ovarian Function ................ 50
Manganese Superoxide Dismutase .................... 50
Biochemical Characterization .................... 50
Regulation of Mn SOD Synthesis by
Oxidative Stress ......... ....... ........... 51
Regulation of Mn SOD by Gonadotropins ............. .... 52
Regulation of Mn SOD Synthesis by
Cytokines and Phorbol Ester ................. 52
Role of Mn SOD in the Ovary .................... 53
Tissue Inhibitors of Metalloproteinases: TIMP-1 and
TIMP-2 ............... .. ........... ........... 54
Tissue Inhibitor of Metalloproteinases-1 ................... 55
Biochemical Characterization .................... 55
Regulation of TIMP-1 Synthesis by Gonadotropins .......... 56
Regulation of TIMP-1 Synthesis by Steroid
Hormones .......... ...... ..... ......... 58
Regulation of TIMP-1 Synthesis by Cytokines and
Growth Factors ............................. 59
Tissue Inhibitor of Metalloproteinases-2 ................. 60
Biochemical Characterization .................... 60
Regulation of TIMP-2 Synthesis by Gonadotropins .......... 61









Regulation of TIMP-2 Synthesis by Steroid
Hormones ......... .................
Regulation of TIMP-2 Synthesis by Cytokines and
Growth Factors ........................

3 PROTEINS SYNTHESIZED AND RELEASED IN CULTURE BY
THE BOVINE CORPUS LUTEUM: THE ESTROUS CYCLE AND
PREGNANCY .......................................


..... 61

..... 62


Introduction .............................................
Materials and M ethods ....................................
M materials ..........................................
Collection of Luteal Tissue ............................
Culture Medium .....................................
Time Course Studies of Incorporation of
R adiolabel ...................................
Culture and Radiolabelling of Luteal Tissue ...............
TCA Precipitation ...................................
Light and Electron Microscopy .........................
Two-Dimensional-SDS-Polyacrylamide Gel
Electrophoresis ..........................
First-dimension: Isoelectric Focusing ...............
Second-dimension: SDS-PAGE ................ ..
Protein Blotting and Amino Acid Sequencing ..............
Progesterone Assay .. ...........................
Statistical Analysis ..................................
R results ................................................ .
Histology of Luteal Tissue .............................
CL Weight and Plasma Progesterone ....................
Incorporation of Radiolabel into TCA-precipitable
Protein ............... .... ........
Luteal Protein Synthesis and N-terminal Amino Acid
M icro Sequencing ............. .................
Protein Synthesis and Release during the Estrous
Cycle ............... ....................
Luteal Protein Synthesis and Release during
Pregnancy ................................
Radiolabelled Culture with 3H-glucosamine and 3S-
methionine ................ ................
Discussion ............................ ....... ........


. 79

. 82

. 86

. 87

. 90
. 93











4 EXPRESSION OF MESSENGER RNA OF APOLIPOPROTEIN
A-1 AND E IN BOVINE LUTEAL TISSUE DURING THE
ESTROUS CYCLE AND PREGNANCY ........................ 100

Introduction .......................................... 100
Materials and Methods ................. ................... 102
M materials ......... ....... ......................... 102
Tissue Collection .................................... 102
Isolation of RNA .................................... 103
Preparation of Plasmid DNA ....................... 104
Restriction Analysis and Isolation of Insert ............ ... 105
Northern Hybridization ........................... 106
Autoradiography .............................. 107
Dot Blot Hybridization .......................... 107
Statistical Analysis .................................. 108
Results ............ ... ... .... ....... ........ 109
Northern Blot Analysis of Apo E and Apo A-1
m RNA ......... .............. ........... 109
Dot Blot Analysis ............................... 109
D discussion ................................. ............ 115

5 EXPRESSION OF MESSENGER RNA OF TISSUE INHIBITOR OF
METALLOPROTEINASES-1 AND -2 IN BOVINE LUTEAL TISSUE
DURING THE ESTROUS CYCLE AND PREGNANCY ............ 120

Introduction .................. ........................ 120
M materials and Methods .................................... 121
M materials ..................... ....... ........... 121
Tissue Collection ............................... 122
Isolation of RNA ................ ............ ...... 123
Preparation of Plasmid DNA ....................... 124
Restriction Analysis and Isolation of Insert ............ ... 125
Northern Hybridization ............................... 126
Autoradiography .................................... 127
Dot Blot Hybridization .......................... 127
Statistical Analysis .................................. 128
Results ................................. ............. 128
Expression of TIMP-1 mRNA ...................... 128
Expression of TIMP-2 mRNA ...................... 132
D discussion ................................. ........... 135










6 CHANGES IN THE EXPRESSION OF MESSENGER RIBONUCLEIC
ACID FOR MANGANESE SUPEROXIDE DISMUTASE IN THE BOVINE
CORPUS LUTEUM DURING THE ESTROUS CYCLE AND
PREGNANCY .............. ............................ 139


Introduction ........ ............
Materials and Methods .............
Tissue Collection ........... .
Isolation of RNA .............
Restriction Analysis and Isolation
Northern Hybridization ........
Autoradiography .............
Dot Blot Hybridization .........
Statistical Analysis ...........
R results .................. ....
Northern Blot Analysis ........
Dot Blot Analysis .............
D discussion ..................... .


........ ....
of Plasmid DNA

. .. .. .
.... .. .. ..


7 GENERAL RESULTS AND DISCUSSION ........


... .... 152


R EFER EN C ES .................................. ..........

APPENDIX 1 ANIMAL CARE AND TISSUE COLLECTION ............

APPENDIX 2 PROGESTERONE IMMUNOASSAY ..................

APPENDIX 3 CULTURE AND RADIOLABELLING OF LUTEAL
T ISS U E ......................... ..........

APPENDIX 4 DETERMINATION OF INCORPORATION OF
RADIOLABEL INTO NEWLY-SYNTHESIZED
PROTEINS .............. ........ ..........

APPENDIX 5 SEPARATION OF PROTEINS IN LUTEAL-
CONDITIONED MEDIUM BY ELECTROPHORESIS.

APPENDIX 6 ELECTRO-BLOTTING OF PROTEINS TO
M EM BRANE .............. ....... ..........

APPENDIX 7 DETERMINATION OF TOTAL PROTEIN BY
METHOD OF LOW RY ........................


. .
. .
. .
.....
Insert
. .
. .
. .
. .
....
. .
. .
. .


139
141
141
141
142
143
144
145
146
146
146
146
149


165


. 192

. 196



.. 200




.. 204



. 206



.. 220



.. 223










APPENDIX 8 MEASUREMENT OF APO A-1 MRNA ..........

APPENDIX 9 MEASUREMENT OF APO E MRNA ............

APPENDIX 10 MEASUREMENT OF TIMP-1 AND TIMP-2 MRNA


APPENDIX 11

APPENDIX 12

APPENDIX 13

APPENDIX 14

APPENDIX 15


APPENDIX 16

APPENDIX 17

APPENDIX 18


RNA ISOLATION AND PURIFICATION ...........

NUCLEIC ACID LABELLING ...................

NORTHERN BLOTTING AND HYBRIDIZATION .

RNA DOT BLOT HYBRIDIZATION ...............

IMMUNOHISTOCHEMICAL LOCALIZATION OF
A P O E ................... ....... ......

RAW DATA-EXPERIMENT 1 ................. .

RAW DATA-EXPERIMENT 2, 3, 4, AND 5 .........

. .... ..... ..


BIOGRAPHICAL SKETCH .............................


280


. 225

... .. 235

...... 237

...... 249

... .. 254

...... 256

...... 264


...... 268

...... 270

...... 273

...... 277














LIST OF TABLES


Table paae

2-1. Summary of factors regulating synthesis of
proteins ............... ........... ........... 63

3-1. Comparison of N-terminal amino acid sequences for
proteins 1, 8, 9, 10 and 11 in luteal-conditioned
medium to sequences in the protein data banks ............. 85














LIST OF FIGURES


Figure

3-1. Electron micrograph of luteal cells. Top panel: luteal cell of tissue
prior to incubation (control), x 9000; Middle panel: luteal cell of
tissue post-24 h incubation without radiolabel (control), x 11,400;
Bottom panel: luteal cell of tissue post-24 h incubation with
50 pCi 3H-leucine, x 11,400 ............................... 76

3-2. Least square means SEM of plasma concentrations of
progesterone (ng/ml) and weight (g) of the corpus luteum
across days 3, 7, 11, 14, 17, and 19 of the estrous cycle
and days 17, 88, 170, and > 240 of pregnancy .............. ... 78

3-3. Time course studies of incorporation of radiolabel into
TCA-precipitable protein ........................... 80

3-4. Percent incorporation of radiolabel into newly-synthesized
proteins .............. ................................. 81

3-5. Representative fluorographs of proteins synthesized de novo
in explant culture and released into the medium by CL on different
days of the estrous cycle .......... ..................... 83

3-6. Representative fluorographs of proteins synthesized de novo in
explant culture and released into the medium by CL on different
days of pregnancy ....................................... 84

3-7. Densitometric analysis of fluorographs of newly-synthesized
proteins in luteal-conditioned medium during the estrous cycle
and pregnancy. Values are least square means standard error
of the mean ........................................ 88









3-8. Representative fluorographs of proteins synthesized de novo in
explant culture and released into medium on day 240 of pregnancy.
Left panel: radiolabelled culture with 3H-leucine. Right panel:
radiolabelled culture with 3H-glucosamine ................ .... 91

3-9. Representative fluorographs of proteins synthesized de novo
in explant culture and released into medium on day 240 of
pregnancy. Left panel: radiolabelled culture with 3H-leucine.
Right panel: radiolabelled culture with 3S-methionine ....... ..... 92

4-1. Northern blot analysis of apolipoprotein E mRNA. The same
blot was probed with Ir-actin that served as a control for the
loading and the integrity of the RNA ................... 110

4-2. Northern blot analysis of apolipoprotein A-1 mRNA. The
same blot was probed with I-actin that served as a control
for the loading and the integrity of the RNA .............. .... 111

4-3. Dot blot analysis of Apo E mRNA. Ten pg total RNA isolated
from CL during the estrous cycle (days 2-3, 16-17, and 20) and
pregnancy (days 17, 90-120, 170-180, and > 215) was loaded
per sample. RNA blots were hybridized with 32P-labelled Apo
E cDNA ............ .............................. 112

4-4. Dot blot analysis of Apo A-1 mRNA. Ten pg total RNA isolated
from CL during the estrous cycle (days 2-3, 16-17, and 20) and
pregnancy (days 17, 90-120, 170-180, and > 215) was loaded
per sample. RNA blots were hybridized with 32P-labelled Apo A-1
cDNA ............................................ 113

4-5. Apo A-1 mRNA expression relative to I-actin is presented as
LSMean SEM ................ ..................... 114

5-1. Northern blot analysis of luteal TIMP-1 mRNA. The same
blot was probed for B-actin that served as a control for the
loading and the integrity of the RNA ................... 129

5-2. Dot blot analysis of TIMP-1 mRNA. Ten pg total RNA isolated
from luteal tissue during the estrous cycle (days 2-3, 16-17,
and 20) and pregnancy (days 17, 90-120, 170-180, and > 215)
was loaded per sample. RNA blots were hybridized with
32P-labelled TIMP-1 cDNA ......................... 130









5-3. TIMP-1 mRNA expression relative to B-actin is presented as
LSMean SEM (n = 2-5) and differences determined by
orthogonal contrasts ................ ..................... 131

5-4. Northern blot analysis of luteal TIMP-2 mRNA. The same
blot was probed for B-actin to act as a control for the loading
and integrity of RNA .................................. 132

5-5. Dot blot analysis of luteal TIMP-2 mRNA. Ten pg total RNA
isolated from luteal tissue during the estrous cycle (days 2-3,
16-17, and 20) and pregnancy (days 17, 90-120, 170-180,
and > 215) was loaded per sample. RNA blots were hybridized
with 32P-labelled TIMP-2 cDNA ...................... 133

5-6. TIMP-2 mRNA expression relative to Ir-actin is presented as
LSMean SEM (n = 2-5) and differences determined by
orthogonal contrasts ......... ........................ 134

6-1. Northern blot analysis of luteal Mn SOD mRNA. The same blot
was probed for B-actin to act as a control for the loading and
integrity of RNA ................. .................... 147

6-2. Dot blot analysis of luteal Mn SOD mRNA .................. 148














Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

IDENTIFICATION AND CHARACTERIZATION OF PROTEINS SECRETED BY
THE CORPUS LUTEUM OF THE COW DURING THE ESTROUS CYCLE AND
PREGNANCY

By

FLORENCE MABOH NDIKUM-MOFFOR

December 1995


Chairperson: Michael J. Fields
Major Department: Animal Science

Experiments were carried out to determine protein synthesis and

secretion by the corpus luteum (CL) of the cow during the estrous cycle and

pregnancy. Temporal changes were observed in the types of proteins

synthesized across days of the estrous cycle, but not across pregnancy. The

bovine CL synthesized many proteins in culture, five of which were further

characterized and identified by N-terminal amino acid sequence analysis.

Proteins identified were apolipoprotein E (Apo E, 35 kDa, pl 5.5), apolipoprotein

A-1 (Apo A-1, 27 kDa, pl 6), tissue inhibitor of metalloproteinases-1 (TIMP-1, 30

kDa, pl 8), tissue inhibitor of metalloproteinases-2 (TIMP-2, 20 kDa, pi 8), and

manganese superoxide dismutase (Mn SOD, 22 kDa, pl 8). Northern and dot









blot analyses revealed presence of mRNA for each of the five identified proteins

within bovine luteal tissue during the estrous cycle and pregnancy, confirming

synthesis of these proteins by the bovine CL.

Synthesis of Apo A-1 and Apo E by the bovine CL is novel and has not

been reported in any species. This is also the first report of expression of Apo

A-1 mRNA by the CL. There have been reports of synthesis of TIMP-1 and

TIMP-2 by the CL during the estrous cycle, but this is the first study to examine

luteal synthesis of TIMP-1 and TIMP-2 during pregnancy. Similarly, this is the

first report of the temporal changes in the synthesis of Mn SOD and expression

of its mRNA during the estrous cycle and pregnancy in the cow.

Apolipoprotein E and Apo A-1, given their association with low density

lipoprotein and high density lipoprotein, respectively, may be involved in

regulating cholesterol availability for luteal membrane synthesis and

steroidogenesis. TIMP-1 and TIMP-2 may be involved in steroidogenesis during

the luteal phase and pregnancy, and tissue remodelling during luteal regression.

Luteal Mn SOD may play a significant role in maintaining luteal function during

the luteal phase and pregnancy by preventing damage of luteal cells by reactive

oxygen species. These luteal proteins are believed to be acting via autocrine

and/or paracrine mechanisms.


xvii















CHAPTER 1
INTRODUCTION


A physiological role for the ovaries was first demonstrated by Frankel

(1903, cited by Short, 1977) when he showed that removal of the ovaries in

rabbits resulted in termination of pregnancy. The corpus luteum of the ovary

was later identified as the component responsible for maintenance of pregnancy.

However, the mechanism through which the CL carries out this function was not

known. Because of its high vascularity, Prenant (cited by Short, 1977)

suggested that the CL is an organ of internal secretion and was capable of

releasing products into systemic circulation.

The advent of new technologies and techniques such as electron

microscopy made it possible to investigate the histology and biochemistry of the

CL during the estrous cycle and pregnancy (Priedkaln and Weber, 1968). As a

result, it is now known that the CL is composed of different types of cells; the

steroidogenic small and large luteal cells, endothelial cells, macrophages,

fibroblasts, and monocytes, and that there are some morphological and

biochemical differences between the CL of the estrous cycle and that of

pregnancy. In addition, the steroidogenic small and large luteal cells have been

identified as responsible for the production of progesterone, the hormone











required for maintenance of pregnancy.

With the use of other modern research techniques in cell biology and

histochemistry, it is common knowledge that luteal cells possess the structural

machinery required for the synthesis and secretion of proteins. Recent research

has shown that apart from the production of progesterone, the CL produces

many different proteins and peptide hormones (Schams, 1989), angiogenic and

mitogenic factors (Grazul-Bilska et al., 1993; Reynolds et al., 1994), and growth

factors (Schams, 1989).

However, there is a paucity of information about the protein-synthetic

potential of the CL across the estrous cycle and pregnancy, the nature of the

proteins synthesized, factors regulating their synthesis and release, and the cell

types responsible for their production. Identification and characterization of

luteal proteins and factors will enable the determination of their roles in luteal

growth, development and steroidogenesis, which will contribute to the

understanding of fertility as well as tissue growth and development. There is an

indication from some studies that other cell types, apart from the steroidogenic

luteal cells, are responsible for synthesis of some luteal proteins.

Objectives of this dissertation were to examine protein synthesis by

bovine luteal explants in culture during the estrous cycle and pregnancy,

identify, isolate and characterize the proteins secreted, and determine presence

of their respective mRNA in luteal tissue during the cycle and pregnancy.














CHAPTER 2
REVIEW OF THE LITERATURE


The Corpus Luteum


The corpus luteum (CL) is formed following the rapid development of the

follicle after ovulation. Regnier de Graaf gave the first detailed description of the

CL which he called "globular bodies" (Jocylyn and Setchel, 1972). Malpighi

(1628-1698) introduced the term corpus luteum or "yellow body" (Short, 1977).

The first attempt to define the role of the CL was made by Prenant (1898) who

suggested that it is a gland of internal secretion because of its high vascularity.

The CL has been shown to play several roles in the reproductive process. Most

of these roles depend upon synthesis of progesterone, which is necessary for

implantation, maintenance of pregnancy, control of the estrous cycle and

parturition. Some of these roles were suggested following observations by

Frankel in 1903 that removal of ovaries in rabbits terminated pregnancy, and this

gave support to Gustav Born's hypothesis that corpora lutea are required for

implantation (cited by Amoroso, 1968). If fertilization does not occur, the CL

regresses in some species with a subsequent decline in circulating

progesterone. The decline in progesterone concentrations leads to an increase

in pulse frequency of luteinizing hormone (LH) increased follicular estrogen










synthesis, and a new wave of follicular development (Karsch et al., 1984).

However, if fertilization occurs, the CL is prevented from regressing by signals

sent by the embryo (Bazer et al., 1991 a; 1991b). The CL could therefore be

regarded as the biological clock for the events of the estrous and menstrual

cycles, pregnancy and parturition.


Histology of the Corpus Luteum


The corpus luteum of several species is made up of different types of

cells, namely, the steroidogenic large luteal and small luteal cells, endothelial

cells, fibroblasts, macrophages and lymphocytes. Earlier studies (Priedkalns

and Weber, 1968; Koos and Hansel, 1981) showed that the mature bovine

corpus luteum contains two distinct steroidogenic cell types: the small and large

luteal cells. The major differences between these two cell types were the

presence of numerous large mitochondria in the large cells, and numerous lipid

bodies in the small cells. Other cell types and stromal cells were also observed

in the corpus luteum. The large luteal cells are larger in size (30-40 p), with

extensively folded plasma membranes which are exposed to other luteal cells

and to inter-cellular vascular areas. The small luteal cells are smaller in size

(15-20 pm in diameter) but with a denser chromatin pattern than the large cells.

More recent studies (O'Shea et al., 1989) showed that in the bovine corpus

luteum, the large luteal cells make up 3.5 % of the total cells/unit area of luteal

tissue while the small luteal cells make up 26.7%, and endothelial cells /











pericytes are the most abundant (52.3%). However, large luteal cells occupy

40.2% of the corpus luteum volume/density, while the small luteal cells occupy

27.7%. Similar observations were made by Lei et al. (1991) who reported that

human and bovine corpora lutea contain more nonluteal than luteal cells, and

the small luteal cells are always greater in number than the large luteal cells

irrespective of the reproductive states.

Ultrastructural and cytochemical studies by Parry et al. (1980) showed

that large granulosa-derived bovine luteal cells are the most common cell types

of the mid-luteal corpus luteum. They, however, did not observe any

morphological differences between small and large cells as reported by

Priedkalns et al. (1968). Large luteal cells are always close to capillaries and

contain large round nuclei (10 pm in diameter), large amounts of agranular

endoplasmic reticulum, and abundant mitochondria scattered throughout the

cytoplasm. Bovine large luteal cells contain numerous Golgi complexes and

electron-dense secretary granules during mid-cycle (Parry et al., 1980). Bovine

luteal secretary granules are single membrane-bound and are 0.2-0.4 pm in

diameter compared to 0.2 pm in sheep (Gemmell et al., 1974). In the cow,

secretary granules are found in a cluster in the cytoplasm (granules make up 2-4

% of luteal cell cytoplasm) close to the Golgi complexes, and sometimes near

the edge of the cells (Parry et al., 1980). The contents of secretary granules are

released by exocytosis to the intercellular space. Lipid droplets are present in

some cells and concentrations varied inversely with the number of electron-










dense granules present. Morphometric analyses have shown that luteal cells

occupy the maximum area of the corpus luteum on day 12 of the cycle while the

protein-synthesizing compartment (ER plus polysomes and/or ribosomes)

increases from day 6 to a maximum on day 17, indicating that protein synthesis

is part of the metabolism of the corpus luteum during the cycle (Parry et al.,

1980). The population of secretary granules is greatest on day 17 of the bovine

cycle and declines thereafter indicating a drop in the secretary process, which

usually precedes the fall in progesterone synthesis (Hansel et al., 1973). More

recent studies in the cow have shown that unlike in sheep, luteal secretary

granules are not dispersed throughout the cytoplasm but are found in a cluster

(Fields et al., 1983; 1992). During the bovine cycle, the percent of large luteal

cells with secretary granules is lowest on day 3 (3%), highest during mid-cycle

(day 7, 84 %, day 11, 64 %) and declines on day 14 (26 %) to lowest level on

days 17 (16 %) and 19 (8 %) (Fields et al., 1992). An earlier study in pregnant

cows also indicated dynamic changes in the population of secretary granules

during the course of pregnancy. In the large luteal cell of pregnancy, the

number of secretary granules are low or undetectable prior to day 45, increases

to maximum around day 200 and declines at the end of gestation (Fields et al.,

1985).

There have been reports of other morphological differences between

small and large luteal cells. Chegini et al. (1984) reported that large (18-45 pm)

cells contain more mitochondria than small (15-18 pm) cells, and both contain










rough and smooth endoplasmic reticulum, lysosomal vesicles, Golgi complexes

and membrane-bound dense granules. Granules vary in shape and form

clusters close to the nucleus (Chegini et al., 1984). In a more recent study,

Chegini et al. (1991) observed that the nuclear volume is greater in small than

large luteal cells during estrous cycle and pregnancy. Small luteal cells are also

more sensitive to hCG-induced increase in nuclear volume than the large luteal

cells. The cytoplasmic:nuclear ratio is greater in large than small cells.

Biochemical analysis of the two cell types have shown that the amount of protein

per cell is lower in small than the large cell, while the protein/DNA ratios are

similar for both. It was suggested that differences observed may have been

caused by transformation of small luteal cells to large luteal cells (Chegini et al.,

1984).


Biochemical Differences Between Small and Large Luteal Cells


Biochemical differences between small and large luteal cells have been

reported. Small luteal cells (which make up 85% of the total luteal cell

population) produce small amounts of basal progesterone and respond to

secretagogues (cAMP, hCG), while the large luteal cells (which make up 8-12%

of luteal cell population) produce greater amounts of basal progesterone but are

not responsive to cAMP and hCG (Koos and Hansel, 1981; Fitz et al., 1982).

The LH-induced increase in progesterone production by small luteal cells is

mediated by an increase in cAMP. Nonhormonal activators of protein kinase A










(forskolin, cholera toxin, dibutyryl cAMP) also selectively stimulate synthesis of

progesterone in small but not in large luteal cells (Wiltbank, 1994). It is not clear

by what mechanism progesterone synthesis by large luteal cells is stimulated,

since they have been shown to produce large amounts of progesterone in

culture even in the absence of luteotropic stimuli (Wiltbank, 1994). Factors

implicated in the LH-independent increase in progesterone production by large

luteal cells include PGE2 (Fitz et al., 1984), insulin (Sauerwein et al., 1992), and

growth factors (Einspanier et al., 1990; Budnik and Mukhopadhyay, 1991;

Miyamoto et al., 1992).

Alila et al. (1989) reported that treatment of bovine luteal cells with LH

causes a rapid increase in intracellular free calcium in both small and large cells.

Alila et al. (1989) reported that the LH-induced rise in intracellular calcium is

biphasic in small cells (initial peak due to mobilization of intracellular calcium,

and a second rise due to influx of extracellular calcium), while a single rise is

observed in the large luteal cells. The increase in calcium is also greater for

small luteal than large luteal cells (Alila et al., 1989). In an earlier study, Alila et

al. (1988) observed that phorbol dibutyrate increases progesterone synthesis in

bovine small, but not large luteal cells in culture. Those observations support

earlier postulations that protein kinase C (PKC) is involved in progesterone

synthesis in the bovine corpus luteum, and that the stimulatory effects of PKC on

progesterone synthesis involves only the small luteal cells (Hansel et al., 1987).

The differential response of large and small luteal cells to secretagogues seems










to be related to the distribution of receptors between the two cell types. Fitz et

al. (1982) showed that small luteal cells of sheep contain 10X more LH receptors

than large cells, while large cells are more enriched with prostaglandin E and F2a

receptors. PGF2a is luteotropic in small luteal cells where it stimulates

phospholipase C activity but does not reduce the LH-stimulated cAMP or

progesterone accumulation (Davis et al., 1989).

Other biochemical differences between small and large luteal cells are in

their abilities to synthesize and secrete proteins and peptide hormones

(Schams, 1989; Rodgers, 1990). Judging from the morphology of luteal cells,

large luteal cells have the intracellular organelles specialized for secretion of

proteins and peptides (Anderson, 1982). Luteal tissue of sheep (Wathes and

Swann, 1982), women (Wathes et al., 1982) and cows (Wathes et al., 1983;

Fields et al., 1983) produces oxytocin and neurophysin. Oxytocin is present in

ovine (Rodgers et al., 1983) and bovine (Fields et al., 1986; 1992) large luteal

cells but not in small cells. Another peptide hormone, relaxin, is produced by

corpora lutea of a multitude of mammals (see Sherwood, 1994, for review).

Bagnell et al. (1989) reported that in pigs, relaxin is localized within the large

luteal cell, but not the small luteal cell.










Ontoqeny of Small and Large Luteal Cells


The origin of small and large luteal cells has been a matter of

controversy. In the late 1800s and early 1900s, some researchers believed that

granulosa cells of the ovarian follicle degenerate following ovulation, and only

cells of theca developed into a corpus luteum. Others, however, thought that

granulosa cells develop into a corpus luteum while thecal cells degenerate.

Loeb (1906) was the first to suggest that the corpus luteum is composed of cells

originating from both theca and granulosa layers. Similar observations were

made in the sow (Corner, 1919), cow (Donaldson and Hansel, 1965; Lobel and

Levy, 1968; Priedkalns et al., 1968), ewe (O'Shea et al., 1980), rat (Pederson,

1951) and human (Guraya, 1971). It is now believed that small luteal cells

originate from theca internal of the follicle, while the large luteal cells originate

from the granulosa cells. Morphological studies in sheep (O'Shea et al. 1980)

and cattle (Donaldson and Hansel, 1965; Priedkalns et al., 1968) indicate that

small luteal cells originate from thecal cells. Alkaline phosphatase, a marker of

theca internal cells, was used to demonstrate that theca cells differentiate into

small luteal cells of the ovine CL (O'Shea et al., 1980). Further support to this

hypothesis was provided by observations that thecal cells incubated with

forskolin and insulin for 9 days become luteinized, have low basal progesterone

secretion, increased LH-induced secretion of progesterone, and do not secrete

oxytocin (Meidan et al., 1990). These physiological characteristics are similar to










those exhibited by small luteal cells. Studies in sheep also suggest that

granulosa cells develop into large luteal cells because the number of granulosa

cells in pre-ovulatory follicles (O'Shea et al., 1985) approximates the number of

large luteal cells (O'Shea et al., 1986), and ovine granulosa cells undergo little

or no mitosis after ovulation (McClellan et al., 1975). Evidence for differentiation

of granulosa cells to large luteal cells was provided by observations that

incubations of bovine granulosa cells with forskolin and insulin for 9 days

resulted in luteinized cells similar to large luteal cells (high basal progesterone

secretion, reduced LH-induced progesterone secretion, and secretion of

oxytocin) (Meidan et al., 1990).

Alila and Hansel (1984) demonstrated that monoclonal antibodies

developed against theca cell membranes bound to small luteal cells while

monoclonal antibodies against granulosa cells bound specifically to the large

luteal cells. As the estrous cycle progressed, the number of large luteal cells

bound to theca antibody was similar to the number of small cells bound to theca

antibody (Alila and Hansel, 1984). They suggested that theca-derived small

luteal cells differentiate into large luteal cells as the estrous cycle progresses.

However, a study by O'Shea et al. (1986) in which comparisons of the cellular

composition of ovine luteal cells of mid- and late estrous cycle were made, did

not agree with the hypothesis that small luteal cells differentiate to large luteal

cells during the estrous cycle. Alila and Hansel (1984) also observed that large

luteal cells bound to granulosa antibody contained more mitochondria and









12
electron-dense granules than those bound to theca antibody. In general, the CL

is a dynamic organ and its morphology changes with the reproductive state of

the animal. Farin et al. (1989) reported a change in cellular composition of the

ovine corpus luteum during the estrous cycle and pregnancy. The number of

small cells increase with no change in size as the cycle progressed, whereas the

size of large cells increased with no change in number. Thus changes in the

relative proportions of the two cell types or interactions between them, may

determine the function of the CL at different periods of the estrous cycle.

In sheep, the number of small steroidogenic luteal cells increases 4-fold

through day 8 and then decreases through day 16 (Niswender et al., 1985).

However, the number of nonsteroidogenic cells > 8 pm increases 2-fold between

days 4 and 8 of the ovine cycle and declines through day 16, while the number

of nonsteroidogenic cells < 8 pm reaches a peak on day 12. Because of the

similarity in pattern of steroidogenic and nonsteroidogenic cells during the

estrous cycle, it was speculated that small nonsteroidogenic cells are a source

of stem cells that give rise to small steroidogenic luteal cells, which later develop

into large luteal cells (Niswender et al., 1985).


Intercellular Communication Among Luteal Cells


The heterogenous nature of the cellular components of the corpus luteum

is an important feature which seems to be necessary for this organ to effectively

carry out its biological functions. Formation of the CL following ovulation










involves incorporation of cells from the theca and granulosa layers of the

ovulating follicle. As earlier discussed, the process of luteinization involves

luteal angiogenesis (Zheng et al., 1993), and an increase in size of theca and

granulosa cells, and number of smooth endoplasmic reticulum and mitochondria.

The corpus luteum is composed of large and small luteal cells, macrophages,

monocytes, fibroblasts and endothelial cells. A substance produced by one cell

may affect the function of another cell and this is referred to as cell-cell

communication (Rodgers, 1990). This communication may be via gap junctions

(Anderson and Little, 1984; Redmer et al., 1991) and adherens-type junctions

(Weber et al., 1987; O'Shea et al., 1990) which have been observed between

luteal cells. Exchange of factors among luteal cells could also occur via the

blood stream humorall). While gap junctions allow passage of very low

molecular weight substances from one cell to another, adherens-type junctions

serve to bind cells together (Rodgers, 1990).


Cell-Cell Communication in Luteal Anqiogenesis and Development


Ovine luteal cells have been shown to produce angiogenic factors

(Redmer et al., 1988; Grazul-Bilska et al., 1992). It has been proposed that

growth factors such as fibroblast growth factor, insulin-like growth factor (IGF)-1,

epidermal growth factor (EGF), and cytokines are involved in luteal

angiogenesis (Koos, 1989). This idea was supported by observations by Zheng

et al. (1993) that bovine large and small luteal cells produce basic fibroblast








14

growth (BFGF) -1 and -2 during the estrous cycle. BFGFs are potent angiogenic

factors and their production by large and small luteal cells follow a pattern

similar to luteal development (Zheng et al., 1993). Thus the steroidogenic large

and small cells produce HBGFs which stimulate proliferation of endothelial cells.

A functional relationship between endothelial cells and luteal cells was

suggested following observations that both endothelial (Mayerhofer et al., 1992)

and luteal cells (Mayerhofer et al., 1991) express a neural cell adhesion

molecule. In addition, endothelial cells produce prostacyclin (PGFI2) (Maclntyre

et al., 1978) which has been shown to increase concentrations of plasma

progesterone in the cow (Milvae et al., 1980). A more recent study by Girsh et

al. (1995) gave further evidence of interactions between endothelial cells and

the steroidogenic cells of bovine CL. Endothelial cells were shown to secrete

PGI2, which in turn stimulates secretion of cAMP and progesterone by bovine

large and small luteal cells. However, it was also observed that the presence of

endothelial cells is required for PGF2a-induced inhibition of progesterone

production by luteal cells (Girsh et al., 1995). Thus endothelial cells of CL may

regulate response of steroidogenic luteal cells to luteotropic and luteolytic

signals.

IGF-I and its mRNA is produced by the bovine CL (Einspanier et al.,

1990) during the estrous cycle and gestation. IGF-I has been shown to stimulate

synthesis and secretion of progesterone and oxytocin by luteal tissue

(Sauerwein et al., 1992). In a more recent study IGF-I was immunolocalized









15

mainly in bovine large luteal cells with little staining in small cells (Amselgruber

et al., 1994). No IGF-I immunoreactivity was observed in pericytes,

macrophages, fibroblasts or smooth muscle cells of blood vessels (Amselgruber

et al., 1994). The differences in distribution of IGF-I immunoreactivity may

indicate differences in cell reactivity and possible paracrine or autocrine

interactions between small and large luteal cells (Amselgruber et al., 1994). A

role of IGF-I in CL function has been suggested following observations that

concentrations of luteal IGF-I increase during early and mid-luteal phases and

decline rapidly after luteolysis (Einspanier et al., 1990).


Cell-Cell Interaction in Luteal Hormone Synthesis


Luteal progesterone is produced mainly by the large luteal cells in most

species. However, progesterone production by isolated populations of porcine

small and large luteal cells has been shown to be greater when both cell

populations are cultured together than when cultured separately (Lemon and

Mauleon, 1982). The increase in progesterone production observed in

cocultures was attributed to stimulation of progesterone production in large cells

by some factors) produced by small cells, suggesting an interaction between

small and large luteal cells. Oxytocin, a peptide hormone produced by CL of

ruminants is thought to be involved in cell-cell interaction between small and

large cells because it is produced only by large luteal cells (Fields and Fields,

1986; Theodosis et al., 1986) and is capable of inhibiting LH-induced P4










production by small luteal cells (Schams, 1989).


Corpus Luteum of the Estrous Cycle


Luteinization and CL Development


Corpora lutea are formed following ovulation of a mature follicle mediated

by gonadotrophin stimulation. Following the preovulatory LH surge, a series of

morphological and biochemical changes take place within the follicle to change

the latter to a corpus luteum. The LH surge serves dual roles of stimulating

ovulation and converting the follicle to a corpus luteum. This luteinization

process involves breakdown of the basement membrane between theca and

granulosa layers, invasion of the follicular antrum space by blood vessels, and

development of an extensive vascular network (Zheng et al., 1993; Niswender et

al., 1994). The invading capillaries are formed via both migration and mitosis of

endothelial cells (Zheng et al., 1993). After ovulation, the follicle grows rapidly

to 10 times its weight in just 7 days. This increased growth is attributed to

hypertrophy and hyperplasia of thecal cells which migrate into the hollow

follicular antrum after ovulation, and integrate among luteinizing granulosa cells

(O'Shea et al., 1980). Formation of the CL is initiated by a series of

morphological and biochemical changes in the theca and granulosa cells of the

preovulatory follicle. This process, referred to as luteinization, changes the

follicle from a predominantly estradiol-producing structure to one that secretes











progesterone.

Morphological changes associated with luteinization include accumulation

of smooth endoplasmic reticulum, mitochondria with tubular cristae, increase in

size of Golgi apparatus, and accumulation of glycogen-containing granules

(Niswender and Nett, 1994). These changes provide the CL with the ability to

efficiently produce progesterone. During the luteinization process, theca and

granulosa cells of the preovulatory follicle differentiate into small and large luteal

cells, respectively (Meidan et al., 1990). Other morphological changes

associated with luteinization include an increase in the cytoplasmic nuclear ratio

and appearance of a large number of lipid droplets containing sterol esters.

After ovulation, there is an increase in gap junctions among developing luteal

cells in rats (Anderson and Little, 1984). O'Shea et al. (1990) reported the

presence of adherens-type junctions between small and large luteal cells of

cattle. In a more recent study, Redmer et al. (1991) reported the presence of

gap junction-like structures in bovine luteal cells from mid-cycle.

Biochemical changes associated with luteinization of the follicle include a

switch from a predominantly estradiol-producing structure to one that secretes

mainly progesterone. During luteal formation, there is an increase in expression

of mRNA and enzyme activity for P4~o side chain cleavage and 3(1-hydroxysteroid

dehydrogenase (3R1HSD). There is also an increase in activity of cholesterol

esterase as the CL becomes fully functional. These changes are consistent with

the CL's role in progesterone synthesis. In contrast, luteinization decreases










estrogen production by decreasing levels of mRNA and protein for 17a-

hydroxylase cytochrome P4., the enzyme that catalyses conversion of

pregnenolone or progesterone to androgen. In preovulatory follicles in cattle

(Rodgers et al., 1987) and rat (Hedin et al., 1987), the expression of mRNA and

protein levels for aromatase cytochrome P4o enzyme also decreases rapidly

after the LH surge. Corpora lutea of the rat express aromatase mRNA and

produce estradiol, while CL of domestic ruminants do not synthesize estradiol

(Savard, 1973). Other biochemical changes observed after ovulation include

reduced expression of genes encoding FSH and LH receptors in granulosa

follicular cells, which results in a down regulation of both receptors. In contrast,

LH receptor levels increase with formation of the CL in ewes (Diekmann et al.,

1978). In support of this observation in rat, expression of the gene encoding the

LH receptor increases with development of the CL (Segaloff et al., 1990).


Luteal Anqiogenesis


Angiogenesis is one of the features of luteinization, and continues after

formation of the CL. Blood flow to the CL increases with luteinization and

accounts for about 90% of the total ovarian blood flow during the mid-luteal

phase. At this time, about 60% of each luteal cell's surface directly faces a

capillary (Keyes and Wiltbank, 1988). In the rat, luteal blood flow and the

number of luteal endothelial cells increase during mid-pregnancy (Bruce et al.,

1984). The ability of the CL to cause angiogenesis was first reported by Jakob










et al. (1977). Corpora lutea of sheep and cattle have been shown to produce

angiogenic factors in culture (Redmer et al., 1988). Gospodarowicz et al. (1985)

also isolated an angiogenic factor from the bovine CL which accounted for 84%

of the angiogenic activity in crude CL extracts, and had amino acid sequence

homology with bovine brain and pituitary fibroblast growth factor. It is not clear

what regulates angiogenesis or the high rate of blood flow to the CL. However,

several factors which affect endothelial cell proliferation are proposed as

regulators of luteal angiogenesis. These include prostaglandin E, epidermal

growth factor (EGF), endothelial growth factor, endothelium-stimulating factor,

angiogenin, insulin and transferring (Findlay, 1986).

Development of luteal vasculature is a dynamic process which varies with

the stage of the estrous cycle. Zheng et al. (1993) observed that capillary

density within luteal tissue is sparse in the early luteal phase (days 1-4 post

ovulation), high in the middle phase (days 5-17), and is reduced dramatically in

the late phase (days 18-21). Since a reduction in tissue function is usually

associated with a decline in blood flow and vascularity, the fall in blood flow and

vascularity during the late luteal phase could indicate degeneration of luteal

cells and a decline in luteal function (Zheng et al., 1993). Redmer et al. (1988)

observed that luteal-conditioned medium from early (days 1-4), mid (days 5-17),

and late (days 18-21) ovine cycle stimulates angiogenesis (mitogenesis and

migration of endothelial cells. Angiogenic activity increases with advancement

of the luteal phase. In that study, LH stimulated the production and/or release of









20

the angiogenic factors, while PGF2, blocked the LH-induced stimulation (Redmer

et al., 1988). Thus, luteal vasculature could be stimulated in an

autocrine/paracrine manner by angiogenic factors produced by the corpus

luteum.

It has also been suggested that growth factors may regulate ovarian

angiogenesis since they are present in the ovary, and have effects on

endothelial cells. Koos (1989) speculated on the possible roles of fibroblast

growth factor, insulin-like growth factors (IGFs), EGF, TGFa and TGFIB in

ovarian angiogenesis. Tumor necrosis factor alpha (TNFa), PGE1, PGE2,

estradiol, plasminogen activator proteolytic enzymes (plasminogen activator,

plasmin, collagenase) also stimulate angiogenesis. However, it is not clear how

the expression and activities of these factors are regulated, and how they

modulate luteal angiogenesis. In a recent study by Zheng et al. (1993), pattern

of immunostaining for heparin binding growth factor (HBGF) (also known as

basic FGF) in bovine luteal tissue was parallel to that of luteal vascular

development throughout the estrous cycle, suggesting a role of HBGF in

vascular development. Follicular granulosa and theca internal cells,

macrophages, endothelial cells, and mast cells have been implicated as involved

in the regulation of angiogenesis in the ovary and the CL (Koos, 1989).

Brannstrom and Norman (1993) proposed that mast cells present in CL of some

species may modulate the luteinization process by producing and secreting

cytokines and proteases involved in tissue remodelling, angiogenesis and










stimulation of progesterone production.

Mechanisms involved in the control of blood flow to the CL are not known

and are still hypothetical. Studies in the rabbit led to rejection of the hypothesis

that luteotropic hormones promote luteal blood flow (Keyes and Wiltbank, 1988).

Rather, it was suggested that blood flow to the CL is not regulated by luteotropic

hormones, and has no correlation with the level of luteal steroidogenesis.


Regulation of Luteal Function During the Estrous Cycle


With its formation, the CL is composed of mainly steroidogenic small and

large luteal cells, and endothelial cells, pericytes, macrophages, smooth muscle

cells and fibroblasts. The number, volume and density of small and large luteal

cells vary throughout the estrous cycle, but volume of CL occupied by each cell

type stays relatively constant (Niswender et al., 1994). Progesterone is the major

hormone produced by the CL during the luteal phase of the cycle and LH is the

major luteotropin that stimulates luteal progesterone production in several

species.

The luteotropic regulation of the CL has been the subject of much

research. Corpus luteum formation is thought to be induced by the preovulatory

LH surge. Dependence of the CL on LH has been tested in several studies

involving hypophysectomy, administration of LH antibodies, and administration

of GnRH antagonists. In sheep, hypophysectomy 5 days post estrus does not

affect serum and luteal progesterone concentrations, although CL weight on day









22
12 was lower than expected (Farin et al., 1990). These results indicated that a

CL deprived of LH may still function although its growth and development could

be compromised. Baird (1992) observed that most LH pulses on days 6-7 and

13-14 of the ovine cycle are followed by a rise in progesterone concentration.

However, changes occur in concentrations of progesterone independent of LH

pulses. Injection of a GnRH antagonist during early luteal phase also causes a

small decline in progesterone production, whereas administration of GnRH

antagonist on day 13 causes a rapid decline in progesterone secretion, and

luteal regression (Baird, 1992).

Thus the CL requires a luteotropic support from LH during early and mid-

luteal phases of the cycle. However, the CL of mid-cycle seems to be less

resistant to withdrawal of luteotropic support than the early luteal phase CL. It

was suggested that high levels of progesterone during the mid-luteal phase

cause a reduction in frequency of endogenous LH pulses. Also, uterine PGF2a,

which is secreted during the mid-luteal phase, may interfere with coupling of LH

to the adenyl cyclase second messenger system, and increase sensitivity of the

CL to PGF2a as a result of the long intervals between LH pulses. These events

would result in luteal regression (Baird, 1992).











Luteal Regression


When the mature follicle is not fertilized, the CL is eventually destroyed

thereby allowing the female another opportunity to start a new cycle, ovulate and

become pregnant. Destruction of the CL is referred to as luteolysis or luteal

regression. The basic features of luteolysis are a decline in progesterone

secretion which is referred to as functional luteolysis, followed by structural

changes which lead to breakdown of luteal cells. The drop in circulating

progesterone reduces the negative feedback of progesterone on the pituitary

and leads to an increase in gonadotropin pulse frequency, a new wave of

follicular growth, and ovulation.

Role of PGF in Luteal Regression

The mechanism of luteal regression has been the subject of numerous

studies. Prostaglandin F2a from the uterus is the natural luteolysin in domestic

farm animals and most rodents (McCracken et al., 1972; Knickerbocker et al.,

1988). Due to the close apposition of the uterine and ovarian blood vessels,

there is a local transport of PGF2a from the uterine vein to the ovarian artery,

making it possible for PGF2a to reach the CL without going through systemic

circulation (Del Campo and Ginther, 1973). The luteolytic effect of PGF2a occurs

through its interaction with ovarian oxytocin. The pulsatile release of oxytocin

from the CL stimulates the pulsatile release of PGF2a from the uterus, which in

turn positively feeds back to further increase luteal oxytocin release. This









24

positive feedback loop continues until the demise of the corpus luteum (Jenkin,

1992a).

The endometrial oxytocin receptor is thought to be the determining factor

as to whether or not luteolysis will occur (Flint et al., 1992a). Inhibition of uterine

oxytocin with a synthetic oxytocin receptor antagonist prevents pulsatile release

of PGF2a,, and luteolysis (Jenkin, 1992a). It is believed that trophoblast

interferons in ruminants prevent luteal regression during early pregnancy via a

similar mechanism (Flint et al., 1992; Jenkin, 1992b). This local effect of PGF2a

is not present in horses because of a different uterine-ovarian vascular anatomy

and the equine CL will regress even in the absence of the ipsilateral uterine

horn. In contrast to the horse, removal of the uterine horn ipsilateral to the CL in

ruminants, pigs and some rodents prevents luteal regression, whereas removal

of the contralateral horn has no effect on luteal lifespan (Ginther, 1974).

Fairclough et al. (1981) reported that passive immunization of cows and ewes

with PGF antibodies prolonged the estrous cycle, demonstrating PGF2a's

luteolytic function. Copelin et al. (1989) reported that cows actively immunized

against PGF2, exhibit prolonged luteal lifespan after first ovulation. Cows with

higher PGF2, antibody titres had longer luteal lifespan and progesterone

secretion (Copelin et al., 1989). In a more recent study, active immunization of

ewes against PGF2, on day 5 postpartum prevented ovulation (Bettencourt et

al., 1993). However, in rhesus monkeys, humans and dogs, the uterus does not

seem to be needed for luteal regression (Neill et al., 1969), and an intraluteal










production of PGF2a has been suggested in these species (Niswender et al.,

1994).

Several mechanisms have been proposed by which PGF2a exerts its

negative effects on luteal function. Nett et al. (1976) proposed that PGF2a

causes degeneration of luteal cells by causing a reduction in blood flow leading

to hypoxia within luteal tissue. Nett et al. (1976) observed degeneration of

capillary endothelial cells during luteal regression. It has been suggested that

the decrease in blood flow could be due to a degeneration of luteal capillaries

rather than vasoconstrictive effects of PGF2a (Wiltbank et al., 1990). The

changes in luteal vasculature observed across the estrous cycle correlated with

luteal growth, development, and regression (Zheng et al., 1993), and supports

earlier reports on the role of blood flow in the control of luteal regression.

Capillary density was low in the early luteal phase (days 1-4), high in mid-cycle

(days 5-17), and decreased dramatically in the late stage (days 18-21). Like in

cattle, PGF2, rapidly decreases luteal blood flow in ewes with a corresponding

decline in circulating progesterone.

It has been proposed that PGF2, decreases LH binding to luteal cells in

vivo and may block stimulation of adenylate cyclase by LH (Niswender and Nett,

1994). In vitro, PGF2, has also been shown to block the LH-induced increase in

cAMP and progesterone production by ovine luteal tissue (Niswender and Nett,

1994). Wiltbank and Niswender (1992) proposed that the luteolytic action of

PGF2a involves binding of PGF2a to a specific membrane receptor on large luteal









26

cells, activation of phosphoinositide-specific phospholipase C, and an increase

in intracellular calcium which activates protein kinase C (PKC). PKC inhibits

intracellular cholesterol transport leading to a decrease in progesterone

production (Wiltbank and Niswender, 1992). Thus the antisteroidogenic effects

of PGF2a, are mediated through the PKC second messenger system (Niswender

et al., 1994). The sustained increase in free intracellular calcium causes

degeneration and death of large luteal cells (Wiltbank et al., 1989).

It has also been proposed that the antisteroidogenic effects of PGF2a on

the CL could be due to a reduction in the number of LH receptors (Behrman et

al., 1978), and the uncoupling of the LH receptors from the adenylate cyclase

second messenger system (Fletcher and Niswender, 1982). It has been shown

that treatment of luteal cells with PGF2a inhibits formation of cAMP by LH in vitro

(Dorflinger et al., 1983).

Apart from intracellular changes, the CL undergoes morphological

changes during regression. The plasma membrane of the regressing CL

contains gap junctions, maculae adherens, coated invaginations and microvilli

(Niswender and Nett, 1994). In the bovine CL, a decrease in amount of smooth

ER, an increase in number of autophagic vacuoles and an increase in number of

lipid droplets in cytoplasm are also observed during regression (Fields et al.,

1992). Other morphological changes observed in regressing bovine CL are a

decrease in the number of secretary granules, presence of numerous swollen

mitochondria, and a decrease in size of the steroidogenic cells (Niswender and









27

Nett, 1994). These morphological changes are common to several species. The

increase in lipid droplets and cytoplasmic vacuoles have been observed in

guinea pigs (Paavola, 1979), humans (Vanlennys and Madden, 1965) and

rabbits (Koering and Thor, 1978).

These observations led to suggestions that the immune system may play

a role in luteal regression. Bovine luteal cells have been shown to express MHC

class II antigens and their expression increase with advancing age of the corpus

luteum (Pate, 1994). Expression of MHC class II antigens is restricted to the

large luteal cells during mid-cycle, but is observed in both the small and large

luteal cells prior to luteal regression (Benyo et al., 1991). Luteal MHC II antigen

expression has also been observed during PGF2a-induced luteolysis, but

expression is absent in pregnant cows (Benyo et al., 1991). Interferon-gamma

has also been shown to induce expression of MHC II in bovine luteal cells

(Fairchild and Pate, 1989). Thus IFN-gamma may contribute to the luteolytic

process by stimulating luteal prostaglandin synthesis and inhibiting

progesterone production (Pate, 1994).

Role of Oxygen Free Radicals in Luteal Regression

The role of oxygen free radicals in the luteolytic process has been the

subject of recent studies. In the rat, luteal levels of superoxide anions and

hydrogen peroxide increase after treatment with PGF2a, and during regression

(Sawada and Carlson, 1989). It has been suggested that oxygen radicals may

cause lipid peroxidation, which in turn, stimulates luteal PGF2a production, which








28

may contribute to luteolysis (Carlson et al., 1993). However, the mechanism by

which superoxide radicals inhibit progesterone secretion in vivo is not known,

although superoxide radicals have been shown to inhibit luteinizing hormone

(LH) stimulation of cAMP, and cAMP-induced progesterone secretion in rat luteal

cell cultures (Gatzuli et al., 1991). Superoxide dismutase, the enzyme that

converts superoxide anions to hydrogen peroxide, increases as the luteal phase

progresses and LH has been shown to induce its synthesis (Laloraya et al.,

1988). These results suggest that oxygen radicals are involved in the luteolytic

process since their removal favors synthesis of progesterone, as opposed to the

decline in progesterone observed during luteal regression.


Corpus Luteum of Pregnancy


When fertilization occurs following ovulation, the CL does not regress but

rather is maintained and becomes responsible for producing progesterone

required to maintain pregnancy. In the cow, the CL is the major source of

circulating progesterone during most of pregnancy. Later in pregnancy, the

placenta can adequately provide progesterone. Thus in the cow, the CL is not

required as a source of progesterone after day 200 of pregnancy. However in

the pig (Nara et al., 1982) and the rat (Steinetz et al., 1976), the CL is required

throughout pregnancy.











CL Morphology During Pregnancy


Morphologically, the CL of pregnancy is similar to that of the estrous cycle

(Fields et al., 1985). It is made up of steroidogenic large and small luteal cells,

macrophages, lymphocytes, endothelial cells, fibroblasts, as discussed earlier.

However, Weber et al. (1987) observed some morphological differences

between steroidogenic cells of the cycle and those of pregnancy. Viability of

small luteal cells is significantly higher in cyclic than in pregnant cows, while

viability of large cells is not different between estrous cycle and pregnancy

(Weber et al., 1987). The significance of these observations is not known.

Biochemically, the CL of pregnancy differs from that of the cycle. Luteal

cells of late pregnancy produce less progesterone than those of nonpregnant

cows (Fields et al., 1985). Weber et al. (1987) observed that large luteal cells of

pregnant cows produce 30 times less progesterone than those of nonpregnant

cows. It has also been shown that small luteal cells of pregnant cows are

unresponsive to exogenous LH in contrast to the small luteal cells of

nonpregnant cows (Weber et al., 1987). More recent studies indicate

differences in protein synthesis and secretion between luteal cells of pregnancy

and the estrous cycle. In the cow, oxytocin and neurophysin are localized within

secretary granules of large luteal cells during the cycle, but are absent in luteal

secretary granules after day 40 of pregnancy (Fields et al., 1992).

When pregnancy occurs, biochemical communications take place










between the concepts and the mother to prevent regression of the corpus

luteum, and thus sustain production of progesterone required to maintain

pregnancy. This phenomenon is referred to as maternal recognition of

pregnancy, whereby the concepts sends signals to the maternal system to

prevent regression of the corpus luteum (Short, 1969). The strategies used for

maternal recognition of pregnancy vary among species and involve different

proteins.


Maternal Recognition of Pregnancy


In cattle, luteolysis is prevented during pregnancy by inhibition of both

basal and oxytocin- or estradiol-stimulated PGF2a secretion via synthesis of an

endometrial prostaglandin synthase inhibitor (Thatcher et al., 1992). In the cow,

presence of the CL is required for maintenance of pregnancy through the first

200 days of gestation. Ovariectomy of pregnant cows prior to day 200 results in

abortions (Estergreen et al., 1967). The bovine placenta does not contribute

significantly to circulating progesterone concentrations even after day 200 of

gestation.

Maternal recognition of pregnancy in cattle occurs on days 16-17

postestrus (Niswender and Nett, 1994). At this time, the concepts secretes

embryonic interferons (interferon-tau) which act via endometrial receptors and

alter the secretion of endometrial PGF2a (Roberts et al., 1992). Earlier studies

have shown that high amplitude pulses of PGF2a occur in nonpregnant heifers









31
during luteolysis, but are absent in pregnant ones (Kindahl et al., 1976). Similar

observations have been reported in ewes (Zarco et al., 1988) and buffaloes

(Batra and Pandey, 1983). Following the transfer of day 15 or 16 bovine

embryos to recipients, recipients with regressed CL were observed to have four

to five spikes of PGFM, whereas recipients in which the CL persisted had

reduced or no PGFM spikes (Betteridge et al., 1984). PGF2a is the uterine

luteolysin in cows. Thatcher et al. (1985) observed that the estradiol-induced

increase in PGFM production is inhibited by the presence of a concepts on day

18 of pregnancy, but is not in cyclic cows on day 18 of the estrous cycle. The

concepts provides greater inhibition of PGFM production on day 20 than the

day 18 concepts because the former has a more extensive contact with the

endometrium (Thatcher et al., 1985). Similarly, the oxytocin-induced increase in

uterine PGFM is significantly less in pregnant than in nonpregnant heifers on

day 19 postestrus (LaFrance and Goff, 1985). These studies suggested an

antiluteolytic-antiPGF effect of the concepts.

The concepts mediates its antiluteolytic-antiPGF effect via the secretion

of proteins. Cyclic cows receiving intrauterine injections containing secretary

proteins found in days 16-18 conceptus-conditioned medium have longer

interestrous intervals than those cows receiving serum proteins (Thatcher et al.,

1985). Intrauterine injections of bovine concepts proteins also reduce

estradiol-stimulated PGFM production (Bazer et al., 1986). The major proteins

secreted by cultured bovine conceptuses have molecular weights of 22-26 kDa









32

and isoelectric points of 5.6-5.8. The protein is referred to as interferon-tau and

it shares 50% amino acid sequence identity with recombinant bovine interferon-

alpha (rblFNa) (Imakawa et al., 1989). Bovine trophoblast protein-1 given at

time of maternal recognition of pregnancy extends luteal function in the cow

(Thatcher et al., 1989) and ewe (Parkinson et al., 1992). Similarly, intrauterine

infusion of recombinant bovine interferon-a extends the length of the estrous

cycle in post-partum cows expected to have short luteal lifespan (Garverick et

al., 1992). The recombinant interferon acts by reducing oxytocin-induced PGFM

release (Plante et al., 1990). Treatment of ewes with recombinant bovine

interferon-alpha I on days 9-19 post-estrus caused a reduction in plasma

concentrations of PGFM when compared to control groups (Parkinson et al.,

1992). Bovine trophoblast protein-1 exerts its antiluteolytic effects by inhibiting

synthesis and/or recycling of endometrial oxytocin receptors directly, or by

inducing synthesis of a PGF2a synthase inhibitor (Bazer et al., 1991). However,

the antiluteolytic signals do not seem to act directly on the CL (Bazer et al.,

1991). Bovine conceptuses also produce PGE2 and small amounts of estradiol

which together stimulate increase in uterine blood flow which may enhance

delivery of antiluteolytic-luteoprotective agents to the ovary (Lewis et al., 1982;

Thatcher et al., 1986).

In the ewe, maternal recognition of pregnancy occurs on days 12-13 post-

estrus. Although serum progesterone concentrations are similar between

pregnant and nonpregnant ewes at this time, the concepts prevents regression









33
of the CL in pregnant ewes (Niswender and Nett, 1994). During the ovine cycle,

prostaglandin F2, is produced by the endometrium and transported to the ovary

where it causes regression of the CL and a decline in progesterone production.

Between days 12-21 of gestation, trophoblast cells of ovine blastocysts secrete

proteins including a major secretary 17 kDa protein (interferon-tau). IFN-tau is

not produced beyond day 21 of gestation, and is the only secretary product

detected on day 13 (Godkin et al., 1982). Intrauterine infusion of IFN-tau on

days 12-18 of the estrous cycle extends corpus luteum lifespan in ewes, while

the CL regresses in untreated ewes. It was also observed that ewes treated with

sheep serum ovulated and formed a new CL while ewes treated with total

concepts proteins did not ovulate (Godkin et al., 1984). Fincher et al. (1984)

showed that IFN-tau inhibits estradiol- and oxytocin-induced uterine production

of PGF2a. Thus IFN-tau is antiluteolytic and anti-PGF. Oxytocin-induced PGF

production is lower in pregnant than in nonpregnant ewes (Fairclough et al.,

1984).

The mechanisms) of action of IFN-tau is not clear but seems to be via its

binding to endometrial receptors (Godkin et al., 1984; Hansen et al., 1989),

causing changes in the secretion of endometrial proteins and prostaglandins,

and extending luteal function (Vallet et al., 1988). Treatment of ewes with ovine

concepts secretary proteins (oCSP) on days 11-15 post-estrus causes a

decline in concentrations of endometrial estrogen receptors, estrogen receptor

mRNA, and progesterone receptor on day 16 when compared with ewes treated










with serum proteins. Ovine concepts secretary proteins also reduce oxytocin

binding and activation of the phosphoinositol second messenger system

(Mirando et al., 1993). In another study, maximum expression of endometrial

progesterone receptor mRNA occurred earlier (days 10-12 post-estrus vs days

14-16) in pregnant than in cyclic ewes, and oxytocin stimulated in vitro

endometrial production of inositol phosphates in cyclic but not in pregnant ewes

(Ott et al., 1993).

Nephew et al. (1989) have shown that intramuscular injections of

recombinant bovine interferon-tau cause an increase in pregnancy rate,

prolificacy and higher survival of conceptuses in ewes. Also, plasma

concentrations of PGE2 in utero-ovarian vein in ewes increase during maternal

recognition of pregnancy (Silvia et al., 1984). IFN-tau and other concepts

proteins also favor production of PGE2 over PGF2a and thus prevent luteal

regression. Ovine IFN-tau, like bovine IFN-tau, acts by binding to endometrial

receptors (Godkin et al., 1984; Hansen et al., 1989), changes secretary patterns

of endometrial proteins and prostaglandins, and extends luteal function (Vallet et

al., 1988). Intrauterine infusions of oCSP and bovine recombinant interferon-

alpha 1 on days 12, 13 and 14 of the estrous cycle cause a decline in

concentrations of endometrial oxytocin receptor, and the oxytocin-induced

increase in PGFM (Vallet and Lamming, 1991).

Ovine IFN-tau genes are expressed specifically by cells of the

trophectoderm and are regulated in a developmental manner. Expression of the








35

gene is present in day 10-11 blastocysts, increases by day 13, declines slightly

by day 15 and sharply thereafter (Guillomot et al., 1990). Roberts et al. (1992)

have also shown that IFN-tau gene expression is undetectable by day 22, a time

when most of the trophoblast is attached to the uterine epithelium. The amount

of protein produced by the blastocysts correlated with the expression of olFN-tau

mRNA (Roberts et al., 1992). Although olFN-tau gene expression ishe amount

developmentally regulated but other factors may affect its production. In sheep

and cattle, an advanced luteal phase enhances concepts development and

earlier expression of IFN-tau (Garrett et al., 1988; Nephew et al., 1991). Xavier

et al (1991) observed a simultaneous expression of c-fos proto-oncogenes and

IFN-tau in ovine trophoblasts. Also in the pig, endometrial expression of c-fos

mRNA increases on day 12 of pregnancy (day of maternal recognition of

pregnancy) and is higher when compared to expression on day 12 of the cycle

(Dubois et al., 1993). Thus c-fos may be induced by IFN-tau or may be involved

in the transcriptional activation of IFN-tau genes. However, no interferon

response elements are present on the c-fos gene, and no AP-1 binding sites

have been found in the promoter region of IFN-tau genes. It is also possible that

both c-fos and IFN-tau genes are regulated by common mechanisms during the

period of maternal recognition of pregnancy (Roberts et al., 1992).










Protein Synthesis by the Corpus Luteum


Apart from its role in the synthesis of progesterone required for the

maintenance of pregnancy and control of the estrous cycle, the CL has been

shown to produce a number of proteins, peptide hormones and factors.

However, most of the proteins have not been fully characterized and attempts to

define their functions have been mainly speculative.

The CL produces oxytocin, neurophysin, relaxin, inhibin, vasopressin, B

endorphin, growth factors, angiogenic factors and protease inhibitors. Results in

this dissertation have shown that the CL also produces many proteins including

apolipoproteins E and A-1, tissue inhibitor of metalloproteinases-1 and 2, and

manganese superoxide dismutase (see table 1-1). The types of proteins

produced by the CL vary with species and the reproductive state of the animal.

Relaxin is produced by CL of human, pig, rat, but not by CL of ruminant species

(Sherwood, 1994). On the other hand, oxytocin and neurophysin are produced

by CL of ruminants but not by nonruminants. These proteins carry out different

functions in the CL, the ovary, and at extraovarian sites. The chemical nature

and the physiological roles of some of these proteins have not been fully

defined.

Luteal proteins and factors are produced by different cell types within the

CL. Cell-cell communication takes place in order to co-ordinate the functions of

luteal cells, and regulate synthesis and release of proteins. Factors regulating










synthesis, release, and biological activity of luteal proteins/factors are not fully

understood.


General Overview of Protein Synthesis and Release


In order to produce a protein, a cell must possess the genetic material

(gene) that codes for that specific protein. The DNA template directs the

synthesis of RNAs which are involved in protein synthesis. Evidence for RNA

involvement in protein synthesis was reported in 1930 when it was observed that

a crude preparation of RNA was rich in protein, and the concentration of RNA-

protein particles ribosomess) correlated with the rate of protein synthesis by the

cell. Francis Crick in 1958 defined the relationship between DNA, RNA and

protein (Crick, 1970). Proteins are synthesized on ribosomes from specific

mRNAs. The mRNA specifies the protein to be synthesized, and associates with

the ribosomes to initiate the process. More recently, Xing et al. (1993) showed

that mRNAs are produced at specific locations in the nucleus and are then

exported to the protein synthesizing machinery in the cytoplasm. RNA

metabolism is also organized in the nucleus in association with the nuclear

matrix (Carter et al., 1993). It has been suggested that the nuclear matrix may

determine what genes are turned on by sequestering and concentrating the DNA

to be transcribed, as well as the transcription factors necessary for a specific

gene expression. Ribosomes associate with transfer RNA and move along the

mRNA to form peptide bonds.










Protein Secretion


Most cells have the ability to secrete proteins. Proteins are secreted via

either a regulatory or a constitutive pathway. Secretion of most eukaryotic

proteins require their transport across the endoplasmic reticulum (ER)

membrane. This process occurs in two steps; targeting, followed by active

transfer across the ER membrane (Rapoport, 1992).


Protein Targeting


Secretory proteins synthesized on ribosomes in the cytoplasm are

targeted to the ER membrane by signal sequences. The signal sequence is

targeted to the ER membrane by the Signal Recognition Particle (SRP). The

nascent peptide chain-ribosome-SRP complex binds to the ER membrane by an

interaction with the membrane-bound SRP receptor or docking protein.

Guanosine triphosphate (GTP) hydrolysis causes the SRP to detach from the

ribosome and signal sequence, then the nascent peptide is transferred to the ER

membrane while the ribosome stays bound to the membrane via a ribosome

receptor. Finally, a GTP hydrolysis reaction causes the SRP to dissociate from

its receptor, and a new targeting cycle can begin (Rapoport, 1992).










Translocation Across ER Membrane


Proteins targeted to the ER membrane are transported across the

membrane at specific sites. The translocation site is a complex and dynamic

structure composed of many proteins and enzymes that catalyze modification of

nascent peptides. Although proteins seem to be translocated through protein-

conducting channels, the mechanism is not clear (Rapoport, 1992). Most

secretary proteins are translocated as precursors with larger masses than the

final mature protein. The mature protein is obtained after cleavage of the signal

sequence during passage across the ER membrane and the Golgi apparatus.


Processing and Sorting of Proteins in the Golqi


After transversing the ER, the protein is transported in vesicles to the

Golgi complex via an energy-dependent process. In the Golgi, the protein may

be post-translationally modified by attachment of functional groups

(glycosylation, acetylation, sulfation). The protein is either concentrated in

secretary granules and released in response to an appropriate stimulus

(regulatory pathway) or is transported to the cell surface in a vesicle and

released directly (constitutive pathway). Exocytosis involves an interaction

between proteins on the cytoplasmic surface of secretary granules vesicless)

and proteins of the inner surface of the plasma membrane (Widnell and

Pfenninger, 1990).









40

One of the objectives of this research was to identify proteins synthesized

and released in culture by the bovine CL during the estrous cycle and

pregnancy. Five of the proteins were identified by N-terminal amino acid

sequence analysis. The next section of this chapter discusses the identified

proteins.


Apolipoprotein A-1


Biochemical Characterization


Apolipoprotein A-1 (Apo A-1) is the major protein component of high

density lipoprotein (HDL), and accounts for 80% or more of the protein moiety of

HDL in the cow and all mammalian species (Sparrow et al., 1992).

Apolipoprotein A-1 has metabolic and structural roles since it contributes to the

size and shape of the lipoprotein particle, solubilizes water-insoluble lipids, and

is a potent activator of lecithin:cholesterol acyltransferase (LCAT).

Apolipoprotein A-1 is also involved in the recognition and modulation of

enzymes involved in lipid metabolism, and binding of lipoproteins to their cellular

receptors (Hopkins et al., 1986).










Apo A-1 gene


Mammalian Apo A-1 gene shows no striking region of evolutionary

conservation, and the bovine Apo A-1 gene is more closely related

phylogenetically to canine than to human and other mammalian lineages

(O'hUigin et al., 1990). The cDNA of Apo A-1 cloned from a bovine cDNA library

(longest insert 963 nucleotides) was shown to contain an open reading frame of

795 nucleotides flanked by 72 and 96 nucleotides at the 5' and 3' end,

respectively (O'hUigin et al., 1990). The 3' flanking region contains a

polyadenylation signal (AATAAA) 14 nucleotides upstream of a poly-A tail.

Based on the cDNA sequence the derived amino acid sequence contains an 18-

residue signal peptide and a 6-residue prosegment.


Metabolism


Apo A-1 is synthesized as a prepro-Apo A-1 mainly by the liver and

intestine, but is also synthesized by other peripheral tissues such as kidney,

adrenal, and testis (Blue et al., 1982), and brain endothelial cells (Guttler et al.,

1990). Sorci-Thomas et al. (1988) showed that the liver and small intestine

contribute to most of plasma Apo A-1, and suggested that other tissues

observed to synthesize Apo A-1 may not contribute significantly to the plasma

Apo A-1 pool, but may play a role in lipid metabolism within these tissues in an

autocrine and /or paracrine manner.









42

The primary structure of Apo A-1 varies with species. Bovine Apo A-1 is

composed of 241 amino acid residues. The propeptide has a sequence (Arg-

His-Phe-Trp-GIn-GIn), and approximately 10% of bovine plasma Apo A-1 is in

the propeptide form (Sparrow et al., 1992). Newly-synthesized Apo A-1 from

different tissues exists in four isoforms (two major and two minor) with isoelectric

points ranging between 5.3 and 5.7, similar to Apo A-1 from the liver (Blue et al.,

1982). The nucleotide and deduced amino acid sequence of bovine Apo A-1

shares 80% homology with the human and rabbit sequences (Gu et al., 1993).

The central region of bovine Apo A-1 is hydrophobic, with highly hydrophilic

regions at the amino and carboxy termini (O'hUigin et al., 1990). The

hydrophobic amphipathic helical regions are necessary for interaction of

apoprotein with phospholipid-cholesterol complexes (Sparrow et al., 1992).

Bovine Apo A-1 contains a single methionine and no cysteine as do the canine

and rabbit proteins (O'hUigin et al., 1990).

The primary translation product is the prepropeptide. It has been

suggested that the bovine Apo A-1 prepropeptide like that of the human (Gordon

et al., 1983) and the rat is post-translationally modified. In humans, intestinal

proapo A-1 contains a hexapeptide extension which ends with GIn-GIn and this

precursor was shown to be secreted by Hep G2, hepatocarcinoma cells in

culture without proteolytic cleavage of the hexapeptide prosegment (Gordon et

al., 1983). Thus it was suggested that Apo A-1 undergoes additional proteolytic

processing before it is integrated into plasma HDL (Gordon et al., 1983). In the









43
cow, cleavage may occur after the conserved Gin-GIn dipeptide to give a mature

Apo A-1 protein with an N-terminal aspartate (O'hUigin et al., 1990). The

conversion of proApo A-1 to mature Apo A-1 is known to occur extracellularly by

an enzyme present in plasma. Edelstein et al. (1988) also showed that this

enzyme produced by a hepatocarcinoma cell line (Hep G2) secretes both Apo A-

1 and the converting enzyme. The converting enzyme is activated by calcium,

inhibited by EDTA, and converts proApo A-1 to Apo A-1 through a first order

kinetic reaction (Edelstein et al., 1988). In another study, Chinese hamster

ovary (CHO) cells transfected with human Apo A-1 secreted Apo A-1.

Furthermore, 90% of the secreted Apo A-1 was the processed mature protein,

and a portion of the secreted protein was associated with lipid (Mallory et al.,

1987). Thus processing of Apo A-1 seems to take place prior to its secretion.


Role of Apo A-1


Apo A-1 is the major protein constituent of high density lipoprotein and it

mediates the binding of HDL to cells. HDL is the major source of circulating

cholesterol in bovine species (Sparrow et al., 1992). Pate and Condon (1989)

showed that both LDL and HDL could be used as a source of cholesterol for

steroidogenesis by bovine luteal cells, and both LDL and HDL enhance luteal

progesterone synthesis in culture. In addition to its ability to solubilize and

transport lipids, Apo A-1 is also a potent activator of lecithin-cholesteryl

acyltransferase, the enzyme that catalyses formation of cholesterol esters from









44

cholesterol (Soutar et al., 1975). It has also been shown that HDL, as opposed

to LDL, causes an increase in the release of placental lactogen by human

placental explants (Handwerger et al., 1987), and from monolayer of trophoblast

cells (Sane et al., 1988). Apo A-1 was implicated for the HDL-mediated

stimulation of placental lactogen release. Wu et al. (1988) showed that HDL

stimulates placental lactogen release by stimulating production of cAMP. Thus

cAMP is a second messenger in HDL-mediated release of hPL, and HDL may

carry out other functions in steroidogenic cells of the ovary by stimulating

adenylate cyclase activity and cAMP production.


Regulation of Apo A-1 Synthesis by Steroid Hormones


There are reports to indicate that estrogen may regulate synthesis of Apo

A-1 by the liver. Archer et al. (1986) reported that treatment of human hepatoma

cell line (HepG2) with estradiol-1713, causes an increase in nuclear estrogen

binding sites, and a parallel increase in the expression of Apo A-1 mRNA and

rate of accumulation of the protein. The increase in mRNA levels accounted for

85-90% of the observed increase in rate of accumulation of secreted protein

(Archer et al., 1986). A study with ovariectomized baboons showed that

baboons treated with estradiol and progesterone had the highest serum

concentrations of Apo A-1, followed by those treated with estrogen alone, and

lowest in the progesterone-treated animals and the untreated controls. Baboons

treated with progesterone alone had similar levels of serum Apo A-1 similar to










those of untreated controls (Kushwaha et al., 1990). Apo A-1 levels were

significantly upregulated by estradiol and progesterone compared to untreated

controls (Kushwaha et al., 1990).

Thyroid hormones also regulate expression of Apo A-1 mRNA. Apo A-1

gene is stimulated by triiodothyronine (T3) and has been shown to contain a

thyroid hormone response element which is critical for the T3-induction of Apo

A-1 mRNA and activity of Apo A-1 promoter (Romney et al., 1992). These

observations were supported by reports from Chan et al. (1993) that the Apo A-1

gene contains a cis-regulatory element which acts on an adjacent site to

increase promoter activity. HNF-4, a new member of the thyroid/steroid

hormone receptor superfamily, was shown to interact with the cis element to

enhance activity of the rat Apo A-1 promoter (Chan et al., 1993).


Effects of Nutrition on Apo A-1 Synthesis


Dietary carbohydrates or fatty acids regulate Apo A-1 gene expression by

altering either gene transcription or mRNA stability. Synthesis and secretion of

Apo A-1 is reduced in hepatocytes from rats fed fish oil (low source of

cholesterol), but the diet did not affect levels of Apo A-1 mRNA (Ribeiro et al.,

1992). Availability of cholesterol has been shown to enhance synthesis of Apo

A-1 by human hepatoma cells (Craig et al., 1988). Similarly, Go et al. (1988)

showed that synthesis of hepatic and intestinal Apo A-1 increases while levels of

Apo A-1 mRNA decrease, following chronic fat and cholesterol feeding.









46

However, hepatic and intestinal synthesis of Apo A-1 is higher in African green

monkeys than in Cynomolgus monkeys fed the same level of cholesterol (Sorci-

Thomas et al., 1988). Thus other factors independent of dietary cholesterol

intake may also regulate hepatic and intestinal Apo A-1 synthesis.


Apolipoprotein E


Biochemical Characterization


Apolipoprotein E (Apo E), sometimes referred to as arginine-rich protein,

is a component of very low density lipoprotein (VLDL), HDL and LDL.

Apolipoprotein E has been shown to have a molecular weight ranging between

33 and 39 kDa on SDS-PAGE (Shelburne and Quarfordt, 1974). Apolipoprotein

E gene is 3597 nucleotides in length and contains four exons and three introns

(Paik et al., 1985), and a similar gene structure is shared by other

apolipoproteins. The primary translation product is a pre-Apo E protein with an

18-residue signal peptide that is cleaved cotranslationally (Zannis et al., 1984),

and the mature protein is secreted.

The primary structure of Apo E ranges in length between 279 and 310

amino acid residues among different species. Apolipoprotein E sequence in the

cow comprises 294 amino acid residues, and the most conserved region is

between residues 28-61 (Yang et al., 1991). The receptor binding region

(residues 130-158) is rich in basic amino acids and is conserved across species,










except for point substitutions in the dog (arginine substituted for lysine at 157)

and cow prolinee substituted for arginine at 145) (Weisgraber, 1994).

Apolipoprotein E has been shown to exhibit heterogeneity in molecular weight

and charge which have been attributed to genetic variation and posttranslational

glycosylation with sialic acid. Sialo-Apo E isoforms comprise 42% of

intracellular Apo E, 81.1% newly-secreted Apo E, and 24% plasma Apo E

(Zannis et al., 1984). Thus sialation may be required for the secretion of Apo E

or glycosylated Apo E is preferentially secreted. There has been evidence to

suggest that Apo E is glycosylated by 0-glycosidic linkage (Zannis et al., 1984).

Liver is the major source of Apo E. However, Apo E and its mRNA is

produced by most organs and by several cell types within the organs including

astrocytes, smooth muscle cells and macrophages (Mahley, 1988).


Effects of FSH, LH, cAMP, and Phorbol Ester on Apo E Synthesis


Apo E is synthesized by the rat ovary and represents 0.15% of the total

protein synthesized in the ovary (Driscoll and Getz, 1984). Secretion of newly-

synthesized Apo E by granulosa cells in culture is stimulated by FSH in a dose-

and time-dependent manner, and the effects of FSH are mediated through cAMP

(Driscoll et al., 1985). Results from that study also suggest that Apo E is

secreted as part of a lipid-protein complex. As the granulosa cells differentiate

in culture, they lose their responsiveness to FSH and cAMP (Driscoll et al.,

1985). Polacek et al. (1992) showed that Apo E mRNA is localized











predominantly in theca cells of rat ovarian follicle, and mRNA levels increase

following treatment of cells with hCG (Polacek et al., 1992).

In another study, Wyne et al. (1989a) demonstrated that BtcAMP and

forskolin (an activator of adenylate cyclase and mediator of kinase A), and

phorbol ester (mediator of kinase C) stimulate production of Apo E by granulosa

cells in culture. BtcGMP (mediator of kinase G) did not stimulate secretion of

Apo E (Wyne et al., 1989a). Kinases A and C had no effect on global protein

synthesis in granulosa cells; incorporation of radiolabel into protein ranged

between 10-15%, suggesting a specific stimulation of a subset of proteins

including Apo E (Wyne et al., 1989a). In addition, cAMP, TPA and cholera toxin

also stimulated expression of Apo E mRNA in rat granulosa cells. These agents

stimulated accumulation of Apo E more than expression of its mRNA, indicating

that kinases A and C may influence both the transcription of Apo E gene and the

translational efficiency of Apo E mRNA. However, it is not yet clear if there is a

crosstalk between the adenylate cyclase pathway (stimulated by cAMP) and the

PKC pathway (stimulated by phorbol ester).

The stimulatory effect of cholera toxin and TPA on Apo E secretion is

inhibited by cycloheximide and actinomycin D, suggesting that new proteins

(such as transcriptional activator proteins AP-1 and AP-2) are required to

mediate the stimulatory effects (Wyne et al., 1989a). The rat Apo E gene does

not contain a cAMP regulatory region (CRE), but contains sequences with 75%

homology to this region. Similarly, the consensus sequence for AP-1 (AP-1









49

responds to phorbol ester) is not present in the upstream region of the rat Apo E

gene, but is found in the first intron. However, the consensus sequence for AP-

2, which responds to both cAMP and phorbol ester, is present in the upstream

region of Apo E gene (Wyne et al., 1989a).


Regulation of Apo E Synthesis by Cell Cholesterol


Cholesterol is the substrate for steroid hormone biosynthesis.

Cholesterol can either be newly synthesized from acetate or is obtained by

uptake of lipoproteins (Schreiber et al., 1980). Wyne et al. (1989b)

demonstrated that inhibition of cholesterol synthesis from acetate with mevinolin,

an inhibitor of HMG-CoA reductase, causes a decline in cholera-stimulated Apo

E synthesis and expression of Apo E mRNA by rat granulosa cells. However, an

inhibitor of the cytochrome P4. side chain cleavage enzyme had no effect on

Apo E synthesis (Wyne et al., 1989b). Human and rat Apo E gene has been

shown to possess the consensus sequence of a sterol regulatory element in

their 5' region. Prack et al. (1991) also reported that depletion of adrenal gland

cholesterol content decreases Apo E mRNA levels. Thus cholesterol together

with stimulators of kinases A and C are required to regulate Apo E production.


Regulation of Apo E Secretion by Cytokines


Macrophage Apo E secretion decreases with macrophage activation

(Zuckerman and O'Neal, 1994). This effect is mediated by macrophage










activating factors such as lipopolysaccharide (LPS) and granulocyte-

macrophage colony stimulating factor (GM-CSF). The LPS-mediated reduction

in Apo E secretion is inhibited by monoclonal antibody to murine tumor necrosis

factor (TNF) (Zuckerman and O'Neal, 1994).


Role of Apo E in Ovarian Function


The role of Apo E in the ovary has not yet been identified. One possibility

is that since Apo E mediates binding of lipoproteins to their receptors, Apo E

could function to provide cholesterol for membrane and steroid hormone

biosynthesis. Apo E may also function in a paracrine fashion to distribute lipid

between ovarian cells and perhaps between compartments of the ovary. It has

been demonstrated that HDL containing Apo E, as opposed to HDL containing

no Apo E, induces rat ovarian theca cells to produce progesterone rather than

androgen (Dyer et al., 1988).


Manganese Superoxide Dismutase


Biochemical Characterization


Superoxide dismutases play critical roles in protecting cells from oxidative

damage by reactive oxygen species. Manganese superoxide dismutase (Mn

SOD) is one of three (the others are Cu SOD and Zn SOD) enzymes that

catalyze the dismutation of superoxide radicals to hydrogen peroxide and










oxygen; 20-2 + 2H = H202 + 02

Manganese SOD is localized in the mitochondrial matrix (Fridovich, 1974)

and is not a secretary protein, while Cu SOD and Zn SOD are secretary proteins

(Rueda et al., 1994). Manganese SOD has a molecular weight of 20 kDa. Two

mRNAs transcripts of 4.0 kb and 1.0 kb encode for human Mn SOD (Melendez

and Baglioni, 1993). However, three Mn SOD mRNA transcripts (1.5, 1.9, and

3.7 kb) have been observed in the bovine CL (Rueda et al., 1995; Ndikum-

Moffor et al. 1995 unpublished data The mRNA transcripts are from the same

gene, have identical coding regions, but differ in length of their 3' untranslated

region (3' UTR) because of polyadenylation (Church, 1990). The 4-kb mRNA is

expressed at a faster rate than the 1-kb mRNA, but the 4-kb transcript has a

shorter half-life (2-4 h in different cells) than the 1-kb transcript (10-12 h) in both

intact cells and a cell-free system (Melendez and Baglioni, 1993). The different

half-lives indicate a post-transcriptional regulation of Mn SOD mRNA, and the

instability of the 4-kb transcript has been attributed to the presence of AU-rich

sequences in the 3' UTR (Melendez and Baglioni, 1993). Manganese SOD

activities are low under normal physiological conditions, but may increase during

differentiation and in response to oxidants and cytokines.


Regulation of Manganese SOD Production by Oxidative Stress


Reactive oxygen species are generated in all cells in vivo, and the toxicity

of oxygen has been shown to be directly related to the production of oxygen-








52

dependent free radicals. Results from a study with yeast indicated that electron

transport is a major source of superoxide anion in vivo (Guidot et al., 1993).

Oxidative stress from the environment has also been shown to increase

production of mitochondrial Mn SOD in plants (Bowler et al., 1991).


Regulation of Manganese SOD Synthesis by Gonadotropins


In the rat, Laloraya et al. (1988) demonstrated a sharp increase in rat

ovarian SOD activity 30 min following an injection of LH, a decline 60 min post-

injection, and no LH-induced SOD activity in rats injected with anti-LH serum.

They also observed changes in ovarian SOD activity across the estrous cycle,

with highest levels at proestrus. However, changes in SOD activity specific to

Mn SOD across the cycle were not discussed (Laloraya et al., 1988). In another

study, Sato et al. (1992) showed that rat ovarian Mn SOD activity decreases

during a hCG-induced ovulation, to a minimum 12 h post-injection, while Mn

SOD mRNA levels increase markedly with time to a maximum 12 h post-hCG

treatment (Sato et al., 1992).


Regulation of Manganese SOD Synthesis by Cytokines and Phorbol Ester


Interleukin-1, TNF, and lipopolysaccharide dramatically increase Mn SOD

mRNA levels in pulmonary epithelial cells (Visner et al., 1990). Similar

observations were reported by White and Tsan (1994) who also showed that

TNF and IL-1 enhance Mn SOD protein and enzyme activity.










Whitsett et al. (1992) showed that TNF-alpha and phorbol ester (TPA)

increase steady state mRNA and rate of transcription of human Mn SOD in

pulmonary adenocarcinoma cells. The time course and extent of increased

manganese SOD gene transcription by TNF-alpha was distinct from that

exhibited by phorbol ester (Whitsett et al., 1992).


Role of Mn SOD in the Ovary


The role of SOD in the ovary has not been defined but there are

indications that it might be involved in ovulation (Laloraya et al., 1988; Sato et

al., 1992) and the luteolytic process (Wu et al., 1992; Rueda et al., 1995). It has

been hypothesized that reactive oxygen species produced during normal

metabolism may be potential mediators of luteal regression. A comparison of

Mn SOD (a scavenger of superoxide radicals) gene expression between a

functional CL (day 21 of pregnancy) and a regressed CL (day 21 of the estrous

cycle) of the cow indicated that Mn SOD mRNA levels are higher in the

functional than the regressed CL (Rueda et al., 1995). A lower expression of Mn

SOD mRNA in the regressed CL suggests that cells within the regressed CL are

less capable of metabolizing the superoxide radical, which may damage the cells

and disrupt luteal function.










Tissue Inhibitors of Metalloproteinases: TIMP-1 and TIMP-2


Tissue inhibitor of metalloproteinases are proteins which inhibit the

activity of enzymes (matrix metalloproteinases) that degrade protein components

of the extracellular matrix. Thus the expression of TIMPs is high in tissues

undergoing remodelling or transformation. Apart from their protease-inhibitory

activity, TIMP-1 (Hayakawa et al., 1992) and TIMP-2 (Stetler-Stevenson et al.,

1992) have also been shown to stimulate growth of erythroid cells, gingival

fibroblasts, and transformed human lung cells. Satoh et al. (1994) also

demonstrated that TIMP-1 stimulates growth of bovine embryos in culture. A

recent study by Boujrad et al. (1995) showed that TIMP-1 secreted by rat Sertoli

cells stimulated steroidogenesis by rat Leydig cells. Activities of matrix

metalloproteinases (collagenases, stromelysins, and gelatinases) may be

controlled at various levels, one of which is by binding to specific inhibitors

(TIMP). Thus a proper balance is required between the amount of inhibitors and

the metalloproteinases to maintain tissue homeostasis or proper remodeling

which occurs during many biological processes. To date, three members of the

TIMP family have been identified, namely TIMP-1, TIMP-2, and TIMP-3.

Comparison of deduced amino acid sequence of TIMP-2 showed that TIMP-1

and TIMP-2 share 37.6% identity of nucleotide and 65.6% similarity of amino

acid at the protein level (Stetler-Stevenson et al., 1990), while human TIMP-3

shares 39 and 46% amino acid sequence identity with human TIMP-1 and TIMP-










2, respectively (Silbiger et al., 1994). The positions of all twelve cysteine

residues and three out of four tryptophans are conserved between TIMP-1 and

TIMP-2.


Tissue Inhibitor of Metalloproteinases-1


Biochemical Characterization


TIMP-1 is a secreted glycosylated protein with molecular weight ranging

from 28-30 kDa. TIMP-1 is expressed in many different tissues and cell types

including monocytes, fibroblasts and macrophages. Ovine TIMP-1 shares 95,

86, and 77% nucleotide sequence with that reported for bovine, human, and

mouse TIMP-1, respectively (Smith et al., 1994). The amino acid sequence of

TIMP-1 deduced from the bovine cDNA sequence shows that the mature protein

contains 12 cysteine residues (conserved among many species) and 2 N-

glycosylation sites (Freudenstein et al., 1990). The nucleotide sequence of

ovine TIMP-1 also indicates the presence of 12 cysteines and 2 N-linked

glycosylation sites (Smith et al., 1994). In addition, TIMP-1 contains a 23-amino

acid signal peptide which contains a core of hydrophobic amino acids (Smith et

al., 1994).

TIMP-1 binds to active collagenase, and to the latent form of the 92 kDa

gelatinase, and proteoglycans. TIMP-1 binds collagenase with high affinity in a

1:1 molar ratio to form an inactive noncovalent enzyme-inhibitor complex










(Welgus et al., 1985).


Regulation of TIMP-1 Synthesis by Gonadotropins


Mann et al. (1991) showed that LH and phorbol ester (TPA) individually

increased metalloproteinase inhibitor activity of granulosa cells in culture in a

dose-dependent manner, and the effects were additive. The inhibitor activity

(identified as TIMP-1) was also stimulated by Br-cAMP and forskolin, and its

mRNA levels increased before ovulation (Mann et al., 1991). In another study

Mann et al. (1993) showed that cycloheximide inhibits basal, LH- and TPA-

stimulated TIMP-1 activity, while indomethacin (an inhibitor of prostaglandin

synthesis) or an antiestrogen did not affect basal or LH-induced rat granulosa

cell inhibitory (TIMP-1) activity. Reich et al. (1991) also reported a lack of effect

of eicosanoid on ovarian expression of TIMP-1. Rat granulosa cell TIMP-1

mRNA is also increased by LH and hCG, but the induced mRNA expression is

not affected by cycloheximide (Mann et al., 1993). Thus de novo protein

synthesis is required for LH- and TPA-induced increase in granulosa cell TIMP-1

activity but protein synthesis is not necessary for stimulation of TIMP-1 mRNA

expression.

Luteal synthesis of TIMP-1 has been reported in the ewe (Smith et al.,

1993; 1994), cow (Freudenstein et al., 1990; Ndikum-Moffor et al., 1995), rat

(Mann et al., 1991), mouse (Edwards et al., 1992), and ferret (Huang et al.,

1993). Results from some of these studies indicate that luteal synthesis of









57

TIMP-1 is triggered and stimulated by the surge in LH. Gonadotropins have also

been shown to regulate expression of TIMP-1 mRNA in Sertoli cells. Treatment

of prepubertal rat Sertoli cells with FSH and 8-bromo cAMP increases activity of

TIMP-1, amount of TIMP-1 protein in conditioned-medium, and expression of

TIMP-1 mRNA (Ulisse et al., 1994). Similar to observations in rat granulosa

cells (Mann et al., 1991), de novo protein synthesis and RNA synthesis are

required for both basal and TPA-, 8-bromo cAMP-, and FSH-stimulated TIMP-1

activity (Ulisse et al., 1994). The effects of phorbol esters on gene transcription

occur through fos and jun containing AP-1 transactivating factors and the latter

is induced by PKC-activating stimuli (Lee et al., 1987). On the other hand,

cAMP enhances gene transcription by stimulating PKA which stimulates

phosphorylation of cAMP response element binding proteins (CREB) (Sassone-

Corsi et al., 1988; Merino et al., 1989).

The mechanisms) through which TIMP-1 synthesis is stimulated has not

been fully characterized. The murine TIMP-1 gene contains cis-acting

regulatory elements upstream of the major transcription start site and also

contains an AP-1 binding site within the cis-acting region (Edwards et al., 1992).

The AP-1 functions as binding site for fos-jun and can stimulate transcription.

Fos and Jun, like the CREB proteins are members of an extended basic region-

leucine zipper (bZIP) superfamily of transcription factors. It has been shown that

oligonucleotides containing a CREB sequence compete for binding of proteins to

TIMP-1 AP-1 site. Thus the CREBP family may be involved in specific binding to









58

TIMP-1 AP-1 site (Edwards et al., 1992). Expression of TIMP-1 is stimulated by

factors that increase intracellular cAMP. Thus TIMP-1 AP-1 site is the cis-acting

regulatory element that mediates the cAMP-induced increase in TIMP-1 gene

expression (Edwards et al., 1992). TIMP-1 enhancer element does not contain a

classical CRE binding site but contains functional AP-1 sites, one of which can

bind fos and jun heterodimers and other transacting factors including the CREB

family (Edwards et al., 1992).

Effects of TPA on TIMP-1 synthesis are mediated through the PKC

second messenger pathway since TPA-stimulated TIMP-1 activity is inhibited by

an inhibitor of PKC (Staurosporine), and TIMP-1 activity is not stimulated by a

non-PKC activating phorbol ester (Mackay et al., 1992).


Regulation of TIMP-1 Synthesis by Steroid hormones


Rajabi et al. (1991 a) demonstrated that estradiol-17f1 stimulates

degradation of collagen type 1 in nonpregnant guinea pig cervix in vitro. In

addition, the cervix has been shown to produce collagenase and its synthesis is

stimulated by estrogens, interleukin-1B, and PGE2 (Rajabi et al., 1991b). Rajabi

et al. (1991c) furthermore showed that activities of collagenase and collagenase

inhibitor are greater in cervical tissue at the time of parturition than in tissues

from nonpregnant animals. The marked increase in inhibitor activity observed at

a time when collagenase activity is increased indicates the presence of a strong

regulatory mechanism to control the extent of collagen degradation beyond the










level required for parturition (Rajabi et al., 1991c).

Progesterone stimulates TIMP-1 production by rabbit uterine cervical

fibroblasts (Imada et al., 1994). Similar observations were reported by Sato et

al. (1991) who showed that progesterone and estradiol-1 7B increases secretion

of TIMP-1 by rabbit uterine cervical fibroblasts in culture and steady state TIMP-

1 mRNA. However, observations by Rajabi et al. (1991c) showed that estradiol

causes a decrease in tissue collagenase activity.

Retinoic acid has been shown to enhance secretion of TIMP-1 by human

fibroblasts in vitro by increasing de novo synthesis of TIMP-1 (Clark et al.,

1987). Retinoic acid also increased TIMP-1 mRNA levels compared to

nontreated controls. Glucocorticoid treatment had no effect on TIMP-1 secretion

(Clark et al., 1987).


Regulation of TIMP-1 Synthesis by Cytokines and Growth Factors


Synthesis of TIMP-1 and collagenase by human fibroblasts is stimulated

by phorbol ester and IL-1 (Murphy et al., 1985). Similar effects of IL-1 were

reported by Rajabi et al. (1991c) for cervical tissue of guinea pig. TIMP-1

activity and expression of its mRNA have been shown to increase in a variety of

normal and tumor cell lines following treatment with IL-1 and tumor necrosis

factor (TNF) (Mackay et al., 1992). Transforming growth factor-Il (TGF-B) has

also been implicated as a regulator of TIMP-1 synthesis. One of the

mechanisms proposed for the control of normal trophoblast proliferation and










invasiveness by TGF-B is via induction of TIMP-1 mRNA expression. This

regulatory mechanism is absent in malignant trophoblast cells (Graham et al.,

1994).

Tissue Inhibitor of Metalloproteinases-2


Biochemical Characterization


Tissue inhibitor of metalloproteinases-2, the second member of the family

of metalloproteinase inhibitors, binds and inactivates all matrix

metalloproteinases but in contrast to TIMP-1 which binds the 92-kDa gelatinase,

TIMP-2 binds the 72-kDa gelatinase. TIMP-2 has a molecular weight of 20-21

kDa and in contrast to TIMP-1, TIMP-2 is not glycosylated. In addition to its

protease-inhibiting activity, TIMP-2 also possesses erythroid potentiating activity

(Stetler-Stevenson et al., 1992), and has been shown to stimulate proliferation of

skin fibroblast cells by stimulating cAMP and activating cAMP-dependent

adenylate cyclase (Corcoran and Stetler-Stevenson, 1995).

Like TIMP-1, TIMP-2 contains 12 conserved cysteine and three

tryptophan residues. The rat TIMP-2 gene encodes a 220 amino acid -long pro-

TIMP-2 protein containing a 26-residue hydrophobic leader sequence, and a

mature 194 amino acid protein (Santoro et al., 1994). TIMP-2 is expressed in a

variety of cells and tissues, and it possesses two mRNA transcripts with

approximate sizes of 3.5 and 1.0 kb (De Clerk et al., 1994).

There is evidence to suggest that although TIMP-1 and TIMP-2 possess









61

similar physiological properties, expression of their activity and mRNA seem to

be differentially regulated in mouse reproductive tissues (Waterhouse et al.,

1993). In contrast to the stimulating effects of phorbol ester on murine TIMP-1

mRNA, phorbol ester does not affect expression of TIMP-2 mRNA (De Clerk et

al., 1994). In another study Waterhouse et al. (1993) reported a differential

expression in TIMP-1 mRNA and TIMP-2 mRNA in the ovary of mice during

gestation; TIMP-1 mRNA is low while TIMP-2 mRNA shows a marginal increase.

Leco et al. (1992) also reported that while TIMP-1 mRNA is highly serum-

inducible in normal murine fibroblasts, expression of TIMP-2 mRNA is mainly

constitutive and is insensitive to transformation while expression of TIMP-1

mRNA is variable (Leco et al., 1992).


Regulation of TIMP-2 Synthesis by Gonadotropins


Like TIMP-1, TIMP-2 activity, TIMP-2 protein and mRNA levels in rat

Sertoli cells are stimulated by FSH through a cAMP-dependent pathway (Ulisse

et al., 1994).


Regulation of TIMP-2 Synthesis by Steroid Hormones


Production of TIMP-2 in culture by rabbit uterine cervical fibroblasts

increases after treatment with physiological concentrations of progesterone

(Imada et al., 1994). TIMP-2 mRNA is expressed constitutively in rat

hepatocytes and its expression is up-regulated following incubation of










hepatocytes with dexamethasone and prostaglandin E2 (Roeb et al., 1995).


Regulation of TIMP-2 Synthesis by Cytokines and Growth Factors


Transforming growth factor-IB has been shown to down-regulate both

mRNA transcripts of TIMP-2 in contrast to its stimulatory effect on TIMP-1 mRNA

expression (Stetler-Stevenson et al., 1990). Mackay et al. (1992) also reported

that TIMP-2 activities are refractory to TPA, IL-1 and TNF-a, in contrast to the

marked stimulation of TIMP-1 activities by all three agents in a variety of human

cell lines. A similar lack of stimulatory effect of TPA on TIMP-2 activity is also

observed in rat Sertoli cells in culture (Ulisse et al., 1994).












Table 1-1. Summary of Factors Regulating Synthesis of Proteins.


Regulation/ Apo A-1 Apo E Mn SOD TIMP-1 TIMP-2

Function
E2, T3, FSH, LH, LH, hCG, LH, cAMP FSH,

HNF-4, cAMP, hCG, TPA, IL-1, TPA, E2, P4 cAMP, P4,

Gene Dietary CT, TPA, TNF, LPS IL-1 PGE2, PGE2,

Expression Cholesterol Cholesterol TNF, TGF-B

TGF- 8,

Vit A
Translation Dietary cAMP, CT, TNF, IL-1 IL-1, TNF, FSH, cAMP

Cholesterol, TPA, TNF, Vit A, P4, E2

E, LPS,

GM-CSF,
Function Cholesterol Cholesterol Prevents Tissue Tissue

metabolism, metabolism, oxidative remodelling, remodelling,

cAMP membrane and stress cell growth, cell growth

stimulation, steroid Steroid

Steroid biosynthesis biosynthesis

biosynthesis














CHAPTER 3
PROTEINS SYNTHESIZED AND RELEASED IN CULTURE BY THE BOVINE
CORPUS LUTEUM: THE ESTROUS CYCLE AND PREGNANCY


Introduction


Apart from its traditional role in progesterone synthesis and maintenance

of pregnancy, the corpus luteum synthesizes and secretes a number of proteins

during the estrous cycle and pregnancy. The corpus luteum of the cow and ewe

synthesizes the peptides oxytocin and neurophysin during the estrous cycle and

stores these proteins in secretary granules of the large luteal cells (Fields et al.,

1986; 1992) and ewe (Fields et al., 1986; Theodosis et al., 1986). The synthesis

and secretion of proteins vary with the physiological status of the animal. For

example, the number of oxytocin-containing secretary granules in the cow

increases from metestrus to diestrus, and then declines prior to luteolysis (Fields

et al., 1992), whereas in pregnancy the population of granules was undetectable

on day 45, then increased to a peak between days 180-210 (Fields et al., 1985).

However, the corpus luteum of mid-pregnancy (after day 45) in the cow contains

neither the mRNA for oxytocin (Ivell et al., 1985) nor do the secretary granules

contain oxytocin (Fields et al., 1992).

Additional proteins identified as secreted by the corpus luteum include the










tissue inhibitor of metalloproteinases-1 and -2 (TIMP-1 and TIMP-2) in the rat

(Parmer et al., 1992), sheep (Smith and Moor, 1991), (Smith et al., 1995; Smith

et al., 1993; 1994 ), cattle (Freudenstein et al., 1990; Juengel et al., 1994), pig

(Smith et al., 1994)) and ferret (Huang et al., 1993), relaxin in humans and non-

ruminants (Sherwood, 1994), inhibin in sheep (Tsonis et al., 1988; Rodgers et

al., 1989; Smith et al., 1991), and insulin-like growth factor-1 (Einspanier et al.,

1990), basic fibroblast growth factor (Stirling et al., 1991) and angiogenic factors

(Redmer et al., 1988; Grazul-Bilska et al., 1992) in cattle.

In addition to its role in maintenance of pregnancy, the corpus luteum of

pregnancy in the cow appears to be necessary for normal parturition since

removal during the third trimester resulted in increased rates of dystocia,

retained fetal membranes (Estergreen, 1967), and greater death loss of calves

(Tanabe, 1966). The synthesis and secretion of proteins by the bovine CL of

pregnancy may play a role in setting the stage for parturition. The ferret corpus

luteum was shown to secrete proteins on days 5-11 of pregnancy, with molecular

masses of 16 to 185 kDa (Huang et al., 1993). Although no qualitative

difference in protein secretion was observed across days of pregnancy studied,

a 32 kDa protein that cross-reacted weakly with a polyclonal antibody to human

TIMP was the most abundantly secreted protein (Huang et al., 1993).

The objectives of this study were to examine for proteins synthesized de

novo by the bovine CL, identify and characterize the newly-synthesized proteins,

and determine quantitative differences in their synthesis and release during the










estrous cycle and pregnancy.


Materials and Methods


Materials


Acrylamide was purchased from ICN Biomedicals Inc. (Cleveland, OH),

bis-acrylamide and agarose from Bio-Rad Laboratories, N,N,N,N,-

tetramethylethylenediamine (TEMED) from Fisher Scientific (Fair Lawn, NJ), and

ampholines from Pharmacia (Piscataway, NJ). Other electrophoretic reagents

were obtained from Bio-Rad Laboratories (Richmond, CA). L-4,5-'H-leucine

(specific activity 164 Ci/mmol) and D-[6-3H]glucosamine (specific activity

20Ci/mmol) were purchased from Amersham (Arlington Heights, IL), and 3S-

methionine (specific activity 1028 Ci/mmol) was purchased from ICN

Biomedicals Inc. Polyvinylidene fluoride (PVDF) was obtained from Millipore

Corporation (Bedford, MA). Tissue culture media including amino acids,

vitamins, insulin and antibiotic-antimycotic mixture, and all other chemicals were

purchased from Sigma Chemical Company (St. Louis, MO).


Collection of Luteal Tissue


Forty-eight Angus and Hereford crossbred beef cows were used for the

study. All procedures in which animals were used were approved by the Animal

Care and Use Committee of the University of Florida. Estrus (day 0) was










defined as that day when a cow would stand to be mounted by a bull. Cows

randomly assigned to the pregnancy group were artificially inseminated at

observed estrus, whereas cows assigned to the cycle group were not bred. Day

17 of pregnancy was confirmed by the presence of an embryo in flushings from

the uterus. Later stages of pregnancy were estimated by measurement of

crown-rump length of the fetus (Winters et al., 1942). Reproductive tracts were

obtained from cows within 5 min after exsanguination at the University of Florida

abattoir. The ovary containing the corpus luteum was collected aseptically from

cows on days 3 (n = 4 cows), 7 (n = 3), 11 (n = 4), 14 (n = 5), 17 (n = 3), and 19

(n = 3) of the estrous cycle, and from cows of early pregnancy (day 17, n = 5),

and the first (day 88, n = 5), second (day 170, n = 7), and third (greater than day

240, n = 9) trimester of pregnancy. Ovaries were immediately transferred to a

sterile Petri dish (100 x 15 mm) containing pre-warmed Eagle's Minimum

Essential Medium (MEM), and the corpus luteum dissected from the ovarian

stroma and weighed.


Culture Medium


Medium was prepared as previously described (Basha et al., 1980) from

Eagle's MEM deficient in leucine, lysine, methionine and sodium bicarbonate.

One liter of stock incomplete MEM was prepared with the addition of glucose (3

g/I), methionine (1.5 mg/1), leucine (5.2 mg/1) and lysine-HCI (72.5 mg/I) to

achieve 4.0, 0.1, 0.1, and 1.0 times, respectively, their usual concentrations in










MEM. Sodium bicarbonate (2.2 g/I), non-essential amino acids (1 %, v/v),

vitamins (1%, v/v) and insulin (200 IU/ll) were added and pH adjusted to 7.1-7.3.

The medium was filter-sterilized (0.22 pm) (Corning Inc, Corning, NY) and stored

at 4 C. For 3H-leucine labelled cultures, methionine (1.5 mg/100 ml) and

antibiotic-antimycotic (ABAM) mixture (1%, v/v) were added to the stock

incomplete MEM to obtain leucine-deficient incomplete modified MEM. Similarly,

for 3S-methionine labelled cultures, methionine-deficient medium was prepared

by adding leucine (5.2 mg/100 ml) and ABAM to the stock incomplete MEM.

Incomplete MEM was used for 3H-glucosamine labelled cultures.


Time Course Studies of Incorporation of Radiolabel


CL from three pregnant cows (two on day 170, and one on day 88) were

used in the time course studies. Slices (0.5 mm in thickness) of luteal tissue

were prepared with a Stadie-Rigg's tissue slicer (Thomas Scientific,

Swedesboro, NJ). Explants were washed three times, each time with 15 ml

incomplete MEM, to reduce serum proteins in the medium during incubation.

After washing, luteal tissue (500 mg/dish) from the same corpus luteum was

placed in four Petri dishes each containing 15 ml of leucine-deficient incomplete

MEM without radiolabel. The dishes were pre-incubated at 37 C on a rocker

platform for 2 h in an atmosphere of 50% N2:47.5% 02:2.5% CO2 (v/v/v). After

pre-incubation, the medium was discarded and replaced with 15 ml leucine-

deficient incomplete MEM containing 50 pCi of 3H-leucine (160 Ci/mmol). Each










of the four Petri dishes of luteal tissue from each CL was then incubated as

described for 6, 18, 24, and 30 h, respectively (ie n = 3 for each time point). At

the end of each incubation, the luteal-conditioned medium (LCM) was separated

from tissue by centrifugation at 2000 x g for 20 min at room temperature.

Medium was dialyzed immediately for 24 h at 4 C using Spectra/por 3 membrane

(molecular weight cut off = 3500; Spectrum Medical Industries Inc., Houston,

TX) against two changes (24 h each) of 4 I Tris-HCI buffer (10 mM, pH 8.2), and

then against deionized water (two changes, 24 h each). Following dialysis,

percent incorporation of radiolabel in the dialyzed LCM was determined for each

incubation time. Percent incorporation was calculated as: post-dialysis

radioactivity (dpm) divided by pre-incubation radioactivity (dpm) x 100%.


Culture and Radiolabelling of Luteal Tissue


The 24 h incubation was determined as the optimal time of incubation

following incubations at 6, 18, 24, and 30 h as described. For each cow at least

two Petri dishes of luteal slices (500 mg tissue/dish) were incubated as

described in the above protocol for 24 h. Following incubation and dialysis, the

total volume of dialyzed LCM (retentate) for each Petri dish was measured and

adjusted to 15 ml with deionized water. Retentates were stored at -20 C until

further analysis. Luteal slices (500 mg) were also incubated with 3H-

glucosamine (50 pCi/15 ml, n = 1 cow, day 240 pregnant) or 3sS-methionine (50

pCi/15 ml, n = 1 cow, day 240 pregnant) to determine if newly-synthesized and










released proteins were glycosylated and/or contained methionine.


TCA precipitation


Following LCM dialysis, samples of dialyzed medium (retentate) for each

corpus luteum were analyzed to determine the amount of 'H-leucine

incorporated into trichloroacetic acid (TCA)-precipitable protein using a

modification of method described by Mans and Novelli (1961). Briefly, a 50 pl

aliquot of LCM retentate was spotted onto a 2.54 cm x 2.54 cm Whatman 3MM

filter paper that was pre-soaked in 20% (w/v) TCA. Each square was air-dried,

soaked in 20% and 5% TCA for 10 and 20 min, respectively, and subsequently

washed twice, for 15 min each, in 95% ethanol. Squares were allowed to air-dry

completely, placed in scintillation cocktail and counted for radioactivity. TCA-

precipitable protein (dpm) was measured in duplicate for each LCM retentate

sample, and values were expressed as least squares means (LSM) for each

group.


Light and Electron Microscopy


To assess the effects of the 24 h-incubation on tissue morphology, luteal

tissues collected before and after 24 h of culture with radiolabel were processed

for evaluation by electron microscopy (Fields et al., 1992). Briefly, the central

part of each CL was dissected into 1-3 mm cubes and fixed in 1% (v/v)

glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4). Cubes were








71
postfixed in 1% (v/v) osmium tetroxide in 0.1 M sodium cacodylate buffer for 30

min, embedded individually in Spurr's medium, and sectioned (0.1 pm).

Sections were observed and photographed using a Philips electron microscope

(Model 301, Philips Electronic Instruments, Mahwah, NJ). For evaluation by

light microscopy, luteal tissue was fixed in Bouins, embedded in paraffin, and

sections were cut and stained with haematoxylin and eosin (Sheehan and

Hrapchak, 1980).


Two-Dimensional-SDS- Polyacrvlamide Gel Electrophoresis


First-dimension: Isoelectric Focusing. Proteins, synthesized and released into

the medium of explant culture, were separated according to their isoelectric

points (pl) by isoelectric focusing (IEF) as previously described (Laemmli, 1970).

Frozen retentates were lyophilized and reconstituted in IEF gel sample buffer [9

M urea, 2% (v/v) NP-40, and 2% (v/v) ampholine (pH 3.5-10)]. Each sample

(100,000 cpm) was loaded and the proteins separated in the IEF tube gel (4%

(w/v) T, 5.4% (w/w) Cbis) by electrophoresis at 400 V for 20 h.

Second-dimension: SDS-PAGE. Following first dimension electrophoresis, the

tube gels (100,000 cpm per gel) were equilibrated in gel-equilibration buffer

[0.0625 M Tris, 5% (w/v) sodium dodecylsulfate (SDS), 10% (v/v) glycerol], for

15 min before they were loaded onto a slab SDS-polyacrylamide gel (14 cm x 16

cm x 0.15 cm; stacking gel 4% (w/v) T, 2.7% (w/w) Cbs; separating gel 10% (w/v)

T, 2.7% (w/w) Cs,). Proteins in the IEF gel were separated on a slab gel by









72
electrophoresis in tank buffer (0.025 M Tris, 0.192 M glycine, 0.1% SDS, pH 8.3)

at 13 mA/slab gel until the dye front reached the end of the gel (O'Farrel, 1975).

Following electrophoresis, gels were stained with 0.1% (w/v) Coomassie blue,

destined in destaining solution [50% (v/v) ethanol, 10% (v/v) acetic acid] and

soaked in deionized water for 30 min. Gels were then treated with 1 M sodium

salicylate for 30 min and dried on a slab gel dryer (Model SE 1150, Hoefer

Scientific Instruments, San Francisco, CA), followed by exposure to x-ray film (X-

OMAT-AR, Eastman Kodak Company, Rochester, NY) for 4 weeks. Intensities

of the radioactive spots on fluorographs were determined by densitometric

scanning (E-C Apparatus Corporation, St. Petersburg, FL). The scanner was

standardized to detect intensities between the lightest (background) and the

darkest spot on fluorographs. The area under the curve for each spot scanned

was measured using the trace mode of an electronic planimeter (Model 1250,

Numonics Corporation, Lansdale, PA). Area measurements in cm2 represent

relative amounts of each newly-synthesized protein.


Protein Blotting and Amino Acid Sequencing


Following detection of newly synthesized proteins, proteins on wet 2-D

gels were transferred to polyvinylidene fluoride (PVDF) membranes (Millipore

Corporation) by electroblotting at a constant voltage (20 V) in transfer buffer (2-

[N-Morpholino]ethanesulfonic acid, 10 mM MES, pH 6, Sigma Chemical Co.) at 4

C for 16 h, according to method of Towbin et al. (1979). Membranes were









73

stained in 0.1% (w/v) Coomassie blue in 50% methanol for 5 min, destined in a

solution of 50% ethanol and 10% acetic acid, rinsed extensively in distilled

water, and then air dried. Proteins on membranes were subjected to N-terminal

amino acid sequence analysis (Protein Sequencer Model 470 A/B, Applied

Biosystems, Foster City, CA) by the Protein Chemistry Core Laboratory of the

Interdisciplinary Center for Biotechnological Research (ICBR) at the University

of Florida. A search of protein, RNA, and DNA data banks was conducted using

the National Center for Biotechnology Information Database Search program

(Devereux et al., 1984).


Progesterone Assay


Trunk blood samples collected into heparinized vacutainers (Becton

Dickinson Vacutainer Systems, Rutherford, NJ) from cows at time of slaughter

were processed, and all plasma samples were analyzed for progesterone by RIA

in a single assay, using the progesterone Coat-A-Count kit (Diagnostic Products

Corporation, Los Angeles, CA). One hundred microliter (100 pl) plasma was

assayed per tube. The kit originally designed for human serum was validated for

use with cow plasma. Progesterone standards were prepared from a stock

solution (5 pg progesterone/ml benzene) by diluting with ovariectomized cow

plasma. A quantitative linear recovery of progesterone was obtained with two

replicates of 0, 7.8, 15.6, 31.2, 62.5, 125, 250, 500, and 1000 pg progesterone/

100 pl ovariectomized cow plasma [Y = 1.595 + 0.864X; Y = amount of










progesterone measured (pg/ml) and X = amount of progesterone added; R =

.99]. The detection limit of the assay, determined as two standard deviations

above the zero dose level, was 9.5 pg/ml. The intra-assay coefficient of

variation for the one assay was 6.5%. The slope (Mean SEM) of the standard

curve using the progesterone standards provided in the kit (0, 10, 50, 200, 1000,

2000, 4000 pg/100 pl) was -0.68 0.13 and the estimated doses at 20, 50, and

80% binding were 10.03, 1.31, and 0.18 ng/ml, respectively.


Statistical Analysis


Data for percent incorporation, radiolabeled TCA precipitable protein, CL

weights and plasma progesterone were analyzed by least squares analysis of

variance using the General Linear Models procedure of the Statistical Analytical

System (SAS, 1988). Data from densitometric scanning were subjected to

analysis of variance to determine differences in protein secretion. The statistical

model had day and reproductive status as the main effects with residuals as the

error term. Values in the text are least square means (LSM) standard error of

the mean (SEM). Orthogonal contrasts were performed to determine differences

among days of the cycle and stages of pregnancy.










Results


Histology of Luteal Tissue


Results of microscopic examination of luteal tissue before and after 24 h-

incubation showed there was no degeneration of the tissue as a result of

incubation; there was no evidence of presence of collagen fibers or loss of

secretary granules (Fig. 3-1). There was presence of secretary granules and

intact mitochondria.


CL Weight and Plasma Progesterone


As expected, weight of the corpus luteum was different (P < 0.004) across

days of the estrous cycle, but not across stages of pregnancy (Fig. 3-2).

Corpora lutea weights were lowest (P < 0.002) on day 3 when compared to the

other days of the estrous cycle examined. Weight of the corpus luteum was

lower (P < 0.03) on day 11 than 14. Plasma progesterone had a similar pattern

as the corpora lutea weight, with concentrations varying (P < 0.006) across days

of the estrous cycle, but not across stages of pregnancy (Fig. 3-2). Plasma

progesterone was lower (P < 0.03) on day 3 when compared to the rest of the

estrous cycle.








Figure 3-1. Electron micrograph of luteal cells.
A) Top panel: Luteal cell of tissue prior to incubation (control). Note
secretary granules (arrow), nucleus at the bottom, and intact mitochondria.
x 9,000; #50627.
B) Middle panel: Luteal cell of tissue post-24 h incubation without
radiolabel (control). Note mature secretary granules (arrow), and nascent
secretary granules associated with the Golgi apparatus, and intact mitochondria.
x 11,400; #50633.
C) Bottom panel: Luteal cell of tissue post-24 h incubation with 50 pCi
3H-leucine. Note secretary granules (arrow) and nucleus at the bottom. Intact
mitochondria is not different from that of the controls. x 11,400; 50630.








~~ed


41


Sr -


.,I;;
V


%F 7
*'67 r4N


M--


A ril l.
." 'J.

Milap































2 4 6 8 10 12 14 16 18
Day


Figure 3-2.


Least square means SEM of plasma concentrations
of progesterone (ng/ml) and weight (g) of the corpus luteum
across days 3, 7, 11, 14, 17, and 19 of the estrous cycle and
days 17, 88, 180, and >240 of pregnancy. Progesterone
concentrations and CL weight increased from day 3 to day
14 of the estrous cycle and then declined to day 19, but
these did not vary across pregnancy.


-8

6


4

2#I
C-,

0

300








79

Plasma progesterone was lower (P < 0.03) during early luteal (day 7) than mid-

luteal phase (days 11 and 14) of the estrous cycle. Similarly, plasma P4 was

lower (P < 0.09) on day 11 than day 14, and day 19 was lower than day 17 (P <

0.05). However, there were no differences between day 17 of the estrous cycle

and day 17 of pregnancy for plasma progesterone (4.79 1.1 vs 5.36 1.3, P <

0.76) and CL weight (4.66 0.7 vs 4.83 0.6, P < 0.87).


Incorporation of Radiolabel into TCA-precipitable Protein


Results of the time course study (n = 3 cows) showed that the 24 h-

incubation time was optimum for incorporation of radiolabel when compared with

the 6, 18, and 30 h incubation time (Fig. 3-3), thus 24 h-incubation time was

used in this study. There was no lag time, the rate of incorporation increased

sharply between 18 h and 24 h of incubation, and did not change or declined

between 24 and 30 h. Analysis of all LCM retentates following dialysis indicated

that there was no difference in the amount of radioactivity associated with TCA-

precipitable proteins released in culture across days of the estrous cycle or

stages of pregnancy. The amount (LSM SEM) of TCA-precipitable

radioactivity was not different (P < 0.23) between day 17 (3104.6 500.5 dpm)

of the estrous cycle and day 17 (2217.6 433.5 dpm) of pregnancy. Percent









80





3000 -
---- 91-0126
2500 -- 91-0036
E 87-0071 /
v 2000 -
S/
150 1S0 -

S1000

S500 .


0 6 12 18 24 30 36
Incubation Time (h)




Figure 3-3. Time course studies of incorporation of radiolabel
(n = 3 cows).


incorporation of radiolabel across the estrous cycle ranged between 2.4 0.8%

(day 11) and 6.2 0.9% (day 19) and approached significance (P < 0.09). The

percent incorporation ranged between 2.0 0.6% and 3.4 0.6% during

pregnancy, did not differ across pregnancy, and was not different (P < 0.35)

between day 17 of the estrous cycle and day 17 of pregnancy.





















C
0

0
0.
0
U


.0
(U
.2


121 Cycle

10. Pregnancy


3 7 11 14 17 19 17 88 170 >240
Day of Cycle/Pregnancy


Figure 3-4. Percent incorporation of radiolabel into newly-synthesized proteins.










Luteal Protein Synthesis and N-terminal Amino Acid Micro Sequencing



The bovine CL synthesized and released a number of different proteins

during the estrous cycle and pregnancy with molecular masses ranging from 12

to 200 kDa. Following 2D-SDS-PAGE and fluorography, eleven discrete

radiolabeled proteins were selected for further study. For convenience, before

the amino acid sequence information was available, each of the eleven major

proteins was assigned a number; proteins 1 (35 kDa, pi 5.5), 2 (30 kDa, pl 5.5),

3 (29 kDa, pl 5.5), 4 (27 kDa, pl 5.5), 5 (70 kDa, pl 5.0), 6 (58 kDa, pi 6.0), 7 (44

kDa, pl 5.0), 8 (30 kDa, pl 8.0), 9 (20 kDa, pl 8.0), 10 (22 kDa, pl 8.0), and 11

(27 kDa, pl 6.0) (Fig. 3-5, 3-6). Five of these proteins were identified from their

N-terminal amino acid sequence. Between 18 and 31 N-terminal amino acid

residues for proteins 1, 8, 9, 10 and 11 were determined following separation by

2D-SDS-PAGE and transfer to PVDF membranes (Table 3-1). Protein 1,

composed of four protein spots, and proteins 8 and 11, each composed of two

protein spots were individually subjected to N-terminal micro-sequence analysis.

A search of protein data banks matched these N-terminal amino acid sequences

to those of bovine apolipoprotein E (Apo E; all four protein spots associated with

protein 1),















pl
4----------9


1
C2
*4-4


D14
e


[j,, t-
19",


Figure 3-5.


D17
5
7 60


D19


I [-' 1--f
1-_19,10


Representative fluorographs of proteins synthesized de novo in
explant culture and released into the medium by CL on different
days of the estrous cycle. Proteins 1, 2, 3, and 4 were observed
only on day 3. Other identified proteins 5-10 were found on all
other days (7-19) and protein 11 was found on days 11-19.


7 L


T'
xi


DII





6.1&0


I. -f1


[ ]9,10




Full Text

PAGE 1

IDENTIFICATION AND CHARACTERIZATION OF PROTEINS SECRETED BY THE CORPUS LUTEUM OF THE COW DURING THE ESTROUS CYCLE AND PREGNANCY By FLORENCE MABOH NDIKUM-MOFFOR A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1995 1 I UNIVERSITY OF aORlDA UBRARIES

PAGE 2

Dedicated to Mandi, Koga, Kongwenebime, Gaston, and my mother and late father.

PAGE 3

ACKNOWLEDGMENTS I would like to thank my major professor, Dr. Michael J. Fields, for his constant guidance throughout the course of my studies. His patience, positive attitude, and constant encouragement, gave me the strength to withstand the difficult and trying moments of this learning process. Dr. Fields did not only supervise my graduate training and research, but also gave me informal lectures about the philosophies of life and the importance to strive for excellence. I am grateful for his leadership and the kindness his family showed towards mine. I also wish to thank Dr. Rosalia CM. Simmen for giving me the opportunity to work and carry out the second part of my research work in her laboratory. The time I spent in her laboratory was very resourceful and reaffirmed the importance for thoroughness and detail, which I appreciate. I wish to thank the other members of my supervisory committee. Dr. William Buhi, Dr. Peter J. Hansen, and Dr. Lynn Larkin for their readiness to discuss my research and give helpful suggestions. I wish to specifically thank Dr. William Buhi for his advice with the first part of my research, and Dr. Peter Hansen for letting me use his laboratory equipment. Dr. Phillip A. Fields is acknowledged for the important contributions to the histological component of this work. Many thanks to Dr. Michael F. Smith and Dr. Harry Nick for providing

PAGE 4

some of the cDNA probes used in this work. I am grateful to Dr. John T. Banser, the Director of the Institute of Animal and Veterinary Research, IRZ, for providing initial financial support, and the Animal Science Department, University of Florida for the Graduate Assistantship. I also wish to acknowledge Dr. Daniel A. Mbah for encouraging me to pursue graduate training at the University of Florida. My sincere thanks go to Dr. Shou-Mei Chang for the technical and moral support she gave me throughout my stay in the laboratory. My thanks go to Ms. Jill Davidson for her help with flushing of the reproductive tracts of cows used in this study. I also wish to thank Mr. Frank Michel, Dr. Lokenga Badinga and Dr. Mike Green for their technical and moral support during the time I spent in their laboratory. Many thanks to Mrs. Glenda Walton for her moral support and help with word processing, and to Christa Jenssen for help with preparing slides. My thanks go to Jack Stokes and Dean Glicco for managing the herd, and to Larry Eubanks and Leeroy Washington for slaughtering the animals used in this work. The graduate school was less stressful because of friends like Dr. Thais Diaz, Mr. Keith Rollyson, Ms. Lannett Edwards, and Mr. Andres Kowalski. I thank all my family friends. Dr. Esther Smith, Ms. Patience Njofang, Mr. Andrew Kweh, Mr. Odemari Mbuya and Ms Marie Gaffney, for their friendship. Special thanks go to my mother, sisters, and brother for their encouragement. IV

PAGE 5

I could not have survived this program without the support, patience, and love from my husband, Gaston, and my kids, Kongwenebime, Koga and Mandi, and I wish to thank them for their love and patience. 1

PAGE 6

TABLE OF CONTENTS ACKNOWLEDGMENTS ii LIST OF TABLES xi LIST OF FIGURES xii ABSTRACT xv CHAPTERS 1 INTRODUCTION 1 *^ 2 REVIEW OF LITERATURE 3 The Corpus Luteum 3 Histology of the Corpus Luteum 4 Biochemical Differences Between Small and Large Luteal Cells 7 I Ontogeny of Small and Large Luteal Cells 10 \ Intercellular Communication Among Luteal Cells 12 Cell-Cell Communication in Luteal Angiogenesis and Development 13 r Cell-Cell Interaction in Luteal Hormone ^ Synthesis 15 Corpus Luteum of the Estrous Cycle 16 Luteinization and CL Development 16 Luteal Angiogenesis 18 Regulation of Luteal Function During The Estrous Cycle 21 Luteal Regression 23 Role of PGF2a in Luteal Regression 23 Role of Oxygen Free Radicals in Luteal Regression 27 Corpus Luteum of Pregnancy 28 CL Morphology During Pregnancy 29 Maternal Recognition of Pregnancy 30 VI

PAGE 7

Protein Synthesis by tine Corpus Luteum 36 General Overview of Protein Synthesis and Release 37 Protein Secretion 38 Protein Targeting 38 Translocation Across ER Membrane 39 Processing and Sorting of Proteins in the Golgi 39 Apolipoprotein A-1 40 Biochemical Characterization 40 Apo A-1 gene 41 Metabolism 41 Role of Apo A-1 43 Regulation of Apo A-1 Synthesis by Steroid Hormones 44 Effects of Nutrition on Apo A-1 Synthesis 45 Apolipoprotein E 46 f^ Biochemical Characterization 46 ^ Effects of FSH, LH, cAMP, and Phorbol Ester on Apo E Synthesis 47 Regulation of Apo E Synthesis by Cell Cholesterol 49 Regulation of Apo E Secretion by Cytokines 49 Role of Apo E in Ovarian Function 50 Manganese Superoxide Dismutase 50 Biochemical Characterization 50 \t Regulation of Mn SOD Synthesis by ^. Oxidative Stress 51 ^ Regulation of Mn SOD by Gonadotropins 52 t Regulation of Mn SOD Synthesis by Cytokines and Phorbol Ester 52 Role of Mn SOD in the Ovary 53 f Tissue Inhibitors of Metalloproteinases: TIMP-1 and TIMP-2 54 Tissue Inhibitor of Metalloproteinases-1 55 Biochemical Characterization 55 Regulation of TIMP-1 Synthesis by Gonadotropins 56 Regulation of TIMP-1 Synthesis by Steroid Hormones 58 Regulation of TIMP-1 Synthesis by Cytokines and Grovvth Factors 59 Tissue Inhibitor of Metalloproteinases-2 60 Biochemical Characterization 60 Regulation of TIMP-2 Synthesis by Gonadotropins 61 VII

PAGE 8

Regulation of TIMP-2 Synthesis by Steroid Hormones 61 Regulation of TIMP-2 Synthesis by Cytokines and Growth Factors 62 3 PROTEINS SYNTHESIZED AND RELEASED IN CULTURE BY THE BOVINE CORPUS LUTEUM: THE ESTROUS CYCLE AND PREGNANCY 64 4 !, Introduction 64 A Materials and Methods 66 i Materials 66 Collection of Luteal Tissue 66 ^ Culture Medium 67 \ Time Course Studies of Incorporation of ^ Radiolabel 68 Culture and Radiolabelling of Luteal Tissue 69 TCA Precipitation 70 Light and Electron Microscopy 70 Two-Dimensional-SDS-Polyacrylamide Gel i Electrophoresis 71 I First-dimension: Isoelectric Focusing 71 Second-dimension: SDS-PAGE 71 Protein Blotting and Amino Acid Sequencing 72 I Progesterone Assay 73 Statistical Analysis 74 Results 75 Histology of Luteal Tissue 75 I CL Weight and Plasma Progesterone 75 Incorporation of Radiolabel into TCA-precipitable Protein 79 • Luteal Protein Synthesis and N-terminal Amino Acid Micro Sequencing 82 Protein Synthesis and Release during the Estrous Cycle 86 Luteal Protein Synthesis and Release during Pregnancy 87 Radiolabeled Culture with ^H-glucosamine and ^^Smethionine 90 Discussion 93 VIII

PAGE 9

^' A EXPRESSION OF MESSENGER RNA OF APOLIPOPROTEIN A-1 AND E IN BOVINE LUTEAL TISSUE DURING THE ESTROUS CYCLE AND PREGNANCY 100 Introduction 100 Materials and Methods 102 Materials 102 • Tissue Collection 102 Isolation of RNA 103 Preparation of Plasmid DNA 104 Restriction Analysis and Isolation of Insert 105 ; Northern Hybridization 106 \ Autoradiography 107 Dot Blot Hybridization 107 ;; Statistical Analysis 108 ^' Results 109 Northern Blot Analysis of Apo E and Apo A-1 mRNA 109 Dot Blot Analysis 1 09 Discussion 115 5 EXPRESSION OF MESSENGER RNA OF TISSUE INHIBITOR OF ^ METALLOPROTEINASES-1 AND -2 IN BOVINE LUTEAL TISSUE DURING THE ESTROUS CYCLE AND PREGNANCY 120 ^ Introduction 120 Materials and Methods 121 Materials 121 Tissue Collection 122 Isolation of RNA 123 Preparation of Plasmid DNA 124 Restriction Analysis and Isolation of Insert 125 Northern Hybridization 126 Autoradiography 127 Dot Blot Hybndization 127 Statistical Analysis 128 Results 128 Expression of TIMP-1 mRNA 128 Expression of TIMP-2 mRNA 132 Discussion 135 IX

PAGE 10

6 CHANGES IN THE EXPRESSION OF MESSENGER RIBONUCLEIC ACID FOR MANGANESE SUPEROXIDE DISMUTASE IN THE BOVINE CORPUS LUTEUM DURING THE ESTROUS CYCLE AND PREGNANCY 139 Introduction 139 Materials and Methods 141 Tissue Collection 141 Isolation of RNA 141 Restriction Analysis and Isolation of Plasmid DNA Insert 142 Northern Hybridization 143 Autoradiography 144 Dot Blot Hybridization 145 Statistical Analysis 146 Results 146 Northern Blot Analysis 146 Dot Blot Analysis 146 Discussion 149 7 GENERAL RESULTS AND DISCUSSION 152 REFERENCES 165 APPENDIX 1 ANIMAL CARE AND TISSUE COLLECTION 192 APPENDIX 2 PROGESTERONE IMMUNOASSAY 196 APPENDIX 3 CULTURE AND RADIOLABELLING OF LUTEAL TISSUE 200 APPENDIX 4 DETERMINATION OF INCORPORATION OF RADIOLABEL INTO NEWLY-SYNTHESIZED PROTEINS 204 APPENDIX 5 SEPARATION OF PROTEINS IN LUTEALCONDITIONED MEDIUM BY ELECTROPHORESIS ... 206 APPENDIX 6 ELECTRO-BLOTTING OF PROTEINS TO MEMBRANE 220 APPENDIX 7 DETERMINATION OF TOTAL PROTEIN BY METHOD OF LOWRY 223

PAGE 11

APPENDIX 8 MEASUREMENT OF APO A-1 MRNA 225 APPENDIX 9 MEASUREMENT OF APO E MRNA 235 APPENDIX 10 MEASUREMENT OF TIMP-1 AND TIMP-2 MRNA 237 APPENDIX 1 1 RNA ISOLATION AND PURIFICATION 249 APPENDIX 12 NUCLEIC ACID LABELLING 254 APPENDIX 13 NORTHERN BLOTTING AND HYBRIDIZATION 256 APPENDIX 14 RNA DOT BLOT HYBRIDIZATION 264 APPENDIX 15 IMMUNOHISTOCHEMICAL LOCALIZATION OF APO E 268 APPENDIX 16 RAW DATA-EXPERIMENT 1 270 APPENDIX 17 RAW DATA-EXPERIMENT 2, 3, 4, AND 5 273 APPENDIX 18 277 BIOGRAPHICAL SKETCH 280 XI

PAGE 12

* LIST OF TABLES Table page 2-1 Summary of factors regulating synthesis of proteins 63 3-1 Comparison of N-terminal amino acid sequences for proteins 1, 8, 9, 10 and 11 in luteal-conditioned medium to sequences in the protein data banks 85 XII

PAGE 13

LIST OF FIGURES Figure 3-1 Electron micrograph of luteal cells. Top panel: luteal cell of tissue prior to incubation (control), x 9000; Middle panel: luteal cell of tissue post-24 h incubation without radiolabel (control), x 1 1 ,400; Bottom panel: luteal cell of tissue post-24 h incubation with 50 pCi 3H-leucine, x 1 1 ,400 76 3-2. Least square means SEM of plasma concentrations of progesterone (ng/ml) and weight (g) of the corpus luteum across days 3, 7, 11, 14, 17, and 1 9 of the estrous cycle and days 17, 88, 170, and > 240 of pregnancy 78 3-3. Time course studies of incorporation of radiolabel into TCA-precipitable protein 80 3-4. Percent incorporation of radiolabel into newly-synthesized proteins 81 3-5. Representative fluorographs of proteins synthesized de novo in explant culture and released into the medium by CL on different days of the estrous cycle 83 3-6. Representative fluorographs of proteins synthesized de novo in explant culture and released into the medium by CL on different days of pregnancy 84 3-7. Densitometric analysis of fluorographs of newly-synthesized proteins in luteal-conditioned medium during the estrous cycle and pregnancy. Values are least square means standard error of the mean 88 XIII

PAGE 14

3-8. Representative fluorographs of proteins synthesized de novo in explant culture and released into medium on day 240 of pregnancy. Left panel: radiolabeled culture with ^H-leucine. Right panel: radiolabelled culture with ^H-glucosamine 91 3-9. Representative fluorographs of proteins synthesized de novo in explant culture and released into medium on day 240 of pregnancy. Left panel: radiolabelled culture with ^H-leucine. ; Right panel: radiolabelled culture with ^^S-methionine 92 4-1. Northern blot analysis of apolipoprotein E mRNA. The same blot was probed with H-actin that served as a control for the loading and the integrity of the RNA 110 ; 4-2. Northern blot analysis of apolipoprotein A-1 mRNA. The same blot was probed with (3.-actin that served as a control for the loading and the integrity of the RNA Ill 4-3. Dot blot analysis of Apo E mRNA. Ten pg total RNA isolated t from CL during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^^P-labelled Apo J EcDNA 112 4-4. Dot blot analysis of Apo A-1 mRNA. Ten pg total RNA isolated from CL during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^^P-labelled Apo A-1 cDNA 113 4-5. Apo A-1 mRNA expression relative to (i-actin is presented as LSMeanSEM 114 5-1. Northern blot analysis of luteal TIMP-1 mRNA. The same blot was probed for fL-actin that served as a control for the ;.;. loading and the integrity of the RNA 129 -:-: 5-2. Dot blot analysis of TIMP-1 mRNA. Ten pg total RNA isolated from luteal tissue during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 1 7, 90-1 20, 1 70-1 80, and > 21 5) ''I was loaded per sample. RNA blots were hybridized with .'J X 32p_|abelled TIMP-1 cDNA 1 30 1 XIV

PAGE 15

5-3. TIMP-1 mRNA expression relative to (1-actin is presented as LSMean SEM (n = 2-5) and differences determined by orthogonal contrasts 131 5-4. Northern blot analysis of luteal TIMP-2 mRNA. The same blot was probed for f3)-actin to act as a control for the loading and integrity of RNA 1 32 5-5. Dot blot analysis of luteal TIMP-2 mRNA. Ten pg total RNA isolated from luteal tissue during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^2p-iabelled TlMP-2 cDNA 133 5-6. TIMP-2 mRNA expression relative to (3.-actin is presented as LSMean SEM (n = 2-5) and differences determined by orthogonal contrasts 1 34 6-1 Northern blot analysis of luteal Mn SOD mRNA. The same blot was probed for fi-actin to act as a control for the loading and integrity of RNA 147 6-2. Dot blot analysis of luteal Mn SOD mRNA 148 i XV

PAGE 16

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy IDENTIFICATION AND CHARACTERIZATION OF PROTEINS SECRETED BY THE CORPUS LUTEUM OF THE COW DURING THE ESTROUS CYCLE AND PREGNANCY By FLORENCE MABOH NDIKUM-MOFFOR December 1995 Chairperson: Michael J. Fields Major Department: Animal Science Experiments were carried out to determine protein synthesis and secretion by the corpus luteum (CL) of the cow during the estrous cycle and pregnancy. Temporal changes were observed in the types of proteins synthesized across days of the estrous cycle, but not across pregnancy. The bovine CL synthesized many proteins in culture, five of which were further characterized and identified by N-terminal amino acid sequence analysis. Proteins identified were apolipoprotein E (Apo E, 35 kDa, pl 5.5), apolipoprotein A-1 (Apo A-1, 27 kDa, pl 6), tissue inhibitor of metalloproteinases-1 (TIMP-1, 30 kDa, pl 8), tissue inhibitor of metalloproteinases-2 (TIMP-2, 20 kDa, pl 8), and manganese superoxide dismutase (Mn SOD, 22 kDa, pl 8). Northern and dot XVI

PAGE 17

blot analyses revealed presence of mRNA for each of the five identified proteins within bovine luteal tissue during the estrous cycle and pregnancy, confirming synthesis of these proteins by the bovine CL. Synthesis of Apo A-1 and Apo E by the bovine CL is novel and has not been reported in any species. This is also the first report of expression of Apo A-1 mRNA by the CL. There have been reports of synthesis of TIMP-1 and TIMP-2 by the CL during the estrous cycle, but this is the first study to examine luteal synthesis of TIMP-1 and TIMP-2 during pregnancy. Similarly, this is the first report of the temporal changes in the synthesis of Mn SOD and expression of its mRNA during the estrous cycle and pregnancy in the cow. Apolipoprotein E and Apo A-1 given their association with low density lipoprotein and high density lipoprotein, respectively, may be involved in regulating cholesterol availability for luteal membrane synthesis and steroidogenesis. TIMP-1 and TIMP-2 may be involved in steroidogenesis during the luteal phase and pregnancy, and tissue remodelling during luteal regression. Luteal Mn SOD may play a significant role in maintaining luteal function during the luteal phase and pregnancy by preventing damage of luteal cells by reactive oxygen species. These luteal proteins are believed to be acting via autocrine and/or paracrine mechanisms. XVII

PAGE 18

CHAPTER 1 INTRODUCTION A physiological role for the ovaries was first demostrated by Frankel (1903, cited by Short, 1977) when he showed that removal of the ovaries in rabbits resulted in termination of pregnancy. The corpus luteum of the ovary was later identified as the component responsible for maintenance of pregnancy. However, the mechanism through which the CL carries out this function was not known. Because of its high vascularity, Prenant (cited by Short, 1977) suggested that the CL is an organ of internal secretion and was capable of releasing products into systemic circulation. The advent of new technologies and techniques such as electron microscopy made it possible to investigate the histology and biochemistry of the CL during the estrous cycle and pregnancy (Priedkain and Weber, 1968). As a result, it is now known that the CL is composed of different types of cells; the steroidogenic small and large luteal cells, endothelial cells, macrophages, fibroblasts, and monocytes, and that there are some morphological and biochemical differences between the CL of the estrous cycle and that of pregnancy. In addition, the steroidogenic small and large luteal cells have been identified as responsible for the production of progesterone, the hormone 1

PAGE 19

required for maintenance of pregnancy. With the use of other modern research techniques in cell biology and histochemistry, it is common knowledge that luteal cells possess the structural machinery required for the synthesis and secretion of proteins. Recent research has shown that apart from the production of progesterone, the CL produces many different proteins and peptide hormones (Schams, 1989), angiogenic and mitogenic factors (Grazul-Bilska et al., 1993; Reynolds et al., 1994), and growth factors (Schams, 1989). However, there is a paucity of information about the protein-synthetic potential of the CL across the estrous cycle and pregnancy, the nature of the proteins synthesized, factors regulating their synthesis and release, and the cell types responsible for their production. Identification and characterization of luteal proteins and factors will enable the determination of their roles in luteal growth, development and steroidogenesis, which will contribute to the understanding of fertility as well as tissue growth and development. There is an indication from some studies that other cell types, apart from the steroidogenic luteal cells, are responsible for synthesis of some luteal proteins. Objectives of this dissertation were to examine protein synthesis by bovine luteal explants in culture during the estrous cycle and pregnancy, identify, isolate and characterize the proteins secreted, and determine presence of their respective mRNA in luteal tissue during the cycle and pregnancy.

PAGE 20

CHAPTER 2 REVIEW OF THE LITERATURE The Corpus Luteum The corpus luteum (CL) is formed following the rapid development of the follicle after ovulation. Regnier de Graaf gave the first detailed description of the CL which he called "globular bodies" (Jocylyn and Setchel, 1972). Malpighi (1628-1698) introduced the term corpus luteum or "yellow body" (Short, 1977). The first attempt to define the role of the CL was made by Prenant (1898) who suggested that it is a gland of internal secretion because of its high vascularity. The CL has been shown to play several roles in the reproductive process. Most of these roles depend upon synthesis of progesterone, which is necessary for implantation, maintenance of pregnancy, control of the estrous cycle and parturition. Some of these roles were suggested following observations by Frankel in 1903 that removal of ovaries in rabbits terminated pregnancy, and this gave support to Gustav Bom's hypothesis that corpora lutea are required for implantation (cited by Amoroso, 1968). If fertilization does not occur, the CL regresses in some species with a subsequent decline in circulating progesterone. The decline in progesterone concentrations leads to an increase in pulse frequency of luteinizing hormone (LH) increased follicular estrogen

PAGE 21

4 synthesis, and a new wave of follicular development (Karsch et al., 1984). However, if fertilization occurs, the CL is prevented from regressing by signals sent by the embryo (Bazer et al., 1991a; 1991b). The CL could therefore be regarded as the biological clock for the events of the estrous and menstrual cycles, pregnancy and parturition. Histology of the Corpus Luteum The corpus luteum of several species is made up of different types of cells, namely, the steroidogenic large luteal and small luteal cells, endothelial cells, fibroblasts, macrophages and lymphocytes. Earlier studies (Priedkalns and Weber, 1968; Koos and Hansel, 1981) showed that the mature bovine corpus luteum contains two distinct steroidogenic cell types: the small and large luteal cells. The major differences between these two cell types were the presence of numerous large mitochondria in the large cells, and numerous lipid bodies in the small cells. Other cell types and stromal cells were also observed in the corpus luteum. The large luteal cells are larger in size (30-40 p), with extensively folded plasma membranes which are exposed to other luteal cells and to inter-cellular vascular areas. The small luteal cells are smaller in size (15-20 pm in diameter) but with a denser chromatin pattern than the large cells. More recent studies (O'Shea et al., 1989) showed that in the bovine corpus luteum, the large luteal cells make up 3.5 % of the total cells/unit area of luteal tissue while the small luteal cells make up 26.7%, and endothelial cells /

PAGE 22

5 pericytes are the most abundant (52.3%). However, large luteal cells occupy 40.2% of the corpus luteum volume/density, while the small luteal cells occupy 27.7%. Similar observations were made by Lei et al. (1991) who reported that human and bovine corpora lutea contain more nonluteal than luteal cells, and the small luteal cells are always greater in number than the large luteal cells irrespective of the reproductive states. Ultrastructural and cytochemical studies by Parry et al. (1980) showed that large granulosa-derived bovine luteal cells are the most common cell types of the mid-luteal corpus luteum. They, however, did not observe any morphological differences between small and large cells as reported by Priedkalns et al. (1968). Large luteal cells are always close to capillaries and contain large round nuclei (10 pm in diameter), large amounts of agranular endoplasmic reticulum, and abundant mitochondria scattered throughout the cytoplasm. Bovine large luteal cells contain numerous Golgi complexes and electron-dense secretory granules during mid-cycle (Parry et al., 1980). Bovine luteal secretory granules are single membrane-bound and are 0.2-0.4 pm in diameter compared to 0.2 pm in sheep (Gemmell et al., 1974). In the cow, secretory granules are found in a cluster in the cytoplasm (granules make up 2-4 % of luteal cell cytoplasm) close to the Golgi complexes, and sometimes near the edge of the cells (Parry et al., 1980). The contents of secretory granules are released by exocytosis to the intercellular space. Lipid droplets are present in some cells and concentrations varied inversely with the number of electron-

PAGE 23

6 dense granules present. Morphometric analyses have shown that luteal cells occupy the maximum area of the corpus luteum on day 12 of the cycle while the protein-synthesizing compartment (ER plus polysomes and/or ribosomes) increases from day 6 to a maximum on day 17, indicating that protein synthesis is part of the metabolism of the corpus luteum during the cycle (Parry et al., 1980). The population of secretory granules is greatest on day 17 of the bovine cycle and declines thereafter indicating a drop in the secretory process, which usually precedes the fall in progesterone synthesis (Hansel et al., 1973). More recent studies in the cow have shown that unlike in sheep, luteal secretory granules are not dispersed throughout the cytoplasm but are found in a cluster (Fields et al., 1983; 1992). During the bovine cycle, the percent of large luteal cells with secretory granules is lowest on day 3 (3%), highest during mid-cycle (day 7, 84 %, day 1 1 64 %) and declines on day 14 (26 %) to lowest level on days 17 (16 %) and 19 (8 %) (Fields et al., 1992). An earlier study in pregnant cows also indicated dynamic changes in the population of secretory granules during the course of pregnancy. In the large luteal cell of pregnancy, the number of secretory granules are low or undetectable prior to day 45, increases to maximum around day 200 and declines at the end of gestation (Fields et al., 1985). There have been reports of other morphological differences between small and large luteal cells. Chegini et al. (1984) reported that large (18-45 pm) cells contain more mitochondria than small (15-18 pm) cells, and both contain

PAGE 24

7 rough and smooth endoplasmic reticulum, lysosomal vesicles, Golgi complexes and membrane-bound dense granules. Granules vary in shape and form clusters close to the nucleus (Chegini et al., 1984). In a more recent study, Chegini et al. (1991) observed that the nuclear volume is greater in small than large luteal cells during estrous cycle and pregnancy. Small luteal cells are also more sensitive to hCG-induced increase in nuclear volume than the large luteal cells. The cytoplasmic:nuclear ratio is greater in large than small cells. Biochemical analysis of the two cell types have shown that the amount of protein per cell is lower in small than the large cell, while the protein/DNA ratios are similar for both. It was suggested that differences observed may have been caused by transformation of small luteal cells to large luteal cells (Chegini et al., 1984). Biochemical Differences Between Small and Large Luteal Cells Biochemical differences between small and large luteal cells have been reported. Small luteal cells (which make up 85% of the total luteal cell population) produce small amounts of basal progesterone and respond to secretagogues (cAMP, hCG), while the large luteal cells (which make up 8-12% of luteal cell population) produce greater amounts of basal progesterone but are not responsive to cAMP and hCG (Koos and Hansel, 1981 ; Fitz et al., 1982). The LH-induced increase in progesterone production by small luteal cells is mediated by an increase in cAMP. Nonhormonal activators of protein kinase A

PAGE 25

8 (forskolin, cholera toxin, dibutyryl cAMP) also selectively stimulate synthesis of progesterone in small but not In large luteal cells (Wiltbank, 1994). It is not clear by what mechanism progesterone synthesis by large luteal cells is stimulated, since they have been shov\/n to produce large amounts of progesterone in culture even in the absence of luteotropic stimuli (Wiltbank, 1994). Factors implicated in the LH-independent increase in progesterone production by large luteal cells include PGEj (Fitz et al., 1984), insulin (Sauerwein et al., 1992), and growth factors (Einspanier et al., 1990; Budnik and Mukhopadhyay, 1991; Miyamoto etal., 1992). Alila et al. (1989) reported that treatment of bovine luteal cells with LH causes a rapid increase in intracellular free calcium in both small and large cells. Alila et al. (1989) reported that the LH-induced rise in intracellular calcium is biphasic in small cells (initial peak due to mobilization of intracellular calcium, and a second rise due to influx of extracellular calcium), while a single rise is observed in the large luteal cells. The increase in calcium is also greater for small luteal than large luteal cells (Alila et al., 1989). In an earlier study, Alila et al. (1988) observed that phorbol dibutyrate increases progesterone synthesis in bovine small, but not large luteal cells in culture. Those observations support earlier postulations that protein kinase C (PKC) is involved in progesterone synthesis in the bovine corpus luteum, and that the stimulatory effects of PKC on progesterone synthesis involves only the small luteal cells (Hansel et al., 1987). The differential response of large and small luteal cells to secretagogues seems

PAGE 26

9 to be related to the distribution of receptors between the two cell types. Fitz et al. (1982) showed that small luteal cells of sheep contain 10X more LH receptors than large cells, while large cells are more enriched with prostaglandin E and Fza receptors. PGFja is luteotropic in small luteal cells where it stimulates phospholipase C activity but does not reduce the LH-stimulated cAMP or progesterone accumulation (Davis et al., 1989). Other biochemical differences between small and large luteal cells are in their abilities to synthesize and secrete proteins and peptide hormones (Schams, 1989; Rodgers, 1990). Judging from the morphology of luteal cells, large luteal cells have the intracellular organelles specialized for secretion of proteins and peptides (Anderson, 1982). Luteal tissue of sheep (Wathes and Swann, 1982), women (Wathes et al., 1982) and cows (Wathes et al., 1983; Fields et al., 1983) produces oxytocin and neurophysin. Oxytocin is present in ovine (Rodgers et al., 1983) and bovine (Fields et al., 1986; 1992) large luteal cells but not in small cells. Another peptide hormone, relaxin, is produced by corpora lutea of a multitude of mammals (see Sherwood, 1994, for review). Bagnell et al. (1989) reported that in pigs, relaxin is localized within the large luteal cell, but not the small luteal cell.

PAGE 27

10 Ontogeny of Small and Large Luteal Cells The origin of small and large luteal cells has been a matter of controversy. In the late 1800s and early 1900s, some researchers believed that granulosa cells of the ovarian follicle degenerate following ovulation, and only cells of theca developed into a corpus luteum. Others, however, thought that granulosa cells develop into a corpus luteum while thecal cells degenerate. Loeb (1906) was the first to suggest that the corpus luteum is composed of cells originating from both theca and granulosa layers. Similar observations were made in the sow (Corner, 1919), cow (Donaldson and Hansel, 1965; Lobel and Levy, 1968; Priedkalns et al., 1968), ewe (O'Shea et al., 1980), rat (Pederson, 1951) and human (Guraya, 1971). It is now believed that small luteal cells originate from theca interna of the follicle, while the large luteal cells originate from the granulosa cells. Morphological studies in sheep (O'Shea et al. 1980) and cattle (Donaldson and Hansel, 1965; Priedkalns et al., 1968) indicate that small luteal cells originate from thecal cells. Alkaline phosphatase, a marker of theca interna cells, was used to demonstrate that theca cells differentiate into small luteal cells of the ovine CL (O'Shea et al., 1980). Further support to this hypothesis was provided by observations that thecal cells incubated with forskolin and insulin for 9 days become luteinized, have low basal progesterone secretion, increased LH-induced secretion of progesterone, and do not secrete oxytocin (Meidan et al., 1990). These physiological characteristics are similar to

PAGE 28

11 those exhibited by small luteal cells. Studies in sheep also suggest that granulosa cells develop into large luteal cells because the number of granulosa cells in pre-ovulatory follicles (O'Shea et al., 1985) approximates the number of large luteal cells (O'Shea et al., 1986), and ovine granulosa cells undergo little or no mitosis after ovulation (McClellan et al., 1975). Evidence for differentiation of granulosa cells to large luteal cells was provided by observations that incubations of bovine granulosa cells with forskolin and insulin for 9 days resulted in luteinized cells similar to large luteal cells (high basal progesterone secretion, reduced LH-induced progesterone secretion, and secretion of oxytocin) (Meidan et al., 1990). Alila and Hansel (1984) demonstrated that monoclonal antibodies developed against theca cell membranes bound to small luteal cells while monoclonal antibodies against granulosa cells bound specifically to the large luteal cells. As the estrous cycle progressed, the number of large luteal cells bound to theca antibody was similar to the number of small cells bound to theca antibody (Alila and Hansel, 1984). They suggested that theca-derived small luteal cells differentiate into large luteal cells as the estrous cycle progresses. However, a study by O'Shea et al. (1986) in which comparisons of the cellular composition of ovine luteal cells of midand late estrous cycle were made, did not agree with the hypothesis that small luteal cells differentiate to large luteal cells during the estrous cycle. Alila and Hansel (1984) also observed that large luteal cells bound to granulosa antibody contained more mitochondria and

PAGE 29

12 electron-dense granules than those bound to theca antibody. In general, the CL is a dynamic organ and its morphology changes with the reproductive state of the animal. Farin et al. (1989) reported a change in cellular composition of the ovine corpus luteum during the estrous cycle and pregnancy. The number of small cells increase with no change in size as the cycle progressed, whereas the size of large cells increased with no change in number. Thus changes in the relative proportions of the two cell types or interactions between them, may determine the function of the CL at different periods of the estrous cycle. In sheep, the number of small steroidogenic luteal cells increases 4-fold through day 8 and then decreases through day 16 (Niswender et al., 1985). However, the number of nonsteroidogenic cells > 8 pm increases 2-fold between days 4 and 8 of the ovine cycle and declines through day 16, while the number of nonsteroidogenic cells < 8 pm reaches a peak on day 12. Because of the similarity in pattern of steroidogenic and nonsteroidogenic cells during the estrous cycle, it was speculated that small nonsteroidogenic cells are a source of stem cells that give rise to small steroidogenic luteal cells, which later develop into large luteal cells (Niswender et al., 1985). Intercellular Communication Among Luteal Cells The heterogenous nature of the cellular components of the corpus luteum is an important feature which seems to be necessary for this organ to effectively carry out its biological functions. Formation of the CL following ovulation

PAGE 30

13 involves incorporation of cells from the theca and granulosa layers of the ovulating follicle. As earlier discussed, the process of luteinization involves luteal angiogenesis (Zheng et al., 1993), and an increase In size of theca and granulosa cells, and number of smooth endoplasmic reticulum and mitochondria. The corpus luteum is composed of large and small luteal cells, macrophages, monocytes, fibroblasts and endothelial cells. A substance produced by one cell may affect the function of another cell and this is referred to as cell-cell communication (Rodgers, 1990). This communication may be via gap junctions (Anderson and Little, 1984; Redmer et al,, 1991) and adherens-type junctions (Weber et al., 1987; O'Shea et al., 1990) which have been observed between luteal cells. Exchange of factors among luteal cells could also occur via the blood stream (humoral). While gap junctions allow passage of very low molecular weight substances from one cell to another, adherens-type junctions serve to bind cells together (Rodgers, 1990). Cell-Cell Communication in Luteal Angiogenesis and Development Ovine luteal cells have been shown to produce angiogenic factors (Redmer et al., 1988; Grazul-Bilska et al., 1992). It has been proposed that grov\^h factors such as fibroblast growth factor, insulin-like growth factor (IGF)-I, epidermal growth factor (EGF), and cytokines are involved in luteal angiogenesis (Koos, 1989). This idea was supported by observations by Zheng et al. (1993) that bovine large and small luteal cells produce basic fibroblast

PAGE 31

14 growth (BFGF) -1 and -2 during the estrous cycle. BFGFs are potent angiogenic factors and their production by large and small luteal cells follow a pattern similar to luteal development (Zheng et al., 1993). Thus the steroidogenic large and small cells produce HBGFs which stimulate proliferation of endothelial cells. A functional relationship between endothelial cells and luteal cells was suggested following observations that both endothelial (Mayerhofer et al., 1992) and luteal cells (Mayerhofer et al., 1991) express a neural cell adhesion molecule. In addition, endothelial cells produce prostacyclin (PGFI2) (Maclntyre et al., 1978) which has been shown to increase concentrations of plasma progesterone in the cow (Milvae et al., 1980). A more recent study by Girsh et al. (1995) gave further evidence of interactions between endothelial cells and the steroidogenic cells of bovine CL. Endothelial cells were shown to secrete PGI2, which in turn stimulates secretion of cAMP and progesterone by bovine large and small luteal cells. However, it was also observed that the presence of endothelial cells is required for PGFjc-induced inhibition of progesterone production by luteal cells (Girsh et al., 1995). Thus endothelial cells of CL may regulate response of steroidogenic luteal cells to luteotropic and luteolytic signals. IGF-I and its mRNA is produced by the bovine CL (Einspanier et al., 1990) during the estrous cycle and gestation. IGF-I has been shown to stimulate synthesis and secretion of progesterone and oxytocin by luteal tissue (Sauerwein et al., 1992). In a more recent study IGF-I was immunolocalized

PAGE 32

15 mainly in bovine large luteal cells with little staining in small cells (Amselgruber et al., 1994). No IGF-I immunoreactivity was observed in pericytes, macrophages, fibroblasts or smooth muscle cells of blood vessels (Amselgruber et al., 1994). The differences in distribution of IGF-I immunoreactivity may indicate differences in cell reactivity and possible paracrine or autocrine interactions between small and large luteal cells (Amselgruber et al., 1994). A role of IGF-I in CL function has been suggested following observations that concentrations of luteal IGF-I increase during early and mid-luteal phases and decline rapidly after luteolysis (Einspanier et al., 1990). Cell-Cell Interaction in Luteal Hormone Synthesis Luteal progesterone is produced mainly by the large luteal cells in most species. However, progesterone production by isolated populations of porcine small and large luteal cells has been shown to be greater when both cell populations are cultured together than when cultured separately (Lemon and Mauleon, 1982). The increase in progesterone production observed in cocultures was attributed to stimulation of progesterone production in large cells by some factor(s) produced by small cells, suggesting an interaction between small and large luteal cells. Oxytocin, a peptide hormone produced by CL of ruminants is thought to be involved in cell-cell interaction between small and large cells because it is produced only by large luteal cells (Fields and Fields, 1986; Theodosis et al., 1986) and is capable of inhibiting LH-induced P4

PAGE 33

16 production by small luteal cells (Schams, 1989). Corpus Luteum of the Estrous Cycle Luteinization and CL Development Corpora lutea are formed following ovulation of a mature follicle mediated by gonadotrophin stimulation. Following the preovulatory LH surge, a series of morphological and biochemical changes take place within the follicle to change the latter to a corpus luteum. The LH surge serves dual roles of stimulating ovulation and converting the follicle to a corpus luteum. This luteinization process involves breakdown of the basement membrane between theca and granulosa layers, invasion of the follicular antrum space by blood vessels, and development of an extensive vascular network (Zheng et al., 1993; Niswender et al., 1994). The invading capillaries are formed via both migration and mitosis of endothelial cells (Zheng et al., 1993). After ovulation, the follicle grows rapidly to 10 times its weight in just 7 days. This increased growth is attributed to hypertrophy and hyperplasia of thecal cells which migrate into the hollow follicular antrum after ovulation, and integrate among luteinizing granulosa cells (O'Shea et al., 1 980). Formation of the CL is initiated by a series of morphological and biochemical changes in the theca and granulosa cells of the preovulatory follicle. This process, referred to as luteinization, changes the follicle from a predominantly estradiol-producing structure to one that secretes

PAGE 34

17 progesterone. Morphological changes associated with luteinization include accumulation of smooth endoplasmic reticulum, mitochondria with tubular cristae, increase in size of Golgi apparatus, and accumulation of glycogen-containing granules (Niswender and Nett, 1994). These changes provide the CL with the ability to efficiently produce progesterone. During the luteinization process, theca and granulosa cells of the preovulatory follicle differentiate into small and large luteal cells, respectively (Meidan et al., 1990). Other morphological changes associated with luteinization include an increase in the cytoplasmic nuclear ratio and appearance of a large number of lipid droplets containing sterol esters. After ovulation, there is an increase in gap junctions among developing luteal cells in rats (Anderson and Little, 1984). O'Shea et al. (1990) reported the presence of adherens-type junctions between small and large luteal cells of cattle. In a more recent study, Redmer et al. (1991 ) reported the presence of gap junction-like structures in bovine luteal cells from mid-cycle. Biochemical changes associated with luteinization of the follicle include a switch from a predominantly estradiol-producing structure to one that secretes mainly progesterone. During luteal formation, there is an increase in expression of mRNA and enzyme activity for P450 side chain cleavage and 3fi-hydroxysteroid dehydrogenase (3(iHSD). There is also an increase in activity of cholesterol esterase as the CL becomes fully functional. These changes are consistent with the CL's role in progesterone synthesis. In contrast, luteinization decreases

PAGE 35

18 estrogen production by decreasing levels of mRNA and protein for 17ahydroxylase cytochrome P450, the enzyme that catalyses conversion of pregnenolone or progesterone to androgen. In preovulatory follicles in cattle (Rodgers et al., 1987) and rat (Hedin et al., 1987), the expression of mRNA and protein levels for aromatase cytochrome P450 enzyme also decreases rapidly after the LH surge. Corpora lutea of the rat expresse aromatase mRNA and produce estradiol, while CL of domestic ruminants do not synthesize estradiol (Savard, 1973). Other biochemical changes observed after ovulation include reduced expression of genes encoding FSH and LH receptors in granulosa follicular cells, v^^hich results in a down regulation of both receptors. In contrast, LH receptor levels increase with formation of the CL in ewes (Diekmann et al., 1978). In support of this observation in rat, expression of the gene encoding the LH receptor increases with development of the CL (Segaloff et al., 1990). Luteal Angiogenesis Angiogenesis is one of the features of luteinization, and continues after formation of the CL. Blood flow to the CL increases with luteinization and accounts for about 90% of the total ovarian blood flow during the mid-luteal phase. At this time, about 60% of each luteal cell's surface directly faces a capillary (Keyes and Wiltbank, 1988). In the rat, luteal blood flow and the number of luteal endothelial cells increase during mid-pregnancy (Bruce et al., 1 984). The ability of the CL to cause angiogenesis was first reported by Jakob

PAGE 36

19 et al. (1977). Corpora lutea of sheep and cattle have been shown to produce angiogenic factors in culture (Redmer et al., 1988). Gospodarowicz et al. (1985) also isolated an angiogenic factor from the bovine CL which accounted for 84% of the angiogenic activity in crude CL extracts, and had amino acid sequence homology with bovine brain and pituitary fibroblast growth factor. It is not clear what regulates angiogenesis or the high rate of blood flow to the CL. However, several factors which affect endothelial cell proliferation are proposed as regulators of luteal angiogenesis. These include prostaglandin E, epidermal growth factor (EGF), endothelial growth factor, endothelium-stimulating factor, angiogenin, insulin and transferrin (Findlay, 1986). Development of luteal vasculature is a dynamic process which varies with the stage of the estrous cycle. Zheng et al. (1993) observed that capillary density within luteal tissue is sparse in the early luteal phase (days 1-4 post ovulation), high in the middle phase (days 5-17), and is reduced dramatically in the late phase (days 18-21). Since a reduction in tissue function is usually associated with a decline in blood flow and vascularity, the fall in blood flow and vascularity during the late luteal phase could indicate degeneration of luteal cells and a decline in luteal function (Zheng et al., 1993). Redmer et al. (1988) observed that luteal-conditioned medium from early (days 1-4), mid (days 5-17), and late (days 18-21) ovine cycle stimulates angiogenesis (mitogenesis and migration of endothelial cells. Angiogenic activity increases with advancement of the luteal phase. In that study, LH stimulated the production and/or release of

PAGE 37

20 the angiogenic factors, while PGFja blocked the LH-induced stimulation (Redmer at al., 1988). Thus, luteal vasculature could be stimulated in an autocrine/paracrine manner by angiogenic factors produced by the corpus luteum. It has also been suggested that growth factors may regulate ovarian angiogenesis since they are present in the ovary, and have effects on endothelial cells. Koos (1989) speculated on the possible roles of fibroblast growth factor, insulin-like growth factors (IGFs), EGF, TGFa and TGFfi in ovarian angiogenesis. Tumor necrosis factor alpha (TNFa), PGE^, PGEj, estradiol, plasminogen activator proteolytic enzymes (plasminogen activator, plasmin, collagenase) also stimulate angiogenesis. However, it is not clear how the expression and activities of these factors are regulated, and how they modulate luteal angiogenesis. In a recent study by Zheng et al. (1993), pattern of immunostaining for heparin binding growth factor (HBGF) (also known as basic FGF) in bovine luteal tissue was parallel to that of luteal vascular development throughout the estrous cycle, suggesting a role of HBGF in vascular development. Follicular granulosa and theca interna cells, macrophages, endothelial cells, and mast cells have been implicated as involved in the regulation of angiogenesis in the ovary and the CL (Koos, 1989). Brannstrom and Norman (1993) proposed that mast cells present in CL of some species may modulate the luteinization process by producing and secreting cytokines and proteases involved in tissue remodelling, angiogenesis and

PAGE 38

21 stimulation of progesterone production. Mechanisms involved in the control of blood flow to the CL are not known and are still hypothetical. Studies in the rabbit led to rejection of the hypothesis that luteotropic hormones promote luteal blood flow (Keyes and Wiltbank, 1988). Rather, it was suggested that blood flow to the CL is not regulated by luteotropic hormones, and has no correlation with the level of luteal steroidogenesis. Regulation of Luteal Function During the Estrous Cycle With its formation, the CL is composed of mainly steroidogenic small and large luteal cells, and endothelial cells, pericytes, macrophages, smooth muscle cells and fibroblasts. The number, volume and density of small and large luteal cells vary throughout the estrous cycle, but volume of CL occupied by each cell type stays relatively constant (Niswender et al., 1994). Progesterone is the major hormone produced by the CL during the luteal phase of the cycle and LH is the major luteotropin that stimulates luteal progesterone production in several species. The luteotropic regulation of the CL has been the subject of much research. Corpus luteum formation is thought to be induced by the preovulatory LH surge. Dependence of the CL on LH has been tested in several studies involving hypophysectomy, administration of LH antibodies, and administration of GnRH antagonists. In sheep, hypophysectomy 5 days post estrus does not affect serum and luteal progesterone concentrations, although CL weight on day

PAGE 39

22 12 was lower than expected (Farin et al., 1990). These results indicated that a CL deprived of LH may still function although its growth and development could be compromised. Baird (1992) observed that most LH pulses on days 6-7 and 13-14 of the ovine cycle are followed by a rise in progesterone concentration. However, changes occur in concentrations of progesterone independent of LH pulses. Injection of a GnRH antagonist during early luteal phase also causes a small decline in progesterone production, whereas administration of GnRH antagonist on day 13 causes a rapid decline in progesterone secretion, and luteal regression (Baird, 1992). Thus the CL requires a luteotropic support from LH during early and midluteal phases of the cycle. However, the CL of mid-cycle seems to be less resistant to withdrawal of luteotropic support than the early luteal phase CL. It was suggested that high levels of progesterone during the mid-luteal phase cause a reduction in frequency of endogenous LH pulses. Also, uterine PGFzc which is secreted during the mid-luteal phase, may interfere with coupling of LH to the adenyl cyclase second messenger system, and increase sensitivity of the CL to PGFja as a result of the long intervals between LH pulses. These events would result in luteal regression (Baird, 1992).

PAGE 40

23 Luteal Regression When the mature follicle is not fertilized, the CL is eventually destroyed thereby allowing the female another opportunity to start a new cycle, ovulate and become pregnant. Destruction of the CL is referred to as luteolysis or luteal regression. The basic features of luteolysis are a decline in progesterone secretion which is referred to as functional luteolysis, followed by structural changes which lead to breakdown of luteal cells. The drop in circulating progesterone reduces the negative feedback of progesterone on the pituitary and leads to an increase in gonadotropin pulse frequency, a new wave of follicular growth, and ovulation. Role of PGF ,^ in Luteal Regression The mechanism of luteal regression has been the subject of numerous studies. Prostaglandin Fja from the uterus is the natural luteolysin in domestic farm animals and most rodents (McCracken et al., 1972; Knickerbocker et al., 1988). Due to the close apposition of the uterine and ovarian blood vessels, there is a local transport of PGF2„ from the uterine vein to the ovarian artery, making it possible for PGFja to reach the CL without going through systemic circulation (Del Campo and Ginther, 1973). The luteolytic effect of PGFja occurs through its interaction with ovarian oxytocin. The pulsatile release of oxytocin from the CL stimulates the pulsatile release of PGF2„ from the uterus, which in turn positively feeds back to further increase luteal oxytocin release. This

PAGE 41

24 positive feedback loop continues until the demise of the corpus luteum (Jenkin, 1992a). The endometrial oxytocin receptor is thought to be the determining factor as to whether or not luteolysis will occur (Flint et al., 1992a). Inhibition of uterine oxytocin with a synthetic oxytocin receptor antagonist prevents pulsatile release of PGFja, and luteolysis (Jenkin, 1992a). It is believed that trophoblast interferons in ruminants prevent luteal regression during early pregnancy via a similar mechanism (Flint et al., 1992; Jenkin, 1992b). This local effect of PGFza is not present in horses because of a different uterine-ovarian vascular anatomy and the equine CL will regress even in the absence of the ipsilateral uterine horn. In contrast to the horse, removal of the uterine horn ipsilateral to the CL in ruminants, pigs and some rodents prevents luteal regression, whereas removal of the contralateral horn has no effect on luteal lifespan (Ginther, 1974). Fairclough et al. (1981) reported that passive immunization of cows and ewes with PGF antibodies prolonged the estrous cycle, demonstrating RGFj^'s luteolytic function. Copelin et al. (1989) reported that cows actively immunized against PGF2a exhibit prolonged luteal lifespan after first ovulation. Cows with higher PGFjc antibody titres had longer luteal lifespan and progesterone secretion (Copelin et al., 1989). In a more recent study, active immunization of ewes against PGFsa on day 5 postpartum prevented ovulation (Bettencourt et al., 1993). However, in rhesus monkeys, humans and dogs, the uterus does not seem to be needed for luteal regression (Neill et al., 1969), and an intraluteal

PAGE 42

25 production of PGFja has been suggested in these species (Niswender et al., 1994). Several mechanisms have been proposed by which PGFjc exerts its negative effects on luteal function. Nett et al. (1976) proposed that PGFja causes degeneration of luteal cells by causing a reduction in blood flow leading to hypoxia within luteal tissue. Nett et al. (1976) observed degeneration of capillary endothelial cells during luteal regression. It has been suggested that the decrease in blood flow could be due to a degeneration of luteal capillaries rather than vasoconstrictive effects of PGF2„ (Wiltbank et al., 1990). The changes in luteal vasculature observed across the estrous cycle correlated with luteal growth, development, and regression (Zheng et al., 1993), and supports earlier reports on the role of blood flow in the control of luteal regression. Capillary density was low In the early luteal phase (days 1-4), high in mid-cycle (days 5-17), and decreased dramatically in the late stage (days 18-21). Like in cattle, PGFjc, rapidly decreases luteal blood flow in ewes with a corresponding decline in circulating progesterone. It has been proposed that PGFja decreases LH binding to luteal cells in vivo and may block stimulation of adenylate cyclase by LH (Niswender and Nett, 1994). In vitro, PGFja has also been shown to block the LH-induced increase in cAMP and progesterone production by ovine luteal tissue (Niswender and Nett, 1994). Wiltbank and Niswender (1992) proposed that the luteolytic action of PGFza involves binding of PGFjc to a specific membrane receptor on large luteal

PAGE 43

26 cells, activation of phosphoinositide-specific phospholipase C, and an increase in intracellular calcium which activates protein kinase C (PKC). PKC inhibits intracellular cholesterol transport leading to a decrease in progesterone production (Wiltbank and Niswender, 1992). Thus the antisteroidogenic effects of PGFja are mediated through the PKC second messenger system (Niswender et al., 1994). The sustained increase in free intracellular calcium causes degeneration and death of large luteal cells (Wiltbank et al., 1989). It has also been proposed that the antisteroidogenic effects of PGFsc on the CL could be due to a reduction in the number of LH receptors (Behrman et al., 1978), and the uncoupling of the LH receptors from the adenylate cyclase second messenger system (Fletcher and Niswender, 1982). It has been shown that treatment of luteal cells with PCFj^ inhibits formation of cAMP by LH in vitro (Dorflingeretal., 1983). Apart from intracellular changes, the CL undergoes morphological changes during regression. The plasma membrane of the regressing CL contains gap junctions, maculae adherens, coated invaginations and microvilli (Niswender and Nett, 1994). In the bovine CL, a decrease in amount of smooth ER, an increase in number of autophagic vacuoles and an increase in number of lipid droplets in cytoplasm are also observed during regression (Fields et al., 1992). Other morphological changes observed in regressing bovine CL are a decrease in the number of secretory granules, presence of numerous swollen mitochondria, and a decrease in size of the steroidogenic cells (Niswender and

PAGE 44

27 Nett, 1994). These morphological changes are common to several species. The increase in lipid droplets and cytoplasmic vacuoles have been observed in guinea pigs (Paavola, 1979), humans (Vanlennys and Madden, 1965) and rabbits (Koering and Thor, 1978). These observations led to suggestions that the immune system may play a role in luteal regression. Bovine luteal cells have been shown to express MHC class II antigens and their expression increase with advancing age of the corpus luteum (Pate, 1994). Expression of MHC class II antigens is restricted to the large luteal cells during mid-cycle, but is observed in both the small and large luteal cells prior to luteal regression (Benyo et al., 1991). Luteal MHC II antigen expression has also been observed during PGFjc-induced luteolysis, but expression is absent in pregnant cows (Benyo et al., 1991). Interferon-gamma has also been shown to induce expression of MHC II in bovine luteal cells (Fairchild and Pate, 1989). Thus IFN-gamma may contribute to the luteolytic process by stimulating luteal prostaglandin synthesis and inhibiting progesterone production (Pate, 1994). Role of Oxygen Free Radicals in Luteal Regression The role of oxygen free radicals in the luteolytic process has been the subject of recent studies. In the rat, luteal levels of superoxide anions and hydrogen peroxide increase after treatment with PGFza, and during regression (Sawada and Carlson, 1989). It has been suggested that oxygen radicals may cause lipid peroxidation, which in turn, stimulates luteal PGFjc production, which

PAGE 45

28 may contribute to luteolysis (Carlson et al., 1993). However, the mechanism by which superoxide radicals inhibit progesterone secretion in vivo is not known, although superoxide radicals have been shown to inhibit luteinizing hormone (LH) stimulation of cAMP, and cAMP-induced progesterone secretion in rat luteal ceil cultures (Gatzuli et al., 1991). Superoxide dismutase, the enzyme that converts superoxide anions to hydrogen peroxide, increases as the luteal phase progresses and LH has been shown to induce its synthesis (Laloraya et al., 1988). These results suggest that oxygen radicals are involved in the luteolytic process since their removal favors synthesis of progesterone, as opposed to the decline in progesterone observed during luteal regression. Corpus Luteum of Pregnancy When fertilization occurs following ovulation, the CL does not regress but rather is maintained and becomes responsible for producing progesterone required to maintain pregnancy. In the cow, the CL is the major source of circulating progesterone during most of pregnancy. Later in pregnancy, the placenta can adequately provide progesterone. Thus in the cow, the CL is not required as a source of progesterone after day 200 of pregnancy. However in the pig (Nara et al., 1982) and the rat (Steinetz et al., 1976), the CL is required throughout pregnancy.

PAGE 46

29 CL Morphology During Pregnancy Morphologically, the CL of pregnancy Is similar to that of the estrous cycle (Fields et al., 1985). It is made up of steroidogenic large and small luteal cells, macrophages, lymphocytes, endothelial cells, fibroblasts, as discussed earlier. However, Weber et al. (1987) observed some morphological differences between steroidogenic cells of the cycle and those of pregnancy. Viability of small luteal cells is significantly higher in cyclic than in pregnant cows, while viability of large cells is not different between estrous cycle and pregnancy (Weber et al., 1987). The significance of these observations is not known. Biochemically, the CL of pregnancy differs from that of the cycle. Luteal cells of late pregnancy produce less progesterone than those of nonpregnant cows (Fields et al., 1985). Weber et al. (1987) observed that large luteal cells of pregnant cows produce 30 times less progesterone than those of nonpregnant cows. It has also been shown that small luteal cells of pregnant cows are unresponsive to exogenous LH in contrast to the small luteal cells of nonpregnant cows (Weber et al., 1987). More recent studies indicate differences in protein synthesis and secretion between luteal cells of pregnancy and the estrous cycle. In the cow, oxytocin and neurophysin are localized within secretory granules of large luteal cells during the cycle, but are absent in luteal secretory granules after day 40 of pregnancy (Fields et al., 1992). When pregnancy occurs, biochemical communications take place

PAGE 47

30 between the conceptus and the mother to prevent regression of the corpus luteum, and thus sustain production of progesterone required to maintain pregnancy. This phenomenon is referred to as maternal recognition of pregnancy, whereby the conceptus sends signals to the maternal system to prevent regression of the corpus luteum (Short, 1969). The strategies used for maternal recognition of pregnancy vary among species and involve different proteins. Maternal Recognition of Pregnancy In cattle, luteolysis is prevented during pregnancy by inhibition of both basal and oxytocinor estradiol-stimulated PGFja secretion via synthesis of an endometrial prostaglandin synthase inhibitor (Thatcher et al., 1992). In the cow, presence of the CL is required for maintenance of pregnancy through the first 200 days of gestation. Ovariectomy of pregnant cows prior to day 200 results in abortions (Estergreen et al., 1967). The bovine placenta does not contribute significantly to circulating progesterone concentrations even after day 200 of gestation. Maternal recognition of pregnancy in cattle occurs on days 16-17 postestrus (Niswender and Nett, 1994). At this time, the conceptus secretes embryonic interferons (interferon-tau) which act via endometrial receptors and alter the secretion of endometrial PCFja (Roberts et al., 1992). Earlier studies have shown that high amplitude pulses of PGFjo, occur in nonpregnant heifers

PAGE 48

31 during luteolysis, but are absent in pregnant ones (Kindahl et a!., 1976). Similar observations have been reported in ewes (Zarco et al., 1988) and buffaloes (Batra and Pandey, 1983). Following the transfer of day 15 or 16 bovine embryos to recipients, recipients with regressed CL were observed to have four to five spikes of PGFM, whereas recipients in which the CL persisted had reduced or no PGFM spikes (Betteridge et al., 1984). PGF2a is the uterine luteolysin in cows. Thatcher et al. (1985) observed that the estradiol-induced increase in PGFM production is inhibited by the presence of a conceptus on day 18 of pregnancy, but is not in cyclic cows on day 18 of the estrous cycle. The conceptus provides greater inhibition of PGFM production on day 20 than the day 18 conceptus because the former has a more extensive contact with the endometrium (Thatcher et al., 1985). Similarly, the oxytocin-induced increase in uterine PGFM is significantly less in pregnant than in nonpregnant heifers on day 19 postestrus (LaFrance and Goff, 1985). These studies suggested an antiluteolytic-antiPGF effect of the conceptus. The conceptus mediates its antiluteolytic-antiPGF effect via the secretion of proteins. Cyclic cows receiving intrauterine injections containing secretory proteins found in days 16-18 conceptus-conditioned medium have longer interestrous intervals than those cows receiving serum proteins (Thatcher et al., 1985). Intrauterine injections of bovine conceptus proteins also reduce estradiol-stimulated PGFM production (Bazer et al., 1986). The major proteins secreted by cultured bovine conceptuses have molecular weights of 22-26 kDa

PAGE 49

32 and isoelectric points of 5.6-5.8. The protein is referred to as interferon-tau and it shares 50% amino acid sequence identity with recombinant bovine interferonalpha (rblFNa) (Imakawa et al., 1989). Bovine trophoblast protein-1 given at time of maternal recognition of pregnancy extends luteal function in the cow (Thatcher et al., 1989) and ewe (Parkinson et al., 1992). Similarly, intrauterine infusion of recombinant bovine interferon-a extends the length of the estrous cycle in post-partum cows expected to have short luteal lifespan (Garverick et al., 1992). The recombinant interferon acts by reducing oxytocin-induced PGFM release (Plante et al., 1990). Treatment of ewes with recombinant bovine interferon-alpha I on days 9-19 post-estrus caused a reduction in plasma concentrations of PGFM when compared to control groups (Parkinson et al., 1992). Bovine trophoblast protein-1 exerts its antiluteolytic effects by inhibiting synthesis and/or recycling of endometrial oxytocin receptors directly, or by inducing synthesis of a PGFjo, synthase inhibitor (Bazer et al., 1991). However, the antiluteolytic signals do not seem to act directly on the CL (Bazer et al., 1991). Bovine conceptuses also produce PGEj and small amounts of estradiol which together stimulate increase in uterine blood flow which may enhance delivery of antiluteolytic-luteoprotective agents to the ovary (Lewis et al., 1982; Thatcher etal., 1986). In the ewe, maternal recognition of pregnancy occurs on days 12-13 postestrus. Although serum progesterone concentrations are similar between pregnant and nonpregnant ewes at this time, the conceptus prevents regression

PAGE 50

33 of the CL in pregnant ewes (Niswender and Nett, 1994). During the ovine cycle, prostaglandin Fjc is produced by the endometrium and transported to the ovary where it causes regression of the CL and a decline in progesterone production. Between days 12-21 of gestation, trophoblast cells of ovine blastocysts secrete proteins including a major secretory 17 kDa protein (interferon-tau). IFN-tau is not produced beyond day 21 of gestation, and is the only secretory product detected on day 13 (Godkin et al., 1982). Intrauterine infusion of IFN-tau on days 12-18 of the estrous cycle extends corpus luteum lifespan in ewes, while the CL regresses in untreated ewes. It was also observed that ewes treated with sheep serum ovulated and formed a new CL while ewes treated with total conceptus proteins did not ovulate (Godkin et al., 1984). Fincher et al. (1984) showed that IFN-tau inhibits estradioland oxytocin-induced uterine production of PGFjaThus IFN-tau is antiluteolytic and anti-PGF. Oxytocin-induced PGF production is lower in pregnant than in nonpregnant ewes (Fairclough et al., 1984). The mechanism(s) of action of IFN-tau is not clear but seems to be via its binding to endometrial receptors (Godkin et al., 1984; Hansen et al., 1989), causing changes in the secretion of endometrial proteins and prostaglandins, and extending luteal function (Vallet et al., 1988). Treatment of ewes with ovine conceptus secretory proteins (oCSP) on days 11-15 post-estrus causes a decline in concentrations of endometrial estrogen receptors, estrogen receptor mRNA, and progesterone receptor on day 16 when compared with ewes treated

PAGE 51

34 with serum proteins. Ovine conceptus secretory proteins also reduce oxytocin binding and activation of the phosphoinositol second messenger system (Mirando et a!., 1993). In another study, maximum expression of endometrial progesterone receptor mRNA occurred earlier (days 10-12 post-estrus vs days 14-16) in pregnant than in cyclic ewes, and oxytocin stimulated in vitro endometrial production of inositol phosphates in cyclic but not in pregnant ewes (Ottetal., 1993). Nephew et al. (1989) have shown that intramuscular injections of recombinant bovine interferon-tau cause an increase in pregnancy rate, prolificacy and higher survival of conceptuses in ewes. Also, plasma concentrations of PGEj in utero-ovarian vein in ewes increase during maternal recognition of pregnancy (Silvia et al., 1984). IFN-tau and other conceptus proteins also favor production of PGEj over PGFja and thus prevent luteal regression. Ovine IFN-tau, like bovine IFN-tau, acts by binding to endometrial receptors (Godkin et al., 1984; Hansen et al., 1989), changes secretory patterns of endometrial proteins and prostaglandins, and extends luteal function (Vallet et al., 1988). Intrauterine infusions of oCSP and bovine recombinant interferonalpha 1 on days 12, 13 and 14 of the estrous cycle cause a decline in concentrations of endometrial oxytocin receptor, and the oxytocin-induced increase in PGFM (Vallet and Lamming, 1991). Ovine IFN-tau genes are expressed specifically by cells of the trophectoderm and are regulated in a developmental manner. Expression of the

PAGE 52

35 gene is present in day 10-11 blastocysts, Increases by day 13, declines slightly by day 15 and sharply thereafter (Guillemot et al., 1990). Roberts et al. (1992) have also shown that IFN-tau gene expression is undetectable by day 22, a time when most of the trophoblast is attached to the uterine epithelium. The amount of protein produced by the blastocysts correlated with the expression of olFN-tau mRNA (Roberts et al., 1992). Although olFN-tau gene expression ishe amount developmentally regulated but other factors may affect its production. In sheep and cattle, an advanced luteal phase enhances conceptus development and earlier expression of IFN-tau (Garrett et al., 1988; Nephew et al., 1991). Xavier et al (1991) observed a simultaneous expression of c-fos proto-oncogenes and IFN-tau in ovine trophoblasts. Also in the pig, endometrial expression of c-fos mRNA increases on day 12 of pregnancy (day of maternal recognition of pregnancy) and is higher when compared to expression on day 12 of the cycle (Dubois et al., 1993). Thus c-fos may be induced by IFN-tau or may be involved in the transcriptional activation of IFN-tau genes. However, no interferon response elements are present on the c-fos gene, and no AP-1 binding sites have been found in the promoter region of IFN-tau genes. It is also possible that both c-fos and IFN-tau genes are regulated by common mechanisms during the period of maternal recognition of pregnancy (Roberts et al., 1992).

PAGE 53

36 Protein Synthesis by the Corpus Luteum Apart from its role in the synthesis of progesterone required for the maintenance of pregnancy and control of the estrous cycle, the CL has been shown to produce a number of proteins, peptide hormones and factors. However, most of the proteins have not been fully characterized and attempts to define their functions have been mainly speculative. The CL produces oxytocin, neurophysin, relaxin, inhibin, vasopressin, Q> endorphin, growth factors, angiogenic factors and protease inhibitors. Results in this dissertation have shown that the CL also produces many proteins including apolipoproteins E and A-1, tissue inhibitor of metalloproteinases-1 and 2, and manganese superoxide dismutase (see table 1-1). The types of proteins produced by the CL vary with species and the reproductive state of the animal. Relaxin is produced by CL of human, pig, rat, but not by CL of ruminant species (Sherwood, 1994). On the other hand, oxytocin and neurophysin are produced by CL of ruminants but not by nonruminants. These proteins carry out different functions in the CL, the ovary, and at extraovarian sites. The chemical nature and the physiological roles of some of these proteins have not been fully defined. Luteal proteins and factors are produced by different cell types within the CL. Cell-cell communication takes place in order to co-ordinate the functions of luteal cells, and regulate synthesis and release of proteins. Factors regulating

PAGE 54

37 synthesis, release, and biological activity of luteal proteins/factors are not fully understood. General Overview of Protein Synthesis and Release In order to produce a protein, a cell must possess the genetic material (gene) that codes for that specific protein. The DNA template directs the synthesis of RNAs which are involved in protein synthesis. Evidence for RNA involvement in protein synthesis was reported in 1 930 when it was observed that a crude preparation of RNA was rich in protein, and the concentration of RNAprotein particles (ribosomes) correlated with the rate of protein synthesis by the cell. Francis Crick in 1958 defined the relationship between DNA, RNA and protein (Crick, 1970). Proteins are synthesized on ribosomes from specific mRNAs. The mRNA specifies the protein to be synthesized, and associates with the ribosomes to initiate the process. More recently, Xing et al. (1993) showed that mRNAs are produced at specific locations in the nucleus and are then exported to the protein synthesizing machinery in the cytoplasm. RNA metabolism is also organized in the nucleus in association with the nuclear matrix (Carter et al., 1993). It has been suggested that the nuclear matrix may determine what genes are turned on by sequestering and concentrating the DNA to be transcribed, as well as the transcription factors necessary for a specific gene expression. Ribosomes associate with transfer RNA and move along the mRNA to form peptide bonds.

PAGE 55

38 Protein Secretion Most cells have the ability to secrete proteins. Proteins are secreted via either a regulatory or a constitutive pathway. Secretion of most eukaryotic proteins require their transport across the endoplasmic reticulum (ER) membrane. This process occurs in two steps; targeting, followed by active transfer across the ER membrane (Rapoport, 1 992). Protein Targeting Secretory proteins synthesized on ribosomes in the cytoplasm are targeted to the ER membrane by signal sequences. The signal sequence is targeted to the ER membrane by the Signal Recognition Particle (SRP). The nascent peptide chain-ribosome-SRP complex binds to the ER membrane by an interaction with the membrane-bound SRP receptor or docking protein. Guanosine triphosphate (GTP) hydrolysis causes the SRP to detach from the ribosome and signal sequence, then the nascent peptide is transferred to the ER membrane while the ribosome stays bound to the membrane via a ribosome receptor. Finally, a GTP hydrolysis reaction causes the SRP to dissociate from its receptor, and a new targeting cycle can begin (Rapoport, 1992).

PAGE 56

39 Translocation Across ER Membrane Proteins targeted to the ER membrane are transported across the membrane at specific sites. The translocation site is a complex and dynamic structure composed of many proteins and enzymes that catalyze modification of nascent peptides. Although proteins seem to be translocated through proteinconducting channels, the mechanism is not clear (Rapoport, 1992). Most secretory proteins are translocated as precursors with larger masses than the final mature protein. The mature protein is obtained after cleavage of the signal sequence during passage across the ER membrane and the Golgi apparatus. Processing and Sorting of Proteins in the Golci After transversing the ER, the protein is transported in vesicles to the Golgi complex via an energy-dependent process. In the Golgi, the protein may be post-translationally modified by attachment of functional groups (glycosylation, acetylation, sulfation). The protein is either concentrated in secretory granules and released in response to an appropriate stimulus (regulatory pathway) or is transported to the cell surface in a vesicle and released directly (constitutive pathway). Exocytosis involves an interaction between proteins on the cytoplasmic surface of secretory granules (vesicles) and proteins of the inner surface of the plasma membrane (Widnell and Pfenninger, 1990).

PAGE 57

40 One of the objectives of this research was to identify proteins synthesized and released in culture by the bovine CL during the estrous cycle and pregnancy. Five of the proteins were identified by N-terminal amino acid sequence analysis. The next section of this chapter discusses the identified proteins. Apolipoprotein A-1 Biochemical Characterization Apolipoprotein A-1 (Apo A-1) is the major protein component of high density lipoprotein (HDL), and accounts for 80% or more of the protein moiety of HDL in the cow and all mammalian species (Sparrow et al., 1992). Apolipoprotein A-1 has metabolic and structural roles since it contributes to the size and shape of the lipoprotein particle, solubilizes water-insoluble lipids, and is a potent activator of lecithin:cholesterol acyltransferase (LCAT). Apolipoprotein A-1 is also involved in the recognition and modulation of enzymes involved in lipid metabolism, and binding of lipoproteins to their cellular receptors (Hopkins et al., 1986).

PAGE 58

41 Apo A-1 gene Mammalian Apo A-1 gene shows no sinking region of evolutionary conservation, and the bovine Apo A-1 gene is more closely related phylogenetically to canine than to human and other mammalian lineages (O'hUigin et al., 1990). The cDNA of Apo A-1 cloned from a bovine cDNA library (longest insert 963 nucleotides) was shown to contain an open reading frame of 795 nucleotides flanked by 72 and 96 nucleotides at the 5' and 3" end, respectively (O'hUigin et al., 1990). The 3' flanking region contains a polyadenylation signal (AATAAA) 14 nucleotides upstream of a poly-A tail. Based on the cDNA sequence the derived amino acid sequence contains an 18residue signal peptide and a 6-residue prosegment. Metabolism Apo A-1 is synthesized as a prepro-Apo A-1 mainly by the liver and intestine, but is also synthesized by other peripheral tissues such as kidney, adrenal, and testis (Blue et al., 1982), and brain endothelial cells (Guttler et al., 1990). Sorci-Thomas et al. (1988) showed that the liver and small intestine contribute to most of plasma Apo A-1 and suggested that other tissues observed to synthesize Apo A-1 may not contribute significantly to the plasma Apo A-1 pool, but may play a role in lipid metabolism within these tissues in an autocrine and /or paracrine manner.

PAGE 59

42 The primary structure of Apo A-1 varies with species. Bovine Apo A-1 is composed of 241 amino acid residues. The propeptide has a sequence (ArgHis-Phe-Trp-Gln-GIn), and approximately 10% of bovine plasma Apo A-1 is in the propeptide form (Sparrow et al., 1992). Newly-synthesized Apo A-1 from different tissues exists In four Isoforms (two major and two minor) with Isoelectric points ranging between 5.3 and 5.7, similar to Apo A-1 from the liver (Blue et al., 1982). The nucleotide and deduced amino acid sequence of bovine Apo A-1 shares 80% homology with the human and rabbit sequences (Gu et al., 1993). The central region of bovine Apo A-1 is hydrophobic, with highly hydrophlllc regions at the amino and carboxy termini (O'hUigin et al., 1990). The hydrophobic amphlpathic helical regions are necessary for interaction of apoprotein with phospholipid-cholesterol complexes (Sparrow et al., 1992). Bovine Apo A-1 contains a single methionine and no cysteine as do the canine and rabbit proteins (O'hUigin et al., 1990). The primary translation product is the prepropeptlde. It has been suggested that the bovine Apo A-1 prepropeptlde like that of the human (Gordon et al., 1983) and the rat Is post-translationally modified. In humans, Intestinal proapo A-1 contains a hexapeptide extension which ends with GIn-GIn and this precursor was shown to be secreted by Hep G2, hepatocarcinoma cells In culture without proteolytic cleavage of the hexapeptide prosegment (Gordon et a!., 1983). Thus it was suggested that Apo A-1 undergoes additional proteolytic processing before it is integrated into plasma HDL (Gordon et al., 1983). In the

PAGE 60

43 cow, cleavage may occur after the conserved GIn-GIn dipeptide to give a mature Apo A-1 protein with an N-terminal aspartate (O'hUigin et al., 1990). The conversion of proApo A-1 to mature Apo A-1 is known to occur extracellularly by an enzyme present in plasma. Edelstein et al. (1988) also showed that this enzyme produced by a hepatocarcinoma cell line (Hep G2) secretes both Apo A1 and the converting enzyme. The converting enzyme is activated by calcium, inhibited by EDTA, and converts proApo A-1 to Apo A-1 through a first order kinetic reaction (Edelstein et al., 1988). In another study, Chinese hamster ovary (CHO) cells transfected with human Apo A-1 secreted Apo A-1. Furthermore, 90% of the secreted Apo A-1 was the processed mature protein, and a portion of the secreted protein was associated with lipid (Mallory et al., 1987). Thus processing of Apo A-1 seems to take place prior to its secretion. Role of Apo A-1 Apo A-1 is the major protein constituent of high density lipoprotein and it mediates the binding of HDL to cells. HDL is the major source of circulating cholesterol in bovine species (Sparrow et al., 1992). Pate and Condon (1989) showed that both LDL and HDL could be used as a source of cholesterol for steroidogenesis by bovine luteal cells, and both LDL and HDL enhance luteal progesterone synthesis in culture. In addition to its ability to solubilize and transport lipids, Apo A-1 is also a potent activator of lecithin-cholesteryl acyltransferase, the enzyme that catalyses formation of cholesterol esters from

PAGE 61

44 cholesterol (Soutar et al., 1975). It has also been shown that HDL, as opposed to LDL, causes an increase in the release of placental lactogen by human placental explants (Handwerger et al., 1987), and from monolayer of trophoblast cells (Sane et al., 1988). Apo A-1 was implicated for the HDL-mediated stimulation of placental lactogen release. Wu et al. (1988) showed that HDL stimulates placental lactogen release by stimulating production of cAMP. Thus cAMP is a second messenger in HDL-mediated release of hPL, and HDL may carry out other functions in steroidogenic cells of the ovary by stimulating adenylate cyclase activity and cAMP production. Regulation of Apo A-1 Synthesis by Steroid Hormones There are reports to indicate that estrogen may regulate synthesis of Apo A-1 by the liver. Archer et al. (1986) reported that treatment of human hepatoma cell line (HepG2) with estradiol-1711, causes an increase in nuclear estrogen binding sites, and a parallel increase in the expression of Apo A-1 mRNA and rate of accumulation of the protein. The increase in mRNA levels accounted for 85-90% of the observed increase in rate of accumulation of secreted protein (Archer et al., 1986). A study with ovariectomized baboons showed that baboons treated with estradiol and progesterone had the highest serum concentrations of Apo A-1, followed by those treated with estrogen alone, and lowest in the progesterone-treated animals and the untreated controls. Baboons treated with progesterone alone had similar levels of serum Apo A-1 similar to

PAGE 62

45 those of untreated controls (Kushwaha et al., 1990). Apo A-1 levels were significantly upregulated by estradiol and progesterone compared to untreated controls (Kushwaha et al., 1990). Thyroid hormones also regulate expression of Apo A-1 mRNA. Apo A-1 gene is stimulated by triiodothyronine (T3) and has been shown to contain a thyroid hormone response element which is critical for the T3-induction of Apo A-1 mRNA and activity of Apo A-1 promoter (Romney et al., 1992). These observations were supported by reports from Chan et al. (1993) that the Apo A-1 gene contains a cis-regulatory element which acts on an adjacent site to increase promoter activity. HNF-4, a new member of the thyroid/steroid hormone receptor superfamily, was shown to interact with the cis element to enhance activity of the rat Apo A-1 promoter (Chan et al., 1993). Effects of Nutrition on Apo A-1 Synthesis Dietary carbohydrates or fatty acids regulate Apo A-1 gene expression by altering either gene transcription or mRNA stability. Synthesis and secretion of Apo A-1 is reduced in hepatocytes from rats fed fish oil (low source of cholesterol), but the diet did not affect levels of Apo A-1 mRNA (Ribeiro et al., 1992). Availability of cholesterol has been shown to enhance synthesis of Apo A-1 by human hepatoma cells (Craig et al., 1988). Similarly, Go et al. (1988) showed that synthesis of hepatic and intestinal Apo A-1 increases while levels of Apo A-1 mRNA decrease, following chronic fat and cholesterol feeding.

PAGE 63

46 However, hepatic and intestinal synthesis of Apo A-1 is higher in African green monkeys than in Cynomolgus monkeys fed the same level of cholesterol (SorciThomas et al., 1988). Thus other factors independent of dietary cholesterol intake may also regulate hepatic and intestinal Apo A-1 synthesis. Apolipoprotein E Biochemical Characterization Apolipoprotein E (Apo E), sometimes referred to as arginine-rich protein, is a component of very low density lipoprotein (VLDL), HDL and LDL. Apolipoprotein E has been shown to have a molecular weight ranging between 33 and 39 kDa on SDS-PAGE (Shelburne and Quarfordt, 1974). Apolipoprotein E gene is 3597 nucleotides in length and contains four exons and three introns (Paik et al., 1985), and a similar gene structure is shared by other apolipoproteins. The primary translation product is a pre-Apo E protein with an 18-residue signal peptide that is cleaved cotranslationally (Zannis et al., 1984), and the mature protein is secreted. The primary structure of Apo E ranges in length between 279 and 310 amino acid residues among different species. Apolipoprotein E sequence in the cow comprises 294 amino acid residues, and the most conserved region is between residues 28-61 (Yang et al., 1991). The receptor binding region (residues 130-158) is rich in basic amino acids and is conserved across species,

PAGE 64

47 except for point substitutions in the dog (arginine substituted for lysine at 157) and cow (proline substituted for arginine at 145) (Weisgraber, 1994). Apolipoprotein E has been shown to exhibit heterogeneity in molecular weight and charge which have been attributed to genetic variation and posttranslational glycosylation with sialic acid. Sialo-Apo E isoforms comprise 42% of intracellular Apo E, 81.1% newly-secreted Apo E, and 24% plasma Apo E (Zannis et al., 1984). Thus sialation may be required for the secretion of Apo E or glycosylated Apo E is preferentially secreted. There has been evidence to suggest that Apo E is glycosylated by 0-glycosidic linkage (Zannis et al., 1984). Liver is the major source of Apo E. However, Apo E and its mRNA is produced by most organs and by several cell types within the organs including astrocytes, smooth muscle cells and macrophages (Mahley, 1988). Effects of FSH. LH. cAMP, and Phorbol Ester on Apo E Synthesis Apo E is synthesized by the rat ovary and represents 0.15% of the total protein synthesized in the ovary (Driscoll and Getz. 1984). Secretion of newlysynthesized Apo E by granulosa cells in culture is stimulated by FSH in a doseand time-dependent manner, and the effects of FSH are mediated through cAMP (Driscoll et al., 1985). Results from that study also suggest that Apo E is secreted as part of a lipid-protein complex. As the granulosa cells differentiate in culture, they lose their responsiveness to FSH and cAMP (Driscoll et al., 1985). Polacek et al. (1992) showed that Apo E mRNA is localized

PAGE 65

48 predominantly in theca cells of rat ovarian follicle, and mRNA levels increase following treatment of cells with hCG (Polacek et al., 1992). In another study, Wyne et al. (1989a) demonstrated that BtcAMP and forskolin (an activator of adenylate cyclase and mediator of kinase A), and phorbol ester (mediator of kinase C) stimulate production of Apo E by granulosa cells in culture. BtcGMP (mediator of kinase G) did not stimulate secretion of Apo E (Wyne et al., 1989a). Kinases A and C had no effect on global protein synthesis in granulosa cells; incorporation of radiolabel into protein ranged between 10-15%, suggesting a specific stimulation of a subset of proteins including Apo E (Wyne et al., 1989a). In addition, cAMP, TPA and cholera toxin also stimulated expression of Apo E mRNA in rat granulosa cells. These agents stimulated accumulation of Apo E more than expression of its mRNA, indicating that kinases A and C may influence both the transcription of Apo E gene and the translational efficiency of Apo E mRNA. However, it is not yet clear if there is a crosstalk between the adenylate cyclase pathway (stimulated by cAMP) and the PKC pathway (stimulated by phorbol ester). The stimulatory effect of cholera toxin and TPA on Apo E secretion is inhibited by cycloheximide and actinomycin D, suggesting that new proteins (such as transcriptional activator proteins AP-1 and AP-2) are required to mediate the stimulatory effects (Wyne et al., 1989a). The rat Apo E gene does not contain a cAMP regulatory region (CRE), but contains sequences with 75% homology to this region. Similarly, the consensus sequence for AP-1 (AP-1

PAGE 66

49 responds to phorbol ester) is not present in the upstream region of the rat Apo E gene, but is found in the first intron. However, the consensus sequence for AP2, which responds to both cAMP and phorbol ester, is present in the upstream region of Apo E gene (Wyne et al., 1989a). Regulation of Apo E Synthesis by Cell Cholesterol Cholesterol is the substrate for steroid hormone biosynthesis. Cholesterol can either be newly synthesized from acetate or is obtained by uptake of lipoproteins (Schreiber et al., 1980). Wyne et al. (1989b) demonstrated that inhibition of cholesterol synthesis from acetate with mevinolin, an inhibitor of HMG-CoA reductase, causes a decline in cholera-stimulated Apo E synthesis and expression of Apo E mRNA by rat granulosa cells. Howeyer, an inhibitor of the cytochrome P450 side chain cleavage enzyme had no effect on Apo E synthesis (Wyne et al., 1989b). Human and rat Apo E gene has been shown to possess the consensus sequence of a sterol regulatory element in their 5' region. Prack et al. (1991 ) also reported that depletion of adrenal gland cholesterol content decreases Apo E mRNA levels. Thus cholesterol together with stimulators of kinases A and C are required to regulate Apo E production. Regulation of Apo E Secretion by Cytokines Macrophage Apo E secretion decreases with macrophage activation (Zuckerman and O'Neal, 1994). This effect is mediated by macrophage

PAGE 67

50 activating factors such as lipopolysaccharide (LPS) and granulocytemacrophage colony stimulating factor (GM-CSF). The LPS-mediated reduction in Apo E secretion is inhibited by monoclonal antibody to murine tumor necrosis factor (TNF) (Zuckerman and O'Neal, 1994). Role of Apo E in Ovarian Function The role of Apo E in the ovary has not yet been identified. One possibility is that since Apo E mediates binding of lipoproteins to their receptors, Apo E could function to provide cholesterol for membrane and steroid hormone biosynthesis. Apo E may also function in a paracrine fashion to distribute lipid between ovarian cells and perhaps between compartments of the ovary. It has been demonstrated that HDL containing Apo E, as opposed to HDL containing no Apo E, induces rat ovarian theca cells to produce progesterone rather than androgen (Dyer et al., 1988). Manganese Superoxide Dismutase Biochemical Characterization Superoxide dismutases play critical roles in protecting cells from oxidative damage by reactive oxygen species. Manganese superoxide dismutase (Mn SOD) is one of three (the others are Cu SOD and Zn SOD) enzymes that catalyze the dismutation of superoxide radicals to hydrogen peroxide and

PAGE 68

51 oxygen; 2O2 + 2H* = H2O2 + O2 Manganese SOD is localized in the mitochondrial matrix (Fridovich, 1974) and is not a secretory protein, while Cu SOD and Zn SOD are secretory proteins (Rueda et al., 1994). Manganese SOD has a molecular weight of 20 kDa. Two mRNAs transcripts of 4.0 kb and 1 .0 kb encode for human Mn SOD (Melendez and Baglioni, 1993). However, three Mn SOD mRNA transcripts (1.5, 1.9, and 3.7 kb) have been observed in the bovine CL (Rueda et al., 1995; NdikumMoffor et al. 1995 unpublished data The mRNA transcripts are from the same gene, have identical coding regions, but differ in length of their 3' untranslated region (3' UTR) because of polyadenylation (Church, 1990). The 4-kb mRNA is expressed at a faster rate than the 1-kb mRNA, but the 4-kb transcript has a shorter half-life (2-4 h in different cells) than the 1-kb transcript (10-12 h) in both intact cells and a cell-free system (Melendez and Baglioni, 1993). The different half-lives indicate a post-transcriptional regulation of Mn SOD mRNA, and the instability of the 4-kb transcript has been attributed to the presence of AU-rich sequences in the 3' UTR (Melendez and Baglioni, 1993), Manganese SOD activities are low under normal physiological conditions, but may increase during differentiation and in response to oxidants and cytokines. Regulation of Manganese SOD Production by Oxidative Stress Reactive oxygen species are generated in all cells in vivo, and the toxicity of oxygen has been shown to be directly related to the production of oxygen-

PAGE 69

52 dependent free radicals. Results from a study with yeast indicated that electron transport is a major source of superoxide anion in vivo (Guidot et al., 1993). Oxidative stress from the environment has also been shown to increase production of mitochondrial Mn SOD in plants (Bowler et al., 1991). Regulation of Manganese SOD Synthesis by Gonadotropins In the rat, Laloraya et al. (1988) demonstrated a sharp increase in rat ovarian SOD activity 30 min following an injection of LH, a decline 60 min postinjection, and no LH-induced SOD activity in rats injected with anti-LH serum. They also observed changes in ovarian SOD activity across the estrous cycle, with highest levels at proestrus. However, changes in SOD activity specific to Mn SOD across the cycle were not discussed (Laloraya et al., 1988). In another study, Sato et al. (1992) showed that rat ovarian Mn SOD activity decreases during a hCG-induced ovulation, to a minimum 12 h post-injection, while Mn SOD mRNA levels increase markedly with time to a maximum 12 h post-hCG treatment (Sato et al., 1992). Regulation of Manganese SOD Svnthesis by Cytokines and Phorbol Ester lnterleukin-1, TNF, and lipopolysaccharide dramatically increase Mn SOD mRNA levels in pulmonary epithelial cells (Visner et al., 1990). Similar observations were reported by White and Tsan (1994) who also showed that TNF and IL-1 enhance Mn SOD protein and enzyme activity.

PAGE 70

63 Whitsett et al. (1992) showed that TNF-alpha and phorbol ester (TPA) increase steady state mRNA and rate of transcription of human Mn SOD in pulmonary adenocarcinoma cells. The time course and extent of increased manganese SOD gene transcription by TNF-alpha was distinct from that exhibited by phorbol ester (Whitsett et al., 1992). Role of Mn SOD in the Ovary The role of SOD in the ovary has not been defined but there are indications that it might be involved in ovulation (Laloraya et al., 1988; Sato et al., 1992) and the luteolytic process (Wu et al., 1992; Rueda et al., 1995). It has been hypothesized that reactive oxygen species produced during normal metabolism may be potential mediators of luteal regression. A comparison of Mn SOD (a scavenger of superoxide radicals) gene expression between a functional CL (day 21 of pregnancy) and a regressed CL (day 21 of the estrous cycle) of the cow indicated that Mn SOD mRNA levels are higher in the functional than the regressed CL (Rueda et al., 1995). A lower expression of Mn SOD mRNA in the regressed CL suggests that cells within the regressed CL are less capable of metabolizing the superoxide radical, which may damage the cells and disrupt luteal function.

PAGE 71

54 Tissue Inhibitors of Metalloproteinases: TIMP-1 and TIMP-2 Tissue inhibitor of metalloproteinases are proteins which inhibit the activity of enzymes (matrix metalloproteinases) that degrade protein components of the extracellular matrix. Thus the expression of TIMPs is high in tissues undergoing remodelling or transformation. Apart from their protease-inhibitory activity, TIMP-1 (Hayakawa et al., 1992) and TIMP-2 (Stetler-Stevenson et al., 1992) have also been shown to stimulate growth of erythroid cells, gingival fibroblasts, and transformed human lung cells. Satoh et al. (1994) also demonstrated that TIMP-1 stimulates growth of bovine embryos in culture. A recent study by Boujrad et al. (1995) showed that TIMP-1 secreted by rat Sertoli cells stimulated steroidogenesis by rat Leydig cells. Activities of matrix metalloproteinases (collagenases, stromelysins, and gelatinases) may be controlled at various levels, one of which is by binding to specific inhibitors (TIMP). Thus a proper balance is required between the amount of inhibitors and the metalloproteinases to maintain tissue homeostasis or proper remodeling which occurs during many biological processes. To date, three members of the TIMP family have been identified, namely TIMP-1, TIMP-2, and TIMP-3. Comparison of deduced amino acid sequence of TIMP-2 showed that TIMP-1 and TIMP-2 share 37.6% identity of nucleotide and 65.6% similarity of amino acid at the protein level (Stetler-Stevenson et al., 1990), while human TIMP-3 shares 39 and 46% amino acid sequence identity with human TIMP-1 and TIMP-

PAGE 72

55 2, respectively (Silbiger et al., 1994). The positions of all twelve cysteine residues and three out of four tryptophans are conserved between TIMP-1 and TIMP-2. Tissue Inhibitor of Metalloproteinases-1 Biochemical Characterization TIMP-1 is a secreted glycosylated protein with molecular weight ranging from 28-30 kDa. TIMP-1 is expressed in many different tissues and cell types including monocytes, fibroblasts and macrophages. Ovine TIMP-1 shares 95, 86, and 77% nucleotide sequence with that reported for bovine, human, and mouse TIMP-1, respectively (Smith et al., 1994). The amino acid sequence of TIMP-1 deduced from the bovine cDNA sequence shows that the mature protein contains 12 cysteine residues (conserved among many species) and 2 Nglycosylation sites (Freudenstein et al., 1990). The nucleotide sequence of ovine TIMP-1 also indicates the presence of 12 cysteines and 2 N-linked glycosylation sites (Smith et a!., 1994). In addition, TIMP-1 contains a 23-amino acid signal peptide which contains a core of hydrophobic amino acids (Smith et al., 1994). TIMP-1 binds to active collagenase, and to the latent form of the 92 kDa gelatinase, and proteoglycans. TIMP-1 binds collagenase with high affinity in a 1:1 molar ratio to form an inactive noncovalent enzyme-inhibitor complex

PAGE 73

56 (Welgusetal., 1985). Regulation of TIMP-1 Synthesis by Gonadotropins Mann et al. (1991) showed that LH and phorbol ester (TPA) individually Increased metalloproteinase inhibitor activity of granulosa cells in culture in a dose-dependent manner, and the effects were additive. The inhibitor activity (identified as TIMP-1) was also stimulated by Br-cAMP and forskolin, and its mRNA levels increased before ovulation (Mann et al., 1991). In another study Mann et al. (1993) showed that cycloheximide inhibits basal, LHand TPAstimulated TIMP-1 activity, while indomethacin (an inhibitor of prostaglandin synthesis) or an antiestrogen did not affect basal or LH-induced rat granulosa cell inhibitory (TIMP-1) activity. Reich et al. (1991) also reported a lack of effect of eicosanoid on ovarian expression of TIMP-1 Rat granulosa cell TIMP-1 mRNA is also increased by LH and hCG, but the induced mRNA expression is not affected by cycloheximide (Mann et al., 1993). Thus de novo protein synthesis is required for LHand TPA-induced increase in granulosa cell TIMP-1 activity but protein synthesis is not necessary for stimulation of TIMP-1 mRNA expression. Luteal synthesis of TIMP-1 has been reported in the ewe (Smith et al., 1993; 1994), cow (Freudenstein etal., 1990; Ndikum-Moffor et al., 1995), rat (Mann et al., 1991), mouse (Edwards et al., 1992), and ferret (Huang et al., 1993). Results from some of these studies indicate that luteal synthesis of

PAGE 74

57 TIMP-1 is triggered and stimulated by the surge in LH. Gonadotropins have also been shown to regulate expression of TIMP-1 mRNA in Sertoli cells. Treatment of prepubertal rat Sertoli cells with FSH and 8-bromo cAMP increases activity of TIMP-1, amount of TIMP-1 protein in conditioned-medium, and expression of TIMP-1 mRNA (Ulisse et al., 1994). Similar to observations in rat granulosa cells (Mann et al., 1991), de novo protein synthesis and RNA synthesis are required for both basal and TPA-, 8-bromo cAMP-, and FSH-stimulated TIMP-1 activity (Ulisse et al., 1994). The effects of phorbol esters on gene transcription occur through fos and jun containing AP-1 transactivating factors and the latter is induced by PKC-activating stimuli (Lee et al., 1987). On the other hand, cAMP enhances gene transcription by stimulating PKA which stimulates phosphorylation of cAMP response element binding proteins (CREB) (SassoneCorsi et al., 1988; Merino et al., 1989). The mechanism(s) through which TIMP-1 synthesis is stimulated has not been fully characterized. The murine TIMP-1 gene contains cis-acting regulatory elements upstream of the major transcription start site and also contains an AP-1 binding site within the cis-acting region (Edwards et al., 1992). The AP-1 functions as binding site for fos-jun and can stimulate transcription. Fos and Jun, like the CREB proteins are members of an extended basic regionleucine zipper (bZIP) superfamily of transcription factors. It has been shown that oligonucleotides containing a CREB sequence compete for binding of proteins to TIMP-1 AP-1 site. Thus the CREBP family may be involved in specific binding to

PAGE 75

58 TIMP-1 AP-1 site (Edwards et al., 1992). Expression of TIMP-1 is stimulated by factors that increase intracellular cAMP. Thus TIMP-1 AP-1 site is the cis-acting regulatory element that mediates the cAMP-lnduced increase in TIMP-1 gene expression (Edwards et al., 1992). TIMP-1 enhancer element does not contain a classical CRE binding site but contains functional AP-1 sites, one of which can bind fos and jun heterodimers and other transacting factors including the CREB family (Edwards et al., 1992). Effects of TPA on TIMP-1 synthesis are mediated through the PKC second messenger pathway since TPA-stimulated TIMP-1 activity is inhibited by an inhibitor of PKC (Staurosporine), and TIMP-1 activity is not stimulated by a non-PKC activating phorbol ester (Mackay et al., 1992). Regulation of TIMP-1 Synthesis by Steroid hormones Rajabi et al. (1991a) demonstrated that estradiol-17(3 stimulates degradation of collagen type 1 in nonpregnant guinea pig cervix in vitro. In addition, the cervix has been shown to produce collagenase and its synthesis is stimulated by estrogens, interleukin-1(3, and PGEj (Rajabi et al., 1991b). Rajabi et al. (1991c) furthermore showed that activities of collagenase and collagenase inhibitor are greater in cervical tissue at the time of parturition than in tissues from nonpregnant animals. The marked increase in inhibitor activity observed at a time when collagenase activity is increased indicates the presence of a strong regulatory mechanism to control the extent of collagen degradation beyond the

PAGE 76

59 level required for parturition (Rajabi et al., 1991c). Progesterone stimulates TIMP-1 production by rabbit uterine cervical fibroblasts (Imada et al., 1994). Similar observations were reported by Sato et al. (1991) who showed that progesterone and estradiol-17fi increases secretion of TIMP-1 by rabbit uterine cervical fibroblasts in culture and steady state TIMP1 mRNA. However, observations by Rajabi et al. (1991c) showed that estradiol causes a decrease in tissue collagenase activity. Retinoic acid has been shown to enhance secretion of TIMP-1 by human fibroblasts in vitro by increasing de novo synthesis of TIMP-1 (Clark et al., 1987). Retinoic acid also increased TIMP-1 mRNA levels compared to nontreated controls. Glucocorticoid treatment had no effect on TIMP-1 secretion (Clark etal., 1987). Regulation of TIMP-1 Synthesis by Cytokines and Growth Factors Synthesis of TIMP-1 and collagenase by human fibroblasts is stimulated by phorbol ester and IL-1 (Murphy et al., 1985). Similar effects of IL-1 were reported by Rajabi et al. (1991c) for cervical tissue of guinea pig. TIMP-1 activity and expression of its mRNA have been shown to increase in a variety of normal and tumor cell lines following treatment with IL-1 and tumor necrosis factor (TNF) (Mackay et al., 1992). Transforming growth factor-!! (TGF-B) has also been implicated as a regulator of TIMP-1 synthesis. One of the mechanisms proposed for the control of normal trophoblast proliferation and

PAGE 77

60 invasiveness by TGF-Ii Is via induction of TIMP-1 mRNA expression. This regulatory mechanism is absent in malignant trophoblast cells (Graham et al., 1994). Tissue Inhibitor of Metalloproteinases-2 Biochemical Characterization Tissue inhibitor of metalloproteinases-2, the second member of the family of metalloproteinase inhibitors, binds and inactivates all matrix metalloproteinases but in contrast to TIMP-1 which binds the 92-kDa gelatinase, TIMP-2 binds the 72-kDa gelatinase. TIMP-2 has a molecular weight of 20-21 l
PAGE 78

61 similar physiological properties, expression of their activity and mRNA seem to be differentially regulated in mouse reproductive tissues (Waterhouse et al., 1993). In contrast to the stimulating effects of phorbol ester on murine TIMP-1 mRNA, phorbol ester does not affect expression of TIMP-2 mRNA (De Clerk et al., 1994). In another study Waterhouse et al. (1993) reported a differential expression in TIMP-1 mRNA and TIMP-2 mRNA in the ovary of mice during gestation; TIMP-1 mRNA is low while TIMP-2 mRNA shows a marginal increase. Leco et al. (1992) also reported that while TIMP-1 mRNA is highly seruminducible in normal murine fibroblasts, expression of TIMP-2 mRNA is mainly constitutive and is insensitive to transformation while expression of TIMP-1 mRNA is variable (Leco et al., 1992). Regulation of TIMP-2 Synthesis bv Gonadotropins Like TIMP-1, TIMP-2 activity, TIMP-2 protein and mRNA levels in rat Sertoli cells are stimulated by FSH through a cAMP-dependent pathway (Ulisse etal., 1994). Regulation of TIMP-2 Synthesis by Steroid Hormones Production of TIMP-2 in culture by rabbit uterine cervical fibroblasts increases after treatment with physiological concentrations of progesterone (Imada et al., 1994). TIMP-2 mRNA is expressed constitutively in rat hepatocytes and its expression is up-regulated following incubation of

PAGE 79

62 hepatocytes with dexamethasone and prostaglandin Ej (Roeb et al., 1995). Regulation of TIMP-2 Synthesis by Cytokines and Growth Factors Transforming growth factor-li has been shown to down-regulate both mRNA transcripts of TIMP-2 in contrast to its stimulatory effect on TIMP-1 mRNA expression (Stetler-Steyenson et al., 1990). Mackay et al. (1992) also reported that TIMP-2 activities are refractory to TPA, IL-1 and TNF-a, in contrast to the marked stimulation of TIMP-1 activities by all three agents in a variety of human cell lines. A similar lack of stimulatory effect of TPA on TIMP-2 activity is also observed in rat Sertoli cells in culture (Ulisse et al., 1994).

PAGE 80

63 Table 1-1 Summary of Factors Regulating Synthesis of Proteins. Regulation/ Apo A-1 ApoE MnSOD TIMP-1 TIMP-2 1 Function E2, T3, FSH, LH, LH, hCG, LH, cAMP FSH, HNF-4, cAMP, hCG, TPA, IL-1, TPA, E2. P, cAMP, P„ Gene Dietary CT, TPA, TNF, LPS IL-1 PGE2, PGE2, Expression Cholesterol Cholesterol TNF, TGFB, VitA TGF-B Translation Dietary cAMP, CT, TNF, IL-1 IL-1, TNF, FSH, cAMP Cholesterol, TPA, TNF, VitA, P„E2 ^2 LPS, GM-CSF, Function Cholesterol Cholesterol Prevents Tissue Tissue metabolism, metabolism. oxidative remodelling, remodelling, CAMP membrane and stress cell grovrth. cell growth stimulation. steroid Steroid Steroid biosynthesis biosynthesis biosynthesis

PAGE 81

CHAPTER 3 PROTEINS SYNTHESIZED AND RELEASED IN CULTURE BY THE BOVINE CORPUS LUTEUM: THE ESTROUS CYCLE AND PREGNANCY Introduction Apart from its traditional role in progesterone synthesis and maintenance of pregnancy, the corpus iuteum synthesizes and secretes a number of proteins during the estrous cycle and pregnancy. The corpus Iuteum of the cow and ewe synthesizes the peptides oxytocin and neurophysin during the estrous cycle and stores these proteins in secretory granules of the large luteal cells (Fields et al., 1986; 1992) and ewe (Fields et al., 1986; Theodosis et al., 1986). The synthesis and secretion of proteins vary with the physiological status of the animal. For example, the number of oxytocin-containing secretory granules in the cow increases from metestrus to diestrus, and then declines prior to luteolysis (Fields et al., 1992), whereas in pregnancy the population of granules was undetectable on day 45, then increased to a peak between days 180-210 (Fields et al., 1985). However, the corpus Iuteum of mid-pregnancy (after day 45) in the cow contains neither the mRNA for oxytocin (Ivell et al., 1985) nor do the secretory granules contain oxytocin (Fields et al., 1992). Additional proteins identified as secreted by the corpus Iuteum include the 64

PAGE 82

65 tissue inhibitor of metalloproteinases-1 and -2 (TIMP-1 and TIMP-2) in the rat (Parmer etal., 1992), sheep (Smith and Moor, 1991), (Smith etal., 1995; Smith at al., 1993; 1994 ), cattle (Freudenstein et al., 1990; Juengel et al., 1994), pig (Smith et a!., 1994)) and ferret (Huang et al., 1993), relaxin in humans and nonruminants (Sherwood, 1994), inhibin in sheep (Tsonis et al., 1988; Rodgers et a!., 1989; Smith etal., 1991), and insulin-like growth factor-1 (Einspanier et al., 1990), basicfibroblast growth factor (Stirling et al., 1991) and angiogenic factors (Redmeretal., 1988; Grazul-Bilska etal., 1992) in cattle. In addition to its role in maintenance of pregnancy, the corpus luteum of pregnancy in the cow appears to be necessary for normal parturition since removal during the third trimester resulted in increased rates of dystocia, retained fetal membranes (Estergreen, 1967), and greater death loss of calves (Tanabe, 1966). The synthesis and secretion of proteins by the bovine CL of pregnancy may play a role in setting the stage for parturition. The ferret corpus luteum was shown to secrete proteins on days 5-1 1 of pregnancy, with molecular masses of 16 to 185 kDa (Huang et al., 1993). Although no qualitative difference in protein secretion was observed across days of pregnancy studied, a 32 kDa protein that cross-reacted weakly with a polyclonal antibody to human TIMP was the most abundantly secreted protein (Huang et al., 1993). The objectives of this study were to examine for proteins synthesized de novo by the bovine CL, identify and characterize the newly-synthesized proteins, and determine quantitative differences in their synthesis and release during the

PAGE 83

66 estrous cycle and pregnancy. Materials and Methods Materials Acrylamide was purchased from ICN Biomedicals Inc. (Cleveland, OH), bis-acrylamide and agarose from Bio-Rad Laboratories, N,N,N,N,tetramethylethylenediamine (TEMED) from Fisher Scientific (Fair Lawn, NJ), and ampholines from Pharmacia (Piscataway, NJ). Other electrophoretic reagents were obtained from Bio-Rad Laboratories (Richmond, CA). L-4,5-^H-leucine (specific activity 164 Ci/mmol) and D-[6-^H]glucosamine (specific activity 20Ci/mmol) were purchased from Amersham (Arlington Heights, IL), and ^^Smethionine (specific activity 1028 Ci/mmol) was purchased from ICN Biomedicals Inc. Polyvinylidene fluoride (PVDF) was obtained from Millipore Corporation (Bedford, MA). Tissue culture media including amino acids, vitamins, insulin and antibiotic-antimycotic mixture, and all other chemicals were purchased from Sigma Chemical Company (St. Louis, MO). Collection of Luteal Tissue Forty-eight Angus and Hereford crossbred beef cows were used for the study. All procedures in which animals were used were approved by the Animal Care and Use Committee of the University of Florida. Estrus (day 0) was

PAGE 84

67 defined as that day when a cow would stand to be mounted by a bull. Cows randomly assigned to the pregnancy group were artificially inseminated at observed estrus, whereas cows assigned to the cycle group were not bred. Day 17 of pregnancy was confirmed by the presence of an embryo in flushings from the uterus. Later stages of pregnancy were estimated by measurement of crown-rump length of the fetus (Winters et al., 1942). Reproductive tracts were obtained from cows within 5 min after exsanguination at the University of Florida abattoir. The ovary containing the corpus luteum was collected aseptically from cows on days 3 (n = 4 cows), 7 (n = 3), 11 (n = 4), 14 (n = 5), 17 (n = 3), and 19 (n = 3) of the estrous cycle, and from cows of early pregnancy (day 17, n = 5), and the first (day 88, n = 5), second (day 170, n = 7), and third (greater than day 240, n = 9) trimester of pregnancy. Ovaries were immediately transferred to a sterile Petri dish (100 x 15 mm) containing pre-warmed Eagle's Minimum Essential Medium (MEM), and the corpus luteum dissected from the ovarian stroma and weighed. Culture Medium Medium was prepared as previously described (Basha et al., 1980) from Eagle's MEM deficient in leucine, lysine, methionine and sodium bicarbonate. One liter of stock incomplete MEM was prepared with the addition of glucose (3 g/l), methionine (1.5 mg/l), leucine (5.2 mg/l) and lysine-HCI (72.5 mg/l) to achieve 4.0, 0.1, 0.1, and 1.0 times, respectively, their usual concentrations in

PAGE 85

68 MEM. Sodium bicarbonate (2.2 g/l), non-essential amino acids (1 %, v/v), vitamins (1%, v/v) and insulin (200 lU/l) were added and pH adjusted to 7.1-7.3. The medium was filter-sterilized (0.22 pm) (Corning Inc, Corning, NY) and stored at 4 C. For ^H-leucine labelled cultures, methionine (1.5 mg/1 00 ml) and antibiotic-antimycotic (ABAM) mixture (1%, v/v) were added to the stock incomplete MEM to obtain leucine-deficient incomplete modified MEM. Similarly, for ^^S-methionine labelled cultures, methionine-deficient medium was prepared by adding leucine (5.2 mg/1 00 ml) and ABAM to the stock incomplete MEM. Incomplete MEM was used for ^H-glucosamine labelled cultures. Time Course Studies of Incorporation of Radiolabel CL from three pregnant cows (two on day 170, and one on day 88) were used in the time course studies. Slices (0.5 mm in thickness) of luteal tissue were prepared with a Stadie-Rigg's tissue slicer (Thomas Scientific, Swedesboro, NJ). Explants were washed three times, each time with 15 ml incomplete MEM, to reduce serum proteins in the medium during incubation. After washing, luteal tissue (500 mg/dish) from the same corpus luteum was placed in four Petri dishes each containing 15 ml of leucine-deficient incomplete MEM without radiolabel. The dishes were pre-incubated at 37 C on a rocker platform for 2 h in an atmosphere of 50% N2:47.5% 02:2.5% CO2 (v/v/v). After pre-incubation, the medium was discarded and replaced with 15 ml leucinedeficient incomplete MEM containing 50 pCi of ^H-leucine (160 Ci/mmol). Each %-

PAGE 86

69 of the four Petri dishes of luteal tissue from each CL was then incubated as described for 6, 18, 24, and 30 h, respectively (ie n = 3 for each time point). At the end of each incubation, the luteal-conditioned medium (LCM) was separated from tissue by centrifugation at 2000 x g for 20 min at room temperature. Medium was dialyzed immediately for 24 h at 4 C using Spectra/por 3 membrane (molecular weight cut off = 3500; Spectrum Medical Industries Inc., Houston, TX) against two changes (24 h each) of 4 I Tris-HCI buffer (10 mM, pH 8.2), and then against deionized water (two changes, 24 h each). Following dialysis, percent incorporation of radiolabel in the dialyzed LCM was determined for each incubation time. Percent incorporation was calculated as: post-dialysis radioactivity (dpm) divided by pre-incubation radioactivity (dpm) x 100%. Culture and Radiolabellino of Luteal Tissue The 24 h incubation was determined as the optimal time of incubation following incubations at 6, 18, 24, and 30 h as described. For each cow at least two Petri dishes of luteal slices (500 mg tissue/dish) were incubated as described in the above protocol for 24 h. Following incubation and dialysis, the total volume of dialyzed LCM (retentate) for each Petri dish was measured and adjusted to 15 ml with deionized water. Potentates were stored at -20 C until further analysis. Luteal slices (500 mg) were also incubated with ^Hglucosamine (50 MCi/15 ml, n = 1 cow, day 240 pregnant) or ^^S-methionine (50 |jCi/15 ml, n = 1 cow, day 240 pregnant) to determine if newly-synthesized and

PAGE 87

70 released proteins were glycosylated and/or contained methionine. TCA precipitation Following LCM dialysis, samples of dialyzed medium (retentate) for each corpus luteum were analyzed to determine the amount of ^H-leucine incorporated into trichloroacetic acid (TCA)-precipitable protein using a modification of method described by Mans and Novelli (1961 ). Briefly, a 50 pi aliquot of LCM retentate was spotted onto a 2.54 cm x 2.54 cm Whatman 3MM filter paper that was pre-soaked in 20% (w/v) TCA. Each square was air-dried, soaked in 20% and 5% TCA for 10 and 20 min, respectively, and subsequently washed twice, for 15 min each, in 95% ethanol. Squares were allowed to air-dry completely, placed in scintillation cocktail and counted for radioactivity. TCAprecipitable protein (dpm) was measured in duplicate for each LCM retentate sample, and values were expressed as least squares means (LSM) for each group. Light and Electron Microscopy To assess the effects of the 24 h-incubation on tissue morphology, luteal tissues collected before and after 24 h of culture with radiolabel were processed for evaluation by electron microscopy (Fields et al., 1992). Briefly, the central part of each CL was dissected into 1-3 mm cubes and fixed in 1 % (v/v) glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4). Cubes were

PAGE 88

71 postfixed in 1% (v/v) osmium tetroxide in 0.1 M sodium cacodylate buffer for 30 min, embedded individually in Spurr's medium, and sectioned (0.1 pm). Sections were observed and photographed using a Philips electron microscope (Model 301, Philips Electronic Instruments, Mahwah, NJ). For evaluation by light microscopy, luteal tissue was fixed in Bouins, embedded in paraffin, and sections were cut and stained with haematoxylin and eosin (Sheehan and Hrapchak, 1980). Two-Dimensional-SDSPolvacrvlamide Gel Electrophoresis First-dimension: Isoelectric Focusing Proteins, synthesized and released into the medium of explant culture, were separated according to their isoelectric points (pi) by isoelectric focusing (lEF) as previously described (Laemmli, 1970). Frozen retentates were lyophilized and reconstituted in lEF gel sample buffer [9 M urea, 2% (v/v) NP-40, and 2% (v/v) ampholine (pH 3.5-10)]. Each sample (100,000 cpm) was loaded and the proteins separated in the lEF tube gel (4% (w/v) T, 5.4% (w/w) Cbis) by electrophoresis at 400 V for 20 h. Second-dimension: SDS-PAGE Following first dimension electrophoresis, the tube gels (100,000 cpm per gel) were equilibrated in gel-equilibration buffer [0.0625 M Tris, 5% (w/v) sodium dodecylsulfate (SDS), 10% (v/v) glycerol], for 15 min before they were loaded onto a slab SDS-polyacrylamide gel (14 cm x 16 cm X 0.15 cm; stacking gel 4% (w/v) T, 2.7% (w/w) Ct,,^; separating gel 10% (w/v) T, 2.7% (w/w) Cbis). Proteins in the lEF gel were separated on a slab gel by

PAGE 89

72 electrophoresis in tank buffer (0.025 M Tris, 0.192 M glycine, 0.1% SDS, pH 8.3) at 13 mA/slab gel until the dye front reached the end of the gel (O'Farrei, 1975). Following electrophoresis, gels were stained with 0.1% (w/v) Coomassie blue, destained in destaining solution [50% (v/v) ethanol, 10% (v/v) acetic acid] and soaked in deionized water for 30 min. Gels were then treated with 1 M sodium salicylate for 30 min and dried on a slab gel dryer (Model SE 1 150, Hoefer Scientific Instruments, San Francisco, CA), followed by exposure to x-ray film (XOMAT-AR, Eastman Kodak Company, Rochester, NY) for 4 weeks. Intensities of the radioactive spots on fluorographs were determined by densitometric scanning (E-C Apparatus Corporation, St. Petersburg, FL). The scanner was standardized to detect intensities between the lightest (background) and the darkest spot on fluorographs. The area under the curve for each spot scanned was measured using the trace mode of an electronic planimeter (Model 1250, Numonics Corporation, Lansdale, PA). Area measurements in cm^ represent relative amounts of each newly-synthesized protein. Protein Blotting and Amino Acid Sequencing Following detection of newly synthesized proteins, proteins on wet 2-D gels were transferred to polyvinylidene fluoride (PVDF) membranes (Millipore Corporation) by electroblotting at a constant voltage (20 V) in transfer buffer (2[N-Morpholino]ethanesulfonic acid, 10 mM MES, pH 6, Sigma Chemical Co.) at 4 C for 16 h, according to method of Towbin et al. (1979). Membranes were

PAGE 90

73 stained in 0.1% (w/v) Coomassle blue in 50% methanol for 5 min, destained in a solution of 50% ethanol and 10% acetic acid, rinsed extensively in distilled water, and then air dried. Proteins on membranes were subjected to N-terminal amino acid sequence analysis (Protein Sequencer Model 470 A/B, Applied Biosystems, Foster City, CA) by the Protein Chemistry Core Laboratory of the Interdisciplinary Center for Biotechnological Research (ICBR) at the University of Florida. A search of protein, RNA, and DNA data banks was conducted using the National Center for Biotechnology Information Database Search program (Devereux et al., 1984). Progesterone Assay Trunk blood samples collected into heparinized vacutainers (Becton Dickinson Vacutainer Systems, Rutherford, NJ) from cows at time of slaughter were processed, and all plasma samples were analyzed for progesterone by RIA in a single assay, using the progesterone Coat-A-Count kit (Diagnostic Products Corporation, Los Angeles, CA). One hundred microliter (100 pi) plasma was assayed per tube. The kit originally designed for human serum was validated for use with cow plasma. Progesterone standards were prepared from a stock solution (5 pg progesterone/ml benzene) by diluting with ovariectomized cow plasma. A quantitative linear recovery of progesterone was obtained with two replicates of 0, 7.8, 15.6, 31.2, 62.5, 125, 250, 500, and 1000 pg progesterone/ 100 pi ovariectomized cow plasma [Y = 1 .595 + 0.864X; Y = amount of

PAGE 91

74 progesterone measured (pg/ml) and X = amount of progesterone added; R = .99]. The detection limit of the assay, determined as two standard deviations above the zero dose level, was 9.5 pg/ml. The intra-assay coefficient of variation for the one assay was 6.5%. The slope (Mean SEM) of the standard curve using the progesterone standards provided in the kit (0, 10, 50, 200, 1000, 2000, 4000 pg/100 \j\) was -0.68 0.13 and the estimated doses at 20, 50, and 80% binding were 10.03, 1.31, and 0.18 ng/ml, respectively. Statistical Analysis Data for percent incorporation, radiolabeled TCA precipitable protein, CL weights and plasma progesterone were analyzed by least squares analysis of variance using the General Linear Models procedure of the Statistical Analytical System (SAS, 1988). Data from densitometric scanning were subjected to analysis of variance to determine differences in protein secretion. The statistical model had day and reproductive status as the main effects with residuals as the error term. Values in the text are least square means (LSM) standard error of the mean (SEM). Orthogonal contrasts were performed to determine differences among days of the cycle and stages of pregnancy.

PAGE 92

76 Results Histology of Luteal Tissue Results of microscopic examination of luteal tissue before and after 24 hincubation showed there was no degeneration of the tissue as a result of incubation; there was no evidence of presence of collagen fibers or loss of secretory granules (Fig. 3-1). There was presence of secretory granules and intact mitochondria. CL Weight and Plasma Progesterone As expected, weight of the corpus luteum was different (P < 0.004) across days of the estrous cycle, but not across stages of pregnancy (Fig. 3-2). Corpora lutea weights were lowest (P < 0.002) on day 3 when compared to the other days of the estrous cycle examined. Weight of the corpus luteum was lower (P < 0.03) on day 1 1 than 14. Plasma progesterone had a similar pattern as the corpora lutea weight, with concentrations varying (P < 0.006) across days of the estrous cycle, but not across stages of pregnancy (Fig. 3-2). Plasma progesterone was lower (P < 0.03) on day 3 when compared to the rest of the estrous cycle.

PAGE 93

Figure 3-1. Electron micrograph of luteal cells. A) Top panel: Luteal cell of tissue prior to incubation (control). Note secretory granules (arrow), nucleus at the bottom, and intact mitochondria. X 9,000; #50627. B) Middle panel: Luteal cell of tissue post-24 h incubation without radiolabel (control). Note mature secretory granules (arrow), and nascent secretory granules associated with the Golgi apparatus, and intact mitochondria. X 11,400; #50633. C) Bottom panel: Luteal cell of tissue post-24 h incubation with 50 [jCi ^H-leucine. Note secretory granules (arrow) and nucleus at the bottom. Intact mitochondria is not different from that of the controls, x 1 1 ,400; 50630.

PAGE 94

77

PAGE 95

78 8 10 12 14 16 18 20 100 200 300 Day Figure 3-2. Least square means SEM of plasma concentrations of progesterone (ng/ml) and weight (g) of the corpus luteum across days 3, 7, 11, 14, 17, and 19 of the estrous cycle and days 17, 88, 180, and >240 of pregnancy. Progesterone concentrations and CL weight increased from day 3 to day 14 of the estrous cycle and then declined to day 19, but these did not vary across pregnancy.

PAGE 96

#•' 79 Plasma progesterone was lower (P < 0.03) during early luteal (day 7) than mldluteal phase (days 1 1 and 14) of the estrous cycle. Similarly, plasma P4 was lower (P < 0.09) on day 1 1 than day 14, and day 19 was lower than day 17 (P < 0.05). However, there were no differences between day 17 of the estrous cycle and day 17 of pregnancy for plasma progesterone (4.79 1.1 vs 5.36 1.3, P < 0. 76) and CL weight (4.66 0.7 vs 4.83 0.6, P < 0.87). Incorporation of Radiolabel into TCA-precipitable Protein Results of the time course study (n = 3 cows) showed that the 24 hincubation time was optimum for incorporation of radiolabel when compared with the 6, 18, and 30 h incubation time (Fig. 3-3), thus 24 h-incubation time was used in this study. There was no lag time, the rate of incorporation increased sharply between 18 h and 24 h of incubation, and did not change or declined between 24 and 30 h. Analysis of all LCM retentates following dialysis indicated that there was no difference in the amount of radioactivity associated with TCAprecipitable proteins released in culture across days of the estrous cycle or stages of pregnancy. The amount (LSM SEM) of TCA-precipitable radioactivity was not different (P < 0.23) between day 17 (3104.6 500.5 dpm) of the estrous cycle and day 1 7 (221 7.6 433.5 dpm) of pregnancy. Percent

PAGE 97

80 3000 r Incubation Time (h) Figure 3-3. Time course studies of incorporation of radiolabel (n = 3 cows). incorporation of radiolabel across the estrous cycle ranged between 2.4 0.8% (day 11) and 6.2 0.9% (day 19) and approached significance (P < 0.09). The percent incorporation ranged between 2.0 0.6% and 3.4 0.6% during pregnancy, did not differ across pregnancy, and was not different (P < 0.35) between day 1 7 of the estrous cycle and day 1 7 of pregnancy.

PAGE 98

81 c 2 o Q. O o ^^ •a to DC 12 I10 8 ^ 6 2 I Cycle Pregnancy I m A i I ^ ^ ^ lilL 7 11 14 17 19 17 88 170>240 Day of Cycle/Pregnancy Figure 3-4. Percent incorporation of radiolabel into newly-synthesized proteins.

PAGE 99

82 Luteal Protein Synthesis and N-terminal Amino Acid Micro Sequencing The bovine CL synthesized and released a number of different proteins during the estrous cycle and pregnancy with molecular masses ranging from 12 to 200 kDa. Following 2D-SDS-PAGE and fluorography, eleven discrete radiolabeled proteins were selected for further study. For convenience, before the amino acid sequence information was available, each of the eleven major proteins was assigned a number; proteins 1 (35 kDa, pi 5.5), 2 (30 kDa, pi 5.5), 3 (29 kDa, pi 5.5), 4 (27 kDa, pi 5.5), 5 (70 kDa, pi 5.0), 6 (58 kDa, pi 6.0), 7 (44 kDa, pi 5.0), 8 (30 kDa, pi 8.0), 9 (20 kDa, pi 8.0), 10 (22 kDa, pi 8.0), and 11 (27 kDa, pi 6.0) (Fig. 3-5, 3-6). Five of these proteins were identified from their N-terminal amino acid sequence. Between 18 and 31 N-terminal amino acid residues for proteins 1 8, 9, 10 and 1 1 were determined following separation by 2D-SDS-PAGE and transfer to PVDF membranes (Table 3-1 ). Protein 1 composed of four protein spots, and proteins 8 and 1 1 each composed of two protein spots were individually subjected to N-terminal micro-sequence analysis. A search of protein data banks matched these N-terminal amino acid sequences to those of bovine apolipoprotein E (Apo E; all four protein spots associated with protein 1),

PAGE 100

83 Pl D D7 5j D11 5 V 6 b 68I — 7-16 45^ 3027N 4^' 24.^4 [ f'^ 9,10 20^^^ '-•' L [• -f i r f [• 3' ir D14 D17 D19 \ 6 Is 6 O 68• I V J ^ X T -' 7 •. 7 • s 45\ i \ • ,8 8 302724' r,-.n*^ [— f •'' 20Figure 3-5. Representative fluorographs of proteins synthesized de novo in explant culture and released into tine medium by CL on different days of the estrous cycle. Proteins 1 2, 3, and 4 were observed only on day 3. Other identified proteins 5-10 were found on all other days (7-1 9) and protein 1 1 was found on days 11-19.

PAGE 101

pi 84 C9 I O 6845302724205 I 6 J D17 [J .1 [ f [-] ,9,10 ["J i; ]9.10 b 68453027D180 \ \ r^ [ ]9. 8 10 D240 5 \ 6 H ^ .9,10 Figure 3-6. Representative fluorographs of proteins synthesized de novo in explant culture and released into the medium by CL on different days of pregnancy. Proteins 5-1 1 were found on all days of pregnancy (17, 88, 180, and > 240) examined.

PAGE 102

o\ 2 c 2 o o cr.S T3 O • C O C a' C/3 03 *-• o o o -a c a o i^ H C3 O e O 3 3 V o 0. 3 3 3 c L. 3 3 3 3 >i 3 3 in u in UJ a < u u c u 3 u I/) Q. >> >J 3 w < s u &. s 3 3 H 3 V s u 5 _3 3 O u 0. ;> 3 3 J 3 3 3 3 a o a < c '> o u < ^ o u I C3 e H a z h < ^ V V CA c s s V. Kl < < Vi 1 >% 1 u V X V A. X _a a. < a b < X u M X H c & 5 3 L. e o u "a U X VI ft. o sa a e I < < a < a < M b < a a < B OS < I u X a. s < 3 B 3 b b M 3 B -S •r J a a M b < >' ;5 V M a B < u u X Bl. A < B 5 B 3 b fi. 3 I b u 73 u b w >> U e Q B s 3 3 G e 3 3 V 4) fl a 3 3 9 a < < c e w m "C < tfi It s 33 b u A. 0. 9 a s a :2 e a < < >> >> 3 3 b b f? f? b b 0. 0. 3 a J3 -S < < o o b b :s js '^ i: u B V 3 O" tl c X a .J Q X s B a E 3 fi a < B '5 2 a. < w s 5 s u 0. B 3 B 5 a u H 41 X u B V 3 T V X a .J fS fS j: a < < a a in e 5 s 5 a H a. a ^ ii o a < V B "> aa 85 m tn o -. 0) *J ^ CB D ^ S CO Jo en C tn .2 01 0) Q O in ^ CO en i CI I S 0) S tn a, H m ^ ^ 73 tn '-' — >% en (U to m ">. o o >. o 0) T3 "O (I) o C N ^ (0 ^ ij > m Tl H 4) fi ;3 J3 JH • rt -^ L^ O I !t; en 0) — '^J (h d a^ 1) in D ~ rj 4) -o fi --i j2 — o x! 03 a o • fi Si fi

PAGE 103

86 bovine TIMP-1 (both protein spots associated with protein 8), bovine TIMP-2 (protein 9), human manganese superoxide dismutase (Mn SOD; protein 10), and bovine apolipoprotein A-1 (Apo A-1; both protein spots associated with protein 1 1 ). Although a few amino acid residues within four of the five sequences were unidentified, all identified residues matched those of the corresponding known sequence. Sequences for other determined proteins were not obtained by Nterminal sequence analysis due to insufficient quantities of protein for transfer. Protein Svnthesis and Release During the Estrous Cycle Analysis of densitometric data of fluorographs indicated qualitative and quantitative differences in protein synthesis and release across days of the estrous cycle. Proteins 1 (Apo E), 2, 3, and 4 were observed only on day 3 (Fig. 3-7). Synthesis of protein 1 1 (Apo A-1 ) was not observed until day 1 1 and levels of release did not vary (P < 0.35) across the remaining days of the estrous cycle although synthesis and release of Apo A-1 appeared to be higher on day 19 than on day 17 {P<0.11). The remaining proteins 5, 6, 7, 8 (TIMP-1), 9 (TIMP-2), and 10 (manganese SOD) were observed on all days except day 3. The relative amount of protein 8 (TIMP-1) synthesized and released into medium did not differ across days of the cycle, but appeared to increase on day 19 (day 1 7 < day 1 9; P < 0. / 1). Synthesis and release of TIMP-2 and manganese SOD analyzed as a combined protein due to overlap on the gel did not vary (P < 0.20) across the estrous cycle. However, orthogonal contrasts showed that synthesis

PAGE 104

87 and release of TIMP-2 and SOD were higher during late rather than mid-luteal phaseof the estrous cycle (days 17, 19 > days 7, 11, ^A,P<0.01). Although there was no overall difference in the release of protein 6 across days of the estrous cycle (P < 0. 13), while protein 7 did vary (P < 0.004) across the estrous cycle. Orthogonal contrast analysis showed that protein 6 was greater (P < 0.01) during the late luteal phase (days 17 and 19) when compared with the midluteal phase (days 7, 1 1 and 14) of the estrous cycle, and greater on day 17 than 19 (P < 0.08). A greater amount of protein 7 was synthesized during late rather than the early and mid-luteal phases (days 17 and 19 > 7, 11 and 14, P < 0.009), and was also greater on day 19 compared to 17 (P < 0.002). The amount of protein 5 synthesized and released was at the minimum detectable limit on all days considered, and did not change during the estrous cycle. Luteal Protein Synthesis and Release During Pregnancy Corpora lutea of pregnancy synthesized and released proteins similar to those released by CL of the estrous cycle, except for Apo E and proteins 2-4 which were not observed during pregnancy (Fig. 3-7). No differences were observed in the synthesis of Apo A-1 (P < 0.42), protein 6 (P < 0.21), protein 7 (P < 0.32), TIMP-1 (P < 0.29), and TIMP-2 and manganese SOD (P < 0.39), released across pregnancy. However, orthogonal contrasts showed that TIMP-1 synthesis tended to be lower on day 1 80 when compared with day 240 or greater of gestation (P < 0. 18). Similarly, synthesis of protein 6 was greater on day >240 than day 170 (P < 0.16), and synthesis of protein 7 on day 17 of

PAGE 105

Figure 3-7. Densitometric analysis of fluorographs of newly-synthesized proteins in luteal-conditioned medium during the estrous cycle and pregnancy. Values are least square means standard error of the mean. Statistical differences were reported for protein 7 across the cycle, and for Apo A-1 across pregnancy. Orthogonal contrasts revealed the following differences: Apo A-1 : day 17 < day 19 (P < 0.11), TIMP-1 : day 1 7 < day 1 9 (P < 0. / y ), TIMP-2: days 7, 1 1 and 14 < days 17 and 19(P<0.0r), protein 6: days 7, 11, 14 < days 17 and 19 (P < 0.01), and day 17 cycle > day 19 (P < 0.08), protein 7: days 7, 11 and 14 < days 17 and 19 {P< 0.009), and day 17 < day 19 (P< 0.002).

PAGE 106

89 10 : ^ cycle < 8 ^H Pregnancy O S: e I •S 4 0) '. tt' \l O i n L(^. 3 7 11 14 17 19 17 88 17O240 Day of Cycle/Pregnancy 20 Ql 16 I12 O >. w c Q 8 Cycle Pregnancy ttttI ^^^H^^H^BLfl^L^^A T -rXl MOj 3 7 11 14 17 19 17 88 170240 Day of Cycle/Pregnancy o w c o O 5|Q O (O 00 4h Dl 31I2 ^ TT Cycle Pregnancy 3 7 11 14 17 19 17 88 170240 Day of Cycle/Pregnancy (D c s a. .^^ V) c o 40 32 9 24 *fe 16 £ 8 i? Cycle Pregnancy fTT-r ^^ It ii 3 7 11 14 17 19 17 88 170240 Day of Cycle/Pregnancy c o i_ Q. ^O W c o O 28 24 20 16 12 8|4 Cycle H Pregnancy I TTTlli m^nttntfVa\ 3 7 11 14 17 19 17 88 170240 Day of Cycle/Pregnancy

PAGE 107

90 gestation tended to be higher than on the other days of pregnancy examined (P < 0. 14). Synthesis of radiolabeled protein 5 was at the minimum detectable limit on all days considered and did not change during pregnancy. Synthesis and release in culture of all proteins, except the TIMP-2/S0D complex, did not differ quantitatively when comparing day 17 of the estrous cycle and day 17 of pregnancy. Qualitatively, the same proteins were observed on both days. Radiolabelled Culture with ^H-qlucosamine and ^^S-methionine Radioactivity patterns, detected by fluorography of gels of LCM retentate of a late pregnant cow, varied with the radiolabel used in the culture. The number of radioactive spots on fluorographs was greater following incubation with ^H-leucine than for incubations with ^H-glucosamine and ^^S-methionine. For example, only protein 6 and TIMP-1 were detected after incubations with ^Hglucosamine suggesting that they are glycosylated, and all the other proteins 5, 7, 9-1 1 were not glycosylated (Fig. 3-8). Protein 7 was the only protein not detected after incubations with ^^S-methionine indicating little or no presence of methionine in this protein (Fig. 3-9). Thus, ^H-leucine was used in all subsequent incubations. Proteins 1-4 were not evaluated since they were not synthesized by the corpus luteum of during pregnancy. .'J

PAGE 108

91 PI rp o X ^ 684530272420! 6 7 Figure 3-8. Representative fiuorographs of proteins synthesized de novo in explant culture and released into medium on day 240 of pregnancy. A) Left panel: Radiolabelled culture with^H-leucine. Proteins 5-1 1 were observed. B) Right panel: Radiolabelled culture with ^Hglucosamine. Proteins 6 and 8 were observed. Proteins 5, 7, 9-1 1 were not observed. J

PAGE 109

92 PI CO o 684530272420-9 5 i 6 1 ,8 r .9,10 Figure 3-9. Representative fluorographs of proteins synthesized de novo in explant culture and released into medium on day 240 of pregnancy. A) Left panel: Radiolabelled culture with^H-leucine. Proteins 5-11 were observed. B) Right panel: Radiolabelled culture with ^^S-methionine. Proteins 5, 6, 8-10 were observed. Proteins 7 and 1 1 were not observed.

PAGE 110

93 Discussion In the present study, the bovine corpus luteum was shown by explant culture techniques and gel electrophoresis and fluorography to synthesize and release proteins differentially across the estrous cycle and pregnancy. Major proteins had molecular masses of 20 to 70 kDa and pis from 5.0 to 8.0. Nterminal amino acid micro sequence analysis identified five of these proteins with significant sequence identity to bovine Apo E, bovine Apo A-1, bovine TIMP-1, bovine TIMP-2, and human manganese SOD. The amounts of protein synthesized and released by the corpus luteum in explant culture was variable during the estrous cycle but and pregnancy. Bovine luteal Apo E was synthesized and released only on day 3 of the estrous cycle. This is the first report of the synthesis and release of Apo E by the bovine CL. Bovine Apo E was synthesized and release as isoforms, suggesting that newly-synthesized and released bovine luteal Apo E may be post-translationally modified. Apo E is synthesized as a preprotein and post-translationally modified with the addition carbohydrate chains such as sialic acid (Zannis et al., 1984). Higher concentrations of the glycosylated form of Apo E in newlysecreted protein than in intracellular forms of Apo E have been described by Zannis et al. (1984). In addition, newly-secreted Apo E has a higher sialic acid content than plasma Apo E (81 11 % vs 24 6%), and these results would suggest the possibility of variable posttranslational modifications which could lead to several

PAGE 111

94 different isoforms and Mr forms. Apo E is a ligand for the low density lipoprotein receptor and may be involved in regulating movement of cholesterol into the corpus luteum. Synthesis and release of Apo E only on day 3 suggests it may regulate cholesterol availability for membrane and/or steroid synthesis during early luteal development. Rat granulosa cells have been shown to produce Apo E (Driscoll et al., 1985; Wyne et al., 1989), and it has been proposed that secretion of Apo E early in the differentiation of the follicle to a corpus luteum may facilitate the shift from androgen production, required for follicular estrogen to progesterone synthesis during luteinization. However there was no evidence for the synthesis of Apo E by luteal cells has been reported (Polacek et al., 1992). A model for autocrine and paracrine involvement of Apo E in peripheral nerve regeneration following injury has been proposed (Mahley, 1988). A parallel model for the ovary could involve a wound healing effect of Apo E, due to ovulation and massive reorganization of the ovulated follicle into a corpus hemorrhagicum. Three other proteins with similar isoelectric points but lower molecular weights have been characterized and were observed only on day 3 of the cycle. To date these proteins have not been identified. Synthesis and release of Apo A-1 was not observed on days 3 and 7 of the estrous cycle, but was observed on all other days of the estrous cycle and pregnancy examined. This is the first report of synthesis of Apo A-1 by the corpus luteum of any species. Apo A-1 is synthesized primarily in the liver and

PAGE 112

95 small intestine but has also been shown to be produced by peripheral tissues and secreted in culture (Blue et al., 1982). The amino acid sequence of the 27 kDa protein has a 100% identity with bovine Apo A-1 (O'hUigin et al., 1990) and has therefore been identified as Apo A-1 in the present study. Apo A-1 is a major protein in the high density lipoprotein complex (HDL) which is the major source of circulating cholesterol in the cow (Sparrow et al., 1992). Thus Apo A-1 may function to regulate the availability of cholesterol to steroidogenic cells within the corpus luteum, ovary, and placenta. In the case of luteal synthesis of Apo A-1 a paracrine role is most likely given the demand for cholesterol by the corpus luteum. There is evidence that Apo A-1 levels are higher in serum from a female with the presence of a corpus luteum than in the serum of an anestrous animal (Oikawa and Katoh, 1995), However, the contribution of synthesized and secreted luteal apolipoproteins to circulating levels, if any, is unknown at this time. Secretion of TIMP-1 was first detected on day 7 of the estrous cycle and was higher during late luteal phase of the cycle, and did not vary in general during pregnancy, although levels tended to be higher during late pregnancy. Studies in sheep have shown that the ovine corpus luteum secretes a number of proteins on day 10 of the cycle, and TIMP-1 was the most abundant (Smith et al., 1991; 1993). In the present study, the 30 kDa protein released into medium has a 100% amino acid sequence similarity to bovine TIMP-1 (Freudenstein et al., 1990). While the role of TIMP in the corpus luteum has not been clarified,

PAGE 113

96 TIMPs in general are specific inhibitors of enzymes that degrade protein components of the extracellular matrix (Werb, 1989). Controlled remodelling of the extracellular matrix has been shown to take place during ovulation (Le Maire, 1989), uterine implantation (Lala et al., 1990) and gestation (Brenner et al., 1989). Therefore, TIMP-1 may be involved in controlling remodeling of luteal tissue during its rapid phase of development and during regression. Additionally, TIMP-1 and TIMP-2 possess erythroid-potentiating activity (StetllerStevenson et al., 1992), while TIMP-1, isolated from bovine granulosa cellconditioned medium, has been shown to promote the growth of bovine embryos in vitro (Satoh et al., 1994). A recent study (Boujrad et al., 1995) has shown that TIMP-1, secreted by rat Sertoli cells, stimulates steroidogenesis by rat Leydig cells, in a dose-dependent manner. TIMP-1 may have a multiplicity of roles in the CL, including those described here but, more importantantly, role(s) which may still be undefined. However, the increasing TIMP-1 secretion by a regressing corpus luteum (plasma progesterone < 0.5 ng/ml) on day 19 was unexpected. In the present study, TIMP-2 and manganese SOD co-migrated in 2-DSDS-PAGE gels. The resultant protein spot was subjected to N-terminal amino acid sequence analysis which indicated the presence of manganese SOD, a primary sequence, and TIMP-2, as the secondary sequence within the protein spot. Synthesis and release of the TIMP-2 and manganese SOD complex varied across the estrous cycle and was greater during later stages of the cycle.

PAGE 114

97 However, synthesis of this complex did not differ across pregnancy. Greater synthesis of TIMP-2 and expression of its mRNA by the ovine corpus luteum was shown earlier rather than later in the cycle (Smith et al., 1995), while in the mouse ovary, TIMP-2 mRNA was shown to increase with advancing gestation (Waterhouse et al., 1993). These differences may reflect species specificity. The physiological role of TIMP-2 in the corpus luteum is not clear. Like TIMP-1 TIMP-2 may be involved in luteal development, tissue remodelling and regression. Manganese SOD is a mitochondrial protein (Church, 1990). There have been no reports to suggest that manganese SOD is a secretory protein. In the present study, synthesis of manganese SOD was undetectable on day 3, but was observed through the rest of the cycle. The bovine large luteal cell has been shown to be densely packed with mitochondria which swell during the cycle and take on dense inclusions as early as day 14 (Fields et al., 1992). This will suggest that luteal manganese SOD synthesized and released in explant culture could have resulted from an involutive process in which mitochondrial membranes become compromised. Another reason for the increase in manganese SOD during the involutive phase of the estrous cycle could be an attempt by the corpus luteum to protect elf from deleterious oxygen radicals. Luteal manganese SOD may protect the corpus luteum from oxidative damage by converting reactive superoxide anions to hydrogen peroxide and oxygen (Laloraya et al., 1988; Hesia et al., 1992).

PAGE 115

98 However, others have shown that hydrogen peroxide inhibits hormone-sensitive steroidogenesis in rat luteal cells by the uncoupling of occupied LH receptors from adenylate cyclase (Musicki et al., 1994), and inhibiting synthesis of proteins involved in the intracellular movement of cholesterol to mitochondrial cytochrome P450 side chain cleavage enzyme (Musicki et al., 1994). A recent study has shown that a functional bovine corpus luteum (day 21 of pregnancy) expresses higher levels of manganese SOD than a regressed corpus luteum on day 21 of the estrous cycle (Rueda et al., 1995). The authors suggested that a decline in the synthesis of manganese SOD and other enzymes that prevent oxidative stress may contribute to the luteolytic process (Rueda et al., 1995). In the present study, the combined amount of newly-synthesized TIMP-2 and manganese SOD present in the medium were higher on day 17 of the estrous cycle than on day 1 7 of pregnancy. Interestingly, protein synthesis and secretion by the corpus luteum was observed as late as day 19 of the estrous cycle, despite the decline in plasma progesterone and regression of the corpus luteum at this time (Fields et al., 1992). Seemingly, the corpus luteum maintains the capability to synthesize proteins independently of steroidogenesis suggesting diffferential regulation of these two functions. These results are consistent with those previously deschbed by Juengel et al. (1994) who also observed protein synthesis and secretion by bovine corpora lutea during luteolysis. In their study, two of the proteins produced, TIMP-1 and TIMP-2, were secreted in amounts observed to

PAGE 116

99 change during the PGFsa-induced luteolysis (Juengel et al., 1994). In the present study, the bovine corpus luteum synthesized and released a number proteins during involution and regression. The roles these proteins play in the luteolytic process may be significant. In conclusion, the bovine CL synthesized and released in culture at least eleven proteins, some of which were identified as Apo E, Apo A-1 TIMP-1 and TIMP-2 and Mn SOD. These may act in an autocrine, paracrine, and/or endocrine manner to regulate luteal development, tissue remodelling, and steroidogenesis.

PAGE 117

CHAPTER 4 EXPRESSION OF MESSENGER RNA OF APOLIPOPROTEINS E AND A-1 IN BOVINE LUTEAL TISSUE DURING THE ESTROUS CYCLE AND PREGNANCY Introduction An earlier study (chapter 3) showed that bovine luteal tissue synthesizes and releases apolipoprotein E (Apo E) and apolipoprotein A-1 (Apo A-1) in explant culture. The apolipoproteins shared 100% amino acid sequence identity with bovine Apo A-1 (O'hUigin et al., 1990) and with bovine Apo E (Weisgraber, 1994). The synthesis and release of Apo A-1 by the corpus luteum of any species, and of Apo E by the corpus luteum of the cow is novel. Apo E is a constituent of very low density lipoproteins (VLDL), low density lipoprotein (LDL), chylomicrons, and certain subclassses of high density lipoprotein (HDL). Apo E mediates the binding of Apo E-containing lipoproteins to the LDL receptor (Mahley, 1988b). Apolipoprotein E is secreted by granulosa and theca cells of the rat (Wyne et al., 1989b), and has been shown to make up approximately 0.15% of the total proteins synthesized in the rat ovary (Driscoll et al., 1985). The theca cells have been shown to be the primary site of Apo E synthesis in the rat ovary (Wyne et al., 1989b). In a more recent study, Polacek et al. (1992) showed that expression of Apo E mRNA within ovarian follicles is 100

PAGE 118

101 differentially localized and developmentally regulated. Apo E mRNA is primarily located in theca cells of immature preantral follicles. The level of Apo E mRNA expression was shown to increase following maturation of the follicle and following treatment with gonadotropins. Granulosa cells of the Graafian follicle show greater expression of Apo E mRNA than cells of the preantral follicle (Polaceketal., 1992). Apo E mRNA expression in granulosa cells is stimulated by agents that stimulate cAMP production (Wyne et al., 1989a) including FSH and cholera toxin that stimulate granulosa cell Apo E secretion in a dose and time-dependent manner (Driscoll et al., 1985). Activation of protein kinases A and C, but not kinase G, independently stimulate synthesis and secretion of Apo E (Wyne et al., 1989a). Polacek et al. (1992) found no evidence of expression of Apo E mRNA in luteal tissue of the rat. Apo A-1 is the major protein of HDL (Sparrow et al., 1992). It is synthesized mainly by the liver but also by other peripheral tissues (Blue et al., 1982; Ferrari et al., 1986). Shackelford and Lebherz (1983) have shown that synthesis of Apo A-1 by the chick liver, skeletal and smooth muscle, is developmentally regulated. It has also been reported that breast muscle of the chick directs more of its protein synthetic efforts to the production of Apo A-1 than does the liver, around the time of hatching (Ferrari et al., 1986). Our previous study showed that Apo A-1 is synthesized by the bovine corpus luteum during the estrous cycle and pregnancy, and the amount synthesized and

PAGE 119

102 released did varied across days the estrous cycle and pregnancy examined (chapter 3). The objectives of the present study were to determine the presence, and temporal expression of mRNA for Apo E and Apo A-1 in bovine luteal tissue across the days of the estrous cycle and pregnancy. Materials and Methods Materials All electrophoresis reagents were of electrophoresis grade and were purchased from Fisher Scientific (Atlanta, GA). Biotrans nylon membrane was obtained from ICN Biomedicals (Costa Mesa, CA). The nick translation labeling kit was obtained from Amersham International pic (Buckinghamshire, England). P^P]dCTP alpha was purchased from ICN Biomedicals. RNA molecular weight markers were from Gibco BRL (Life Technologies Inc., Gaithersburg, MD). XAR5 film was obtained from Eastman Kodak (Rochester, NY). Prostaglandin Fja was obtained from Upjohn (Kalamazoo, Ml). Tissue Collection Angus and Hereford crossbred cows were used for the study. All procedures in which animals were used were approved by the Animal Care and Use Committee of the University of Florida. Estrus (day 0) was determined by observing cows twice daily for standing to be mounted by a bull rendered

PAGE 120

103 incapable of mating. Cows were bred by artificial insemination at observed estrus. Day 17 pregnancy was confirmed by the presence of an embryo in flushings from the uterus. Later stages of pregnancy were estimated by measurement of crown-rump length of the calves (Winters et al., 1942). Reproductive tracts were collected within 5 min after exsanguination from cows on days 2-3 (n = 4), 7 (n = 1 ), 1 6-1 7 (n = 4) and 20 (n = 2) of the estrous cycle, and on days 17 (n = 5), 90-120 (n = 2), 170-180 (n = 3) and > 215 (n = 5) of pregnancy. The corpus luteum was dissected from surrounding ovarian stroma, snap-frozen in liquid nitrogen, and then stored at 80 C until further analysis. Isolation of RNA Total cellular RNA was isolated from luteal tissue by the acid guanidinium thiocyanate-phenol-chloroform extraction method (Puissant and Houdebine, 1990). Fresh or frozen tissue (1 g) was homogenized on ice in a solution containing 10 ml 4 M cold guanidinium thiocyanate, 1 ml 2 M sodium acetate (pH 5.0) and 78 pi 2-mercaptoethanol, using a polytron tissue homogenizer (Tissumizer'^, Tekmar Company, Cincinnati, OH). Total RNA was extracted from homogenate with a phenol-chloroform mixture (10 ml:2 ml) and centrifuged at 4000 x g for 1 5 min. The upper phase was treated with isopropanol and centrifuged to precipitate the RNA. The pellet was resuspended in 4 M lithium chloride, repelleted by centrifugation, and dissolved in RNA buffer (10 mM Tris, 1 mM EDTA, 0.5% (w/v) SDS, pH 7.5). RNA in solution was treated with

PAGE 121

104 chloroform and the upper phase collected, treated with 2 M sodium acetate and isopropanol, and stored at 80 C to reprecipitate the RNA. Pure RNA pellet was retrieved after centrifugation at 4000 x g for 1 min and dissolved in sterile water. The concentration and purity (260 : 280 nm ratio) of the RNA were determined, integrity of the RNA preparations was assessed by electrophoresis on a 1 .2% (w/v in 1X TAE buffer) agarose gel. Total RNA samples were stored at 80 C until further analysis. Preparation of Plasmid DNA Plasmid DNA containing cDNA of Apo A1 (clone pAI-1 1 3) and Apo E (clone pE-301) from human liver (Breslow et al., 1982) was purchased from American Type Cell Culture (ATCC, Rockville, MD). Each plasmid DNA was reconstituted in Tris-EDTA buffer (10 mM Tris, 1 mM EDTA, pH 7.5), and spread on tet^ LB plates at 37 C overnight. Single bacteria clones were picked and added to 200 ml LB broth containing tetracycline (0.1 mg/ml), and then incubated at 37 C overnight with shaking. The plasmid DNA in the overnight culture was purified using the QIAGEN protocol (QIAGEN Inc., Chatsworth, CA), based on the optimized alkaline lysis method of Birnboim and Doly (1979). The incubation mixture was centrifuged at 5000 x g for 15 min at 4 C. The plasmid DNA in the pellet was lysed by suspending the pellet in an alkaline solution. The lysate was neutralized with acidic potassium acetate. The mixture was centrifuged to precipitate denatured proteins, chromosomal DNA and cellular

PAGE 122

105 debris. The supernatant containing the plasmid DNA was then loaded on a resin (QIAGEN-tip column). The DNA was eluted from the column with elution buffer containing 1.25 M NaCI (pH 8.5). The eluted plasmid DNA was desalted and concentrated by isopropanol precipitation. The solution was centrifuged at 20,000 X g for 45 min at 4 C. The DNA pellet was washed with 70% ethanol and centrifuged again. The purified plasmid DNA pellet was lyophilized and dissolved in TE buffer (10 mM Tris, 1 mM EDTA, pH 7.5). Yield was determined by measuring DNA concentration on a spectrophotometer at 260 nm, and the purity of the plasmid cDNA was assessed on a 1 .2% agarose gel. Restriction Analysis and Isolation of Insert Plasmid preparations from the clones of Apo E and Apo A1 were digested with Psti at 37 C for 16 h. Plasmid digests were separated by electrophoresis on a 1.2% agarose gel stained with ethidium bromide. Fragments were transferred by electroblotting to DEAE cellulose membrane (Schleicher and Schuell, Keene, NH), and then extracted from membrane with a high salt buffer (1 M NaCI, 0.1 mM EDTA, 20 mM Tris, pH 8). Extracted inserts were reprecipitated with 3 M sodium acetate and 100% ethanol, then centrifuged at 12,000 X g to repellet. The DNA pellet was freeze dried and dissolved in TrisEDTA buffer (10 mM Tris, 1 mM EDTA, pH 7.5). Size of purified inserts were checked on a 1.2% agarose gel by comparison to a series of RNA markers. Digestion of the Apo E cDNA-containing plasmid with PstI yielded two

PAGE 123

106 fragments of about 500 and 300 base pairs. Fragments were pooled before labeling by nick translation. Digestion of Apo A-1 cDNA-containing plasmid (pAI-1 13) with PstI yielded a single fragment of 600 base pairs. Northern Hybridization To determine the size of mRNA for Apo E and Apo A-1 in luteal tissue of different reproductive states, total RNA (30 pg) was denatured in denaturing buffer (24 mM HEPES, 6 mM sodium acetate, 1.2 mM EDTA, 50% (v/v) deionized formamide, 2.2 M formaldehyde) for 1 h on ice, followed by incubation at 65 C for 15 min. Denatured samples were separated by electrophoresis at 20 V for 16 h on a 1.5% agarose-formaldehyde gel in IX running buffer (0.5 M NaH2P04, 0.5 M Na2HP04, pH 7). The gels were stained with ethidium bromide, and the presence of distinct bands of the 28 S and 18 S ribosomal RNA was used to assess the integrity of the RNA (Sambrook et al., 1989). Fractionated RNA was transferred by capillary blotting to nylon membrane (Biotrans, ICN) using the TurboBlotter™ (Schleicher and Schuell) in 20X SSC (single-strength SSC = 0.15 M sodium chloride, 0.015 M sodium citrate). Filters were exposed to ultraviolet light for 90 sec to immobilize the RNA, then prehybridized for 2 h at 42 C in prehybridization buffer (5X Denhardt, 4X SSC, 0.5 M sodium phosphate pH 6.5, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) followed by hybridization at 42 C for 16 h in buffer (IX Denhardt, 4X SSC, 0.5 M sodium phosphate, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v)

PAGE 124

107 formamide) containing ^^P-cDNA probes (10^ cpm/ ml of hybridization buffer) labeled by nick translation (Amersham International pic). Blots were washed twice in double-strength SSC-0.1% SDS at 42 C for 15 min each, with shaking, followed by two washes in 0.1 -strength SSC-0.1% SDS at 42 C for 15 min each. Autoradiography Blots were exposed to x-ray film (Kodak XAR, Eastman Kodak) and an intensifying screen at 80 C to detect hybridization signals. For hybridization with a different probe, blots were washed with hot (90 C) 10% SDS for 1 h. In this study, blots were rehybridized with a f3.-actin cDNA probe at 42 C to assess the amount and integrity of total RNA loaded in each lane. Dot Blot Hybridization Total RNA (10 pg) was denatured in denaturation buffer (20 mM Tris, pH 7.0, 50% formamide, 6% formaldehyde), followed by incubation at 65 C for 5 min. Denatured samples were immobilized on membrane (Biotrans, ICN) using a microsample filtration unit (Schleicher and Schuell). Filters were exposed to ultraviolet light for 90 sec to crosslink the RNA to the membrane, then prehybridized in buffer (5X Denhardt, 4X SSC, 0.5 M sodium phosphate pH 6.5, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) at 42 C for 2 h, followed by hybridization in buffer (IX Denhardt, 4X SSC, 0.5 M sodium phosphate, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide)

PAGE 125

108 containing the ^^P-cDNA (lO^cpm/ ml of hybridization buffer) labeled by nick translation (Amersham International pic) at 42 C for 16 h. Blots were washed twice with 2X SSC, 0.1% SDS at 42 C for 15 min each on a shaker, followed by two washes with 0.1X SSC, 0.1% SDS at 42 C for 15 min each. Blots were exposed to x-ray film (Eastman Kodak) and/or phosphorimaging cassettes, and hybridization signals were quantified on a phosphorimager (ImageQuant and Phosphorimager, Molecular Dynamics Inc., CA), The relative expression of each mRNA to that of ll-actin was used to calculate least square means. Statistical Analysis Data from dot blot analyses were analysed by Least Squares Analysis of Variance of the General Linear Models of the Statistical Analytical System (SAS, 1988). The statistical model had day and reproductive status as the main effects with residuals as error term. Values in the text are least square means (LSM) standard error of the mean (SEM). Orthogonal contrasts were used to determine differences among days of the estrous cycle and stages of pregnancy.

PAGE 126

109 Results Northern Blot Analysis of Ado E and Apo A-1 mRNA Northern blot analysis revealed the presence of a single Apo E mRNA (1.0 kb) in bovine luteal tissue. Apolipoprotein E mRNA was expressed only during the early luteal phase (days 2-3). No hybridization was detected on the other days of the cycle or during pregnancy (Fig. 4-1). Apolipoprotein A-1 cDNA hybridized with a single mRNA transcript of approximately 1 kb in size (Fig. 4-2). Northern blots showed negligible expression of Apo A-1 mRNA on days 2-3 of the cycle. However, expression of Apo A-1 mRNA was detected on day 7, continued throughout the remainder of the estrous cycle and during pregnancy. Dot Blot Analysis In agreement with results obtained with northern blots, dot blot analysis showed expression of Apo E mRNA only on days 2-3 of the estrous cycle (Fig. 4-3). Although negligible in northern blots, dot blots revealed expression of Apo A-1 mRNA on days 2-3 (Fig. 4-4). Although expression of Apo A-1 mRNA was not different across days of the cycle, it did vary with stage of pregnancy (P < 0.01). Expression of Apo A-1 mRNA was at its lowest on days 90-120, and increased to significantly higher levels during the second half of pregnancy

PAGE 127

110 CM CM CYCLE (D l^ O CO 1^ ^ P c5 PREG -s 1 J: 1kb-ApoE -ACTIN Figure 4-1 Northern blot analysis of apolipoprotein E mRNA. The same blot was probed with fi-actin that served as a control for the loading and the integrity of the RNA. Expression of Apo E mRNA (1 .0 kb) was observed only on days 2-3 of the estrous cycle, and was absent on the other days of the estrous cycle (7, 16, 17, and 20) and pregnancy (17 and 90) examined.

PAGE 128

111 CYCLE PREGNANCY I 1 I — r 1 ^ ^ O O CM CMCMCOCO fs.i=^i^ CmPO) t-i= CM i^ffHRnUI* -ACTIN Figure 4-2. Northern blot analysis of apolipoprotein A-1 mRNA. The same blot was probed with B-actIn that served as a control for the loading and the integrity of the RNA. Expression of Apo A-1 mRNA (1 .0 kb) was not observed on days 2 and 3 of the estrous cycle, was expressed on day 7 and continued through the other days of the estrous cycle (7, 16, 17, and 20) examined. Apo A-1 mRNA was expressed on days 17, 90, 170, 180, and day 272 of pregnancy.

PAGE 129

112 CMC0I^i-CMt-O)i-CM QQQQOQOQQ V Figure 4-3. Dot blot analysis of Apo E mRNA. Ten pg total RNA isolated from CL during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^^P-labelled Apo E cDNA. Expression of Apo E mRNA was observed only on days 2-3 of the estrous cycle, was not observed on the other days of the cycle (16-17, and 20) and pregnancy (17, 90, 170, and 272) examined.

PAGE 130

113 (days 90-120 vs days 170-180 and 215 or greater; P < 0.002), with no difference (P < 0.47) in expression between days 170-180 and days 215 or greater of gestation (Fig. 4-5). Q. O ^ O Q. <^ 6 o 1^ 6 ^ CM 1CM T0) •^ i\ Q O Q O O O Q I I I I I I • # # 't' # Figure 4-4. Dot blot analysis of Apo A-1 mRNA. Ten fjg total RNA isolated from CL during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with '^P-labelled Apo A-1 cDNA. Expression of Apo A-1 mRNA was observed on all days of the estrous cycle (2-3, 16-17, and 20) and pregnancy (17, 90-120, 170-180, and > 215) examined.

PAGE 131

^ 0.35 ^ "S 0.30 i 0.25 „ 0.20 < w < ^ j2 0.15 0.10 s 0.05 ra 0.00 114 T777?i Cycle Pregnancy 16-17 20 17 90-120 170-180 >215 Day of Cycle/Pregnancy Figure 4-5. Apo A-1 mRNA expression relative to f3.-actin is presented as LSMean SEM. Levels of Apo A-1 mRNA expression were similar across days of the estrous cycle, but were higher during the second compared with the first half of pregnancy.

PAGE 132

115 Discussion In addition to the presence of radiolabeled apolipoprotein E and apolipoprotein A-1 in iuteal-conditioned medium following incubations with ^Hleucine (chapter 3), the presence of mRNA for apolipoproteins A-1 and E in bovine luteal tissue is further evidence for synthesis of these proteins by luteal tissue. This is the first report of expression of mRNA for apolipoproteinn A-1 by the corpus luteum of any species, and for apolipoprotein E by the bovine corpus luteum. Nicosia et al. (1992) reported the presence of mRNA for apolipoprotein E in regressed corpus luteum of the rat. In the present study, bovine luteal tissue showed a single hybridization signal with the apolipoprotein E cDNA probe, with size of the Apo E mRNA transcript (1 .0 kb) similar to that expressed in rat ovarian follicles (Polacek et al., 1992). Apolipoprotein E mRNA was expressed only during the early luteal phase (days 2-3). No expression of Apo E mRNA was observed on the other days of the estrous cycle and pregnancy examined. Studies in the rat have shown that Apo E mRNA is expressed primarily by theca cells of the ovarian follicle. Levels of expression of Apo E mRNA have been shown to be lower in granulosa cells, and become undetectable with formation of the corpus luteum (Polacek et al., 1992). However, Nicosia et al. (1992) observed expression of Apo E mRNA in theca cells and regressed corpus luteum, but not in granulosa cells of the rat. In the present study, Apo E mRNA

PAGE 133

116 was expressed only during early luteal phase (days 2-3) of the estrous cycle, the only period when the bovine corpus luteum has been observed to synthesize and release Apo E in explant culture (chapter 3). Apolipoprotein E is a constituent of liver-synthesized very low density lipoprotein (VLDL) and of a subclass of high density lipoprotein (HDL) (Mahley, 1988). The significance of luteal synthesis Apo E, and its production only during the early luteal phase of the estrous cycle is not understood. We suggest that it may act in an autocrine and /or paracrine fashion to provide cholesterol for membrane biosynthesis and cell proliferation at a time when the corpus luteum is undergoing rapid reorganization and growth following the trauma of ovulation (Zheng et al., 1993). A similar role of apolipoprotein E in nerve regeneration following injury has been suggested (Mahley, 1988). Apolipoprotein E also functions as a heparin binding protein, thus may compete with other heparin binding growth factors (fibroblast growth factor -2, FGF-2), influencing luteal cell interaction with the extracellular matrix (Mahley, 1988). Zheng et al. (1993) reported the presence of FGF-2 in the boviine corpus luteum during the estrous cycle. Immunohistochemical staining showed apolipoprotein E was restricted to the large steroidogenic luteal cells of days 3, 7, 1 1 and 14 of the estrous cycle. No staining was observed in the corpus luteum of pregnancy (F.M. NdikumMoffor, P.A. Fields, M.J. Fields unpublished data). It appears that apolipoprotein E may be localized in secretory granules since immunostaining

PAGE 134

117 was specifically localized in a region close to the nucleus, characteristically occupied by a cluster of secretory granules (Fields et al., 1992). Thus, even though synthesis of apolipoprotein E is restricted to the first days of the forming CL, it appears the protein is stored in the large luteal cell for release later in the estrous cycle. The temporal relationship of mRNA, de novo synthesis, and immunohistochemical staining is in a manner very similar to that observed for bovine luteal oxytocin (Ivell et al., 1987; Fields et al., 1992). Since macrophages and monocytes have been shown to produce large quantities of Apo E (Driscoll et al., 1985), these cells may also contribute to luteal synthesis of Apo E. Synthesis of apolipoprotein E and presence of its mRNA only during the early luteal phase indicate presence of a tightly controlled regulatory factor. A good candidate would be the gonadotropins and agents that stimulate cAMP and protein kinase A, which have been shown to increase the production of Apo E (Driscoll et al., 1985) and its mRNA (Polacek et al., 1992). Dot blot analysis revealed the presence of Apo A-1 mRNA in bovine luteal tissue in which expression was similar across the estrous cycle, but differed across pregnancy. Level of expression was higher with advancing stages of pregnancy. A single hybridization signal (1 .0 kb) was observed similar to that expressed by liver and other tissues (Ferrari et al., 1986). There is precedence for extrahepatic secretion of apolipoprotein A-1 in a variety of tissues including the yolk sac, gut, adrenal, kidney, human brain (Hopkins et al., 1986), pig brain (Guttler et al., 1990), liver, kidney, muscle, intestine, and testis

PAGE 135

118 of the rooster (Blue et al., 1982), baboon hepatocytes (White et al., 1994), and genital ridge of human embryos (Vorob'eV and Perevozchikov, 1992). Apo A-1 is the major protein component of high density lipoprotein (HDL), the primary form of circulating cholesterol in the bovine species (Sparrow et al., 1992). The role of Apo A-1 in the CL can only be speculative at this point of our investigation. Pate and Condon (1989) showed that bovine luteal cells require lipoproteins for steroidogenesis. Although LDL is 3.5 times more potent than HDL, the latter is more important in vivo in the cow because of its higher concentrations in blood (Pate and Condon, 1989). It has also been shown that HDL is required in maintaining bovine luteal cells in culture, and that luteal cells utilize both LDL and HDL as a source of cholesterol for steroidogenesis (Pate and Condon, 1989). Luteal Apo A-1 may play an autocrine and/or paracrine role in the CL and ovary to provide cholesterol for steroidogenesis. Prostaglandin Fjc, the natural luteolysin in cattle has been shown to inhibit the lipoprotein-induced increase in progesterone production by bovine luteal cells (Pate and Condon, 1989), not by inhibiting cholesterol entry into the cells, but by inhibiting its utilization for steroidogenesis. In a recent study, Grusenmeyer and Pate (1992) reported that PGF2a does not affect lipoproteininduced increase in cellular or mitochondrial cholesterol content but instead, has an effect after cholesterol transport to mitochondria, but prior to cholesterol side chain cleavage.

PAGE 136

119 It has been suggested that Apo A-1 interacts with HDL to regulate placental biosynthesis of cholesterol and other events involved with onset of labor (Del Priore et al., 1991). The observed increase in luteal Apo A-1 mRNA as pregnancy progressed is congruent with a role for this protein during pregnancy. Serum concentrations of Apo A-1, HDL, and cholesterol in women have been shown to increase as pregnancy progressed, decline to normal levels during the periparturient period, and decline further with onset of labor (Del Priore et al., 1991). Apolipoprotein A-1 has also been shown to increase production of placental lactogen by human trophoblast cells by stimulating cAMP (Wu et al., 1988). An increase in luteal apolipoprotein A-1 during pregnancy could have a systemic effect on the placenta, although it is more likely that it is involved in an autocrine and/or paracrine role at the level of the ovary. The cell types responsible for luteal synthesis of apolipoprotein A-1 have not been identified, but studies to investigate that are underway. Identification of the cells may give additional insight into the function of luteal apolipoprotein A-1. In conclusion, the bovine corpus luteum expressed mRNA for apolipoproteins E and A-1 during the estrous cycle and pregnancy. Although the role of these proteins in the corpus luteum is undefined, they may act in an autocrine and/or paracrine manner to regulate luteal development and/or steroidogenesis.

PAGE 137

CHAPTER 5 EXPRESSION OF MESSENGER RNA OF TISSUE INHIBITOR OF METALLOPROTEINASES1 AND -2 IN BOVINE LUTEAL TISSUE DURING THE ESTROUS CYCLE AND PREGNANCY Introduction Luteal cells of domestic ruminants have tine structural machinery required for the synthesis and secretion of proteins or peptide hormones. Tissue inhibitors of metalloproteinases (TIMP) are produced by different cells and tissues. In the cow (Freudenstein et al., 1990; Juengel et al., 1994), ewe (Smith et al., 1993; 1995), and pig (Tanaka et al., 1992), the corpus luteum (CL) has been shown to synthesize TIMP-1 and TIMP-2. In an earlier study (chapter 3), we showed that bovine luteal tissue synthesizes and releases tissue inhibitor of metalloproteinases-1 and 2 during the estrous cycle and pregnancy. TIMPs are inhibitors of enzymes that degrade the protein component of the extracellular matrix. Apart from their protease inhibitory activity, TIMP-1 (Hayakawa et al., 1992) and TIMP-2 (Stetler-Stevenson et al., 1992) have been shown to promote the growth of different types of cell. TIMP-1 enhances growth of bovine embryos in culture (Satoh et al., 1994). More recently, Boujrad et al. (1995) have shown that TIMP-1 of testicular origin stimulates the synthesis of pregnenolone by Leydig cells, and progesterone by Sertoli cells of the rat testis. 120

PAGE 138

121 TIMP-1 is a 28-30 kDa glycosylated protein whereas TIMP-2 has a molecular mass of 20 kDa and is not glycosylated. In addition, TIMP-1 has a sequence homology with the granulocyte macrophage colony stimulating factor (GM-CSF) in the N-terminal region whereas TIMP-2 has no such sequence (Patarca and Haseltine, 1985). Waterhouse et al. (1993) have also reported that the temporal expression of TIMP-1 is different from that of TIMP-2 in reproductive tissues of the rat, suggesting that these proteins are independently regulated (Waterhouse et al., 1993). Earlier reports in sheep (Smith et al., 1994) and cattle (Freudenstein et al., 1990) examined changes in expression of TIMP-1 and TIMP-2 mRNA during the estrous cycle. There is a paucity of information about the patterns of expression of mRNA for TIMP-1 and TIMP-2 during pregnancy. The objectives of this study were to determine expression of TIMP-1 and TIMP-2 in bovine luteal tissue, and differences in expression across the estrous cycle and pregnancy. Materials and Methods Materials All electrophoresis reagents were of electrophoresis grade and were purchased from Fisher Scientific (Atlanta, GA). Biotrans nylon membrane was obtained from ICN Biomedicals (Costa Mesa, CA). The nick translation labeling

PAGE 139

122 kit was obtained from Amersham International pic (Buckinghamshire, England). [^^PJdCTP alpha was purchased from ICN Biomedicals (Costa Mesa, CA). RNA molecular weight markers were from Gibco BRL (Life Technologies Inc., Gaithersburg, MD). XAR-5 film was obtained from Eastman Kodak (Rochester, NY). Prostaglandin Fjc was obtained from Upjohn (Kalamazoo, Ml). Tissue Collection Angus and Hereford cross-bred cows were used for the study. All procedures in which animals were used were approved by the Animal Care and Use Committee of the University of Florida. Estrus was synchronized in cows by injection of 25 mg prostaglandin Fja (Lutalyse, Upjohn). Cows were observed for estrous behavior twice daily with assistance of a bull. Some cows were bred by artificial insemination at observed estrus (estrus = day 0). Day 17 pregnancy was confirmed by the presence of an embryo in flushings from the uterus. Later stages of pregnancy were estimated by measurement of crown-rump length of the calves (Winters et al., 1942). Reproductive tracts were collected within 5 min after exsanguination from cows on days 2-3 (n = 3), 7 (n = 1), 16-17 (n = 4) and 20 (n = 2) of the estrous cycle, and on days 17 (n = 5), 90-120 (n = 2), 170180 (n = 3) and > 215 (n = 5) of pregnancy. The CL was dissected from surrounding ovarian stroma, snap-frozen in liquid nitrogen, and then stored at 80 C until further analysis.

PAGE 140

i 123 Isolation of RNA ••< Total cellular RNA was isolated from luteal tissue by the acid guanidinium thiocyanate-phenol-chioroform extraction method (Puissant and Houdebine, 1990). Fresh or frozen tissue (1 g) was homogenized on ice in a solution containing 10 ml 4 M cold guanidinium thiocyanate, 1 ml 2 M sodium acetate (pH 5.0) and 78 |j| 2-mercaptoethanol, using a polytron tissue homogenizer (Tissumizer'^, Tekmar Company, Cincinnati, OH). Total RNA was extracted from homogenate with a phenol-chloroform mixture (10 ml:2 ml) and centrifuged at 4,000 X g for 15 min. The upper phase was treated with isopropanol and centrifuged to precipitate the RNA. The pellet was resuspended in 4 M lithium chloride, repelleted by centrifugation, and dissolved in RNA buffer (10 mM Tris, 1 mM EDTA, 0.5% (w/v) SDS, pH 7.5). The RNA in solution was treated with chloroform and the upper phase collected, treated with 2 M sodium acetate and isopropanol, and stored at 80 C to reprecipitate the RNA. Pure RNA pellet was retrieved after centrifugation at 4,000 x g for 10 min and dissolved in sterile water. The concentration and purity (260 : 280 nm ratio) of the RNA were determined. Integrity of the RNA preparations was assessed by electrophoresis on a 1.2% (w/v) in IX TAE buffer (Tris, Acetic acid, EDTA) agarose gel. Total RNA samples were stored at 80 C until further analysis.

PAGE 141

124 Preparation of Plasmid DNA Plasmid DMAs containing cDNA of TIMP-1 (clone 6-2) and TIMP-2 (clone MMI) In agar stabs were a gift from Dr. Michael F. Smith of the University of Missouri, Columbia, MO (Smith et al., 1995). Each plasmid cDNA (100 pi) was suspended in 1 ml LB medium containing ampicillin (100 mg/l), and placed at 37 C overnight with shaking. The bacteria culture was grown on LB-Amp plates at 37 C overnight. A single bacterial colony was picked from the LB-Amp plate and suspended in 200 ml LB medium containing ampicillin at a final concentration of 100 pg/ml, and then incubated with shaking at 37 C overnight with shaking. The culture was centrifuged at 7,000 x g for 10 min at 4 C, and the plasmid DNA in the pellet was lysed by suspending the pellet in an alkaline mixture containing 5.8 ml ST buffer (25% sucrose, 50 mM Tris, pH 8.0), 1.95 ml lysozyme solution (10 mg/ml in ST) for 5 min. The mixture was centrifuged briefly, placed on ice for 5 min. Then 4.75 ml 0.2 M EDTA was added, followed by addition of 12.5 ml Triton Lysis buffer (1 ml 10% Triton X-100, 5 ml 1 M Tris (pH 8.0), 62.5 pi 0.1 M EDTA to 100 ml with sterile water). The mixture was centrifuged at 12,000 x g for 30 min to precipitate denatured proteins, chromosomal DNA and cellular debris. The supernatant containing the plasmid DNA was treated with 0.026X its volume of 8 M ammonium acetate. The DNA in the supernatant was extracted in a phenol-chloroform mixture and repelleted by centrifugation. The DNA pellet was lyophilized and dissolved in TE buffer (10 mM Tris, 1 mM EDTA, pH 7.5).

PAGE 142

125 Yield was determined by measuring DNA concentration on a spectrophotometer at 260 nm, and the purity of the plasmid cDNA was assessed on a 1 .2% agarose gel. Restriction Analysis and Isolation of Insert Plasmid DNA of TIMP-1 and TIMP-2 were digested following a double digestion protocol. For TIMP-1 the digestion reaction was carried out with Xho I restriction enzyme for 16 h, followed by EcoR1 at 37 C for 16 h. For TIMP-2, the cleavage reaction was done with Bam HI at 37 C for 16 h, followed by EcoRI at same temperature for 16 h. Plasmid digests were treated with RNase for 30 min at 37 C, and separated by electrophoresis on a 1 .2% agarose gel stained with ethidium bromide. The fragments were transferred by electroblotting to DEAE cellulose membrane (Schleicher and Schuell, Keene, NH), and then extracted from membrane with a high salt buffer (1 M NaCI, 0.1 mM EDTA, 20 mM Tris, pH 8). The extracted inserts were reprecipitated with 3 M sodium acetate and 100% ethanol, then centrifuged at 12,000 x g to repellet. The DNA pellet was freeze dried and dissolved in Tris-EDTA buffer (10 mM Tris, 1 mM EDTA, pH 7.5). Sizes of purified inserts were checked on a 1 .2% agarose gel by comparing with a series of RNA markers.

PAGE 143

126 Northern Hybridization Size of mRNA of TIMP-1 and TIMP-2 in luteal tissue was determined by northern blot analysis (Sambrook et al., 1989). Total RNA (30 [jg) was denatured in buffer (24 mM HEPES, 6 mM sodium acetate, 1.2 mM EDTA, 50% (v\v) deionized formamide, 2.2 M formaldehyde) for 1 h on ice, followed by incubation at 65 C for 15 min. Denatured samples were separated by electrophoresis at 20 V for 16 h on a 1 .5% agarose-formaldehyde gel in IX running buffer (0.5 M NaH2P04, 0.5 M Na2HP04, pH 7). The gels were stained with ethidium bromide, and the presence of distinct bands of the 28s and 18s ribosomai RNA was used to assess the integrity of the RNA. Fractionated RNA was transferred by rapid downward capillary blotting to nylon membrane (Biotrans, ICN Biochemicals) using the TurboBlotter™ (Schleicher and Schuell) in 20X SSC (single-strength SSC = 0.15 M sodium chloride, 0.015 M sodium citrate). Filters were exposed to ultraviolet light for 90 sec to immobilize the RNA, then prehybridized for 2 h at 42 C in 5X Denhardt (IX = 0.002% w/v ficoll, 0.02% w/v polyvinylpyrrolidone, 0.02% w/v BSA), 4X SSC, 0.5 M sodium phosphate pH 6.5, 0.1% w/v sodium dodecyl sulfate (SDS), yeast RNA (0.25 mg/ml), 50% (v/v) formamide). Filters were hybridized at 42 C for 16 h in buffer (IX Denhardt, 4X SSC, 0.5 M sodium phosphate, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) containing ^^P-TIMP-I or ^^p. TIMP-2 cDNA (10^ cpm /ml hybridization buffer) labeled by nick translation

PAGE 144

127 (Amersham International). Filters were washed twice in double-strength SSC0.1% SDS at 42 C for 15 min each, followed by two washes in 0.1 -strength SSC0. 1 % SDS at 42 C for 1 5 min each. Autoradiography Filters were exposed to x-ray film (XAR, Eastman Kodak) and an Intensifying screen for 4 days at 80 C to detect hybridization signals. For hybridization with a different probe, blots were washed with hot (90 C) 1% SDS for 1 h. Blots were rehybridized with a R.-actin cDNA probe at 42 C to assess the amount and integrity of total RNA loaded in each lane. Dot Blot Hybridization Total RNA (10 pg) was denatured in denaturation buffer (20 mM Tris, pH 7.0, 50% formamide, 6% formaldehyde), followed by incubation at 65 C for 5 min. The denatured samples were immobilized on membrane (Biotrans, ICN) using a microsample filtration unit (Schleicher and Schuell). The filters were exposed to ultraviolet light for 90 sec to crosslink the RNA to the membrane, then prehybridized in buffer (5X Denhardt, 4X SSC, 0.5 M sodium phosphate pH 6.5, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) at 42 C for 2 h, followed by hybridization in buffer (1X Denhardt, 4X SSC, 0.5 M sodium phosphate, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) containing the ^^P-cDNA TIMP-1 or TIMP-2 (ICN Biomedicals) labeled by nick

PAGE 145

* 128 translation (Amersham International pic) at 42 C for 16 h. The blots were washed twice with 2X SSC, 0.1% SDS at 42 C for 15 min each time, followed by two washes with 0.1X SSC, 0.1% SDS at 42 C for 15 min each time. The blots were exposed to x-ray film (Eastman Kodak) and phosphorimaging cassettes, and hybridization signals were detected and quantified on a phosphorimager (ImageQuant and Phosphorimager, Molecular Dynamics Inc., CA). The relative expression of each mRNA to that of ll-actin was used to calculate least square means. Statistical Analysis Data from dot blot analyses were analysed by least squares alysis of variance using the General Linear Models procedure of the Statistical Analytical System (SAS, 1988). The statistical model had day and reproductive status as the main effects with residuals as the error term. Values in the text are least square means (LSM) standard error of the mean (SEM). Orthogonal contrasts were used to determine differences in mRNA expression among days of the estrous cycle, and stages of pregnancy. Results Expression of TIMP-1 mRNA Bovine luteal tissue showed a major hybridization signal (0.9 kb) with ovine TIMP-1 cDNA probe, and a weaker signal at 3.0 kb (Fig. 5-1 ). Dot blot .^t

PAGE 146

129 CYCLE PREGNAN CY 18Sn •• • U M 19* -3-0kb • • -0-9kb TIMP-1 •Itl*W#lteii ii -ACTIN Figure 5-1 Northern blot analysis of luteal TIMP-1 mRNA. The same blot was probed for (1-actin that served as a control for the loading and the integrity of the RNA. The major transcript appeared at 0.9 kb and a minor transcript at 3.0 kb.

PAGE 147

130 analysis showed that expression of TIMP-1 mRNA varied across the estrous cycle (Fig 5-2). Orthogonal contrasts indicated that TIMP-1 mRNA levels were lower on days 2-3 compared to the other days examined (days 2-3 vs days IBIT and 20, P < 0.05) (Fig. 5-3). Q. Q. O O CO Q. N Q. CM T1 10 CO CVJ CD o CVJ N 8 o N CM A Q G G Q G G G • • • • Figure 5-2. Dot blot analysis of TIMP-1 mRNA. Ten pg total RNA isolated from luteal tissue during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^^P-labelled TIMP-1 cDNA at 42 C.

PAGE 148

131 There was no difference (P < 0.34) in TIMP-1 mRNA expression between days 16-17 and day 20 of the estrous cycle. Expression of TIMP-1 mRNA varied with stage of pregnancy, reaching its highest level during late pregnancy when compared with the other days of pregnancy examined (P < 0.007) (Fig. 5-3). Expression of TIMP-1 mRNA was not different between days 16-17 of the estrous cycle and day 1 7 of pregnancy. 0.35 o 0.30 (0 i Z 0-25 t£. P 0.20 (V m 0.15 s 1 jZ ^ 0.10 s 0.05 0.00 7ZP72 Cycle Pregnancy A 1 ^ 2-3 16-17 20 17 90-120 170-180 >215 Day of Cycle/Pregnancy Figure 5-3. TIMP-1 mRNA expression relative to B-actin is presented as LSMean SEM (n = 2-5) and differences determined by orthogonal contrasts. Days 2-3 vs days 16-17, 20 cycle (P < 0.05). Days 17, 90-120, 1 70-1 80 pregnant vs days > 21 5 pregnant (P<0.00f). Days 17, 90-120 vs days 170-180, > 215 pregnant {P<0.03).

PAGE 149

M 132 Expression of TIMP-2 mRNA Northern blot analysis revealed two transcripts (1 kb and 3.5 kb) of TIMP2 mRNA following hybridization with an ovine TIMP-2 cDNA probe (Fig. 5-4). CO 18S CYCLE P 1 I -3-5kb TIMP-2 Ikb -ACTIN Figure 5-4. Northern blot analysis of luteal TIMP-2 mRNA. The same blot was probed for f3>-actin to act as a control for the loading and integrity of RNA. The major transcript appeared at 1.0 kb and a minor transcript at 3.5 kb. The 1.0 kb-transcript was predominant.

PAGE 150

133 Results of dot blot analysis (Fig. 5-5) showed a strong effect of day on expression of TIMP-2 mRNA across the estrous cycle (P < 0.004) (Fig. 5-6). Q. Q. o O 00 o N s % CM \^ Q. J, 9 CM 6 o CM N 8 S O Q o a o o Q. 10 CM A O Figure 5-5. Dot blot analysis of luteal TIMP-2 mRNA. Ten pg total RNA isolated from luteal tissue during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^2p-labelled TIMP-2 cDNA at 42 C.

PAGE 151

134 Levels of TIMP-2 mRNA were lower on days 2-3 than on the other days examined (days 2-3 vs days 16-17 and 20, P < 0.002) (Fig. 5-6). ^ 0.8 ^ 0.7 re < 2 0.6 2 0) 0^ -^ 0.5 E ^ £ 0.4 c 0.3 H ^ 0.2 S 0.1 ~ 0.0 g^^ Cycle Pregnancy ^ M 2-3 16-17 20 17 90-120 170-180 > 215 Day of Cycle/Pregnancy Figure 5-6. TIMP-2 mRNA expression relative to B-actin is presented as LSMean SEM (n = 2-5) and differences determined by orthogonal contrasts. Days 2-3 vs days 16-17, 20 cycle (P < 0.002). Day 17 vs days 90-1 20 (P < 0. 10). Day 1 70-1 80 vs day > 21 5 (P < 0. 12)

PAGE 152

135 No difference {P <0.17) was observed in expression of TIMP-2 mRNA between day 16-17 and day 20 of the estrous cycle. Altliough expression of TIMP-2 mRNA did not change overall across pregnancy (P < 0.20), orthogonal contrasts indicated that expression tended to be lower during mid-pregnancy when compared with early (P < 0. 10) and late (P < 0. 12) pregnancy. The level of expression of TIMP-2 mRNA was similar between days 16-17 of the cycle and day 1 7 of pregnancy. Discussion The presence of mRNA for TIMP-1 and TIMP-2 within bovine luteal tissue in the present study confirms earlier observations of synthesis of these proteins by this tissue following incubations with ^H-leucine (Ndikum-Moffor et al., 1995). Previous studies have shown that TIMP-1 mRNA is expressed within CL of the cow (Freudenstein et al., 1990; Juengel et al., 1994), ewe (Smith et al., 1994), sow (Tanaka et al., 1992), rat (Mann et al., 1993), and mouse (Nomura et a!., 1989). In the present study, a 0.9 kb-TIMP-1 mRNA transcript was predominant, and a minor transcript was observed at 3.0 kb. The 0.9 kb-transcript is the only size reported in the studies mentioned above. However, M. F. Smith (personal communication) has also observed two TIMP-1 mRNA transcripts in the bovine CL. There was a significant effect of day on the expression of TIMP-1 and TIMP-2 mRNA during the estrous cycle, with levels higher during late than early

PAGE 153

136 luteal phase. Expression of TIMP-1 mRNA was similar between the first and second trimesters, and increased dramatically during the third trimester of pregnancy. These results agree with previous observations of greater synthesis of TIMP-1 protein during the late luteal phase of the estrous cycle, and during late pregnancy (Ndikum-Moffor et al., 1995), indicating that synthesis of TIMP occurs because of the presence of specific TIMP mRNA. However, these results differ from those of Freudenstein et al. (1990) who observed highest expression of TIMP-1 mRNA on days 1 -4 and 11-17, low levels on days 5-1 and 1 8-20 of the bovine cycle, and no expression after day 60 of gestation. In the ovine CL, levels of TIMP-1 mRNA did not vary across the estrous cycle (Smith et al., 1994). We are unable to offer a reason for the major discrepancies between the high levels of expression of TIMP-1 mRNA early in the estrous cycle, no expression after day 60 of pregnancy reported by Freudenstein et al. (1990) in contrast to the lowest expression of TIMP-1 mRNA early in the estrous cycle, and high expression late in pregnancy that was observed in this study. Northern blot analysis revealed two TIMP-2 mRNA species (approximately 1 .0 kb and 3.5 kb) within bovine luteal tissue, similar to that reported for the ewe (Smith et al., 1995), rat (Santoro et al., 1994), and human (Stetler-Stevenson et al., 1990). However, unlike in human, the 1.0 kb-transcript was predominant. Expression of TIMP-2 mRNA varied across the cycle, and was highest on day 16-17. Contrary to these results. Smith et al. (1995) reported in sheep a greater concentration of TIMP-2 mRNA during the early

PAGE 154

137 (days 3 and 7) than the late luteal phase (day 16), and Waterhouse et al. (1993) observed an increase in TIMP-2 mRNA In the rat ovary as pregnancy progressed. These differences may reflect species specificity in the synthesis of TIMP, and may indicate differential roles. The role of TIMP in the CL has not been clearly defined. However, tissue inhibitors of metalloproteinases are specific inhibitors of enzymes that degrade protein components of the extracellular matrix. Additionally, TIMP-1 promotes growth of a wide range of bovine and human cells (Hayakawa et al., 1992), and both TIMP-1 and TIMP-2 have been shown to possess erythroid-potentiating activity (Stetler-Stevenson et al., 1992). Thus TIMP-1 and TIMP-2 may be involved in remodeling of luteal tissue during luteal development as well as at the time of its regression. Factors regulating expression of mRNA of TIMP-1 and TIMP-2 have not been fully identified. However, evidence suggests that TIMP-1 and TIMP-2 are regulated through different mechanisms (Waterhouse et al., 1993). Expression of TIMP-1 has been shown to increase following the gonadotropin surge in the ewe (Smith et al., 1993; 1994), cow (Freudenstein et al., 1990) and rat (Reich et al., 1991). Mann et al. (1991) also reported that expression of TIMP-1 mRNA in rat ovarian granulosa cells is stimulated by luteinizing hormone, phorbol ester (TPA), and cAMP. The effects of TPA and LH were additive suggesting that each acts through separate intracellular second messenger systems (Mann et al., 1991). Follicle stimulating hormone (FSH) also increases the activity of TIMP-1 and TIMP-2, and expression of their respective mRNA in prepubertal rat

PAGE 155

138 Sertoli cells (Ulisse et al., 1994). Progesterone has also been shown to increase the production of TIMP-1 and TIMP-2 by rabbit uterine cervical fibroblasts in culture (Imada et al., 1994). The cell types responsible for the synthesis of TIMP-1 and TIMP-2 in the i ^ bovine CL have not been determined. In sheep, TIMP -1 mRNA is present in isolated small and large steroidogenic luteal cells and in cells within the luteal connective tissue, with greater expression observed in the large cells (Smith et ,, al., 1994). Similarly, TIMP-2 mRNA is expressed by isolated small and large ovine luteal cells, and other cells within luteal tissue (Smith et al., 1995). TIMP-1 and TIMP-2 mRNA are expressed by many cell types including fibroblasts, monocytes, and macrophages (Stetler-Stevenson et al., 1990). Apart from the steroidogenic large and small luteal cells, the CL also contains monocytes, macrophages, endothelial cells and fibroblasts. Thus the non-steroidogenic luteal cells may also contribute to luteal synthesis of TIMP-1 and TIMP-2. The role of TIMP in luteal function is not yet understood but has been suggested to be involved in tissue remodelling during formation of the CL following ovulation. However, the role of TIMP-1 and TIMP-2 in the CL during pregnancy has not been discussed. In the present study, expression of TIMP-1 mRNA and TIMP-2 mRNA was highest during the late luteal phase of the estrous cycle and late pregnancy. TIMP-1 may act in an autocrine and/or paracrine fashion to maintain steroidogenesis and CL function during the estrous cycle and pregnancy, and control tissue remodelling during involution and regression.

PAGE 156

CHAPTER 6 CHANGES IN THE EXPRESSION OF MESSENGER RIBONUCLEIC ACID FOR MANGANESE SUPEROXIDE DISMUTASE IN THE BOVINE CORPUS LUTEUM DURING THE ESTROUS CYCLE AND PREGNANCY Introduction Results from a previous experiment (chapter 3) showed that the bovine corpus luteum synthesizes and releases manganese superoxide dismutase (Mn SOD) in culture. Manganese SOD is one of three superoxide dismutases, including copper SOD and zinc SOD. Manganese SOD is encoded by a nuclear gene and is found predominantly in the mitochondria (Church, 1990). Superoxide dismutases are believed to protect cells from reactive oxygen species (such as superoxide anions) which may damage the cells. Superoxide dismutase catalyses the conversion of superoxide anions to hydrogen peroxide and oxygen; 20v + 2H+ = H2O2 + O2, and has been referred to as a scavenger of superoxide radicals. Damage to luteal cells by reactive oxygen species has been proposed as one of the mechanisms of luteal regression (Riley and Behrman, 1991). Production of Mn SOD can be stimulated by oxidants, cytokines, and lipopolysaccharides (Melendez and Baglioni, 1993). Laloraya et al. (1988) observed that LH induces production of superoxide dismutase, suggesting a role of SOD in luteal function. It was also observed that levels of 139

PAGE 157

140 ovarian superoxide dismutase and superoxide radicals change inversely during the rat estrous cycle. Prostaglandin Fja is the natural luteolysin in many domestic species. Treatment of rats with prostaglandin Fja in vivo has been shown to cause a rapid transient increase in superoxide radical production and a decrease in fluidity of plasma membranes from luteinized rat ovaries (Sawada and Carlson, 1991). In a later study Sawada and Carlson (1994) suggested that superoxide radicals may disrupt LH-stimulated progesterone secretion by rat CL. Sato et al. (1992) showed that treatment of rats with hCG causes a decrease in ovarian superoxide dismutase activity, especially that of Mn SOD, and an increase in expression of Mn SOD mRNA. However, Laloraya et al. (1988) observed a transient increase (declined by 60 min post-PGF injection) in ovarian total SOD activity, but the changes specific to Mn SOD were not discussed. Sato et al. (1992) also observed that long-acting SOD blocks hCG-induced ovulation while HjOj-inactivated SOD has no effect, and suggested the inhibition could be due to dismutation of superoxide radicals by SOD. Rueda et al. (1995) proposed that the increased oxidative stress during luteolysis was due to reduced activity of enzymes that metabolize superoxide radicals. This hypothesis was supported by observations that functional bovine CL (day 20 of pregnancy) expresses significantly higher levels of mRNA encoding Mn SOD than regressed CL (day 20 of the estrous cycle) (Rueda et al., 1995). • Little is known about the pattern of expression of Mn SOD during the bovine estrous cycle and pregnancy. The objective of the present study was to

PAGE 158

^"m 141 examine the changes in expression of mRNA encoding Mn SOD across day of the cycle and during pregnancy. Materials and Methods Tissue Collection Angus and Hereford crossbred cows were used for the study. Estrus (day 0) was determined by observing cows twice daily for estrous behavior with assistance of a bull surgically modified to be incapable of mating. Cows in the pregnant group were bred by artificial insemination at observed estrus. Day 17 pregnancy was confirmed by the presence of an embryo in flushings from the uterus. Later stages of pregnancy were estimated by measurement of crownrump length of the calves (Winters et al., 1942). Reproductive tracts were collected within 5 min after exsanguination from cows on days 2-3 (n = 3), 7 (n = 1), 16-17 (n = 4) and 20 (n = 2) of the estrous cycle, and on days 17 (n = 5), 90120 (n = 2), 170-180 (n = 3) and > 215 (n = 5) of pregnancy. The CL was dissected from surrounding ovarian stroma, snap-frozen in liquid nitrogen, and then stored at 80 C until further analysis. Isolation of RNA Total cellular RNA was isolated from luteal tissue by the acid guanidinium thiocyanate-phenol-chloroform extraction method (Puissant and Houdebine,

PAGE 159

142 1990). Fresh or frozen tissue (1 g) was homogenized on ice in a solution containing 10 ml 4 M cold guanidinium thiocyanate, 1 ml 2 M sodium acetate (pH 5.0) and 78 pi 2-mercaptoethanol, using a polytron tissue homogenizer (Tissumizer"^, Tekmar Company, Cincinnati, OH). Total RNA was extracted from homogenate with a phenol-chloroform mixture (10 ml:2 ml) and centrifuged at 4000 X g for 1 5 min. The upper phase was treated with isopropanol and centrifuged to precipitate the RNA. The pellet was resuspended in 4 M lithium chloride, repelleted by centrifugation, and dissolved in RNA buffer (10 mM Tris, 1 mM EDTA, 0.5% (w\v) SDS, pH 7.5). The RNA in solution was treated with chloroform and the upper phase collected, treated with 2 M sodium acetate and isopropanol, and stored at 80 C to reprecipitate the RNA. Pure RNA pellet was retrieved after centrifugation at 4000 x g for 10 min and dissolved in sterile water. The concentration and purity (260 : 280 nm ratio) of the RNA were determined. Integrity of the RNA preparations was assessed by electrophoresis on a 1 .2% (w/v in IX TAE buffer) agarose gel. Total RNA samples were stored at 80 C until further analysis. Restriction Analysis and Isolation of Plasmid DNA Insert Plasmid DNA containing human Mn SOD cDNA was a gift from Dr. Harry Nick of the Department of Biochemistry and Molecular Biology, University of Florida. The plasmid DNA (pUc 19) was digested with EcoRI at 37 C for 16 h. Plasmid digests were separated by electrophoresis on a 1.2% agarose gel

PAGE 160

143 stained with ethidium bromide. The fragments were transferred by electroblotting to DEAE cellulose membrane (Schleicher and Schuell, Keene, NH), and then extracted from membrane with a high salt buffer (1 M NaCI, 0.1 mM EDTA, 20 mM Tris, pH 8). The extracted insert was reprecipitated with 3 M sodium acetate and 100% ethanol, then centrifuged at 12,000 x g to repellet. The DNA pellet was freeze dried and dissolved in Tris-EDTA buffer (10 mM Tris, 1 mM EDTA, pH 7.5). The size of the purified insert was checked on a 1.2% agarose gel by comparing with a series of RNA markers. Digestion of plasmid DNA containing Mn SOD cDNA yielded one fragment about 1.5 kb in size. Northern Hybridization Size of mRNA for Mn SOD in luteal tissue was determined by northern blotting (Sambrook et al., 1989). Total RNA (30 pg) was denatured in denaturing buffer (24 mM HEPES, 6 mM sodium acetate, 1.2 mM EDTA, 50% (v/v) deionized formamide, 2.2 M formaldehyde) for 1 h on ice, followed by incubation at 65 C for 15 min. Denatured samples were separated by electrophoresis at 20 volts for 16 h on a 1 .5% agarose-formaldehyde gel in IX running buffer (0.5 M NaH2P04, 0.5 M Na2HP04, pH 7). The gels were stained with ethidium bromide, and the presence of distinct bands of the 28 s and 18 s ribosomal RNA was used to assess the integrity of the RNA. Fractionated RNA was transferred by capillary blotting to nylon membrane (Biotrans, ICN, Irvine, CA) using the TurboBlotter™ (Schleicher and Schuell) in

PAGE 161

144 20X SSC (single-strength SSC = 0. 1 5 M sodium chloride, 0.01 5 M sodium citrate). Filters were exposed to ultraviolet light for 90 sec to immobilize the RNA, then prehybridized for 2 h at 42 C in prehybridization buffer (5X Denhardt, 4X SSC, 0.5 M sodium phosphate pH 6.5, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) followed by hybridization at 42 C for 16 h in buffer (IX Denhardt, 4X SSC, 0.5 M sodium phosphate, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) containing ^^PMn SOD cDNA (ICN Biomedicals) (10^ cpm /ml of hybridization buffer) labeled by nick translation (Amersham International pic). Blots were washed twice in double-strength SSC-0.1% SDS at 42 C for 15 min each, with shaking, followed by one wash in 0.1 -strength SSC-0. 1 % SDS at 42 C for 1 min. Autoradiography Blots were exposed to x-ray film (Kodak XAR, Eastman Kodak) and an intensifying screen at 80 C to detect hybridization signals. For hybridization with a different probe, blots were washed with hot 1% SDS for 1 h. In this study, blots were rehybridized with a (1-actin cDNA probe at 42 C to assess the amount and integrity of total RNA loaded in each lane.

PAGE 162

145 Dot Blot Hybridization Total RNA (10 pg) was denatured in denaturation buffer (20 mM Tris, pH 7.0, 50% formamide, 6% formaldehyde), followed by incubation at 65 C for 5 min. The denatured samples were immobilized on membrane (Biotrans, ICN) using a microsample filtration unit (Schleicher and Schuell). The filters were exposed to ultraviolet light for 90 sec to crosslink the RNA to the membrane, then prehybridized in buffer (5X Denhardt, 4X SSC, 0.5 M sodium phosphate pH 6.5, 0.1% SDS, yeast RNA (0,25 mg/ml), 50% (v/v) formamide) at 42 C for 2 h, followed by hybridization in buffer (IX Denhardt, 4X SSC, 0.5 M sodium phosphate, 0.1% SDS, yeast RNA (0.25 mg/ml), 50% (v/v) formamide) containing the ^^P-Mn SOD cDNA (106 cpm/ml of hybridization buffer) labeled by nick translation (Amersham International pic) at 42 C for 16 h. The blots were washed twice with 2X SSC, 0.1% SDS at 42 C for 15 min each on a shaker, followed by two washes with 0.1X SSC, 0.1% SDS at 42 C for 15 min each. The blots were exposed to x-ray film (Eastman Kodak) and/or phosphorimaging cassettes, and hybridization signals were detected and quantified on a phosphorimager (ImageQuant and Phosphorimager, Molecular Dynamics Inc.). The relative expression of each mRNA to that of (3-actin was used to calculate least square means.

PAGE 163

146 Statistical Analysis Data from dot blot analyses were analysed by least squares analysis of variance using the General Linear Models procedure of the Statistical Analytical System (SAS, 1988). The statistical model had day and reproductive status as the main effects with residuals as the error term. Values in the text are least square means (LSM) standard error of the mean (SEM). Orthogonal contrasts were used to determine differences among days of the cycle, and stages of pregnancy. ^^ Results Northern Blot Analysis Northern blot analysis revealed the presence of three mRNA transcripts of approximately 1.5, 1.9, and 3.7 kb in size in bovine luteal tissue. The 1.5 kbtranscript was the most predominant (Fig. 6-1 ). Dot Blot Analysis Dot blot analysis showed that levels expression of Mn SOD mRNA varied across the estrous cycle (P < 0.06), and orthogonal contrasts indicated that levels were higher during the late compared to the early luteal phase of the estrous cycle (P < 0.01) (Fig. 6-2).

PAGE 164

147 CYCLE PR wcMr^^i999o r::;^ 28S18S— "37 4-1-5 MnSOD 4-ACTIN Figure 6-1 Northern blot analysis of luteal Mn SOD mRNA. The same blot was probed for fJ-actin that served as a control for the loading and the integrity of the RNA. The major transcript appeared at 1 .5 kb and two minor transcripts at 1 .9 kb and 3.7 kb.

PAGE 165

148 2-3 16-17 20 17 90-120 170-180 >215 Day of Cycle/Pregnancy Figure 6-2. Dot blot analysis of Mn SOD mRNA. Total RNA (10 pg) isolated from luteal tissue during the estrous cycle (days 2-3, 16-17, and 20) and pregnancy (days 17, 90-120, 170-180, and > 215) was loaded per sample. RNA blots were hybridized with ^^P-labelled Mn SOD cDNA at 42 C. Expression on days 2-3 < days 1 6-1 7 and 20 {P<0.01).

PAGE 166

m.. 149 Although expression of Mn SOD mRNA did not vary overall across pregnancy (P < 0.27), orthogonal contrasts showed that levels were lower during early pregnancy compared to the other days of pregnancy examined (P < 0.07). Discussion Results of this study confirmed synthesis of manganese superoxide dismutase by bovine CL as indicated in an earlier study (Ndikum-Moffor et al., 1995). Bovine luteal tissue expressed three mRNA transcripts of approximately 1.5, 1.9, and 3.7 kb following hybridization with human Mn SOD cDNA. The 1.5 kb-transcript was the most predominant, followed by the 3.7 kb-transcript. Melendez and Baglioni (1993) observed two Mn SOD mRNA transcripts (1.0 kb and 4.0 kb) in human cell lines following TNF-alpha-induced expression; the 4kb mRNA is less stable and has a shorter half-life than the 1-kb mRNA. Three Mn SOD mRNA transcripts of similar sizes as observed in the present study have been reported in luteal tissue of the cow (Rueda et al., 1995) and rat (Sato etal., 1992). Manganese SOD is believed to play a role in ovulation (Laloraya et al., 1988; Sato et al., 1992) and luteal regression (Riley and Behrman, 1991; Sawada and Carlson, 1994; Rueda et al., 1995). Superoxide dismutase has been shown to inhibit vascular permeability of injured tissues (Ando et al., 1990), while hCG has been observed to increase vascular permeability of the ovary. Thus SOD is thought to inhibit hCG-induced ovulation in the rat by dismutation

PAGE 167

:*?. 150 of superoxide radicals, which may be responsible for the increase in vascular permeability, promoting disorganization of cells of follicular walls and ovulation -^ (Satoetal., 1992). In the present study changes in the expression of Mn SOD mRNA in the corpus luteum (post-ovulation) revealed higher levels of expression during the late than early luteal phase of the estrous cycle. Manganese SOD is produced by the mitochondria; thus changes in expression of Mn SOD mRNA may reflect changes in the mitochondrial status of luteal cells. The large luteal cell of the cow contains an abundance of mitochondria which swell and take on dense inclusions as early as day 14 of the estrous cycle (Fields et al., 1992). In this study, lower expression of Mn SOD mRNA was detected on day 2-3 (when no swelling of mitochondria is present), and greatest expression during the late luteal phase when swelling of mitochondria is at its highest (Fields et al., 1992). An increase in expression of Mn SOD mRNA during the involutive phase of the estrous cycle agrees with increased synthesis and release of Mn SOD protein observed at this time (Ndikum-Moffor et al., 1995). This may serve as a defense mechanism by the CL to counteract the first stages of luteal regression caused by reactive oxygen species. This is plausible since by day 20 of the estrous cycle, the CL is completely regressed and levels of Mn SOD mRNA decline (Rueda et al., 1995). Evidence for the possible role of Mn SOD in preventing luteal regression was provided in a study by Rueda et al. (1995) in which regressed bovine CL (day 20 of the estrous cycle) exhibited lower expression of

PAGE 168

151 Mn SOD mRNA as opposed to functional CL (day 20 of pregnancy). Little has been reported about changes in expression of Mn SOD mRNA during pregnancy. Results of this study revealed that expression of Mn SOD mRNA did not vary overall across stages of pregnancy, although levels of expression tended to be lower during early pregnancy compared to the other stages examined. Maintenance of the CL at that time is crucial to maintain pregnancy (Niswender and Nett, 1994). Thus manganese SOD mRNA could be one of many genes expressed during pregnancy to counteract a luteal insult of PGF2alpha that might preempt pregnancy. It is interesting that in the day 14 to day 17 pregnant cow, mitochondria of the large luteal cells do not swell and are devoid of dense inclusions (Fields et al., 1992). Thus changes in Mn SOD gene expression may be indicative of luteal function during the estrous cycle and pregnancy.

PAGE 169

)n ;^CHAPTER 7 GENERAL DISCUSSION AND CONCLUSIONS Adequate luteal function In cattle is important for control of the estrous cycle, the maintenance of pregnancy (Tanabe, 1966; Chew et al., 1979), and normal parturition and expulsion of fetal membranes (Estergreen et al., 1967; Pimentel et al., 1987). Identification of intraluteal factors that may be involved in regulation of luteal function will be useful in the management of cows with suboptimal luteal function and low fertility rates. Histological studies have shown that the CL has the structural machinery required for synthesis and I secretion of proteins (Fields et al., 1985; 1992). Apart from its role in ,f progesterone synthesis, recent studies indicate that the CL synthesizes proteins ":"'i> and peptide factors which may act in an autocrine and /or paracrine manner to regulate luteal and ovarian function (Schams, 1989). Objectives of this dissertation were to investigate de novo protein synthesis by luteal explants in ^ ; culture, determine the pattern of synthesis and release during the estrous cycle and pregnancy, identify and characterize the newly-synthesized proteins, and determine presence and pattern of expression of their respective mRNA during the estrous cycle and pregnancy. "3 152

PAGE 170

153 In experiment 1 the protein synthetic capability of the CL was investigated during the estrous cycle (days 3, 7, 11, 14, 17 and 19) and pregnancy (days 17, 88, 170, and 240 or greater). Luteal explants were incubated with ^H-leucine and the level of incorporation of the radiolabel into proteins was measured by estimating percent incorporation and TCAprecipitable radioactivity (chapter 3). Data from this experiment showed that the bovine CL is capable of incorporating amino acids for de novo protein synthesis in vitro during the estrous cycle and pregnancy. In general, the synthetic ability of the CL did not vary with reproductive status as percent incorporation was similar between day 1 7 of the estrous cycle and day 1 7 of pregnancy. To determine the pattern of protein synthesis and release, proteins released into luteal-conditioned medium were concentrated by dialysis and lyophilization, and were then separated by lEF followed by 2D-PAGE. Gels were analyzed by fluorography. Analysis of fluorographs showed that the CL synthesized many proteins in culture. For convenience, the major proteins were numbered from 1 to 11. The newly-synthesized proteins had molecular weights ranging from 12 to > 200 kDa. The percent incorporation of radiolabel into protein varied across the estrous cycle and pregnancy. Fewer proteins were synthesized by the young CL of day 3 than by CL of the other days of the cycle examined. As the cycle progressed, the CL seemed to attain its full protein-synthetic ability on day 1 1 since all the major proteins synthesized were present on day 1 1 Although the CL on day 3 synthesized fewer proteins, it was unique in its synthetic profile

PAGE 171

154 since the proteins observed on day 3 (proteins 1-4) were not observed on the other days of the cycle and during pregnancy. Similarly, proteins synthesized on the other days of the cycle and pregnancy were not observed on day 3 (chapter 3). Following identification of newly-synthesized proteins by fluorography, the next phase of experiment 1 was to further characterize the proteins by Nterminal amino acid sequence analysis. Proteins separated by 2D-PAGE were transferred to PVDF membrane and subjected to N-terminal amino acid sequencing. Protein 1 (35 kDa, pi 5.5) was identified as apolipoprotein E (Apo E), protein 8 (30 kDa, pi 8.0) was identified as tissue inhibitor of metalloproteinases-1 (TIMP-1), protein 9 (20 kDa, pi 8.0) as TIMP-2, protein 10 (22 kDa, pl 8.0) as manganese superoxide dismutase (SOD), and protein 11 (27 kDa, pl 6.0) as Apo A-1 Luteal synthesis of TIMPs has been reported in the ewe (Smith et al., 1993; 1994), cow (Freudenstein et al., 1990; Juengel et al., 1994), sow (Tanaka et a!., 1992), rat (Mann et al., 1993) and mouse (Nomura et al., 1989). However, this is the first study to evaluate synthesis of TIMP-1 by the bovine CL throughout gestation. To our knowledge, luteal synthesis of Apo A-1 has not been reported in any species and this is the first report of Apo E synthesis by the bovine CL. Studies in the rat showed that Apo E is produced predominantly by theca cells of the ovarian follicle, and synthesis of Apo E was not observed in the corpus luteum (Polacek et a!., 1992). However, Nicosia et al. (1992) reported presence of Apo E mRNA in rat CL.

PAGE 172

155 In the present study (experiment 1 chapter 3), synthesis of Apo E was observed only on day 3 of the estrous cycle, was not observed on the other days of the cycle examined and during pregnancy. Similarly, expression of Apo E mRNA was detected within CL only on days 2 and 3 of the estrous cycle (experiment 2, chapter 4). No Apo E mRNA expression was detected on the other days of the cycle examined and during pregnancy. Thus luteal synthesis of Apo E protein during the early luteal phase (day 3) was confirmed by the presence of Apo E mRNA at that period of the estrous cycle, and the absence of Apo E during the other periods of the cycle examined and pregnancy was due to absence of its mRNA at those times. A single mRNA transcript (1 .0 kb) in bovine luteal tissue hybridized with the Apo E cDNA probe. The size of the transcript was similar to that expressed in rat ovarian follicles (Polacek et al., 1992). The significance of synthesis of Apo E only during the early luteal phase of the estrous cycle is not known, but may suggest presence of a regulatory (stimulatory) factor at that time which is absent or inactivated during the remaining part of the cycle and during pregnancy. A good candidate would be the gonadotropins and agents that activate adenylate cyclase and protein kinase A, which have been shown to increase production of Apo E (Driscoll et al., 1985) and Apo E mRNA (Polacek et al., 1992) in rat granulosa cells. Driscoll et al. (1985) showed that FSH stimulates secretion of newly-synthesized Apo E by granulosa cells in culture, in a doseand time-dependent manner, and the

PAGE 173

156 effects of FSH were mediated through cAMP. However, as the granulosa cells differentiated in culture, they lost responsiveness to FSH and cAMP and Apo E mRNA was not expressed in luteinized granulosa cells (Driscoll et al., 1985). The preovulatory surge of gonadotropins may stimulate Apo E synthesis observed on days 2 and 3 of the estrous cycle, and ceases during diestrus and pregnancy when plasma levels of gonadotropins are low or negligible. The role of Apo E in the CL is not known. However, Apo E is a constituent of VLDL, LDL, and HDL, and has been shown to mediate binding of lipoproteins to the LDL receptor (Mahley, 1988). Thus luteal Apo E may have an autocrine and /or paracrine role to provide cholesterol for membrane biosynthesis and cell proliferation at a time when the CL is undergoing rapid reorganization and growth following the trauma of ovulation. A similar role of Apo E in nerve regeneration following injury has been suggested (Mahley, 1988). Apo A-1 was not observed on days 3 and 7, but was observed on the other days of the estrous cycle examined and throughout pregnancy (experiment 1). This is a novel finding because to our knowledge, there has been no report on luteal synthesis of Apo A-1 in any species. Apo A-1 synthesized and released by luteal explants in this study shared 100% amino acid sequence identity (31 residues) with bovine Apo A-1 (O'hUigin et al., 1990). Synthesis of Apo A-1 by bovine luteal tissue was confirmed by results from experiment 2 (chapter 4) which showed that bovine luteal tissue expressed Apo A-1 mRNA. Bovine luteal tissue exhibited a single hybridization signal (1 .0 kb) with a human

PAGE 174

157 Apo A-1 cDNA probe. In agreement with the protein data in experiment 1, northern blot analysis showed negligible expression of Apo A-1 mRNA during the early luteal phase (days 2 and 3) of the estrous cycle (experiment 2). However, dot blot analysis indicated expression of Apo A-1 mRNA on days 2 and 3 and levels did not differ when compared with the other days of the estrous cycle examined (experiment 2). Newly-synthesized Apo A-1 released in culture was similar among days 11,14 and 17 of the estrous cycle, but increased on day 19. Thus there appears to be translational regulation of the Apo A-1 mRNA in favor of greater protein synthesis during the late luteal phase of the estrous cycle as observed in experiment 1 Although there was no difference overall in the synthesis and release in culture of Apo A-1 by CL across pregnancy (experiment 1 ), levels tended to be greater on day 170 when compared with the other days of pregnancy examined. Expression of Apo A-1 mRNA varied with stage of pregnancy and was higher during the second than the first half of gestation (chapter 4). Apo A-1 is produced mainly by the liver and small intestine, but is also synthesized by many tissues in humans and several animal species (O'hUigin et al., 1990; Sparrow et al., 1992). However, there have been no reports on luteal synthesis. The present study has shown for the first time that a CL synthesizes Apo A-1 The role of Apo A-1 in the CL is not known at this point. Apo A-1 is the major protein component of HDL. High density lipoprotein is more important in vivo in the cow than LDL because of its higher concentrations in blood. In

PAGE 175

158 addition, it has been shown that HDL is required to maintain bovine luteal cells in culture, and HDL is utilized by luteal cells as a source of cholesterol for steroidogenesis (Pate and Condon, 1989). Thus luteal Apo A-1 may play an autocrine and/or paracrine role in the CL and ovary to provide cholesterol for steroidogenesis, since HDL is known to enhance synthesis of progesterone by bovine luteal cells in culture (Pate and Condon, 1989). There is evidence that a female with a functional CL has higher levels of serum Apo A-1 than an anestrous female (Oikawa and Katoh, 1995). However, the contribution of luteal Apo A-1 to circulating levels is unknown and is under investigation. SorciThomas et al. (1988) suggested that Apo A-1 produced by peripheral tissues may not contribute significantly to the plasma Apo A-1 pool but may play a role (autocrine and /or paracrine) in lipid metabolism within the tissue of secretion. Kushwaha et al. (1990) showed that Apo A-1 content of HDL is greater in baboons treated with a combination of estradiol and progesterone, compared to those treated with estradiol alone. Treatment with progesterone alone had no effect on Apo A-1 synthesis (Kushwaha et al., 1990). Although the effects of progesterone on luteal synthesis of Apo A-1 was not specifically investigated in this study, changes were observed in plasma progesterone across days of the estrous cycle in this study (experiment 1 ), with no synthesis of Apo A-1 on days 3 and 7, elevated synthesis on day 19 (experiment 1), and similar expression of Apo A-1 mRNA (experiment 2) across days of the estrous cycle examined. Thus luteal synthesis of Apo A-1 does not seem to depend on circulating levels of

PAGE 176

159 progesterone. It would be useful to identify the cell types responsible for the synthesis of Apo A-1 as this may give an indication of its role in the corpus luteum. Results from experiment 1 also showed that the bovine CL synthesized and released tissue inhibitor of metalloproteinases-1 and -2 in culture during the estrous cycle and pregnancy. Synthesis and release in culture of TIMP-1 was absent on day 3, and increased on day 19 of the estrous cycle. Luteal synthesis of TIMP-1 and TIMP-2 was confirmed in experiment 3 (chapter 5) by the presence of mRNA for TIMP-1 and TIMP-2 within bovine luteal tissue. Bovine luteal tissue showed a major hybridization signal at 0.9 kb and a weaker signal at 3.0 kb (experiment 3). Expression of TIMP-1 mRNA was highest during the late luteal phase of the estrous cycle and increased with advancing pregnancy (experiment 3). Synthesis of TIMP-1 followed a similar pattern with expression of its mRNA during the estrous cycle. Although no significant differences in synthesis of TIMP-1 were observed during pregnancy, levels tended to be greater during late pregnancy (day 240 or greater), and similarly, expression of TIMP-1 mRNA was significantly greater during late pregnancy (day 215 or greater) when compared with the other days of pregnancy examined. Northern blot analysis also revealed two TIMP-2 mRNA transcripts (1.0 kb and 3.5 kb) following hybridization to ovine TIMP-2 cDNA probe similar to that reported for the ewe (Smith et al., 1995), rat (Santoro et al., 1994), and human (Stetler-Stevenson et al., 1990). Similar to the ewe (Smith et al., 1995) but

PAGE 177

160 unlike in the human, the 1.0 kb-TIMP-2 transcript was predominant in this study. In contrast to the changes observed in this study Smith et al. (1994) observed no differences in expression of TIMP-1 mRNA during the ovine estrous cycle. Another contradiction is Freudenstein et al. (1990) observed high TIMP-1 mRNA expression during the early luteal phase of the bovine estrous cycle, and no expression after day 60 of gestation. We are unable to offer a reason for the discrepancies between this study and that of Freudenstein et al. (1990) conducted with bovine tissue. However, our observations of elevated expression of TIMP-1 mRNA during regression agrees with those reported by Juengel et al. (1994). This is the first study to deschbe the changes in expression of TIMP-1 and TIMP-2 mRNA during pregnancy in the cow. The function(s) of TIMP-1 and TIMP-2 in the CL has not been defined, but it has been suggested that TIMPs may be involved in tissue remodelling during formation of the CL following ovulation (Smith et al., 1994). However, in the present study, TIMP-1 was synthesized by bovine luteal tissue and expression of its mRNA increased as the CL matured and regressed. TIMP-1 may serve in an autocrine and/ or paracrine manner to maintain steroidogenesis (Boujrad et al., 1995) and CL lifespan by preventing proteolytic digestion of the extracellular matrix of the CL. The large increase during luteal regression is consistent with an orderly demise of the CL (Juengel et al., 1994). In experiment 1 (chapter 3) and experiment 4 (chapter 6) data was presented to show that the bovine CL synthesized manganese SOD during the

PAGE 178

•.<•'161 estrous cycle and pregnancy. During the estrous cycle, expression of manganese SOD mRNA was lower during the early luteal phase than at the time of luteal involution (day 16-17) and regression (day 20). During pregnancy, levels of manganese SOD mRNA were lower during early pregnancy compared to the other days of pregnancy examined. There is a paucity of information on the temporal changes in gene expression and activity of manganese SOD in the CL during the reproductive cycles. Superoxide dismutase isoenzymes are scavengers of reactive superoxide anions and thus may protect cells from oxidative damage. Manganese SOD has been implicated as involved in ovulation and luteal regression. Laloraya et al. (1988) showed that activity of total superoxide dismutase varies inversely with levels of superoxide anion in the rat ovary during the estrous cycle. The SOD activity contributed by manganese SOD was not discussed in that study. Sato et al. (1992) showed that ovarian levels of manganese SOD mRNA increase after treatment of rats with hCG, and demonstrated that manganese SOD inhibited hCG-induced ovulation. Results from experiment 1 and 4 of this dissertation showed an increase in manganese SOD mRNA at a time of the estrous cycle (day 16) when the CL is entering a regressive phase, and a decline at time of complete regression (day 20). An increase manganese SOD during the involutive phase of the estrous cycle proceeding regression may serve as a mechanism by the CL to counteract the first stages of luteal regression caused by reactive oxygen species. A fully regressed CL (day 20 of the bovine estrous cycle) has been shown to express

PAGE 179

162 significantly lower levels of manganese SOD mRNA when compared to a functional CL (day 20 of pregnancy) (Rueda et al., 1995). Luteal expression of manganese SOD mRNA during the estrous cycle may indicate a CL undergoing an insult. Expression during pregnancy may reflect a rescue effect. It was observed in this study that synthesis and release in culture of all the proteins discussed, except the TIMP-2/S0D complex, did not differ quantitatively between day 17 of the estrous cycle and day 17 of pregnancy. In order to confirm day of the estrous cycle (observed estrus was day 0) total CL weight and plasma progesterone were monitored (chapter 3). As expected, CL weight varied across days of the estrous cycle being lowest on day 3, Increased to day 14, and declined thereafter. Plasma progesterone followed a similar pattern and levels declined after day 14 to low levels on day 19, indicative of luteal regression. CL weight and plasma progesterone did not vary across pregnancy starting on day 17. The protein synthetic ability of the CL during luteal regression observed on days 17 and 19 was unexpected and suggests that the CL maintained its protein synthetic ability independently of the decline in steroidogenesis. Conclusions In summary, results of the four experiments show that the bovine CL synthesizes many proteins. Eleven proteins were characterized as synthesized and released by the CL across the estrous cycle and pregnancy. Five of these proteins were identified by N-terminal amino acid sequence analysis as Apo E, Apo A-1 TIMP-1 TIMP-2 and Mn SOD. The gene

PAGE 180

163 expression and temporal changes of their respective mRNA across the estrous cycle and pregnancy were also investigated. Two of these proteins, Apo E and Apo A-1 are novel as there have been no reports of their synthesis by the CL. Apolipoprotein E is associated with VLDL and LDL and apolipoprotein A-1 is associated with HDL. HDL and LDL are sources of cholesterol for the cow (Pate and Condon, 1989). Apolipoproteins mediate uptake and binding of lipoproteins to the lipoprotein receptor. Thus they may be involved in regulating cholesterol availability for membrane and progesterone biosynthesis. Two other proteins identified, (TIMP-1 and TIMP-2) have been reported in luteal tissue of several species including the cow during the estrous cycle. However, this is the first report of these proteins being synthesized and secreted by the CL of the pregnant cow. TIMP are produced by many cell types in several tissues where they regulate degradation of the extracellular matrix (Werb, 1989). Although their role in the CL has not been defined, they are believed involved in tissue remodelling during rapid development of the CL post-ovulation, the demise and reorganization of the CL that would allow for ovulation. We also believe that they play a role in maintaining luteal function during pregnancy as TIMP-1 has been shown to stimulate steroid hormone synthesis in the testis (Boujradetal., 1995). Suggested roles of these proteins in luteal function are speculative at this point. Determination of factors that regulate luteal production of these proteins and determination of their functions will be critical to better reproductive

PAGE 181

f K 164 management of the cow and related species. For example, the significance of synthesis of Apo E only during early luteal development (chapter 3 and 4) needs to be investigated; factors regulating Apo E gene expression that are present only during the small physiological window need to be identified. Due to the heterogenous nature of the cellular components of the CL, identification of the cell types responsible for production of these proteins would provide more insight into what their functions might be. Finally, a better understanding of the regulatory role of these proteins on luteal development and function may serve as a model to study factors that regulate tissue growth under normal and pathological conditions in humans. ROLE OF LUTEAL PROTEINS O ^^"'^"" (C^ k Apo E Follicle LH FSH E2 Young CL LH E2 /i„^K.=r,<. Cw^fh^oic Membrane Synthesis Mature CL Cholesterol TIMP 4— 4JI— ^ SOD^ a Regressing CL Tissue Remodelling Progesterone

PAGE 182

REFERENCES Alila HW, Corradino RA, Hansel W 1989 Differential effects of luteinizing hormone on intracellular free Ca^"" in small and large bovine luteal cells. Endocrinology 124:2314-2320 Alila HW, Hansel W 1984 Origin of different cell types in the bovine corpus luteum as characterized by specific monoclonal antibodies. Biol Reprod 31:1015-1025 Amoroso EC 1968 Reproductive Physiology: from a distinguished past to a promising future. J Anim Sci 27 (SuppI 1):214 Amselgruber W, Sinowatz F, Schams D, Skottner A 1994 Immunohistochemical aspects of insulin-like growth factors I and II in the bovine corpus luteum. J Reprod Fertil 101:445-451 Anderson E, Little B 1984 The ontogeny of the rat granulosa cell. In: D.O. Toft and R.J. Ryan (Ed.). Proceedings of the Fifth Ovarian Workshop, pp 203. Ovarian Workshop, Champaign, Illinois Anderson LL 1982 Relaxin localization in porcine and bovine ovaries by assay and morphological techniques. Adv Exp Med Biol 143:1-67 Anderson LL, Bast JD, Melampy RM 1973 Relaxin in ovarian tissue during different reproductive stages in the rat. J Endocrinol 59:371-372 Ando Y, Inoue M, Morino Y 1990 Inhibition of carrageenin-induced paw edema by a superoxide dismutase derivative that circulates bound to albumin. Biochem Biophys Acta 1073:374-379 Archer TK, Tam SP, Deeley RG 1986 Kinetics of estrogen-dependent modulation of apolipoprotein A-1 synthesis in human hepatoma cells. J Biol Chem 261:5067-5074 Bagnell CA, Ayan E, Downey BR, Tsang BK, Ainsworth L 1989 Localization of relaxin during formation of the porcine corpus luteum. Biol Reprod 40:835-841 165

PAGE 183

166 Baird DT 1992 Luteotrophic control of the corpus luteum. Anim Reprod Sci 28:95-102 Basha SMM, Bazer FW, Roberts RM 1 980 Effect of the conceptus on quantitative and qualitative aspects of uterine secretion in pigs. J Reprod Ferti I 60:41-48 Batra SK, Pandey RS 1983 Relative concentration of 13, 14-dihydro-15-ketoprostaglandin Fja in blood and milk of buffaloes during the estrous cycle and early pregnancy. J Reprod Fertil 67:191-195 Bazer FW, Simmen RCM, Simmen FA 1991b Comparative aspects of conceptus signals for maternal recognition of pregnancy. Anna! NY Acad Sci 622:202-211 Bazer FW, Thatcher WW, Hansen PJ, Mirando MA, Ott TL, Plante C 1 991 a Physiological mechanisms of pregnancy recognition in ruminants. J Reprod Fertil SuppI 43:39-47 Bazer FW, Vallet JL, Roberts RM, Sharp DC, Thatcher WW 1 986 Role of conceptus secretory products in establishment of pregnancy. J Reprod Fertil 76:841-850 Behrman H, Grinwich DH, Hichens M, MacDonald GJ 1978 Effect of A, hypophysectomy, prolactin and prostaglandin Fza on gonadotrophin binding in vivo and in vitro in the corpus luteum. Endocrinology 103:349357 ^^^ Benyo DF, Haibel GF, Laufman HB, Pate JL 1991 Expression of major histocompatibility complex antigens on the bovine corpus luteum during the estrous cycle, luteolysis and early pregnancy. Biol Reprod 45:229-234 Bettencourt CM, Moffat RJ, Keisler DH 1993 Active immunization of ewes against prostaglandin F2 alpha to control ovarian function. J Reprod Fertil 97:123-131 Betteridge KJ, Randal GCB, Eaglesome MD, Sugden EA 1984 The influence of pregnancy on PGFja secretion in cattle. Concentrations of 15-keto-13,14. dihydroprostaglandin Fjo, and progesterone in peripheral blood of recipients of transferred embryos. Anim Reprod Sci 7:195-216 Birnboim HC, Doly J 1979 A rapid alkaline extraction procedure for screening recombinant plasmid DNA. NucI Acids Res 9:1513-1522

PAGE 184

167 Blobel G, Dobberstein B 1975 Transfer of proteins across membranes. I. Presence of proteolytically processed and unprocessed nascent immunoglobulin light chains on membrane-bound ribosomes of murine myeloma. J Cell Biol 67:835-851 Blue ML, Ostapchuk P, Gordon JS, Williams DL 1982 Synthesis of apolipoprotein A1 by peripheral tissues of the rooster: a possible mechanism of cellular cholesterol efflux. J Biol Chem 257: 1 1 1 51 -1 1 1 59 Boujrad N, Ogwuegbu SO, Gamier M, Lee CH, Martin BM, Papadopoulos V 1995 Identification of a stimulator of steroid hormone synthesis isolated from testis. Science 268:1609-1612 Bowler C, Slooten L, Vandenbranden S, Rycke R de, Botterman J 1991 Manganese superoxide dismutase can reduce cellular damage induced by oxygen radicals in transgenic plants. EMBO J 10:1723-1732 Brannstrom M, Norman RJ 1993 Involvement of leukocytes and cytokines in the ovulatory process and corpus luteum function. Human Reprod 8:17621775 Brenner CA, Adier RR, Rappolee DA, Pedersen RA, Werb Z 1989 Genes for extracellular matrix-degrading metalloproteinases and their inhibitors, TIMP, are expressed during early mammalian development. Genes Dev 3:848-859 Breslow JL, McPherson J, Nussbaum AL, Williams HW, Lofquist-Kahl F, Karathanasis SK and Zannis VL 1982 Identification and DNA sequence of a human apolipoprotein E cDNA clone. J Biol Chem 257:14639-14641 Bruce NW, Meyer GT, Dharmarajan AM 1 984 Rate of blood flov\/ and growth of the corpora lutea of pregnancy and of previous cycles throughout pregnancy in the rat. J Reprod Fertil 71:445-452 Budnik LT, Mukhopadhyay AK 1991 Epidermal growth factor, a modulator of luteal adenylate cyclase. J Biol Chem 266:13908-13913 Carlson JC, Wu XM, Sawada M 1993 Oxygen radicals and the control of ovarian corpus luteum function. Free Radical Biol Med 14:79-84 Carr BR, MacDonald PC, Simpson ER 1982 The role of lipoproteins in the regulation of progesterone secretion by the human corpus luteum. Fertil Steril 38:303-311

PAGE 185

168 Carter KC, Bowman D, Carrington W, Fogarty K, McNeil JA, Fay FS Lawrence JB 1993 A three-dimensional view of precursor messenger RNA metabolism within the mammalian nucleus. Science 259: 1 330-1 334 Chan J, Nakabayashi H, Wong HC 1993 HNF-4 increases activity of the rat Apo A-1 gene. NucI Acids Res 21:1205-1211 Chegini N, Lei ZM, Rao CV 1991 Nuclear volume and chromatin conformation of small and large luteal cells: effects of gonadotrophins and prostaglandins and dependence on luteal phase. Cell Tissue Res 264:453-460 Chegini N, Ramani N, Rao CV 1984 Morphological and biochemical characterization of small and large bovine luteal cells during pregnancy. Mo! Cell Endocrinol 37:89-102 Chew BP, Erb RE, Fessler JF, Callahand CJ, Malven PV 1979 Effects of ovariectomy during pregnancy and of prematurely induced parturition on progesterone, estrogens, and calving traits. J Dairy Sci 62:557-566 Church SL 1990 Manganese superoxide dismutase: nucleotide and deduced amino acid sequence of a cDNA encoding a new human transcript. Biochim Biophys Acta 1087:250-252 Clark SD, Kobayashi DK, Welgus HG 1987 Regulation of the expression of tissue inhibitor of metalloproteinases and collagenase by retinoids and glucocorticoids in human fibroblasts. J Clin Invest 80:1280-1288 Copelin JP, Smith MF, Keisler DH, Garverick HA 1989 Effect of active immunization of pre-partum and post-partum cows against prostaglandin F2 alpha on lifespan and progesterone secretion of short-lived corpora lutea. J Reprod Fertil 87:199-207 Corcoran ML, Stevenson WG 1995 Tissue inhibitor of metalloproteinases-2 stimulates fibroblast proliferation via a cAMP-dependent mechanism. J Biol Chem 270:13453-13459 Corner GW 1919 On the origin of the corpus luteum of the sow from both granulosa and theca interna. Am J Anat 26:117-183 Craig WY, Nutik R, Cooper AD 1988 Regulation of apoprotein synthesis and secretion in the human hepatoma Hep G2. The effect of exogenous lipoprotein. J Biol Chem 263:13880-13890

PAGE 186


PAGE 187

170 Driscoll DM, Getz GS 1984 Extrahepatic synthesis of apolipoprotein E. J Lipid Res 25:1368-1379 Driscoll DM, Schreiber JR, Schmit VM, Getz GS 1985 Regulation of apolipoprotein E synthesis in rat ovarian granulosa cells. J Biol Chem 260:9031-9038 Dubois DH, Simmen RC, Fliss A, Smith LC, Bazer FW 1993 Expression of c-fos in porcine endometrium during the estrous cycle and early pregnancy. Biol Reprod 49:943-950 Dyer C, Curtiss L 1988 Apolipoprotein E-rich HDLs inhibit ovarian androgen synthesis. J Biol Chem 263:10965-10973 Edelstein C, Kaiser M, Piras G, Scanu AM 1988 Demonstration that the enzyme that converts precursor of apolipoprotein A-1 to apolipoprotein A-1 is secreted by the hepatocarcinoma cell line Hep G2. Arch Bichem Biophys 267:23-30 Edwards DR, Heath JK, Hogan BL, Nomura S, Wills AJ 1992 Expression of TIMP in fetal and adult mouse tissues studied by in situ hybridization. Matrix SuppI 1 :286-293 Edwards DR, Rocheleau H, Sharma RR, Wills AJ, Cowie A, Hassell JA, Heath JK 1992 Involvement of AP-1 and PEA3 binding sites in the regulation of murine tissue inhibitor of metalloproteinases-1 (TIMP-1 ) transcription. Bichim Biophys Acta 1171:41 -55 Einspanier R, Miyamoto A, Schams D, Muller M, Brem G 1990 Tissue concentration, mRNA expression and stimulation of IGF-I in luteal tissue during the oestrous cycle and pregnancy of cows. J Reprod Fertil 90:439445 Estergreen VL, Froste OL, Gomes WR, Erb RE, Bullard JF 1967 Effects of ovariectomy on pregnancy maintenance and parturition in dairy cows. J Dairy Sci 50:1293-1295 Fairchild DL, Pate JL 1989 Interferon-gamma induction of major histocompatibility complex antigens on cultured bovine luteal cells. Biol Reprod 40:453-457

PAGE 188

171 Fairclough RJ, Smith JF, McGowen LT 1981 Prolongation of the estrous cycle in cows and ewes after passive immunization with PGF antibodies. J Reprod Fertil 62:213-219 Farin CE, Nett TM, Niswender GD 1990 Effects of luteinizing hormone on luteal cell populations in hypophysectomized ewes. J Reprod Fertil 88:61-70 Farin CE, Sawyer HR, Niswender GD 1989 Analysis of cell types in the corpus luteum of the sheep. J Reprod Fertil SuppI 37:181-187 Ferrari S, Drusiani E, Calandra S, Tarugi P 1986 Isolation of a cDNA cone for chick intestinal apolipoprotein A1 (Apo-AI) and its use for detecting apo A1 mRNA in several chick tissues. Gene 42:209-214 Fields MJ, Dubois W, Fields PA 1985 Dynamic features of luteal secretory granules: ultrastructural changes during the course of pregnancy in the cow. Endocrinology 117:1675-1682 Fields MJ, Barros CM, Watkins WB, Fields PA 1992 Characterization of large luteal cells and their secretory granules during the estrous cycle of the cow. Biol Reprod 46:535-545 Fields MJ, Fields PA 1986 Luteal neurophysin in nonpregnant cow and ewe: immunocytochemical localization in membrane-bounded secretory granules of the large luteal cell. Endocrinology 118:1723-1725 Fields PA, Eldridge RK, Fuchs AR, Roberts RF, Fields MJ 1983 Human placental and bovine corpora luteal oxytocin. Endocrinology 112:15441546 Fincher KB, Bazer FW, Hansen PJ, Thatcher WW, Roberts RM 1986 Proteins secreted by the sheep conceptus suppress induction of uterine prostaglandin Fja release by oestradiol and oxytocin. J Reprod Fertil 76:425-433 Findlay JK 1986 Angiogenesis in reproductive tissues. J Endocrinol 111:357-366 FitzTA, Mock EJ, Mayan MH, Niswender GD 1984 Interactions of prostaglandins with subpopulations of ovine luteal cells. II. Inhibitory effects by PGFzc and protection by PGE2. Prostaglandins 28:127-138

PAGE 189

172 FitzTA, Mayan MH, Sawyer HR, Niswender GD 1982 Characterization of two steroidogenic cell types in ovine corpus luteum. Biol Reprod 27:703-71 1 Fletcher PW, Niswender GD 1982 Effects of PGFza on progesterone secretion and adenylate cyclase activity in ovine luteal tissue. Prostaglandins 20:803-818 Flint AP, Stewart HJ, Lamming GE, Payne JH 1992 Role of oxytocin receptor in the choice between cyclicity and gestation in ruminants. J Reprod Fertil SuppI 45:53-58 Freudenstein J, Wagner S, Luck RM, Einspanier R, Scheit KH 1990 mRNA of bovine tissue inhibitor of metalloproteinase: sequence and expression in bovine ovarian tissue. Biochem Biophys Res Comm 171:250-256 Fridovich I 1974 Superoxide dismutase. Adv Enzymol 41:38-97 Garrett JE, Geisert RD, Zavy MT, Morgan GL 1988 Evidence for maternal regulation of early conceptus growth and development in beef cattle. J Reprod Fertil 84:437-446 Garverick HA, Moser MT, Keisler DH, Hamilton SA, Roberts RM, Smith MF 1992 Luteal function after intrauterine infusion of recombinant bovine interferon-Qil into postpartum beef cows expected to have short or normal luteal phases. J Reprod Fertil 94:319-325 Gatzuli E, Aten RF, Behrman HR 1991 Inhibition of gonadotropin action and progesterone synthesis by xanthine oxidase in rat luteal cells. Endocrinology 128:2253-2258 Gemmell RT, Stacy BD, Thoburn GD 1974 Ultrastructural study of secretory granules in the corpus luteum of the sheep during the estrous cycle. Biol Reprod 1 1 :447-462 Ginther OJ 1974 Internal regulation of physiological processes through local venoarterial pathways: a review. J Anim Sci 39:550-564 Girsh E, Greber Y, Meidan R 1995 Luteotrophic and luteolytic interactions between bovine small and large luteal-like cells and endothelial cells. Biol Reprod 52:954-962

PAGE 190

173 Go MF, Schonfeld G, Pfleger B, Cole TG, Sussman NL, Alpers DH 1988 Regulation of intestinal and hepatic apoprotein synthesis after chronic fat and cholesterol feeding. J Clin Invest 81:1615-1620 Godkin JD, Bazer FW, Moffat J, Sessions F, Roberts RM 1982 Purification and properties of a major low molecular weight protein released by the trophoblast of sheep blastocysts at day 13-21. J Reprod Fertil 65:141-150 Godkin JD, Bazer FW, Thatcher WW, Roberts RM 1984 Proteins released by cultured day 15-16 conceptuses prolong luteal maintenance when introduced into the uterine of cyclic ewes, J Reprod Fertil 71 :57-64 Gordon JJ, Sims HF, Lentz SR, Edelstein C, Scanu AM 1983 Proteolytic processing of of human preproapolipoprotein A-1 : a proposed defect in the conversion of apo A-1 to A-1 in Tangier's disease. J Biol Chem 258:4037-4044 Gospodarowicz D, Cheng J, Lui GM, Baird A, Esch F, Bohlen P 1985 Corpus luteum angiogenic factor is related to fibroblast growth factor. Endocrinology 117:2383-2391 Graham CH, Connelly I, Mac Dougall JR, Kerbel RS, Stetler-Stevenson WG, Lala PK 1994 Resistance of malignant trophoblast cells to both the antiproliferative and anti-invasive effects of transforming growth factor-beta. Expt Cell Res 214:93-99 Grazul-Bilska AT, Redmer DA, Killilea SD, Kraft KC, Reynolds LP 1992 Production of mitogenic factor(s) by ovine corpora lutea throughout the estrous cycle. Endocrinology 130:3625-3632 Grusenmeyer DF, Pate JL 1992. Localization of prostaglandin F2 alpha inhibition of lipoprotein use by bovine luteal cells. J Reprod Fertil 94:31 1318 Gu ZW, Xie YH, Yang M, Sparrow JT, Wang K, Li Y, Li WH, Gotto Jr AM, Yang CY 1993 Primary structure of Beijing duck apolipoprotein A-1. J Prot Chem 12:585-591 Guidot DM, McCord JM, Wright RM, Repine JE 1 993 Absence of electron transport (Rho (o)state restores growth of a manganese superoxide dismutase-deficient Saccharomyces cerevisiae in hypoxia: evidence for electron transport as a major source of superoxide generation in vivo. J Biol Chem 268:26082-26084

PAGE 191

r)' 174 Guillomot M, Michel C, Gaye P, Charlier N, Trojan J, Martal J 1990 Cellular localization of an embryonic interferon, ovine trophoblastin and its mRNA in sheep embryos during early pregnancy. Biol Cell 68:205-21 1 Guraya SS 1971 Morphology, histochemistry, and biochemistry of human ovarian compartments and steroid hormone synthesis. Physiol Rev 51:785-807 Guttler HW, Sommerfeldt M, Papandrikopoulou A, Mischek U, Bonitz D, Frey A, Grupe M, Scheerer J, Gassen HG 1990 Synthesis of apolipoprotein A-1 in pig brain microvascular endothelial cells. J Neurochem 54:444-450 Handwerger S, Quarfordt S, Barrett J, Harman I 1987 Apolipoproteins A-1, A-11, and C-1 stimulate placental lactogen release from human placental tissue. A novel action of high density lipoprotein apolipoproteins J Clin Invest 79:625-628 Hansel W, Alila HW, Dowd JP, Yang XZ 1987 Control of steroidogenesis in small and large bovine luteal cells. Aust J Biol Sci 40:331-347 Hansel W, Concannon PW, Lukav^zewska JH 1973 Corpora lutea of large domestic animals. Biol Reprod 8:222-245 Hansen TR, Kazemi M, Keisler DH, Malathy PV, Imakawa K, Roberts RM 1989 Complex binding of the embryonic interferon, ovine trophoblast protein-1, to endometrial receptors. J Interferon Res 9:215-225 Hayakawa T, Yamashita K, Tanzawa K, Uchijima E, Iwata K 1992 Growthpromoting activity of tissue inhibitor of metalloproteinases-1 (TIMP-1) for a wide range of cells: a possible growth factor in serum. FEBS 298:29-32 Hedin L, Rodgers RJ, Simpson ER, Richards JS 1987 Changes in content of cytochrome P450i7„, cytochrome P450scc. and 3-hydroxy-3-methylglutaryl CoA reductase in developing rat ovarian follicles and corpora lutea: correlation with theca cell steroidogenesis. Biol Reprod 37:21 1-223 Hesia JS, Miyazaki T, Dasko LM, Wallach EE, Dharmarajan AM 1992 Superoxide dismutase activity, lipid peroxide production and corpus luteum steroidogenesis during natural luteolysis and regression induced by oestradiol deprivation of the ovary in pseudopregnant rabbits. J Reprod Fertil 95:915-924

PAGE 192

175 Hopkins B, Sharpe CR, Baralle FE, Graham CF 1986 Organ distribution of apolipoprotein gene transcripts in 6-12 week postfertiiization human embryos. J Embryo! exp Morph 97:177-187 Huang J, Powell M, Mead RA 1993 Luteal protein secretion during preimplantation in the ferret. Biol Reprod 48:647-654 Imada K, Ito A, Itoh Y, Nagase H, Mori Y 1994 Progesterone increases the production of tissue inhibitor of metalloproteinase-2 in rabbit uterine cervical fibroblasts. FEBS Lett 341 : 1 09-1 1 2 Imakawa K, Hansen TR, Malathy PV, Anthony RV, Polites HG, Marotti KR, Roberts RM 1989 Molecular cloning and characterization of complementary deoxyribonucleic acids corresponding to bovine trophoblast protein-1 : a comparison with ovine trophoblast protein-1 and bovine interferon-a,,. Mol Endocrinol 3:127-139 Ivell R, Brackett KH, Fields MJ. Richter D 1985 Ovulation triggers oxytocin gene expression in the bovine ovary. FEBS Lett 190:263-267 Jakob W, Jentzch KD, Mauerberger B, Oehme P 1977 Demonstration of angiogenesis activity in the corpus luteum of cattle. Exp Pathol 13:231236 Jenkin G 1992a Interaction between oxytocin and prostaglandin F2 alpha during luteal regression and early pregnancy in sheep. Reprod Fertil Dev 4:321328 Jenkin G 1992b Oxytocin and prostaglandin interactions in pregnancy and at parturition. J Reprod Fertil SuppI 45:97-1 1 1 Jocylyn HD, Setchell BP 1972 Regnier de Graaf on the human reproductive organs. J Reprod Fertil SuppI 17:85-209 Juengel JL, Smith GW, Smith MF, Youngquist RS, Garverick HA 1994 Pattern of protein production by bovine corpora lutea during luteolysis and characterization of expression of two major secretory products of regressing corpora lutea. J Reprod Fertil 100:515-520 Karsch FJ, Bittman EL, Foster DL, Goodman RL, Legan SJ, Robinson JE 1984 Neuroendocrine basis of seasonal reproduction. Recent Prog Hormone Res 40:185-232

PAGE 193

176 Keyes PL, Wiltbank MC 1988 Endocrine regulation of the corpus luteum. Ann Rev Physiol 50:465-482 Kindahl H, Edquist LE, Bane A, Granston E 1976 Blood levels of progesterone and 15-keto-13,14-dihydro-prostaglandin Fjc during the normal oestrous cycle and early pregnancy in heifers. Acta Endocrinol 82:134-149 Knickerbocker JJ, Wiltbank MC, Niswender GD 1988 Mechanisms of luteolysis in domestic livestock. Domestic Anim Endocrinol 5:91-107 Koering MJ, Thor MJ 1978 Structural changes in the regressing corpus luteum of the rabbit. Biol Reprod 17:719-733 Koos RD 1989 Potential relevance of angiogenic factors to ovarian physiology. Sem Reprod Endocrinol 7:29-40 Koos RD, Hansel W 1981 The large and small cells of the bovine corpus luteum: ultrastructural and functional differences. In: Dynamics of Ovarian Function (N.B. Schv\/artz and M. Hunzicker-Dunn, eds.). Raven Press, New York, pp 197-203 Kushwaha RS, Foster DM, Murthy VN et al. 1990 Metabolic regulation of apoproteins of high density lipoproteins by estrogen and progesterone in the baboon (Papio sp). Metabolism 39:544-552 Laemmli UK 1970 Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685 LaFrance M, Goff AK 1985 Effect of pregnancy on oxytocin-induced release of prostaglandin F2a in heifers. Biol Reprod 33: 1 1 1 3-1 1 1 9 Lala PK, Graham CH 1990 Mechanisms of trophoblast invasiveness and their control: the role of proteases and protease inhibitors. Cancer Metast Rev 9:369-379 Lalaloya M, Pradeep KG, Laloraya MM 1988 Changes in the levels of superoxide anion radical and superoxide dismutase during the estrous cycle of Rattus norvegicus and induction of superoxide dismutase in rat ovary by luteotropin. Biochem Biophys Res Commun 157:146-153 Leco KJ, Hayden LJ, Sharma RR, Rocheleau H, Greenberg AH, Edwards DR 1992 Differential regulation of TIMP-1 and TIMP-2 mRNA expression in normal ras-transformed murine fibroblasts. Gene 117:209-217

PAGE 194

177 Lee W, Mitchel P, Tjian R 1987 Purified transcription factor AP-1 interacts with TPA-inducible enhancer elements. Cell 49:741 -752 Lei ZM, Chegini N, Rao CV 1991 Quantitative cell composition of human and bovine corpora lutea from various reproductive states, Biol Reprod 44:1148-1156 Le Maire WJ 1989 Mechanism of mammalian ovulation. Steroids 54:455-4669 Lemon M, Mauleon P 1982 Interaction between two luteal cell types from the corpus luteum of the sow in progesterone synthesis in vitro. J Reprod Ferti I 64:315-323 Lewis GS, Thatcher WW, Bazer FW, Curl JS 1982 Metabolism of arachidonic acid in vitro by bovine blastocysts and endometrium. Biol Reprod 27:431439 Lobel BL, Levy E 1968 Enzymic correlates of developing secretory function and regression of follicles and corpora lutea in the bovine ovary. Acta Endocrinol SuppI 132:7-63 ^ Loeb L 1906 The formation of the corpus luteum of the guinea pig. J Am Med Assoc 46:416 Maclntyre DE, Pearson JD, Gordon JL 1978 Localization and stimulation of prostacyclin production in vascular cells. Nature 271:549-551 Mackay AR, Ballin M, Pelina MD, Farina AR, Nason AM, Hartzler JL, Thorgeirsson UP 1992 Effect of phorbol ester and cytokines on matrix metalloproteinase and tissue inhibitor of metalloproteinase expression in tumor and normal mouse cell lines. Invasion Metastasis 12:168-184 Mahley R 1988 Apolipoprotein E: cholesterol transport protein with expanding role in molecular biology. Science 240:622-629 Mallory JB, Kushner PJ, Protter AA, Cofer CL, Appleby VL, Lau K, Schilling JW, VIgne JL 1987 Expression and characterization of human apolipoprotein A-1 in Chinese hamster ovary cells. J Biol Chem 262:4241-4247 Mann JS, Kindy MS, Edwards DR, Curry Jr. TE 1991 Hormonal regulation of matrix metalloproteinase inhibitors in rat granulosa cells and ovaries. Endocrinology 1 28: 1 825-1 832

PAGE 195

178 Mann JS, Kindy MS, Hyde JF, Curry Jr. TE 1993 Role of protein synthesis, prostaglandins, and estrogen in rat ovarian metal loproteinase inhibitor production. Biol Reprod 48:1006-1113 Mans RJ, Novelli GD 1961 Measurement of the incorporation of radioactive amino acids into protein by a filter-paper disc method. Arch Biochem 94:48-53 Mayerhofer A, Lahr G, Gratzl M 1991 Expression of the neural cell adhesion molecule in endocrine cells of the ovary. Endocrinology 129:792-800 Mayerhofer A, Spanel-Borowski K, Watkins S, Gratzl M 1992 Cultured microvascular endothelial cells derived from bovine corpus luteum possess NCAM-1 40. Exp Cell Res 201 :545-548 Mazzone T, Basheeruddin K, Poulos C 1987 Regulation of macrophage apolipoprotein E gene expression by cholesterol. J Lipid Res30:10551064 McClellan MC, Diekman MA, Abel JH, Niswender GD 1975 Luteinizing hormone, progesterone, and the morphological development of normal and superovulated corpora lutea in sheep. Cell Tissue Res 164:291-307 McClellan MC, Abel Jr JH, Niswender GD 1977 Function of lysosomes during luteal regression in normal cycling and PGFja treated ewes. Biol Reprod 10:499-512 McCracken JA, Carlson JC, Glew ME, Goding JR, Baird DT, Green K, Samuelsson B 1972 Prostaglandin Fja identified as a luteolytic hormone in sheep. Nature 238:129-134 Meidan RO, Girsh E, Blum 0, Aberdam E 1990 In vitro differentiation of bovine theca and granulosa cells into small and large luteal-like cells: morphological and functional characteristics. Biol Reprod 43:913-921 Melendez JA, Baglioni C 1993 Differential induction and decay of manganese superoxide dismutase mRNAs. Free Radical Biol Med 14:601-608 Merino A, Buckbinder L, Mermelstein FH, Reinberg D 1989 Phosphorylation of cellular proteins regulates their binding to the cAMP response element. J Biol Chem 264:21266-21276

PAGE 196

179 Milvae RA, Hansel W 1980 The effects of prostacyclin (PGI2) and 6-keto-PGFi„ on bovine plasma P4 and LH concentrations. Prostaglandins 20:641-647 ;; Mirando MA, Harney JP, Zhou Y, Ogle TF, Ott TF, Moffatt RJ, Bazer FW 1993 Changes in progesterone and oestrogen receptor mRNA and protein and oxytocin receptors in endometrium of ewes after intrauterine injection of trophoblast interferon. J Mol Endocrinol 10:185-192 Miyamoto A, Okuda K, Schweigert FJ, Schams D 1992 Effects of basic fibroblast growth factors, transforming growth factor-li and nerve growth factor on the secretory function of the bovine corpus luteum in vitro. H Endocrinol 135:103-114 Murphy G, Reynolds J J, Werb Z 1985 Biosynthesis of tissue inhibitor of metalloproteinases by human fibroblasts in culture: stimulation by 12-0tetradecanoylphorbol 13-acetate and interleukin-1 in parallel with collagenase. J Biol Chem 260:3079-3083 Musicki B, Aten RF, Behrman HR 1994 Inhibition of protein synthesis and hormone-sensitive steroidogenesis in response to hydrogen peroxide in rat luteal cells. Endocrinology 1 34:588-595 Nara BS, Welk FA, Rutherford JE, Sherwood OD, First NL 1982 Effect of relaxin on parturition and frequency of live births in pigs. J Reprod Fertil 66:359365 Ndikum-Moffor FM, Rollyson MK, Chang S-M, Buhi W, Fields MJ 1995 Proteins secreted and released in culture by the bovine corpus luteum: the estrous cycle and pregnancy. Biol Reprod (SuppI 1) 52:215 (Abstract) Nephew KP, McClure KE, Ott TL, Dubois DH, Bazer FW, Pope WF 1991 Relationship between variation in conceptus development and differences in estrous cycle duration in ewes. Biol Reprod 44:536-539 Neill JD, Johansson EDB, Knobil E 1969 Failure of hysterectomy to influence the normal pattern of cyclic progesterone secretion in the rhesus monkey. Endocrinology 84:464-465 Nett TM, McClellan MC, Niswender GD 1976 Effects of prostaglandins on the ovarian corpus luteum: blood flow, secretion of progesterone and morphology. Biol Reprod 15:66-78

PAGE 197

180 Nicosia M, Moger WH, Dyer CA, Prack MM, Williams DL 1992 Apolipoprotein E messenger RNA in rat ovary is expressed in theca and interstitial cells and presumptive macrophage, but not in granulosa cells. Mol Endocrinol 6:978-988 Niswender GD, Juengel JL, Mcguire WJ, Belfiore, CJ, Wiltbank MC 1994 Luteal Function: the estrous cycle and early pregnancy. Biol Reprod 50:239-247 Niswender GD, Nett TM 1994 Corpus luteum and its control in infraprimate species. In: Knobil, E., Neill, J. (Eds.), The Physiology of Reproduction. Raven Press Ltd, New York, pp 781-816 Niswender GD, Schwall RH, Fitz TA, Farin CE, Sawyer HR 1985 Regulation of luteal function in domestic ruminants: new concepts. Recent Prog Horm Res 41:101-152 Nomura S, Hogan BLM, Wills AJ, Heath JK, Edwards DR 1989 Developmental expression of tissue inhibitor of metalloproteinase (TIMP) RNA. Development 105:575-583 O'hUigin C, Chan L, Li WH 1990 Cloning and sequencing of bovine apolipoprotein A-1 cDNA and molecular evolution of apolipoproteins A-1 and B-100. Mol Biol Evol 7:327-339 O'Farrell, PH 1975 High-resolution two-dimensional electrophoresis of proteins. J Biol Chem 250:4007-4021 Oikawa S, Katoh N 1995 Enzyme-linked immunosorbent assay for apolipoprotein A-1 in the serum of cattle. Am J Vet Res 56:409-414 O'Shaughnessy PJ 1989. The control of the development of corpus luteum and its maintenance in cattle. Tierarztl Prax SuppI 4:7-1 1 O'Shea JD, Cran DG, Hay MF 1980 Fate of the theca interna following ovulation in the ewe. Cell Tissue Res 210:305-319 O'Shea JD, Rodgers RJ, D'Occhio MJ 1989 Cellular composition of the cyclic corpus luteum of the cow. J Reprod Fertil 85:483-487 O'Shea JD, Rodgers, RJ, McCoy K, D'Occhio MJ 1990 Ultrastructural cytology of the cyclic corpus luteum of the cow. Acta Anat 138:154-165

PAGE 198

3j* 181 O'Shea JD, Rodgers RJ, Wright PJ 1986 Cellular composition of the sheep corpus luteum in the midand late luteal phases of the estrous cycle. J Reprod Fertil 76:685-691 Ott TL, Zhou Y, Mirando MA, Stevens C, Harney JP, Ogle TF, Bazer FW 1993 Changes in progesterone and oestrogen receptor mRNA and protein fduring maternal recognition of pregnancy and luteolysis in ewes. J Mol Endocrinol 10:171-183 Paavola LG 1979 The corpus luteum of the guinea pig. IV. Fine structure of macrophages during pregnancy and postpartum luteolysis and the phagocytosis of luteal cells. Am J Anat 154:337-364 Paik YK, Chang DJ, Reardon CA, Davies GE, Mahley RW, Taylor JM 1985 Nucleotide sequence and structure of the human apolipoprotein E gene. Proc Natl Acad Sci USA 82:3445-3449 Parkinson TJ, Lamming GE, Flint AP, Jenner LJ 1992 Administration of bovine interferon-alpha 1 at the time of maternal recognition of pregnancy inhibits prostaglandin F2 alpha secretion and causes luteal maintenance in cyclic ewes. J Reprod Fertil 94:489-500 Parmer TG, McLean MP, Duan WR, Nelson SE, Albarracin CT, Khan I, Gibori G 1992 Hormonal and immunological characterization of the 32 kilodalton ovarian-specific protein. Endocrinology 131:2213-2221 Parry DM, Willcox DL, Thorburn GD 1980 Ultrastructural and cytochemical study of the bovine corpus luteum. J Reprod Pert 60:349-357 Patarca R, Haseltine WA 1985 A major retroviral core protein related to EPA and TIMP. Nature 318:390 Pate JL 1994 Cellular components involved in luteolysis. J Anim Sci 72:18841890 Pate JL, Condon WA 1989. Regulation of steroidogenesis and cholesterol synthesis by prostaglandin F2 alpha and lipoproteins in bovine luteal cells. J Reprod Fertil 87:439-446 Pederson ES 1951 Histogenesis of lutein tissue of the albino rat. Am J Anat 88:397-428

PAGE 199

182 Pimentel S, Evans G, Wagner WC 1987 Placental synthesis of estrogens at parturition and during placental retention in the cow. Theriogenology 28:755-765 Plante C, Hansen PJ, Thathcher WW 1990 Effect of recombinant bovine interferon-al type 1 (IFN) on ovarian function, oxytocin-induced release of prostaglandin Fja metabolite (PGFM) and body temperature in cows. Biol Reprod 42 (SuppI 1 ): 80 (Abstract) Polacek D, Beckmann MW, Schreiber JR 1992 Rat ovarian apolipoprotein E: localization and gonadotropic control of messenger RNA. Biol Reprod 46:65-72 Prack MM, Nicosia M, Williams DL, Gwynne J 1991 Relationship between apolipoprotein E mRNA expression and tissue cholesterol content in adrenal gland. J Lipid Res 32:1611-1618 Priedkalns J, Weber AF 1968 Ultrastructural studies of the bovine Graafian follicle and the corpus luteum. Z Zellforsch Mikrosk Anat 91 :554-573 Priedkalns J, Weber AF, Zemjanis R 1968 Qualitative and quantitative morphological studies of the cells of the membrana granulosa, theca interna, and corpus luteum of the bovine ovary. Z Zellforsch Mikrosk Anat 85:501-520 Puissant C, Houdebine LM 1990 An improvement of the single-step method of RNA isolation by acid-guanidinium thiocyanate-phenol-chloroform extraction. Biotechniques 8:148-149 Rajabi MR, Dodge G, Solomon S, Poole AR 1991a Immunochemical and immunohistochemical evidence of estrogen-mediated collagenolysis as a mechanism of cervical dilatation in the guinea pig at parturition. Endocrinology 128:371-378 Rajabi MR, Solomon S, Poole R 1991b Biochemical evidence of collagenasemediated collagenolysis as a mechanism of cervical dilatation at parturition in the guinea pig. Biol Reprod 45:764-772 Rajabi MR, Solomon S, Poole AR 1991c Hormonal regulation of interstitial collagenase in the uterine cervix of the pregnant guinea pig. Endocrinology 128:863-871

PAGE 200

183 Rapoport TA 1992 Transport of proteins across the endoplasmic reticulum membrane. Science 258:931-935 Redmer DA, Grazul AT, Kirsch JD, Reynolds LP 1988 Angiogenic activity of bovine corpora lutea at several stages of luteal development. J Reprod Fertil 82:627-634 Redmer DA, Grazul-Bilska AT, Reynolds LP 1991 Contact-dependent intercellular communication of bovine luteals in culture. Endocrinology 129:2756-2766 Reich R, Daphna-lken D, Chun SY, Popliker M, Slager R, Adelmann-Grill BC, Tsafriri A 1991 Preovulatory changes in ovarian expression of collagenases and tissue metalloproteinase inhibitor messenger ribonucleic acid: role of eicosanoids. Endocrinology 129:1869-1875 Riley JC, Behrman HR 1991 Oxygen radicals and reactive oxygen species in reproduction. Proc Soc Exp Biol Med 198:781-791 Roberts RM, Leaman DW, Cross JC 1992 Role of interferons in maternal recognition of pregnancy in ruminants. Proc Soc Exp Biol Med 200:7-18 Rodgers RJ, O'Shea JD, Findlay JK, Flint APF, Sheldrick EL 1983 Large luteal cells the source of luteal oxytocin in the sheep. Endocrinology 1 13:23022304 Rodgers RJ 1990 Cell-cell communication in corpora lutea. Reprod Fertil Dev 2:281-289 Rodgers RJ, Waterman MR, Simpson ER 1987 Levels of messenger ribonucleic acid encoding cholesterol side-chain cleavage cytochrome P-450, 17ahydroxylase cytochrome P-450, adrenoxin, and low density lipoprotein receptor in bovine follicles and corpora lutea throughout the ovarian cycle. Mol Endocrinol 1 :274-279 Rodgers RJ, Stuchbery SJ, Findlay JK 1989 Inhibin mRNA in ovine and bovine ovarian follicles and corpora lutea throughout the estrous cycle and gestation. Mol Cell Endocrinol 62:95-101 Roeb E, Rose-John S, Erren A, Edwards DR, Matern S, Graeve L, Heinrich PC 1995 Tissue inhibitor of metalloproteinases-2 (TIMP-2) in rat liver cells is increased by lipopolysaccharide and prostaglandin Ej. FEBS Lett 357:3336

PAGE 201

184 Romney JS, Chan J, Carr FE, Mooradian AD, Wong NC 1992 Identification of a thyroid hormone-responsive mRNA spot 1 1 as apolipoprotein A-1 mRNA and effects of the hormone on the promoter. Mol Endocrinol 6:943-950 Rueda BR, Tilly Kl, Hansen TR, Hoer PB, Tilly JL 1995 Expression of superoxide dismutase, catalase and glutathione peroxidase in the bovine corpus luteum: evidence supporting a role for oxidative stress in luteolysis. Endocrine 3:227-332 Sambrook J, Fritsch EF, Maniatis T 1989 In: Molecular Cloning: A Laboratory Manual, 2nd Ed., vol 1, Cold Spring Harbor Laboratory, Cold Spring Harbor, New York, 1.21-1.29 Sane A, Harman I, Quarfordt S, Costello A, Handwerger S 1988 Characterization of placental lactogen released from perifused human trophoblast cells. Placenta 9:129-138 Santoro M, Battaglia C, Zhang L, Carlomagno F, Martelli ML, Salvatore D, Fusco A 1994 Cloning of the rat tissue inhibitor of metalloproteinases type 2 (TIMP-2) gene: analysis of its expression in normal and transformed thyroid cells. Exp Cell Res 213:398-403 Sassone-Corsi P, Visvader J, Ferland L, Mellon PL, Verma IM 1988 Induction of proto-oncogenes fos transcription through the adenylate cyclase pathv\/ay: characterization of a cAMP responsive element. Genes Dev 2:1529-1538 Sato T, Ito A, Mori Y, Yamashita K, Hayakawa T, Nagase H 1991 Hormonal regulation of collagenolysis in uterine cervical fibroblasts: modulation of synthesis of procollagenase, prostromelysin and tissue inhibitor of metalloproteinase (TIMP) by progesterone and oestradiol-17li. Biochem J 275:645-650 Sato EF, Kobuchi H, Edashige K, Takahashi M, Yoshioka T, Utsumi K, Inoue M 1992 Dynamic aspects of ovarian superoxide dismutase isoenzymes during the ovulatory process in the rat. FEBS Lett 303:121-125 Satoh T, Kobayashi K, Yamashita S, Kikuchi M, Sendai Y, Hoshi H 1994 Tissue inhibitor of metalloproteinases (TIMP-1 ) produced by granulosa and oviductal cells enhances in vitro development of bovine embryos. Biol Reprod 50:835-844

PAGE 202

185 Sauerwein H, Miyamoto A, Gunther J, Meyer HHD, Schams D 1992 Binding and action of insulin-like growth factors and insulin in bovine luteal tissue during the estrous cycle. J Reprod Fertil 96:103-115 Savard K 1973 The biochemistry of the corpus luteum. Biol Reprod 8:183-202 Sawada M, Carlson JC 1989 Superoxide radical production in plasma membrane samples of regressing rat corpora lutea. Can J Physiol Pharmacol 67:465471 Schams D 1989 Ovarian peptides in the cow and sheep. J Reprod Fertil SuppI 37:225-231 Schreiber J, Hsueh A, Weinstein D, Erickson G 1980 Plasma lipoproteins stimulate progestin production by rat ovarian granulosa cell cultured in serum-free medium. J Steroid Biochem 13:1009-1014 Segaloff DL, Wang H, Richards JS 1990 Hormonal regulation of luteinizing hormone/chorionic gonadotropin receptor mRNA in rat ovarian cells during follicular development and luteinization. Mol Endocrinol 4:18561865 Shakelford JE, Lebherz HG 1983 Synthesis and secretion of apolipoprotein A-1 by chick breast muscle. J Biol Chem 258:7175-7180 Sheehan DC, Hrapchak BB In: Theory and Practice of Histotechnology. The C.V. Mosby Company, St. Louis, MO, pp 143 Shelburne FA, Quarfordt SH 1974 A new apoprotein of human plasma low density lipoproteins. J Biol Chem249: 1428-1 433 Sherwood OD 1994 Relaxin. In: The Physiology of Reproduction, Second Edition, edited by E. Knobil and JD Neill, Raven Press, Ltd., New York, pp 861-1009 Short RV 1969 Implantation and maternal recognition of pregnancy. In: GEW Wolztenholme and M O'Connor (Ed). Fetal autonomy, pp 2-26, Churchill, London Short RV 1977 The discovery of the ovaries. In: S. Zuckerman and B.J. Weir (Ed.). The ovary: I. General Aspects, pp 1-39 Academic Press, New York

PAGE 203

186 Silbiger SM, Jacobsen VL, Cupples RL, Koski RA 1994 Cloning of cDNAs encoding human TIMP-3, a novel member of the tissue inhibitor of metalloproteinase family. Gene 141:293-297 Smith MF, Moor RM 1991 Secretion of a putative metalloproteinase inhibitor by ovine granulosa cells and luteal tissue. J Reprod Fertil 91:627-635 Smith GW, Moor RM, Smith MF 1993 Identification of a 30,000 M(r) polypeptide secreted by cultured ovine granulosa cells and luteal tissue as tissue inhibitor of metalloproteinases. Biol Reprod 48:125-132 Smith GW, Goetz TL, Anthony RV, Smith MF 1994 Molecular cloning of an ovine ovarian tissue inhibitor of metalloproteinases: ontogeny of messenger ribonucleic acid expression and in situ localization within preovulatory follicles and luteal tissue. Endocrinology 134:344-352 Smith GW, McCrone S, Petersen SL, Smith MF 1995 Expression of messenger ribonucleic acid encoding tissue inhibitor of metalloproteinases-2 within ovine follicles and corpora lutea. Endocrinology 136:570-576 Smith MF, Kemper CN, Smith GW, Goetz TL, Jarrel VL 1994 Production of tissue inhibitor of metalloproteinases-1 by porcine follicular and luteal cells. J Anim Sci 72:1004-1012 Smith KB, Millar MR, McNeilly AS, lllingworth PJ, Eraser MH, Baird DT 1991 Immunocytochemical localization of inhibin a-subunit in the human corpus luteum J Endocrinol 129:155-160 Sorci-Thomas M, Prack MM, Dashti N, Johnson F, Rudel LL, Williams DL 1988 Apolipoprotein (apo) A-1 production and mRNA abundance explain plasma Apo A-1 and high density lipoprotein differences between two nonhuman primate species with high and low susceptibilities to dietinduced hypercholesterolemia. J Biol Chem 263:5183-5189 Soutar Ak, Garner CW, Baker HN, Sparrow JT, Jackson RL, Gotto AM, Smith LC 1975 Effect of the human plasma apolipoproteins and phosphatidylcholine acyl donor on the activity of lecithin:cholesterol acyltransferase. Biochem 14:3057-3063 Sparrow DA, Lee BR, Laplaud PM, Auboiron S, Bauchart D, Chapman MJ, Gotto Jr. AM, Yang CY, Sparrow JT 1992 Plasma lipid transport in the preruminant calf. Bos spp: primary structure of bovine apolipoprotein A-1. Bioch Biophys Acta 1123:145-150

PAGE 204

187 Statistical Analysis System 1988. SAS User's Guide, Statistics, Statistical Analysis Institute Inc., Gary, NC Steinetz BG, Butler MC, Sawyer WK, O'Byrne FM 1976 Effects of relaxin on early pregnancy in rats. Proc Soc Exp Biol Med 152:419-422 Stetler-Stevenson WG, Bersch N, Golde DW 1992 Tissue inhibitor of metalloproteinase-2 (TIMP-2) has erythroid-potentiating activity. FEBS 296:231-234 Stetler-Stevenson WG, Brown PD, Onisto M, Levy AT, Liotta LA 1990 Tissue inhibitor of metalloproteinases-2 (TIMP-2) mRNA expression in tumor cell lines and human tumor tissues. J Biol Chem 265:13933-13938 Stirling D, Waterman MR, Simpson ER 1991 Expression of mRNA encoding basic fibroblast growth factor (bFGF) in bovine corpora lutea and cultured luteal cells. J Reprod Fertil 91:1-8 Stoelk E, Chegini N, Lei ZM, Rao CHV, Bryant-Greenwood G, Sanfilippo J 1991 Immunocytochemical localization of relaxin in human corpora lutea: cellular and subcellular distribution and dependence on reproductive state. Biol Reprod 44: 1 1 40-1 1 47 Tanabe TY 1966 Essentiality of the corpus luteum for maintenance of pregnancy in dairy cows. (Abstr.) J Dairy Sci 49:731 Tanaka T, Andoh N, Takeya T, Sato E 1992 Differential screening of ovarian cDNA libraries detected the expression of the porcine collagenase inhibitor gene in functional corpora lutea. Mol Cell Endocrinol 8:65-71 Thatcher WW, Bazer FW, Sharp DC, Roberts RM 1986 Interrelationship between uterus and conceptus to maintain corpus luteum function during early pregnancy: sheep, cattle, pigs and horses. J Anim Sci 67 (SuppI 2):47-61 Thatcher WW, Danet-Desnoyers G, Wetzels C 1992 Regulation of bovine endometrial prostaglandin secretion and the role of bovine trophoblast protein-1 complex. Reprod Fertil Dev 4: 329-334 Thatcher WW, Hansen PJ, Gross TS, Helmer SD, Plante C, Bazer FW 1989 Antiluteolytic effects of bovine trophoblast protein-1. J Reprod Fertil SuppI 37:91-99

PAGE 205

188 Thatcher WW, Knickerbocker JJ, Bartol FF, Bazer FW, Roberts RM, Drost M 1985 Maternal recognition of pregnancy in relation to the survival of transferred embryos: endocrine aspects. Theriogenology 23:129-144 Theodosis DT, Wooding FBP, Sheldrick EL, Flint APF 1986 Ultrastructural localization of oxytocin and neurophysin in the ovine corpus luteum. Cell Tissue Res 243:129-135 Tov\/bin H, Staehelin T, Gordon J 1979 Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350-4354 Tsonis CG, Baird DT, Campbell BK, Leask R, Scaramuzzi RJ 1988 The sheep corpus luteum secretes inhibin. J Endocrinol 116:R3-R5 Ulisse S, Farina AR, Piersanti D, Tiberio A, Cappabianca L, D'Orazi G, Jannini EA, Malykh 0, Stetler-Stevenson WG, D'Armiento M, Gulino A, Mackay AR 1994 Follicle-stimulating hormone increases the expression of tissue inhibitor of metalloproteinases TIMP-1 and TIMP-2 and induces TIMP-1 AP-1 site binding complexe(s) in prepubertal rat Sertoli cells. Endocrinology 135:2479-2487 Vallet JL, Bazer FW, Fliss MF, Thatcher WW 1988 Effect of conceptus ovine secretory proteins and purified ovine trophoblast protein-1 on interestrous interval and plasma concentrations of prostaglandin Fja, and E and 13, 14-dihydro-15-keto prostaglandin Fja in cyclic ewes. J Reprod Fertil 84:493-504 Vallet JL, Lamming GE 1991 Ovine conceptus secretory proteins and bovine recombinant interferon alpha (1)-1 decrease endometrial oxytocin receptor concentrations in cyclic and progesterone-treated ovariectomized ewes. J Endocrinol 131:475-482 Vanlennys EW, Madden LM 1965 Electron microscopic observations of the involution of the human corpus luteum of menstruation. A. Zellforsch Mikrosk Anat 66:365-380 Visner GA, Dougall WC, Wilson JM, Burr lA, Nick HS 1990 Regulation of manganese superoxide dismutase by lipopolysaccharide, interleukin-1, and tumor necrosis factor. Role in the acute inflammatory response. J Biol Chem 265:2856-2864

PAGE 206

189 Vorob'ev EV, Perevozchikov AP 1992 The expression of apolipoprotein A-1 gene in the early stages of human embryogenesis studied by hybridization in situ. Ontogenez 23:469-479 Waterhouse P, Denhardt DT, Khokha R 1993 Temporal expression of tissue inhibitors of metalloproteinases in mouse reproductive tissues during gestation. Mol Reprod Dev 35:219-226 Wathes DC, Pickering BT, Swann RW, Porter DG, Hull MGR, Drife JO 1982 Neurohypophysial hormones in the human ovary. Lancet 2:410-412 Wathes DC, Swann RW 1982 Is oxytocin an ovarian hormone? Nature (London) 297:225-227 Wathes DC, Swann RW, Birkett SD, Porter DG, Pickering BT 1983 Characterization of oxytocin, vasopressin and neurophysin from the bovine corpus luteum. Endocrinology 113:693-698 Weber DM, Fields PA, Romrell LJ, Tumwasorn S, Ball BA, Drost M, Fields MJ 1987 Functional differences between small and large luteal cells of the late-pregnant versus nonpregnant cow. Biol Reprod 37:685-697 Weisgraber KH 1994 Apolipoprotein E: structure-function relationships. Adv Prot Chem 45:247-302 Welgus HG, Jeffrey JJ, Eisen AZ, Roswit WT, Stricklin GP 1985 Human skin fibroblast collagenase: interaction with collagen and collagenase inhibitor. Collagen Relat Res 5:167-179 Werb Z 1989 Proteinases and matrix degradation. In: Kelley WN, Harris ED Jr, Ruddy S, Sledge CB (eds): Textbook of Rheumatology, Ed 3. Philadelphia, WB Saunders, pp 300-321 White JE, Tsan MF 1994 Induction of pulmonary Mn superoxide dismutase mRNA by interleukin-1. Am J Physiol 266:L664-671 Whitsett JA, Clark JC, Wispe JR, Pryhuber GS 1992 Effects of TNF-alpha and phorbol ester on human surfactant protein and manganese superoxide dismutase transcription in vitro. Am J Physiol 262:L688-693 Widnell CC, Pfenninger KH 1990 The Cytoplasm In: Essential cell biology, (ed) Williams and Wilkins, Baltimore, MD, pp 75-131

PAGE 207

190 Wiltbank MC 1994 Cell types and hormonal mechanisms associated with midcycle corpus luteum function. J Anim Sci 72:1873-1883 Wiltbank MC, Gallagher KP, Christensen AK, Brabec RK, Keyes PL 1990 Physiological and immunocytochemical evidence for a new concept of blood flow regulation in the corpus luteum. Biol Reprod 42:139-149 Wiltbank MC, Guthrie PB, Mattson MP, Kater SB, Niswender GD 1989 Hormonal regulation of free intracellular calcium concentrations in small and large ovine luteal cells. Biol Reprod 41:771-778 Wiltbank MC, Niswender GD 1992 Functional aspects of differentiation and degeneration of the steroidogenic cells of the corpus luteum in domestic ruminants. Anim Reprod Sci 28:103-110 Winters IM, Green WW, Comstock RE 1942 Prenatal development of the bovine. Minn Tech Bull 151 Wu XM, Sawada M, Carlson JC 1992 Stimulation of phospholipase A2 by xanthine oxidase in the rat corpus luteum. Biol Reprod 47:1053-1058 Wu YQ, Jorgensen EV, Handwerger S 1988 High density lipoproteins stimulate placental lactogen releaseand adenosine 3',5'-monophosphate (cAMP) production in human trophoblast cells: evidence for cAMP as a second messenger in human placental lactogen release. Endocrinology 123:1879-1884 Wyne LW, Schreiber JR, Larsen AL, Getz GS 1989a Regulation of apolipoprotein E biosynthesis by cAMP and phorbol ester in rat ovarian granulosa cells. J Biol Chem 264:981-989 Wyne LW, Schreiber JR, Larsen AL, Getz GS 1989b Rat granulosa cell apolipoprotein E secretion: regulation by cell cholesterol. J Biol Chem 264:16530-16536 Xavier F, Guillemot M, Charlier M, Martal J, Gaye P 1991 Co-expression of the proto-oncogene FOS (c-fos) and an embryonic interferon (ovine trophoblastin) by sheep conceptuses during implantation. Biol Cell 73:2733 Xing Y, Johnson CV, Dobner PR, Lawrence JB 1993 Higher level organization of individual gene transcription and RNA splicing. Science 259:1328-1330

PAGE 208

191 Yang YW, Chan L, Li WH 1991 Cloning and sequencing of bovine apolipoprotein E complementary DNA and molecular evolution of apolipoprotein E, C-!, and C-ll. J Mol Evol 32:469-475 Zannis VI, McPherson J, Goldberger G, Karanthanasis SK, Breslow JL 1984 Synthesis, intracellular processing, and signal peptide of human apolipoprotein E. J Biol Chem 259:5495-5499 Zarco L, Stabenfeldt GH, Basu S, Bradford GE, Kindahl H 1988 Modification of prostaglandin Fja synthesis and release in the ewe during the initial establishment of pregnancy. J Reprod Fertil 83:527-536 Zheng J, Redmer DA, Reynolds LP 1993 Vascular development and heparinbinding growth factors in bovine corpus luteum at several stages of the estrous cycle. Biol Reprod 49: 1 1 77-1 1 89 Zuckerman SH, O'Neal L 1984 Endotoxin and GM-CSF-mediated downregulation of macrophage Apo E secretion is inhibited by a TNF-specific monoclonal antibody. J Leukoc Biol 55:743-748

PAGE 209

APPENDIX 1 ANIMAL CARE AND TISSUE COLLECTION Experimental Protocol Animals Forty-eight cross-bred beef cows were used in the study. Estrus (day 0) was determined by monitoring cows twice daily for standing to be mounted by a vasectomized bull. Some cows were bred by artificial insemination at observed estrus. Early pregnancy was confirmed by the presence of an embryo in flushings of the uterus. Later stages of pregnancy were determined by measurement of crown-rump length of the calves (Winters et al., 1942). Preparation for Tissue Collection Autoclave all materials to be used. These include: Towels Forceps Scissors Stadie Rigg's tissue slicer (Thomas Scientific, Swedesboro, NJ) Tissue slicer handle Autopipet tips. 15 X 100 mm Petri dishes (Fisher Scientific Co., Springfield, NJ) 192

PAGE 210

193 Individually wrapped sterile disposable serological pipets. Preparation of Culture Medium The medium was prepared as described by Basha et al. (1980) from Eagle's Minimum Essential Medium (MEM, Sigma Chemical Co., St. Louis, MO), that was deficient in leucine, methionine and sodium bicarbonate. One liter of stock incomplete MEM was prepared as follows: 1. Empty the bottle containing lyophilized Eagle's MEM (9.4 g/l) into a 1 I beaker. 2. Then add: Glucose (3 g) Methionine (1.5 mg) Leucine (5.2 mg) Lysine-HCI (72.5 mg) to achieve 4.0, 0. 1 0. 1 and 1 .0 times, respectively, their usual concentrations in MEM. Sodium bicarbonate (2.2 g) 10 ml non-essential amino acids solution 10 ml vitamins mixture Insulin (200 lU) 3. Adjust pH to 7.1-7.3 with 1 M HCI. Make up to 1 I with deionized water. 4. Filter-sterilize the medium (0.22 pm) into sterile bottles in a laminar flow hood. 5. Store medium at 4 C.

PAGE 211

194 For labeling with ^H-leucine, methionine (1.5 mg/100 ml) and antibioticantimycotic mixture (ABAM) (1%, v/v) were added to the stock (modified) incomplete MEM to obtain leucine-deficient incomplete MEM. Similarly, for the ^^S-methionine labeling incubations, methionine-deficient medium was prepared by adding leucine (5.2 mg/100 ml) to the stock modified incomplete MEM. For ^H-glucosamine culture, incomplete modified MEM was used. Blood Collection Trunk blood was collected at slaughter into heparinized tubes. Blood samples were centrifuged at 3000 rpm for 15 min at 4 C. Plasma was collected and stored at -20 C until analysed for progesterone by radioimmunoassay (see appendix 2). Tissue Collection Reproductive tracts were obtained from cows within 5 min after exsanguination. The ovary containing the corpus luteum (CL) was collected aseptically from cows on days 3 (n = 4), 7 (n = 3), 1 1 (n = 4), 1 4 (n = 5), 1 7 (n = 3) and 19 (n = 3) of the estrous cycle and from cows of early (day 17, n = 5), first (day 88, n = 5), second (day 170, n = 7), and third (> day 240, n = 9) trimester of pregnancy. The ovaries were harvested and transferred immediately to a sterile Petri dish containing Eagle's incomplete Minimum Essential Medium (modified MEM) pre-warmed at 37 C. Subsequent processing of tissue was done in a laminar flow hood. The CL was dissected out of the ovaries and trimmed of any excess fat and surrounding ovarian stroma. Each CL was then weighed and

PAGE 212

195 transferred to another Petri dish containing fresh pre-warmed incomplete MEM. Luteal tissue for RNA isolation were snap-frozen in liquid nitrogen after collection, and then stored at 80 C until further analysis. Preparation of Luteal Tissue 1. Prepare 0.5 mm slices of luteal tissue with the use of sterile forceps and scissors, using the Stadie Rigg's tissue slicer. 2. Wash the slices 3 times with incomplete MEM to reduce serum proteins in the medium during incubation. 3. For culture in each 15 x 100 mm Petri dish, weigh 500 mg of luteal slices directly in the Petri dish containing 15 ml of culture medium (see appendix 3).

PAGE 213

APPENDIX 2 PROGESTERONE RADIOIMMUNOASSAY Introduction The DPC Coat-A-Count (Diagnostic Products Corporation, Los Angeles, CA) procedure is a solid-phase radioimmunoassay in which ^^^l-progesterone competes for a fixed time with progesterone in the experimental sample for antibody sites. Since the antibody is immobilized to the wall of a polypropylene tube, decanting the supernatant is sufficient to terminate competition and to isolate the antibody-bound fraction of the radiolabeled progesterone. Counting the tube in a gamma counter yields a number, which converts by way of a calibration curve to a measure of the progesterone present in the sample. Materials Supplied in Kit Progesterone antibody-coated tubes lodinated progesterone (^^^l-progesterone) supplied in liquid form. Progesterone Standards (0, 0.1, 0.5, 2, 10, 20 and 40 ng progesterone/ml, equivalent to 0, 10, 50, 200, 1000, 2000 and 4000 pg progesterone/100 pl/tube). Materials Prepared in Lab Plain uncoated 12 x 75 mm polypropylene tubesfor use as NSB tubes. Plasma from an ovariectomized cow. 196

PAGE 214

197 Standard solutions of progesterone prepared in ovariectomized plasma via a serial dilution as follows from a stock solution of progesterone (Sigma Chemical Co.) in benzene (50 pg/ml): X: 1 000 pg/1 00 pi = 1 pi stock + 5 ml OVX cow plasma A: 500 pg/1 00 pi = 1 ml X + 1 ml OVX cow plasma B:250pg/100pl = 1mlA+1ml C: 125pg/100pl = 1 ml B + 1 ml D: 62.5 pg/1 00 pi = 1 ml C + 1 ml E: 31.25 pg/1 00 pi = 1 ml D + 1 ml 100 pi plasma/tube was used in the assay. Assay Procedure 1. Label four uncoated 12 x 75 mm polypropylene tubes T (total counts) and NSB (nonspecific binding) in duplicate. 2. Label tubes coated with progesterone antibody for Zero binding (Bo), the standard curve (31.25, 62.5, 125, 250, 500, 1000 pg/1 00 pl/tube in OVX cow plasma, the samples and reference controls in duplicate. 3. Pipet 100 pi OVX cow plasma into the NSB and Bo tubes. 4. Pipet 100 pi of the standards and samples into appropriate tubes. If sample is less than 100 pi, adjust to 100 pi with OVX plasma. 5. Add 1 .0 ml of ^^^l-progesterone to every tube and vortex. No more than 10 min should elapse between the first and last tubes during addition of the tracer.

PAGE 215

198 6. Set the T tubes aside for counting at end of assay; they require no further processing. 7. Incubate all tubes at room temperature for 3 h. 8. Decant contents of all tubes (except the T tubes) and allow to drain for 23 min. Strike the tubes sharply on absorbent paper to shake off all residual droplets. 9. Count all tubes for 1 min in a gamma counter. Progesterone RIA Protocol Tube ovx standard/ m Sample (gl) Total NSB 100 Zero 100 Standards 100 Sample 100 Reference 100 125 l-P. ml)

PAGE 216

199 Calculation of Results Progesterone concentrations are calculated from a logit-log representation of the calibration curve. Calculate the average NSB-corrected counts per minute for each pair of tubes, and then determine the binding of each pair of tubes as a percent of maximum binding (MB), with the NSB-corrected counts of the A tubes taken as 1 00%: Percent Bound = Net Counts/ Net MB Counts x 100 Validation of Assay The progesterone kit originally designed for human serum was validated for use with cow plasma, as described in chapter 3.

PAGE 217

APPENDIX 3 CULTURE AND RADIOLABELLING OF LUTEAL TISSUE Time Course Studies of Incorporation of Radiolabel Introduction A time course study was carried out with CL from 3 animals to determine the time of incubation required to obtain optimum incorporation of radiolabel into proteins released into culture medium. Five Petri dishes of luteal tissue (500 mg/dish) were prepared from each corpus luteum. Each of the five was given the same treatment except for the length of incubation. Procedure Pre-incubation of Luteal Tissue without Radiolabel 1. Prepare five culture dishes of luteal slices (500 mg/ Petri dish) for each CL 2. Incubate luteal slices (500 mg/dish) in 15 ml leucine-deficient incomplete MEM without radiolabel for 2 h at 37 C in a gas chamber mixing equal volumes of 100% Nj and a 90:10, OsiCOj mixture, and maintain the temperature at 37 C by using a Fisher Isoterm incubator model 255D (Springfield, NJ). 3. After pre-incubation, take Petri dishes out of the incubator, and discard 200

PAGE 218

201 the spent medium. Incubation of Luteal Tissue with ^H-leucine 4. Add 15 ml of pre-warmed incomplete MEM (15 ml/Petri dish) to tissue slices in each dish. 5. Add 50 |jCi of L-[4,5-^H-leucine (160 Ci/mmol) to each Petri dish and swirl gently to mix. 6. Pipet 10 pi of medium from each dish into a scintillation vial, add 3 ml scintillation cocktail and count to obtain pre-incubation counts. 7. Then place Petri dishes in the gas chamber and incubate under same atmospheric conditions. 8. Take out one of the five Petri dishes after 6, 1 2, 1 8, 24 and 30 h, respectively of incubation. 9. Transfer tissue and medium from Petri dish into a 15 ml centrifuge tube. 10. Centrifuge at 2000 x g for 20 min to separate luteal-conditioned medium (LCM) from the tissue. 11. Transfer the supernatant (conditioned-medium) from each Petri dish into a separate dialysis tubing (Spectra/Por'^ 3 membrane, molecular weight cutoff = 3500) (Spectrum Medical Industries Inc., Houston, TX). 12. Dialyze medium against 2 changes (24 h each) of 4 I Tris-HCI buffer (10 mM pH 8.2) at 4 C, and then against 2 changes (24 h each) of 4 I deionized water. 13. Following dialysis, transfer the retentate (dialyzed luteal-conditioned

PAGE 219

202 medium) to a 50 ml centrifuge tube, and measure total volume of retentate. Volume should be 15 ml or close to that. 14. Pipet 10 |jl of each retentate in triplicate into a scintillation vial and count by scintillation spectroscopy to obtain post-dialysis counts for each incubation period (see appendix 4). 1 5. Store the retentates at -20 "C. Radiolabelinq with^H-leucine Results of the time course study shov^ed that the 24 h-incubation time was optimal for incorporation of radiolabel into proteins. Thus 24 h was used for subsequent incubations. Procedure 1. Prepare two Petri dishes of luteal slices (500 mg/dish) from each CL. 2. Carry out pre-incubation without radiolabel for 2 h as described earlier. 3. Discard spent medium and carry out incubation with radiolabel for 24 h as described above. 4. Before incubation, take 10 pi of medium (in triplicate) and count by scintillation spectroscopy (pre-incubation counts). 5. Following incubation and dialysis, take 10 pi of retentate and count by scintillation spectroscopy (post-dialysis counts). 6. Measure volume of dialyzed LCM (retentate) from each Petri dish and adjust to 15 ml with deionized water if < 15 ml. 7. Store retentates at 20 C until further analysis.

PAGE 220

203 Radiolabeling with^H-glucosamine and ^^S-methionine In addition to radiolabelling with ^H-leucine, luteal slices (500 mg/dish) were also incubated with ^H-glucosamine (50 (jCi/15 ml, n = 1 cow) or ^^Smethionine (50 |jCi/15 ml, n = 1 cow) following the same protocol, to determine if newly synthesized and secreted proteins were glycosylated and/or contained methionine. Liver tissue was also incubated in the same manner with ^H-leucine to evaluate the tissue specificity in protein synthesis and secretion. Retentates from these incubations were also stored at 20 C until further analysis. Lyophilization of Retentates 1. Freeze-dry each frozen retentate at 50 C under vacuum. 2. Weigh and record weight of each lyophilized sample. 3. Store samples at 20 C until further analyses.

PAGE 221

APPENDIX 4 DETERMINATION OF INCORPORATION OF RADIOLABEL INTO NEWLYSYNTHESIZED PROTEINS A. Percent Incorporation 1. Add 3 ml scintillation cocktail to 10 pi of culture medium (in triplicate) prior to addition of ^H-leucine, and measure radioactivity (pre-incubation counts) by scintillation spectroscopy (2200CA TRI-CARB scintillation analyzer, Packard Instrument Company). 2. After dialysis, take 10 pi of retentate (in triplicate) and count by scintillation spectroscopy (post-dialysis counts). 3. Calculate Percent Incorporation of ^H-leucine as post-dialysis counts divided by pre-incubation counts x 100%, for each sample. B. Measurement of TCA-Precipitable Radioactivity 1. Cut one-inch squares of Whatman 3MM filter paper (Whatman Ltd., Maidstone, England) and number using a pencil. Squares should be prepared in duplicate for each sample. 2. Soak numbered squares in 20% TCA and let air dry completely (could use heat lamp to promote drying). 3. Prepare a series of 500 ml plastic beakers (reservoir beakers) containing 20% TCA, 5% TCA, 5% TCA, 95% ethanol, 95% ethanol, respectively. 204

PAGE 222

205 4. Cut holes into the bottom of a beaker small enough to fit Into the reservoir beakers. 5. Place a radioactive tape on all equipment used. 6. Bring out post-dialysis retentates from the 20 C freezer to warm to room temperature. 7. After the squares are completely dry, spot 50 \j\ of each retentate onto each square. Dispense the samples slowly to avoid loss. Spot each sample in duplicate. 8. Place the spotted squares on a tray double-lined with bench paper, and allow to air dry completely. It would take at least 30 min. 9. When completely air dry, pick up the squares with a pair of tweezers and place in the beaker with holes at the bottom. 10. Soak the squares successively in 20% TCA, 5% TCA, 5% TCA, 95% ethanol and 95% ethanol for 10, 20, 10, 10, and 5 min, respectively, by fitting the beaker with holes into each reservoir beaker. 1 1 Retrieve the squares and place on a tray for 1 h to completely air dry. 12. Pick up the dried squares using tweezers and place in scintillation vials. 13. Add 3 ml scintillation cocktail to each vial and count radioactivity on a beta counter. 14. Obtain and record counts (dpm and cpm) in duplicate for each sample.

PAGE 223

APPENDIX 5 SEPARATION OF PROTEINS IN LUTEAL-CONDITIONED MEDIUM BY ELECTROPHORESIS First-dimension : Isoelectnc Focusing (lEF) in Tube Gels was carried out according to method of Laemmli (1970). Preparation of Glassware and Reagents for lEF Glass Tubes: 1. Soak lEF glass tubes (inner diameter 2.5 mm) in concentrated nitric acid for at least 24 h. 2. Remove the glass tubes from the acid and soak in deionized water to rinse off excess acid. 3. Soak tubes in 0.2% (w/v) potassium hydroxide (2 g KOH/I in 95% ethanol) for at least 2 h. 4. Rinse tubes extensively with deionized water and place in a beaker lined with paper napkins to drain. Ensure complete dryness by placing the beaker in a drying oven overnight. 5. Prior to use, seal one end of each tube with parafilm (at least three layers). 6. Mark each tube at 12 cm from the sealed end to maintain same length in 206

PAGE 224

207 tube gels. 7. Place the tubes In the gel casting stand with sealed ends down. Pouring of Tube Gels: 1. Fill the lEF tubes to the mark with the lEF gel solution using a 10 cc syringe and a 20 gauge 5 inch-long needle with tubing at the tip. 2. Overlay tube gels with 20 pi 8 M urea. 3. Overlay the urea layer with 30 pi deionized water. 4. Cover the tubes with clear plastic and leave at room temperature for gels to polymerize. Isoelectric Focusing lEF Pre-run: 1. After the gels polymerize, carefully aspirate the gel overlays using a Pasteur pipet and a long pipet tip. 2. Set up the lEF gel apparatus. Fill the lower reservoir with anolyte solution. Stand the tube holder (upper reservoir) inside the lower tank to separate the upper and lower reservoirs. 3. Take each gel tube out of the casting stand, and remove the parafilm from the sealed end. 4. Dip that end into glycerol (to eliminate air and facilitate insertion of tube through bushings in the tube holder), and insert tubes through the holes into the lower reservoir, making sure the top of each gel is visible over the

PAGE 225

208 bushings. 5. Fill the gel tubes to the brim with catholyte solution. Then pour more catholyte solution into the upper reservoir so that all the tubes are submerged. 6. Connect the gel apparatus to a cooling apparatus pre-set at 20 C, and a power supply unit set at constant voltage. 7. Carry out an isoelectric focusing pre-run for 15 min at 200 V, then for 30 min at 300 V and for 30 min at 400 V. 8. In the meantime, prepare the samples for loading. Sample Preparation/Loading: 1. Take lyophilized retentates out of the freezer. 2. Weigh the amount of each sample equivalent to 500,000 cpm. 3. Dissolve each sample in 200 pi lEF sample buffer. Spin solution in microfuge for 2 min to precipitate any insoluble material. 4. Pipet 10 pi of supernatant into a scintillation vial, add 3 ml scintillation cocktail and count radioactivity. 6. Based on the counts for the 10 pi aliquot, calculate the volume of sample equivalent to 100,000 cpm for each sample. 6. After the pre-run, turn off power supply and cooling units. 7. Siphon off the catholyte solution in the upper tank. 8. Aspirate (using autopipet and long pipet tip) the layer of catholyte overlay on the gels.

PAGE 226

209 9. Overlay each tube gel with 10 pi of lEF sample buffer. 1 0. Then load the sample (1 00,000 cpm) on top of the tube gel. 1 1 Overlay the samples with catholyte solution to the brim of the glass tubes. 12. For every lEF run, load one tube gel with only sample buffer, to serve as the pH gradient of the run. 13. After loading the samples, fill the upper reservoir with fresh catholyte solution. 14. Reconnect the cooling unit and power supply. 15. Set the voltage at 400 V and carry out isoelectric focusing for a minimum of 8000 Vh. 16. After electrofocusing, remove lEF tubes from the apparatus, and extrude the JEF gel by attaching a tube at the end of a syringe filled with deionized water, to one end of the lEF tube and pushing the plunger. 17. Nick the acidic end (bottom) of the lEF gel for orientation. 18. Either use the lEF tube gels immediately for the second-dimension slab gel electrophoresis or store tube gels individually in 18 x 150 mm glass tubes at 20 C until further analysis. Determination of pH Gradient: 1. Extrude the lEF tube gel containing no sample (control gel) from the glass tube after electrofocusing. 2. Cut the control gel into 0.5 cm pieces sequentially from the acidic to the basic end.

PAGE 227

210 3. Place each piece in a 1 .5 ml bullet tube vial containing 600 pi of deionized water. 4. Place the vials on a rocker platform at room temperature overnight, to allow sippage of ampholine out of gels into the water. 5. Measure pH of the solution in each vial, beginning at the acidic end. Second Dimension Electrophoresis Preparation of Acrvlamide Separating Gel (10% T. 2.7% C^. ^^ In a 125 ml side arm vacuum flask with a small magnetic stir bar, the separating gel solution was prepared by mixing the following: 20 ml Acrylamide Stock Monomer solution (30% T, 2.7% CtJ 15 ml 4X Running Gel Buffer stock 0.6ml10%(w/v)SDS 24.1 ml deionized water. Stopper the flask and deaerate the mixture by applying vacuum for several minutes. Add 300 pi 1 % (w/v) ammonium persulfate, mix, and add 20 |jl TEMED to 60 ml. Swirl the flask gently to mix, avoid generating air bubbles.

PAGE 228

*, 211 Pouring of Separating Gel Mixture: 1. Pour the separating gel mixture into the sandwich to the mark using a 60 ml syringe. 2. Overlay the gels with 3 ml deionized water and allow to polymerize (about 1h). 3. Pour off the overlay by tilting the casting stand. 4. Overlay (equilibrate) gels with 3 ml Running Gel Overlay Solution for at least 2 h. Preparation of the Stacking Gel (4% T. 2.7% Ct, ,,) In a 125 ml side arm flask containing a magnetic stir bar add: 2.66 ml Acrylamide Stock solution (30% T, 2.7% Cbis) 5.0 ml 4X Stacking Gel Buffer 0.2 ml 10%(w/v)SDS 12.2 ml deionized water. Deaerate the solution, then add 100 pi 1% (w/v) ammonium persulfate, and 10MlTEMEDto20ml. Swirl to mix. Pouring of Stacking Gel Mixture After equilibration, pour off the Running Gel Overlay. 1. Clamp the stacking gel caster on top of the separating gel assembly. 2. Rinse the surface of the polymerized separating gel with 1-2 ml of

PAGE 229

212 stacking gel solution. 3. Rock the casting stand manually to wash the surface, and pour off the liquid. 4. Overlay the polymerized separating gel with the remaining stacking gel solution until latter is visible above the grooves of the casting gel stand, using a Pasteur pipet and pipet tip. 5. Overlay the stacking gel with deionized water and allowed to polymerize (45 min). 2-D Electrophoresis: Following isoelectric focusing, 1. Equilibrate the tube gels (100,000 cpm/IEF gel) in gel equilibration buffer (0.0625 M Tris, 5% (w/v) SDS, 10% (v/v) glycerol) for 15 min. 2. Lay the equilibrated lEF gel in the grooves. Take note of the orientation (acidic/basic ends). 3. Place a molecular weight marker worm to the left of the lEF gel. 4. Dispense hot 0.1 % (w/v) agarose into the groove with a transfer pipet. This will connect the lEF gel to the stacking gel when the agar sets. 5. Place the assembly in the lower electrophoresis tank, fill the upper reservoir with tank buffer, place the lid, and connect the assembly to the power supply unit. 6. Carry out electrophoresis at a constant current of 1 3 mA/gel until the dye front reaches the end of the glass plates.

PAGE 230

213 7. Remove gels from the apparatus, and detach the stacking from the separating gel. 8. Nick the acidic end of the separating gel for orientation. 9. Stain separating gel with 0.1 % (w/v) Coomassie Blue for at least 2 h on a rocker platform. 10. Destain gels in destaining solution (see below) overnight on a rocker platform. Fluorography and Densitometry 1. Soak destained gels in deionized water for at least 30 min. 2. Soak gels in 1 M sodium salicylate for 30 min. 3. Place gels on filter paper (8 in x 10 in) pre-wet with deionized water. 4. Cover gel with cling film, and dry for 3 h in a vacuum slab gel dryer (Model SE 1150, Hoefer Scientific Instruments, San Francisco, CA). 5. Mark positions of the molecular weight markers on the filter paper with pencil. Also write the sample number. 6. Expose dried gels to x-ray film (X-OMAT-AR, Eastman Kodak Company, Rochester, NY) for 4 wk at 80 C. 7. Identify radiolabeled proteins by matching spots on the fluorographs with those on the Coomassie-stained gel. 8. Determine the intensity of the spots by densitometric scanning (E-C Apparatus Corporation, St Petersburg, FL). 9. Measure the area under the curve for each spot after scanning with a

PAGE 231

214 planimeter (Model 1250, Numonics Corporation, Lansdale, PA). Second-Dimension : Preparation of Glassware. Equipment and Reagents: 1. Soak glass plates (8 cm x 15 cm) in concentrated nitric acid for at least 24 h. 2. Rinse the plates extensively with deionized water, wipe dry with kinwipes, and wipe with 95% ethanol. 3. Also wipe the spacers (1 .25 mm in thickness), rubber bushings, clamps and the gel caster with 95% ethanol. 4. Pair the glass plates, separated by 2 spacers to make a sandwich. 5. Clamp the sandwich and mount it on the gel caster lined with rubber bushings which block the bottom end of the sandwich, to prevent leakage of the gel solution before it polymerizes. 6. Mark the glass plates 3 cm from the top, to serve as the top of the separating gel. Solutions Acrvlamide Monomer Stock Solution (18% T. 5.4% C bi,): 43.8 ml of 40% (w/v) acrylamide 50.0 ml of 2% (w/v) Cbis acrylamide 6.2 ml deionized water to 100 ml. Store at 4 C in a dark bottle.

PAGE 232

215 lEF Tube Gel Solution (4% T. 5.4% ^.) In a 125 ml flask with side arm, add: 3.24 ml acrylamide monomer stock (18% T, 5.4% bis) 8.25 g urea 1.17 ml deionized water 3.0 ml 10%NP-40 0.75 ml ampholine (pH range 3.5-10.0) Swirl to dissolve urea. Degas the solution, then add 15 iJl of 10% (w/v) ammonium persulfate Mix by swirling, and add 11|jlTEMED. Mix by swirling. lEF Gel Sample Buffer (9 M urea. 2% NP-40, 2% ampholine) Into a 50 ml graduated centrifuge tube, add: 2.85 g urea 1.0 ml10%(v/v) NP-40 250 pi ampholine (pH range 3.5-10). Swirl to mix. Deionized water to 5 ml. Aliquot (500 pi) into 1.5 ml bullet tubes and store at 20 C. lEF Anolvte Stock Solution (1 M Phosphoric Acid): 17ml85%(v/v)H3P04 Deionized water to 250 ml.

PAGE 233

216 lEF Anolvte Working Solution 40 ml of anolyte stock solution Deionized water to 4 I. Prepare just before use. lEF Catholyte Stock Solution (1 M Sodium Hydroxide) 10 g sodium hydroxide Deionized water to 250 ml. lEF Catholyte Working Solution 36 ml catholyte stock Dilute to 1.8 I with deionized water. Prepare prior to use in a 2-liter flask with side arm. Degas extensively before use. Gel Overlay Solution (8 M urea) 4.8048 g urea Deionized water to 10 ml. Aliquot (100 pi) in 500 pi bullet tubes and store at 20 C. Monomer Stock Solution (30% T. 2.7% C ^i,) 58.4 g acrylamide 1.6 g bis acrylamide Deionized water 200 ml. Store at 4 C in a dark bottle.

PAGE 234

217 4X Running Gel Buffer (1.5 M Tris-CI. pH 8.8) 36.3 g Tris (pH 8.8) Deionized water to 200 ml. Adjust pH to 8.8 with 1 M HCI. 4X Stacking Gel Buffer (0.5 M Tris-CI, pH 6.8) 3.0 g Tris Deionized water to 50 ml Adjust the pH to 6.8 with 1 M HCI. Sodium Dodecvl Sulfate (10% SDS) lOgSDS Deionized water to 100 ml. Initiator (10% ammonium persulphate) 100 mg ammonium persulfate Deionized water to 1 ml. Prepare just before use. Running Gel Overlay (0.375 M Tris-CI. 0.1% SDS. pH 8.8 25 ml of 4X Running Gel Buffer 1.0 ml 10%(w/v)SDS Deionized water to 100 ml. 2X Treatment Buffer (0.125 M Tris-CI, 4% SDS. 20% glycerol. pH 6.8) 4X stacking gel buffer (2.5 ml) 10% (w/v) SDS (4.0 ml)

PAGE 235

218 Glycerol (1.0 ml) Deionized water 10 ml. Aliquot (1 ml) into 1.5 ml bullet tubes and store at 20 C. Tank Buffer (0.025 M Tris. 0. 1 92 M glycine. 0. 1 % SDS, pH 8.3) 2.5 g Tris 72 g glycine 5gSDS Deionized water to a final volume of 5 I. It is not necessary to check the pH of this solution. Stain (0.1% Coomassie Blue R-250, 40% ethanol, 7% acetic acid) 1 g Coomassie Blue 400 ml 95% (v/v) ethanol (stir to dissolve) 70 ml acetic acid Deionized water to a final volume of 1 liter. Filter before use. Destaininq Solution I (50% ethanol. 10% acetic acid) Ethanol (500 ml) Acetic acid (100 ml) Deionized water to 1 liter. Destaininq Solution II (7% acetic acid, 5% ethanol) Acetic acid (700 ml) Ethanol (500 ml)

PAGE 236

219 Deionized water to 10 liters. Bromophenol Blue (0.5% in 10% ethanol) 50 mg Bromophenol Blue Ethanol to 10 ml.

PAGE 237

APPENDIX 6 ELECTRO-BLOTTING OF PROTEINS TO MEMBRANE Introduction Proteins of interest were blotted to polyvinylidene difluoride (PVDF) membranes according to procedure by Towbin et al. (1979). Procedure 1. Carry out two-dimensional SDS polyacrylamide gel electrophoresis as previously described. 2. Place the gel (equilibrate) in blotting buffer (morpholino-ethane sulfonic acid (MES, pH 6.0) for 20-30 min. 3. Pre-wet a piece of PVDF membrane (same size as gel) in a small volume of 100% methanol for 1-2 sec. 4. Rinse PVDF membrane in deionized water to remove the excess methanol. 5. Place membrane in the blotting buffer to equilibrate for 1 0-20 min prior to use. 6. Equilibrate Whatman™ type 3MM chromatography paper (Whatman Ltd., Maidstone, England) and support foam sponges in blotting buffer prior to use. 7. Arrange the blotting assembly in a sandwich configuration (-) fibre 220

PAGE 238

221 sponge-2 Whatman paper-gel-PVDF membrane-2 Whatman paper-fibre sponge (+), and clamp in sandwich holder. 8. Place the sandwich assembly with the membrane facing the anode in the Bio-Rad Model TRANS-BLOT™ CELL blotting unit (Bio-Rad Laboratories, Richmond, CA). 9. Place assembly in a cold room (4 C) and carry out electrotransfer at a constant voltage (20 V) for 16 h with stirring. Staining and destaining of blots 1. Disassemble the blotting unit and retrieve the PVDF membrane. 2. Place blots in staining solution for 2-5 min. 3. Destain blots in destaining solution (50% (v/v) ethanol, 10% (v/v) acetic acid) for 3-5 min. 4. Wash blots extensively with many changes of distilled water. 5. Place blots between Whatman No. 3 paper to completely air dry. 6. Place dried blots in plastic bags, wrap the bags in aluminium foil, and store at 20 C until further analysis. Solutions Transfer Buffer (10 mM MES Buffer, pH 6.0, 20% Methanol) 1.95 gMES(pH 6.0) 200 ml methanol Make up to 1 liter with distilled water.

PAGE 239

222 Staining Solution (0.1% (w/v) Coomassie Blue R-250. 50% (v/v) ethanol. 7% (v/v) acetic acid) 1 g Coomassie Brilliant Blue R-250 500 ml 100% ethanol 70 ml glacial acetic acid Make up to 1 liter with distilled water. Destaininq Solution (50% ethanol. 10% acetic acid) 500 ml 100% ethanol 100 ml glacial acetic acid Make up to 1 liter with distilled water.

PAGE 240

APPENDIX? DETERMINATION OF TOTAL PROTEIN IN RETENTATE BY METHOD OF LOWRY Procedure 1. Add samples and standards to appropriate tubes (200 pi/tube). If sample volume is less than 200 pi, make up to that volume with double-distilled water. 2. Add 3.0 ml of solution I (see below) to each tube, mix briefly and let stand at room temperature for 1 5 min. 3. Add 0.3 ml of solution II (see below) to each test tube, mix immediately and let stand for 35 min at room temperature. 4. Turn on spectrophotometer (use red filter). 5. Read optical density at 750 nm under UV light. 6. Plot a standard curve of optical density versus concentration. Read off concentration of unknown samples from the standard curve. Solutions Protein Standards The standards used were prepared fresh by diluting bovine serum albumin (BSA) in water to obtain: (blank), 5, 1 0, 20, 50, 1 00, 200 pg BSA. 223

PAGE 241

224 Stock Solutions Solution A 2%Na2C03 0.1 NNaOH 20 g NajCOa 4.0gNaOH Dissolve in double-distilled water to 1 I. Solution B 1%CuS04.5H20 1 g dissolved to 100 ml with double-distilled water Solution C 2% Sodium Potassium Tartrate 2 g dissolved to 100 ml with water. Working Solutions Mix equal volumes of solutions B and C and let stand at room temperature for 10 min. Solution I Add 50 ml of solution A to 1 ml of the mixture of solutions B and C, mix immediately and let stand for 15 min. Solution II Dilute Folin Reagent (2 N, Sigma Chemical Co.) 1;1 with distilled water to obtain 1 N Folin Reagent.

PAGE 242

APPENDIX 8 MEASUREMENT OF APO A-1 MRNA Preparation of Apo A-1 Plasmid DNA Apo A-1 plasmid DNA in E coii was purchased from ATCC. The plasmid DNA was in lyophilized form with the following specifications: freeze-dried E coli containing plasmid medium: 1273, LB plus tetracycline insert contains Apo A-1 cloned from adult human liver detects sequence: human Apo A-1 11 q23-q24 insert size: 0.6 KBtotal size: 4.4 KB source of insert DNA: cDNA name of vector: pKT218 insert site: Pst1 insert ends: Psti excise with: Psti markers: tet'' sequence corresponds to amino acids 94-243, contains 3' untranslated sequence plus poly (A) tail. 225

PAGE 243

226 Procedure A. Reconstitution 1. Add 200 pi sterile TE buffer (pH 7.4 or pH 8) to the vial containing the lyophilized plasmid DNA (stock). Store the stock at -20 C for long-term, or at 80 C for very long-term storage. B. Cultivation of Plasmid DNA 1. Take out 50 |jI transformed bacteria competent cells (JM 109, Promega Corporation, Madison, Wl) from the stock (usually stored at 80 C) to thaw on ice. Return stock transformed cells to 80 C immediately after use. 2. Place 5 |jl of reconstituted plasmid DNA in a snap cap tube and placed on ice for 30 min. 3. In the meantime, take out LB-tet'^ plate from 20 C and place in incubator upside down at 37 C to pre-warm. Label the plates. 4. Heat shock the mixture in 2 by placing in a water bath at 42 C for 90 sec. 5. Place mixture on ice for 2 min. 6. Add 100 \j\ SOC medium (see below) and place in an incubator with a shaker at 37 C for 1 h. 7. Take out mixture from incubator and transfer to a laminar flow bacteria hood. 8. Turn on the blower and light in the hood. 9. Then take out the tet*^ plate from the incubator to the hood.

PAGE 244

227 10. Carefully open the plate under the hood, and in a single motion dump the DNA mixture from # 7 onto the plate. 11. Spread out the DNA-competent cell mixture on the plate using the sides of a sterile autopipet tip (large). 12. Leave the plate half-covered and allow to air dry for about 5 min in the hood. 13. Take plate out of hood and place in the incubator (37 C) upside down, overnight. Turn off blower and light in hood. Isolation of plasmid DNA 14. Next day, take out plate from incubator and examine for presence of colonies. 15. Seal off ends of the plates with parafilm and store at 4 C until ready for use. 16. With use of an autopipet, pick one colony from the plate and add to 200 ml sterile LB medium (see below) in a sterilized 500 ml culture bottle. Add 75 |jl of tetracycline stock (10 mg tetracycline in 50% ethanol, 50% water) for every 50 ml of LB medium i.e. add 300 pi tetracycline stock for 200 ml LB medium. For ampicillin resistant clones, use 20 mg ampicillin (Na salt) for 200 ml LB medium (weigh 20 mg ampillicin and add directly to LB prior to use. 17. Cover the bottle loosely, place in a styrofoam box padded with paper towels, and place assembly in an incubator at 37 C overnight with

PAGE 245

228 shaking. 18. Next day, take out 1 ml of the overnight culture of E coli containing plasmid DNA and put in a snap cap tube. Add 1 ml freeze broth to tube and place in incubator (37 C) on a shaker overnight to grow. Freeze broth is usually stored at -80 C as a stock in case it is needed in future. 1 9. Pour out the rest (1 99 ml) of the overnight culture in # 1 8 into a v\/idemouthed 250 ml plastic bottle with a screw cap. 20. Centrifuge mixture at 4000 x g (Sorvall model RC-5B, HS-4 rotor head) for 15minat4C. 21. Decant supernatant back into the 500 ml culture bottle that was used in step #16, add bleach and let sit for 10 min before pouring down the drain. Save the pellet (containing the plasmid DNA) for further processing. 22. From this step, the plasmid DNA was prepared using QIAGEN-tip 1 00 plasmid kit (Qiagen Inc., Chatsworth, CA) midi protocol as follows: 23. Add 4 ml of buffer P1 (resuspension buffer) to the pellet obtained in step # 21 Use a sterile Pasteur pipet to squirt the P1 buffer over the pellet to obtain a suspension. 24. Tranfer the suspension to a smaller plastic tube with screw cap, using a sterile transfer pipet. 25. Add 4 ml buffer P2 (lysis buffer), mix by vortexing and incubate at room temperature for 5 min. 26. Add 4 ml chilled buffer P3 (neutralization buffer), mix, and incubate on ice

PAGE 246

229 for 15 min. 27. Centrifuge at 4 C for 1 5 min at 47,800 x g (Sorvali model RC-5B, ss 34 rotor head). 28. Equilibrate a QIAGEN-tip 1 00 column with 4 ml of buffer QBT (equilibration buffer). Let the column suspend on holders over the 250 ml plastic bottle used in step # 1 9. 29. Apply the supernatant from step # 27 onto the QIAGEN-tip 100 column. 30. Wash the QIAGEN-tip 1 00 twice with 1 ml of buffer QC (wash buffer). Discard eluate. 31. Transfer and place the column and holder over a sterile hard plastic tube. 32. Elute the DNA bound to the column with 5 ml of buffer QF (elution buffer). Save the eluate. 33. Add 3.5 ml (0.7 volumes of reconstituted plasmid DNA started with) isopropanol to precipitate the DNA. 34. Centrifuge at 47,800 x g for 45 min (Sorvali ss 34 rotor head) at 4 C. 35. Discard the supernatant and carefully look for an opaque spot around the curved region of glass tube. That opaque spot is the DNA pellet. 36. Add 5 ml of cold 70% ethanol (stored at -20 C) to the pellet to wash the DNA, centrifuge at 47,800 x g (Sorvali ss 34 rotor head) for 10 min. 37. Place the tube face down on paper towels to drain and air dry for 5 min. 38. Redissolve DNA pellet in 50 \j\ of Tris-EDTA buffer (see below). This is the purified plasmid DNA stock. Use some of the plasmid DNA to

PAGE 247

230 determine concentration and size of the plasmid. 39. Store the rest of the plasmid DNA at 4C until further analysis. Measurement of concentration of Apo A-1 plasmid DNA 1. Pipet 995 pi of sterile water into a 12 x 75 mm test tube. 2. Add 5 pi of purified plasmid DNA to the tube. Vortex to mix. 3. Measure optical density of mixture at 260 nm on a spectrophotometer. 4. Calculate DNA concentration. Digestion (lysis) of Apo A-1 plasmid DNA The plasmid DNA is digested in order to release the Apo A-1 DNA insert from the plasmid. Procedure 1. Pipet 2 pi of purified plasmid DNA into a 0.5 ml bullet tube. 2. Add 3 pi Pst1 buffer, 22 pi sterile water, and 3 pi Psti enzyme, to obtain a final volume of 30 pi. Pipet up and down to mix. 3. Incubate mixture in a water bath at 37 C overnight. 4. Separate plasmid digest by electrophoresis. Separation of Plasmid Digest bv Agarose Gel Electrophoresis Agarose gel Preparation 1. Dissolve 0.6 g agarose in 50 ml TAE buffer (see below) (1 .2 % gel). 2. Heat to a boil in a microwave until agarose is completely dissolved. 3. Add ethidium bromide (4 pi) to the gel mixture, mix, allow gel to cool, and then pour onto a mini gel casting tray with comb. Pour from end of plate

PAGE 248

231 away from the comb. Avoid air bubbles in the gel. 4. Allow gel to set, and in the meantime, assemble mini gel apparatus (small horizontal gel system, Fisher Biotech Electrophoresis Systems, Fisher Scientific, Pittsburgh, PA). 5. When set, place the gel tray in the tank containing 1X TAE buffer. 6. Gently pull out the comb, take off the rubber bungs and pour more 1X TAE buffer to cover gel. Loading of Samples and Electrophoresis Preparation of Lamda Markers Pipet 20 pi TE buffer 4 pi lamda markers 5 pi DNA loading dye, in a bullet tube. Place mixture in a water bath (62 C) for 2-5 min. Preparation of Samples Add 5 pi of DNA loading dye to sample digest. Mix well using autopipet. Electrophoresis 1. Load lamda markers in first lane, and plasmid digest on subsequent lanes. 2. Cover the electrophoresis tank, and set voltage between 81-92 V (E-C Apparatus Corporation, St. Petersburg, PL) with voltage setting range set at 'low'.

PAGE 249

232 3. Run electrophoresis until dye front is about midway in gel. The wells should be at the black terminal of the tank cover. 4. Turn off power, take out gel and examine under UV light. 6. Compare migration of sample with that of DNA markers to determine if insert is of the right size. Take a picture of the gel. 6. If insert is of the right size, carry out a digestion of plasmid DNA on a larger scale, separate digests on agarose gel as before, and tranfer inserts by electroblotting from the gel to DEAE paper. Solutions SOC Medium Bacto tryptone (2 %) Bacto yeast extract (0.5%) Sodium chloride (10 mM) Potassium chloride (2.5 mM) Magnesium chloride (10 mM) Magnesium sulfate (10 mM) Glucose (20 mM) Combine tryptone, yeast extract, NaCI, and KCI in sterile water and autoclave 30-40 min. Make a 2 M stock of Mg^*, comprised of 1 M MgClzand 1 M MgS04. Filter sterilize (22 p membrane). Prepare a 2 M stock of glucose similarly and store at 20 C.

PAGE 250

233 Prior to use, combine the media with Mg and glucose and filter sterilize. Freeze Broth 22 g Glycerol 5 g Tryptone 2.5 g Yeast Extract 2.5gNaCI S.ISgKjHPO^ 0.9 g KH2PO4 0.23 g Sodium citrate 0.45 g (NH4)2S04 Dilute to 500 ml, and autoclave. Add 0.25 ml 1 M MgSO^ (filter-sterile). LB-medium, 1 liter : 1 g Tryptone 5 g Yeast Extract 5gNaCI pHto7.5with10MNaOH Dilute to 1 I with water Autoclave and store at 4 C LB Agar plates LB medium + 15 g Bacto-agar Autoclave

PAGE 251

234 When cool, add 100 mg ampicillin (or tetracycline), plus 1 ml X-Gal (20 mg/ml in NN dimethyl formamide). TE buffer. pH 8.0 10mMTris-HCI 1 mM EDTA

PAGE 252

APPENDIX 9 MEASUREMENT OF APO E MRNA Preparation of Apo E Plasmid DNA Apo E plasmid DNA in E coli was purchased from ATCC. The plasmid DNA was in lyophilized form with the following specifications: freeze-dried E coli containing plasmid medium: 1273, LB plus tetracycline insert contains Apo E cloned from adult human liver detects sequence: human Apo E, 1 1 q23-q24 insert size: 0.6 KBtotal size: 4.4 KB source of insert DNA: cDNA name of vector: pKT21 8 insert site (s): Pst1 insert size: 0.9 kb excise with: Pst1 markers: tet'^ sequence encodes amino acids 81-299, and contains the 158 bp 3' untranslated region and 44 bp poly (A). The insert has 5 internal Pst1 sites. 235

PAGE 253

236 Procedure Reconstitution 1. Add 200 |jl sterile TE buffer (pH 7.4 or pH 8) to the vial containing the lyophilized Apo E plasmid DNA (stock). 2. Store the stock at -20 C for long-term, or at 80 C for very long-term storage. Following reconstitution, Apo E plasmid DNA was cultivated, isolated, digested with Pst I enzyme, and separated on agarose gel, following the same protocol as with Apo A-1 Apo E plasmid DNA was stored at 4 C until further analysis.

PAGE 254

APPENDIX 10 MEASUREMENT OF TIMP-1 AND TIMP-2 MRNA TIMP-1 and TIMP-2 plasmid cDNAs were a kind gift from Dr. Michael F. Smith of the University of Missouri, Columbia, MO. The plasmid DNAs were sent in the form of agar stabs in snap cap tubes. Ovine TIMP-1 900 base pair TIMP-1 cDNA (clone 6-2) cloned into EcoRI/Xhol site of PBIuescript SK linearize with BamHI and transcribe from 11 promoter to generate antisense cRNA linearize with Kpnl and transcribe off T3 promoter to generate sense cRNA plasmid carries ampicillin resistance gene. Ovine TIMP-2 438 base pair TIMP-2 cDNA (MMI) cloned into BamHI/EcoRI site of PBIuescript SK linearize with Xbal and transcribe off 11 promoter to generate antisense cRNA linearize with Kpn and transcribe off T3 promoter to generate sense cRNAplasmid carries ampicillin resistance gene. 237

PAGE 255

238 Growth of Bacterial Cultures 1. Add 1 ml of LB-amp (ampicillin at 100 mg/l) to the agar stabs of TIMP-1 and TIMP-2 plasmids. 2. Incubate tubes at 37 C overnight with shaking. 3. Next day, pipet 50 |jl of overnight culture and plate onto pre-warmed LBamp-Gal plates in a bacteria hood, as described earlier. 4. Incubate the plates at 37 C overnight. Purification of DNA Isolation of Plasmid DNA Introduction Isolation of plasmid DNA is performed in essentially three stages. The bacterial cell wall is first weakened by the action of lysozyme, and then lysed by EDTA and a detergent at high pH. Finally, the insoluble cell debris consisting of genomic DNA and protein is precipitated with high salt and centrifuged down, leaving the plasmid DNA in solution. Small-scale isolation of plasmid DNA: Miniprep Procedure. 1. Pick a single bacteria colony from each LB-amp plate and suspend in 2 ml of LB-broth containing ampicillin at a final concentration of 100|jg/ml. Incubate at 37 C overnight with vigorous shaking in an air incubator.

PAGE 256

239 2. Place 1 .5 ml of the overnight culture into a microcentrifuge tube and centrifuge at 12,000 x g for 1 min. The remainder of the overnight culture can be stored at 4 C. 3. Discard the supernatant and resuspend the pellet by vortexing in 100 ul solution I (see below). 4. Incubate for 5 min at room temperature. 5. Add 200 pi of a freshly prepared solution II (see below). Mix by Inversion (DO NOT VORTEX). 6. Incubate for 5 min at room temperature. 7. Add 150 pi of 3 M sodium acetate (pH 5.0). Mix by inversion. 8. Incubate for 5 min at room temperature. 9. Centrifuge at 1 2,000 x g for 5 min. 10. Transfer the supernatant to a fresh tube, avoiding the white precipitate. 1 1 Add 200 pi of water-saturated phenol and 1 00 pi of chloroform: isoamy I alcohol (24:1). Vortex and centrifuge at 12,000 x g for 5 min. 12. Transfer the upper, aqueous phase to a fresh tube and add 2.5 volumes of ethanol. Mix and allow the samples to precipitate for 1 h at -80 C. 13. Centrifuge at 12,000 x g for 15 min. Remove the supernatant and wash the pellet with 200 pi of prechilled 70% ethanol. Dry pellet

PAGE 257

240 under vacuum for 5 to 1 min. 14. Dissolve the dried pellet in 25 pi of TE buffer. 15. Add 3 |jl of appropriate endonuclease buffer and 2 |j| of enzyme to the DNA solution. Incubate for 16 h in a 37 C water bath. 16. Centrifuge briefly and add 5 |jI of RNase A to each tube. Incubate 'a for 30 min at 37 C. 17. Add 5 |jl of DNA dye mix and separate the DNA fragments by electrophoresis on an agarose gel. Solutions Solution I. pH 8.0 : 25mMTris-HCI, pH8.0 i lOmMEDTA 50 mM glucose Solution II : 0.2 N NaOH 1%SDS 3 M Sodium acetate. pH 5.2 : / ^ Dissolve 246 g sodium acetate in 800 ml HjO Adjust the pH to 5.2 vi'ith glacial acetic acid Add H2O to 1 I Water-saturated phenol : Melt the phenol at 65 C

PAGE 258

241 Add an equal volume of H2O and mix vigorously Allow the phases to separate overnight at 4 C Large-scale isolation of plasmid DNA Procedure 1. Inoculate 200 ml of LB-medium containing 0.1 mg/ml of ampicillin with 100 ul of stock plasmid (glycerol stock or miniprep culture). 2. Incubate overnight at 37 C with vigorous shaking. 3. Transfer the medium to 250 ml centrifuge tubes and centrifuge for 15 min at 4000 x g (HS-4 rotor). 4. Discard the supernatant and resuspend the pellet in 3 ml of ST buffer. Pipet up and down to dissolve the pellet. Transfer to 50 ml polypropylene high speed tubes. 5. Add 1 ml of lysozyme (10 mg/ml in ST buffer) to each tube and vortex gently. 6. Incubate for 1 min on ice. 7. Add 2.5 ml of 0.2 M EDTA, vortex gently and let sit on ice for 5 min. 8. Add 6 ml of triton-lysis buffer and cap the tubes. Mix by inversion and let sit on ice for 8 min. 9. Invert again and let sit for another 8 min. 10. Centrifuge for 20 min at 47,800 x g (SS-34 Rotor). 11. Transfer the supernatant to a fresh tube, add 400 pi of 8 M

PAGE 259

242 ammonium acetate, and vortex gently. 12. Add 12 ml of phenol and 4 ml of chloroform-isoamyl alcohol (24:1), cap and mix thoroughly by shaking. 1 3. Centrifuge for 5 min at 4000 x g. 14. Transfer the aqueous layer to a fresh tube and repeat steps # 12 and #13. 15. Transfer the aqueous phase to a 50 ml disposable tube. Add 0.1 vol of 8 M ammonium acetate to each tube followed by 2 volumes of absolute ethanol. 16. Mix by inverting a few times and incubate at -80 ''C for 1 h. 17. Remove the tubes from the deep freezer, let the samples thaw at RT and centrifuge for 20 min at 4000 rpm. 18. Discard the supernatant and rinse the pellet with 3 ml of 70% ethanol by centrifuging at 4000 rpm for 10 min. 19. Discard the supernatant and invert to drip-dry (approx. 10 min). 20. Resuspend the DNA pellet in 300 pi of TE buffer. 21. Add 20 (jl RNase (20 mg/ml) and incubate in a 37 X water bath for 30 min. 22. Add 150 pi of water-saturated phenol and 75 pi chloroform: isoamyl alcohol (24:1 ) to the tube. Mix and centrifuge at 12,000 g for 5 min. 23. Transfer the supernatant to a fresh tube and add 0.1 volume of 8 M ammonium acetate and 2.5 volume of ethanol. Allow sample to

PAGE 260

243 precipitate for 1 h at -80 C. 24. Centrifuge at 12,000 g for 15 min. Remove the supernatant and wash the pellet with 200 |jl of prechilled 70% ethanol. Dry the pellet for 5 to 10 min under vacuum. 25. Resuspend the pellet (TIMP-1 or TiMP-2 plasmid DNA) in 300 pi of TE buffer and store at 4C. 26. Determine concentration of the plasmid DNA. Solutions ST Buffer. pH 8.0 : 50 mM Tris 25 % sucrose Triton lysis buffer, pH 8.0 : 1 ml 10%Triton-X100 5 ml 1 M Tris, pH 8.0 62.5 ul 0.1 MEDTA, pH 8.0 HjOtolOOm! RNase : 20 mg/ml of RNAse A Boil for 15 min. 8 M Ammonium acetate : Add 61.6 g ammonium acetate H2Oto100ml

PAGE 261

244 Digestion of TIMP-1 and TIMP-2 Plasmid DNA Procedure TIMP-1 and TIMP-2 plasmids were digested according to the following double digestion protocol: For TIMP-1 add: For TIMP-2, add: 24 [j\ plasmid DNA 24 pi plasmid DNA 3 pi 1 0X high buffer (H) 3 pi 1 0X buffer (B) 3 pi Xhol enzyme 3 pi BamHI enzyme Mix by pipeting and place in a water bath at 37 C overnight. Next day, spot spin and add; 15 pi sterile water 15 pi sterile water 2 pi 1 0X high buffer (H) 2 pi 1 0X high buffer (H) 3 pi EcoRI enzyme 3 pi EcoRI enzyme Mix and place in water bath at 37 C overnight. Spot spin tubes and add 5 pi low RNase to all tubes. Incubate in water bath at 37 C for 30 min. Spot spin plasmid digests, add 5 pi DNA loading dye to each tube and load on a 1 .2 % agarose gel.

PAGE 262

245 Electrophoresis of Nucleic Acids Introduction Smalland medium-sized nucleic acids are best separated using polyacrylamide gel electrophoresis while larger molecules are separated on gels of agarose which have the largest pore size. Agarose Gels Preparation of Gel 1. Assemble the gel mold. 2. Add the weighed amount of agarose to the volume of buffer (1X TAE) needed to fill the mold. 3. Heat in a microwave oven until the agarose has dissolved. 4. Cool the agarose solution to about 50 C, add ethidium bromide and swirl to mix, then pour into the mold and immediately place the comb in position. 5. When the gel has completely cooled and set (30 min), remove the comb and place the gel in the electrophoresis apparatus. 6. Add sufficient buffer to fill the electrode chamber and cover the gel to a depth of about 1 mm. 7. Load the samples and electrophorese at appropriate voltage. 1 0X Loading Buffer 50% (v/v) glycerol 0.5% (w/v) bromophenol blue

PAGE 263

246 0.5% (w/v) xylene cyanol SOX TAE Buffer 242 g Tris base 57.1 ml Glacial acetic acid 100ml0.5MEDTApH8.0 Dilute to 1 I Ethidium Bromide 10 mg/ml To prestain gel, add 1 pi /20 ml gel mixture. To prestain gel buffer, add 10 pi to 3 I. Recovery of DNA from Agarose Gels onto DEAE Membrane Introduction An incision is made in the gel just ahead of the band of interest and a strip of DEAE paper put in place. Electrophoresis is resumed until all of ethidium-staining material has transferred from the gel to the paper. The DNA is eluted with a high salt buffer. Preparation of DEAE Membrane 1. Cut the DEAE membrane into strips as long as your band and 6 mm wide. 2. Soakin 10mM EDTApH 7.6for 10min. 3. Pour off EDTA solution and replace it with 0.5 M NaOH for 5 min. 4. Wash 5 times with double-distilled water. Membrane can be

PAGE 264

247 prepared in advance and stored in water at 4 C for later use. Transfer Procedure 1. Prepare and run the agarose gel in a UV-transparent mold. 2. Identify the band(s) of interest and make an incision in the agarose on the anode side of the band using a scalpel blade. Open the slit in the gel and slide in a piece of prewetted DEAE paper. 3. Resume electrophoresis and monitor DNA migration by observing with the transluminator every few minutes. This is critical If several bands are closely spaced and unique fragments are to be isolated. The fluorescent band will disappear from the gel and be adsorbed onto the paper. 4. Place the DEAE paper to which DNA is bound in a 1 .5 ml microcentrifuge tube and add 500 ul of high salt buffer to cover the area of paper where DNA is bound. 5. Incubate for 20 min in a 65 C water bath. 6. Transfer the supernatant containing the DNA to a fresh 1 .5 ml tube. 7. Add 500 ul of high salt buffer to the tube with DEAE paper and incubate for another 20 min at 65 C. 8. Transfer the supernatant to a fresh 1 .5 tube and extract with 250 ul of water-saturated phenol and 125 ul of chloroform-isoamyl alcohol (24:1). 9. Centrifuge for 4 min and transfer the aqueous layer to a fresh 1 .5

PAGE 265

*v 248 ml tube. 10. Add 0.1 vol of 8 M ammonium acetate and 2 volumes of absolute ethanol and incubate at -80 C for 1 h. 1 1 Remove the tubes from the deep freezer, let sit at RT for a few minutes, and centrifuge for 10-15 min. 12. Discard the supernatant and rinse the pellet with 200 ul of 70% ethanol (4 min). 13. Decant ethanol, blot on paper towel and dry in speedvac for approx. 10 min. 14. Resuspend the DNA insert in 20 ul of TE buffer and combine the fractions containing the same insert. 15. Electrophorese 5 ul DNA solution on a mini gel to assess the ( amount of DNA recovered. Solution High Salt Buffer, pH 7.4 : 20 mM Tris pH 8.0 1 M NaCI 0.1 mMEDTA This is the final concentration of the buffer.

PAGE 266

APPENDIX 11 ISOLATION AND PURIFICATION OF RNA Introduction The most important consideration in the preparation of RNA is to rapidly and efficiently inhibit the endogenous ribonucleases which are present in virtually all living cells. Another important concern in the preparation of RNA is to avoid accidental introduction of trace amounts of ribonucleases from hands, glassware and solutions. It is therefore imperative to: 1 use gloves throughout all procedures. 2. autoclave reagents that need to be sterile. 3. use only sterile tubes, pipettes, tips and glassware. Isolation of Total Cellular RNA from Luteal Tissue Procedure 1. Label 50 ml conical centrifuge tubes and let sit on ice. 2. Add 10 ml of cold guanidinium thiocyanate to each tube. 3. Add 1 ml of 2 M sodium acetate (pH 5.0) to each tube and mix gently. 4. Add 1 g of fresh or frozen tissue (cut into pieces) to each tube. 5. Add 78 ul of 2-mercaptoethanol to each tube. 249

PAGE 267

250 6. Homogenize tissue with a polytron tissue homogenizer for three to four 5 sec bursts. Keep tissue on ice during homogenization. Rinse the polytron head between samples. 7. Add 10 ml of water-saturated phenol and 2 ml of chloroform: isoamyl alcohol (24:1). Cap and mix by shaking. 8. Centrifuge at 4000 rpm for 15 min. 9. Transfer upper phase to a fresh 50 ml tube. Be careful not to take any of the bottom layer. 10. Add 10 ml of isopropanol to precipitate RNA. Mix gently and place in a -SOX freezer for 1 h. 11. Centrifuge at 4000 rpm for 10 min and discard the supernatant. 12. Resuspend the pellet in 2 ml of 4 M LiCI. Dislodge RNA pellet from bottom of the tube using a sterile transfer pipette. RNA does not go into solution. 13. Centrifuge at 4000 rpm for 1 0-1 5 min to repellet. 14. Pour off the supernatant and blot on a clear paper towel. 15. Resuspend the pellet in 2 ml of RNA buffer (10 mM Tris, 1 mM EDTA, 0.5% SDS, pH 7.5) and let sit at RT until RNA goes into solution. It takes about 15-40 min to go into solution. 16. Transfer content of the tube into a 15 ml conical tube. 17. Add 2 ml of chloroform and vortex for 15 sec. 18. Centrifuge at 4000 rpm for 1 min.

PAGE 268

251 19. Collect upper phase with sterile pipet and transfer into a fresh 15 ml conical tube. 20. Add 200 ul of 2 M sodium acetate (pH 5.0) and mix gently. 21. Add 2 ml of isopropanol and mix. 22. Store overnight in a 80 C freezer. 23. Centrifuge at 4000 rpm for 30 min. Insoluble pellet should contain pure RNA after centrifugation. 24. Pour off supernatant, blot and add approximately 1 ml of cold 70% ethanol to wash pellet. Recentrifuge at 4000 rpm for 10 min. Discard supernatant and blot on adsorbent paper. Air dry 12-15 min. 25. Resuspend the pellet in 100-200 ul sterile H2O. Let sit at RT for 10 min or until it dissolves. Mix gently and transfer to 1 .5 ml bullet tubes. 26. Pipet 5 ul of sample to a 1 2 X 75 test tube. 27. Add 995 ul sterile water to 5 ul sample and read absorbance at 260/280 nm on a spectrophotometer (Milton Roy Spectronic 21 D, Fisher Scientific, Atlanta, GA). ** To calculate ug RNA use the following equation: O.D. 260 ( ) X Dilution ( ) X 33 ug/ml X Total sample volume = Total ug RNA. At 260 nm, RNA has a refractive index of 33 (ie 33 pg for each OD

PAGE 269

252 reading). DNA has a refractive index of 40. ** To determine RNA purity: calculate 260:280 ratio. a. Ratios between 1.8 2.0 are good. b. Ratios less than 1.2 may indicate substantial protein contamination. Solutions for RNA Extraction 4 M Guanidinium thiocvnate : 4 M guanidinium thiocynate (236.4 g) 25 mM sodium citrate (3.7 g) 0.5 % (w/v) sarkosyl (2.5 g) pH to 7.0 Distilled water to 500 ml Autoclave and store in dark bottle at 4 C 4MLiCI : Add 42.4 g LiCI Distilled HjO to 250 ml Autoclave and store at 4 C RNA buffer, pH 7.5 : lOmMTris-HCI, pH 7.5 1 mM EDTA 0.5% (w/v) SDS Autoclave and store at -20 C

PAGE 270

253 2 M Sodium acetate, pH 5.0 : 16.4 g Sodium acetate Add 80 ml of acetic acid pH to 4.0 (difficult to get pH to 4) Add water to 1 00 ml Autoclave and store at -20 C Water Saturated Phenol Melt phenol at 65 C. Add an equal volume of distilled water. Store at 4 C. The phenol settles at the bottom.

PAGE 271

APPENDIX 12 NUCLEIC ACID LABELLING Nick Translation Introduction E. coli DNA polymerase I adds nucleotide residues to the 3' hydroxyl terminus that is created when one strand of a double-stranded DNA molecule is nicked. In addition, by virtue of its 5'-> 3' exonucleolytic activity, the enzyme can remove nucleotides from the 5' side of the nick. The simultaneous elimination of nucleotides from the 5' side and the addition of nucleotides to the 3' side results in movement of the nick along the DNA. Procedure 1. Add the following components to a 500 ul Eppendorf tube: DNA to be labeled 500 ng Nucleotide mix buffer (kit) 20 ul [a-^2p]dCTP 5 ul Enzyme (kit) 10 ul H2O (to a total volume of 100 ul) 254

PAGE 272

255 2. Mix lightly and incubate at 14 C for 1 h. 3. Add an equal volume of phenol (100 ul) and vortex. 4. Run the entire fraction over a Sephadex G-50 column which has been equilibrated in TE buffer. 5. Check radioactivity of fractions v^^ith a Geiger counter. ] i, 6. Save fractions containing the first peak of radioactivity for hybridization. <

PAGE 273

APPENDIX 13 NORTHERN BLOTTING AND HYBRIDIZATION Introduction RNA is separated according to size by electrophoresis through a denaturing agarose gel and is then transferred to nitrocellulose or nylon membrane. The RNA of interest is then located by hybridization with radiolabeled DNA or RNA followed by autoradiography. A. RNA Denaturation 1 Aliquot total RNA (1 0-30 ug) into 0.5 ml tubes. 2. Lyophilize samples in speed vac for 15 to 20 min. 3. Add 1 5 ul sample denaturation buffer to each tube. 4. Let samples sit on ice for at least one hour. While denaturing the samples, set up 1.5 % agarose gel as described in step B. B. Setting up a 1.5% Agarose Gel 1, Weigh out 3.0 g of high melt agarose and place in a 500 ml glass beaker. 2, In a graduated cylinder, mix 8 ml running buffer (25X), 158 ml distilled water and add this to the beaker with agarose. 3, Heat the solution in a microwave for 1 to 2 min. Repeat this step 256

PAGE 274

257 until agarose is completely dissolved. 4. Add 32.1 ml 37% formaldehyde to the agarose solution and mix. 5. When agarose is sufficiently cool, pour into the gel holder. Make sure there are no air bubbles within the gel or near the comb. 6. Allow the gel to polymerize for at least 1 h in the hood. C. Loading and Running RNA samples 1. Pour the running buffer {1X) in electrophoresis unit and place the gel in the caster. 2. Carefully remove the comb and add enough buffer to cover the gel. 3. Turn on the pump and get all air bubbles out of the line and then turn off. 4. Heat denature RNA samples at 65 C for 1 5 min. 5. Cool quickly on ice and add 5 ul of loading buffer. 6. With the pump turned off, add samples to the wells. The circulation unit should be connected so that the buffer runs from the anode (black) to the cathode (red). 7. Electrophorese samples at 100 V until the dye front is in the gel (20 min). 8. Turn on the pump and reduce the voltage to 24 V to run overnight. 9. The next morning, with the pump still running, add 10-15 ul of ethidium bromide (10 mg/ml) and let stain for 30 to 60 min. 10. Turn the pump off and check RNA quality on UV light box. The 28s

PAGE 275

>-• 258 and 18s ribosomal RNA bands should be intact. D. Northern Blotting RNA transfer was done using the TurboBlotter and Blotting Assembly. 1. RNA gels run rinsed in 2.2 M formaldehyde should be rinsed four times in deionized water. 2. Measure the gel and cut a piece of Nylon filter paper to the exact size, using a pair of sharp scissors. Soak transfer membrane in distilled water for 1 5 min. 3. Pour off water and soak the membrane in 20X SSC. Cut another piece of Whatman paper to the same size as the nylon membrane. 4. Place "stack tray" of transfer device on bench, making sure it is level. 5. Place 20 sheets of dry GB004 blotting paper (thick) in stack tray. 6. Place 4 sheets of dryGB002 blotting paper (thin) on top stack. 7. Place one sheet of GB002 blotting paper, prewet in transfer buffer on stack. 8. Place transfer membrane on stack. 9. Cover the membrane with agarose gel, cut the gel to the size of the membrane, making sure there are no air bubbles between the gel and the membrane. 10. Wet the top of the gel with transfer buffer and place 3 sheets of GB002 blotting paper, presoaked in transfer buffer on top of the

PAGE 276

259 gel. 1 1 Attach the "buffer tray" of the transfer device to the bottom tray using the circular allignment buttons to align both trays. 12. Fill the buffer tray with transfer buffer. 13. Start transfer by connecting the gel stack with the buffer tray using the pre-cut "buffer wick" (included in each blotter stack), presoaked in transfer buffer. Place the wick across the stack so that the short dimension of the wick completely covers the blotting stack and both ends of the long dimension extend into the buffer tray. Place the "wick cover" on top of the stack to prevent evaporation. Make sure the edges of the wick are immersed in the transfer buffer. 14. Continue the transfer for at least 5 h or overnight. 15. Dismantle the blot and check the efficiency of RNA transfer on a UV light box. Mark the position of the wells and ribosomal RNA bands on the nylon membrane with a blunt pencil. 10. To immobilize the RNA, place the nylon membrane between two sheets of Whatman paper and bake at 80 C for 2 h in a vacuum oven, or expose blot to ultraviolet light on a light box for 90 sec. The membrane can be prehybridized immediately or stored dry in a cool place.

PAGE 277

260 E. Hybridization and Autoradiography 1. Prehybridize the nylon membrane for 1 to 2 h at 42 C in prehybridization buffer (see below). 2. Discard the prehybridization solution and add the hybridization buffer (see below). 3. Add the denatured radiolabelled probe (boiled for 5 min) directly to the hybridization solution and continue the incubation for 24 h. 4. Wash the membrane twice (30 and 1 5 min) at 42 C in 2 X SSC, 0.1% SDS, followed by two washes of 15 min each at 42 C in 0.1 XSSC, 0.1%SDS. 5. Establish an autoradiograph by exposing the membrane for 24-48 h to X-ray film (Dupont, ) at 70 C with an intensifying screen. Solutions 25X Running Buffer, pH 7.0 Make 0.5 M stock solutions of mono and dibasic phosphate: A: 60 g NaH2P04 (mono, FW=120.0) B: 71 g Na2HP04 (di, FW=141.96) Mix 39 ml of solution A and 61 ml of solution B HjO to 200 ml. Autoclave and store at room temperature.

PAGE 278

261 10X Denaturing Buffer, pH 7.0 6.76 g Hepes (FW=238.3) 0.68 g Sodium citrate (FW=1 36. 1 ) 0.37 g EDTA HjOtolOOmI Autoclave and store at -20 C wrapped in foil Sample Denaturing Buffer, pH 7.5 0.72 ml 10X denaturing buffer 3.40 ml 90% formamide 1.07 ml 37% formaldehyde 0.81 ml H2O Filter-sterilize and store at -20 "C Loading Buffer 0.4 ml 25X runnung buffer (20 mM) 5.0 ml glycerol (50%) 4.6 ml H2O 5.0 mg bromophenol blue (0.05% w/v) 20X SSC, pH 7.0 3 M Sodium chloride (175 g/l) 0.3 M Sodium citrate (88 g/l) pH to 7.0 with 1 MHCI

PAGE 279

262 SOX Denhardt Solution, 500 ml 5 g FIcoll 5 g Polyninylpyrrolidone SgBSA Filter-sterilize and store at -20 C Prehybridization Buffer, 50 ml 5.0 ml 50X Denhardt solution 10.0ml20XSSC 4.0 ml 0.5 M phosphate, pH 6.5 0.5ml10%SDS 1.0 ml Yeast RNA (12.5 mg/ml) 25.0 ml Formamide 4.5 ml H2O Filter-sterilize and store at -20 C Hybridization Buffer, 50 ml 1.0 ml 50X Denhardt solution 10.0ml20XSSC 4.0 ml 0.5 M phosphate, pH 6.5 0.5 ml 10%SDS 1.0 ml Yeast RNA (12.5 mg/ml) 25.0 ml Formamide 8.5 ml H,0

PAGE 280

263 Filter-sterilize and store at -20 C Wash Solutions Low stringency 2X SSC, 0.1% SDS High stringency 0.1X SSC, 0.1% SDS All reagents were of electrophoresis grade (Fisher Biomedicals).

PAGE 281

APPENDIX 14 RNA DOT BLOT HYBRIDIZATION A. Sample Preparation 1. Pipet 2.5, 5, 10 or 20 |jg of each RNA sample into sterile bullet tubes. Speed vac the samples. 2. Dilute the RNA with 250 [^\ of Denaturation Buffer. 3. Incubate the samples in 65 C water bath for 5 min. 4. Add 250 pi of 20X SSC to each tube. Vortex mix briefly for 5 to 1 sec. The sample is ready to load. B. Filter Preparation 1. Cut a piece of BioTrans nylon membrane to the appropriate size and soak the membrane in water and then in 10X SSC. 2. Soak the Schleicher & Schiell baking paper for the dot apparatus in water and then in 10X SSC. 3. Place one layer of baking paper on the blotting unit. Place the Bio Trans membrane on top of the baking paper. 4. Clamp the blotting unit together. 0. Loading RNA Samples 1. Prefilter the wells with 500 pi of 10X SSC. Turn on the vacuum so 264

PAGE 282

265 the buffer moves through the wells slowly. 2. Load the RNA samples and elute slowly. Turn on the vacuum only after all RNA samples have been loaded. 3. Once all samples have filtered through, wash each well with 500 [jj of20XSSC. 4. Disassemble the apparatus, mark filter for orientation and allow it to air dry. 5. Place the membrane between two sheets of Whatman paper and bake it at 80 C for 2 h or expose the membrane to ultraviolet light for 90 sec. D. Hybridization 1. Prehybridize each filter in prehybridization buffer at 42 C for 2 h. 2. Hybridize each filter with hybridization buffer and the appropriate probe. Incubate at 42 C overnight. 3. Visualize RNA dots by autoradiography at -70 C. E. Solutions Denaturation Buffer, pH 7.0 : 20 mM Tris 50 % Formamide 6 % Formaldehyde This is the final concentration. Filter-sterilize.

PAGE 283

266 Yeast RNA 1. Weigh 400 yeast RNA (Ribonucleinsaure, Boehringer Mannheim) and place in a 50 ml conical tube. Prepare 4 tube simultaneously. 2. Add 20 ml of TE (pH 7.5) to obtain a 20 mg/ml solution. 3. Add 0.026 volumes (520 pi) 8 M ammonium acetate. Mix and place at 65 C in a water bath for 2-3 min until RNA is in solution. 4. Add 0.75 volumes phenol (15 ml) and 0.25 volumes chloroformisoamyl alcohol (5 ml), mix and spin at 4000 rpm for 15 min (Sorvall RC5B with HS-4 rotor). 5. Remove upper phase with a sterile transfer pipet and place in a fresh 50-ml conical tube. 60% of the volume (24 ml) is usually recovered as the upper phase. 6. Add 0.75 volumes of phenol (18 ml) and 0.25 volumes of chloroform-isoamyl alcohol (6 ml), mix and centrifuge at 4000 rpm for 15 min. 7. Remove upper phase (usually about 24 ml) and divide into halves and place in a new set of 50-ml conical tubes. Add 0.1 volume (1.2 ml) 8 M ammonium acetate and mix. Freeze at -80 C for 1 h. 8. Centrifuge at 4000 rpm for 30 min. 9. Discard supernatant invert tube and allow to dry for 5 min. 10. Washwith2ml 70%ethanol. Centrifuge at 4000 rpm for 10 min.

PAGE 284

267 1 1 Resuspend in original volume with TE (1 ml/tube). Once in solution, combine all fractions, mix and aliquot in 15 ml conical tubes. 12. Store at20 C.

PAGE 285

APPENDIX 15 IMMUNOHISTOCHEMICAL LOCALIZATION OF APO E Procedure Immunostaining was done using the VECTASTAIN ABC kit (Vector Laboratories, Burlingame, CA). 1 Deparaffinize and hydrate tissue sections through xylenes or other clearing agents and graded alcohol series. 2. Rinse for 5 min in distilled water. 3. If quenching of endogenous peroxidase activity is required, incubate sections for 30 min in 0.3% H2O2 in methanol. Skip this step if endogenous peroxidase is not a problem. 4. Wash in buffer for 20 min. 5. Incubate sections for 20 min with diluted normal serum from the species in which the secondary antibody is made. In cases where non-specific staining is not a problem, skip steps 5 and 6. 6. Blot excess serum from sections. 7. Incubate sections overnight with primary Apo E antiserum diluted 1 500 in buffer. 8. Wash slides for 10 min in buffer. 268

PAGE 286

269 9. Incubate sections for 30 min with diluted biotinylated antibody solution. 10. Wash slices for 10 min in buffer. 1 1 Incubate sections for 30-60 min with VECTASTAIN ABC reagent. 12. Wash slides for 10 min in buffer. 13. Incubate sections for 2-7 min in peroxidase substrate solution. 14. Wash sections for 5 min in tap water. 15. Counterstain, clear and mount. Solutions Bouins Fixativefor Light Microscopy : 750 ml saturated picric acid 250 mF3
PAGE 287

270 Q O CO CM s OOOOHOHHHrHHHH HH HHHO^ OOOOOOOOOOOOO -OO OVOOVOO CD ^ >< Q Z UJ Q. Q. < < O H s LU Di 111 vo Q. Pu X LU < H ^ < Pu Q ^ < o a: cu < u Eh U OOOOOOOHi-li-liHHrH r-i r-\ OOOOOOOOOOOOO OO .H H H n in O rH OOOOHHrHrHHU3rHrHrH OOOOOOOOOOOOO O <-{ rH O CTl rH rH n rH VD O O [^ O en rH >* in rH O CTl Ol CO fSJ rH (N VO oo(xir~-'a'rHcoinfNjcor>)cooor~rOC0'^'*(NrHrHin'^rHr^-<*rH>* O O rH rH (N rH r^cNnin'^'in'XJinr^eNinvooo r-inr-Hr)rHincrifMCNjco^(Mf\vDr~cooint^HrHra(yi voco(M(TirH>r^'^cocoocovD'*cN)cx)CT\ininvo vor^noo '3<>X)n^OOncri(N^r^rHnnvOOrH(MCT\ VDVDrH> cnco'S'incnO'^or-iHfot^oofNrHcoujO'^ inninvo io>or^(NininoorH0'^co r^ogoovo (NncM^CMfNjcM roo ^(NCMrH cTiOfNOOooint^rHinnrHinroofNjcNvor^r^rocot^ coom'*(Ti'*cor-roO'*'3"vorHn(Nn>o>i>u50^co incooovovot^inrH Ln(Mn(NOr^>i)rHr^nn'^r--ocorHtO'^Mnnr^t^(M OOrHOOOOrHOOOr-rHOCNrHOOOrHOOOCO OOOOO'd'OOO-^OrHininOOOOOOOOrHrH oooo(Ti
PAGE 288

271 rA rA r^ t-i t-i D O O O O O (M O O o y3 CO H "^r o rH H H H CO • • • o o • O • o • O • O • • o • • • O o • • o • o o • • o o H H H H H H H rH VO H H C>0 ^ H CO O O O O O O O O in O O o ro n H CT\ H H H H H H CO VO H H l-l H r-i H a\ O H O O O O O H in H O O o O O O H H iH H CM H CTt i-i ^ !H (N r~^ (Ti cH O O O O O O VO O VO O O CTi rH • • O • O • CN) in o VO • 'I' • 00 • • in 00 • • O CN • CO o • •>* VO VO C\J VO (71 • • inmininoi---*co'^v0'3' VO Hr^'S'roiTir^oovo H r-> r-> t^ > H n 00 r* H CNJ CM r^ H i^cofMrHinr^ncocTitncrirvjrorHvD a>rNj[^vorNjinr)Hocri'3*iHco'3'r~ n m fN n n "J" rj H n CM H rH O H O fO *n in'*ocofOCTirNjnor~HOOpHinrHr^voncoinvovo voHHcvjcvjoO'^i^i^ovorH'^rot^cNnr^cOf-irvjrr)n'a*rHi-in'toroHHCM(N)nnOi-irHHoin'3'n ^r)cvjncT>oorHr)i^ooconrvjinr^incMcriin'3'roc\i or^'^roTi'Lnr^incoojnfNCNjnincNinr^Ht^HfNirsiin t^'i'n->i*in'a'fMininininin>*vO'>a'invo'*^'^vo^incn eooocococococooO'^'i''^'^'^"^"*oooooooooo HHHcococococor^r^r^r^t^r~-t^^'^'^^^'=J''*'>^'i''^ HrHrHHrHrHHfsJf\]CNrM(NfN]fM(NCNOJ CM(NtNJCMfMCM(NCN]OgfMr\JCNrv)OgfMf\]fM(N COCOC0C0C000C0C0COCOC0C0C0COO^(T\

PAGE 289

272 X5 Q) a +J (0 (0 ^^ +J & C (U M -P Q) 3 M iH (4-1 g c \ (u a e c-x: £ £1 j:: Q) £1 -d \ a a tt a>H ^-^ (0 — iT>(-;(OtO(0(C&i >i^.Z, 0> CP en tJi >i cn >i — o c on of radi dioactivit entration ::oncentrat; on fluoro on fluoro on fluoro on fluoro MP-2/S0D c C (0 g (0 C 3 C CP Q) D^ Q) -P -i 3 M ttrH 0< U W II 3 a-H (0 "'i^r^HrHH • • OOO^I+JV^Cni lii Q rH U Q) O rH 15 >i in o (0 o d x: O-H+J 5j CuCkO 0-1 c a o ei ^ uauoo+j-po) a O 5 II M u 4-1 (W rH W O -P 0) o, -H -H Q H-t Q) W ocLfOfa+J-PCC II >-< W tH W 0) 3 (0 -P S J >i T) P (d II II 2 Q Q \ II II "^ H r^ c^ C en Q Q) p +j 11 II II 0^ rt: eu < o SOS G II II IS Q) 1-1 C U i-i ^vo rM DjH >^ W Q U HEHa<^(^AH
PAGE 290

273 Q Z < CO cm" ><^ Q LU LLi CL < a: UJ Q_ X LJU < Q CM ldoo iX)'^criiX)m'3x)nr~ii)rHnr~oocDLf)nx>c\iLn'9'invDLn'rLn'3'rO'3'^i'rOLO'^iDLO oo oooooooooooooooooooooo CM E-i iHO ^^lX)c^^<^^^c^lVDI^l^^x>Lnooor-ocM•^o^>3*Lnor^ 4j vD'O' inrHcriLnrHn*x)r~crvoocNrocsi.-ir^unnininLn u i< OiH ooi^'^aiOtHncrirHi^ncomfMOOovDrHO.H'3'rr~o ^Ln^r)Lntr)(-DLn^i-nLr)r~-"=3'iOLn'^rO'3xioo>^cT>cMr-cri'^.D Eh usvo o(JiCN)vocorHr~cncNi^ooi^cMr~cMU3iHinnoor-m CC; oo r\liHCMi-liHrOHCM>-ICMrHCNJCMCMCMrOCO oo oooooooooooooooooooooo Eh cocsi u3cNCM-sroo'^CMrorO(N>ximocTiocriLnf0^iHC\ioo 4-) ujld cElC^^o^H^oc^l'3^cricric\ico.H'9'crinoocM'^nrocorHCD"^ro-^rC2 >H^CMrslrHtHiHx>-gxi<-t'X)^tr)T'X)CDr~r-*x)criLnoooiH u <; r^,-ioo'^'^>xioocMr-criix>'^ooroir)C\i>X)Lnoor-csi>x>c\i>X)oo crioromLnLn'^LnLn^'3'LnLnoo'3''^'^nn'3''sr'3'ir)'3'n x-i CM Qj vDcTi oc7icriLnLn>x)^roiO[^^crir-cricri r-mvDtHO>X)concT)'3'cr>o'X)<-HLn(^ooocMOOoo'xic\i S U3(£i '3"nroiHo-g'Oooc5^>X)rO'^vDiHrHvx)LnocNjcsjcriro M Eh ro^ [^cTii— icocD'>Doooooooi— iu3cri[^QOLnooi— irHroinn <; r^ooo^-CNCM^J'0^ocDLn^3'c^Joooo'TLncMcrl^^^-c^l^X)r^^H O oocgrHrOiHoroLnLr)'x>^rr-r~Lncri(Nncrii-io>-(M'Troir) a < csou3cTiooOr-iroiHOO.HoroLna)iX)>xirHr~Lnc\i>-ir\jiHO CO vHiHiHx>U3(^r^r-r^rC/3 iHiHiHi-HiHrH^rHi-HiHCXJCMCMCslCMCMCMCMCNCMCMCSJCMCMCM r-LD'^oocTir-r-cococricrir-innovDcriOrHCOLnr-'— irocM OrHOi-icrivOrHcr>i-H^or-oi£)-^oOrHrHOOOOOO 5 CNJCMOrOiHCM
PAGE 291

274 m Q TO > D C o (D II "^ o' CM in >Tnj CM o (0 I c2 CO 03 o < E E CD 1^ O CO I C2 o CD I c2 D < -9 a: E TCM T" O CO 0) Q. > Q) TO ^ •= a Q) o Q. Q. Q. < Q> „' -H^ n' -? /— w/ *— J5 05 TC "n 0) II r^ o < H ^ o < o 03 I o CD > 0) o o o H— M— M— CM -^ O --: CO II 00 >^ CMCN-g of ^ II O 03 CD >^-o CJ o II CM O O O O "o -a O O o o o >— O o D O . >^ 5 03 ^ i: : C C/3 II CO 03 T3 03 CO II o) II CO CO in < Q. < Ic II CSJ CO c CD c c c r: o < (O w II II < c5 CO S 0) CO c c II ;z t ^ f— o I— q:: < Q^ II CM •;.

PAGE 292

275 (z: Q O CO z o u. o OQ Io Q i a I Q O Q O -P U < Q O Q CO o •5j''^crio'XiiHOOcriooniDC\)CNj'3'Lncrio>X)c\i'^Lr)Csi rHi-irorocM'^n'xi'x>o'3'CMiHLnocriOrHmr-oocri'53'rocsir^ C\jrOiH<-liH^iHi-(HrHrHi-lrH-l,HiH ^5x)nor~(Tiocriooo-i cMCM>x)vDix>r-oor-r-r-r-r-ocNioooLnoooo rHi-liHiHrHrHr-l.-ICM0slC\lOJCMC\)
PAGE 293

276 o c TO C D) 0) o o < QI Q CO O c c i_ 5 P CD CD Q CD a: CNJ o u o o o > c 0) o 0) CO ^ c c 0) E Q o CO c £ c il II Q o CO a:

PAGE 294

277 UJ u> CO Io o o lO 01 LU Q. 00 g Cl LU Q Qi 2 O LU LU QW < ? LU Jg Q. < Q ^ oooor-r-'^Mr~--r~.— icri r^^nooro o ooLDiDCNir) "c oooominuJrHt^LnocTiLr) ix>rOrHinior^ to est en o iD r~Cu T3 OOOOrOrHCNJCMOOCMrHrsl CMt-H CslC\ir~CTirO OiH 00 o iH n CM iH i-l n oo oooor^-r-'TM'criLn^Dr-rrHcri .— iro-^n ^iolocsji-h iT) ooooinLn>x>rHi^'^oH * i-i in OT oooor^r~-'^^crioou3r^r^ iHrr-'^rocNcNj o '^ m in oi it> (N ooooLnLr)>X)r-Hi^LnocriLn ^ a\ oosrocricTi ld oorOLDLnr^ Eh X) ooooniHinir)CNjir) iH ooooLnLn>x>iHr— inocTiun >^ro^ncTicM ld ooroiDLnrH X) OOOOrOiHCMCMOOCNiHCM (MiH CsJCNjrOr^CN r-l iHOgOOrH rH iH oooooooiHcrioorHr~-r.hcti r-^r\i.HrHLn^iHr-^tH^ vornro-srog M Dj ooi-HrHCNjrH r^cNjnin'^'Ln'xiLnr-csiLn'Xioo mininino aiotMOOooinr-iHLnniHinroocNjcNUJr-r-rooor^LO'^ooofocyi u coon'^cyi'a(Ticric\ir~>x>i-HiHCMCs)o c M cD'^inoiincsiinmj-ir-\mm^Tin^rm'^mcninoDencnc^cncn'^<-\^cn -P ooroinr~rniHror-'^Lnroooor--(>j>x)ooor--vD>-(roro'Troc\irocrio 3 r-cTiooor^r-oocouj'^oocN^r-oomrsjinLnrocMiHoor-'^pn'^in H Cj oOi-HOM''=rro'^roroooLnvD-=r"3'LniH'3'n^3'VDr~'3'n'a'Ln'^ 4J w O u mronnr-r-r^'HiH— liHiHiHi— IrH.— Ii— IrHtHiHi-HCOOOOO rHiHi-iX)r-r~r-r---r-oooooo rHiHiH<-liHrHrH^rHiH<-HrHrHx)X) incNjfOc^io[^u3rHr--rnrO'3'r~oc\)iHin'^rororor~r--o\jrO'>3'Lnc\i'^csi ooiHoooo
PAGE 295

278 CM Csl o LD Ol K£> CTl ^ I^ Csl CM CM CTl 03 CD [^ iH CM 00 U5 •^ ^ o r'S' iH r• '3' CTl VD en (Tl o in CM iH r~ CTl VD iH <— 1 '3' VD iH CM H VDCTl CTl CTl vT CTlinCTl^ rooo VD CTl r~ cr^r-oot rHOO Hr^^focTir~oovD 2. ^ ^ O 0) oorHiniHr-voroooinvDVD CO vDiH'j'Or^cMror^ooiHCMr^ cMCMCoroOi-ir-irHOin'^ro -icriVDcMoocri T5 C
PAGE 296

g ^ 500 p O 3 43 W 400 P ,(0 g O) 300 ^ O 200 h 1^ 600 I7 11 14 17 19 17 88 170>240 Day of Cycle/Pregnancy Cycle Pregnancy 279 lull vwi ryyA iM/i ryyM uyy 7 11 14 17 19 17 88 170>240 Day of Cycle/Pregnancy V o 200 r g 12 150 Q ? O c (/> o ed lo QQ. 100 50 Cycle Pregnancy 3 7 11 14 17 19 17 88 17O240 Day of Cycle/Pregnancy c .2 3 800 Io a> W 600 (0 ^ O) 400 h Co 200 h '5 u> 2 2 Cycle Pregnancy / gbg^ gb gb 7 11 14 17 19 17 88 170>240 Day of Cycle/Pregnancy 1000 I800 600 400 200 Cycle Pregnancy ikih^^m m 7 11 14 17 19 17 88 170>240 Day of Cycle/Pregnancy

PAGE 297

BIOGRAPHICAL SKETCH Florence Maboh Ndikum-Moffor is a Cameroonian and was bom o Mutia and Rufina Manyi Mutia in Enugu, Nigeria. She received her Bachelor of Science degree in chemistry from the University of Ibadan, Nigeria in November 1980, and a Master of Science degree in chemical pathology from the same university. The author then worked as a research officer with the Institute of Animal and Verterinary Research, Cameroon from 1983 to 1991 when she enrolled in the doctoral program in the Animal Science Department at the University of Florida under the supervision of Dr. Michael J. Fields (Professor). After completion of her degree she would like to continue her training in the biological sciences as a postdoctoral fellow. Florence is married to Gaston Ndikoum Moffor and they are blessed with three children, Kongwenebime 11, Koga 8, and Mandi 3. 280

PAGE 298

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of^hilosopl' Mfchael J. Fields, (5hair Professor of-^nimal Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. in H. Larkin Professor of Anatomy and Cell Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Rosalia CM. Simmen Professor of Animal Science certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate^ in scope and quality, as a dissertation for the degree ^ff D^^tor <^ ^cy William C. Buhi ^-^ Associate Professor of Biochemistry and Molecular Biology I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree oft53ztr of Phiilosophy. 3r ij. Hansen Professor of Dairy and Poultry Sciences

PAGE 299

=;lf|HPai-r=p^ f This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. December, 1995 ji^ / Jiy Dean, College of Agriculture Dean, Graduate School V'

PAGE 300

UNIVERSITY OF FLORIDA 3 1262 08554 5472


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ED4OPQJIB_9EHOMG INGEST_TIME 2012-02-20T21:47:03Z PACKAGE AA00009036_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES