Identification, characterization, and regulation of porcine oviductal plasminogen activator inhibitor-1 and functional a...

MISSING IMAGE

Material Information

Title:
Identification, characterization, and regulation of porcine oviductal plasminogen activator inhibitor-1 and functional analysis of oviductal-specific secretory glycoprotein
Physical Description:
vii, 220 leaves : ill. ; 29 cm.
Language:
English
Creator:
Kouba, Andrew J., 1968-
Publication Date:

Subjects

Subjects / Keywords:
Animal Science thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Animal Science -- UF   ( lcsh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1999.
Bibliography:
Includes bibliographical references (leaves 193-219).
Statement of Responsibility:
by Andrew J. Kouba.
General Note:
Printout.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 021563300
oclc - 43757289
System ID:
AA00014255:00001


This item is only available as the following downloads:


Full Text











IDENTIFICATION, CHARACTERIZATION, AND REGULATION OF PORCINE
OVIDUCTAL PLASMINOGEN ACTIVATOR INHIBITOR-1 AND FUNCTIONAL
ANALYSIS OF OVIDUCTAL-SPECIFIC SECRETARY GLYCOPROTEIN












By

ANDREW J. KOUBA


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
























This Dissertation is dedicated to Brenda and Jerry Kouba.














ACKNOWLEDGMENTS

First and foremost, I need to thank God for the accomplishment of this

dissertation. Many times when the path became difficult and the end seemed so far away,

His gentle hand would continue to push me forward. Throughout my graduate career,

His graciousness has provided me with countless blessings. My relationship with and

faith in God is the responsibility of two wonderful individuals to whom I owe a great deal

of gratitude. My parents, Brenda and Jerry Kouba, have been a constant source of

support, encouragement, and love. They were my first teachers and continue to play such

a role in my life. At a very young age, they encouraged me to make observations,

formulate questions, and test my hypothesis; hence, my parents were the first to teach me

the scientific method. I also owe a great deal of thanks to my siblings, Bonnie and Joann

Kouba, who have always been supportive of my endeavors.

Second, I would like to thank my mentor, Bill Buhi, for his advisement during my

time at the University of Florida. I am deeply grateful for his acceptance of me into his

laboratory and the time that was involved in my education. I will always treasure his

guidance and our numerous conversations on what it means to become a scientist. I also

need to thank members of my committee-Pete Hansen, Ken Drury, Frank Simmen, and

Bill Thatcher-for their assistance throughout my program. Their suggestions in the

design and analysis of my experiments are greatly appreciated and I look forward to

future collaborations with all of them.








Next, I would like to thank Idania Alvarez for the often difficult task of training

me in numerous aspects of protein biochemistry. During my graduate career, several

students, post-docs, and secretarial staff have aided, and I thank them for all their

assistance-Lalantha Abeydeera, Brant Burkhardt, Dan Arnold, Morgan Peltier, Lannett

Edwards, Michael Porter, Carrie Vance, Mary Ellen Hissem, and all those students who

heat-checked every morning with me. I also thank Billy Day and Maureen Goodenow

for allowing me to conduct experiments in their laboratories for the completion of this

dissertation. Lastly, heartfelt thanks go to all those friends who have constantly been

there for me and supported me throughout this journey you know who you are.
















TABLE OF CONTENTS


ACKN OW LEDGM ENTS ................................................ .....................iii

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

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

ABSTRACT................... .. ............... ........ .. x

CHAPTERS

1 INTRODUCTION............................. ......... ... ............... 1

2 LITERATURE REVIEW

Biology of the Oviduct........... ............................... ..... ......... .............
Proteins, Growth Factors, and Cytokines of the Oviduct..............................19
Oviduct Secretions and Co-culture on Fertilization and Embryonic
Development....................... ................... .................. 26
Sperm-Oviduct Interactions........... ........................................ 29
Structure, Regulation, and Biological Actions of Plasminogen Activator
Inhibitor-1 ..................................... ................ ........ 31
Structure, Regulation, and Biological Actions of Porcine Oviductal-Specific
Secretory Glycoprotein....................................... ....... ............ 41

3 IDENTIFICATION AND LOCALIZATION OF PLASMINOGEN
ACTIVATOR INHIBITOR-1 (AI-1) IN THE PORCINE OVIDUCT

Introduction... ....... .................................... ....58
M materials and M ethods ................... ......... .......... .......... ... ... .........60
Results....................................... ............. .... ..... ........... 66
D discussion .......................... ....... ...............83









4 OVIDUCTAL PLASMINOGEN ACTIVATOR INHIBITOR-1 (PAI-1)
PROTEIN, mRNA, AND HORMONAL REGULATION

Introduction... ............. ... ... ........................... ...................... 89
Materials and Methods........... ............. ...... .. ...... ... ..............91
Results.......................... ... ........... ... ...... 98
D discussion .......1......... .............................................. .......109

5 PLASMINOGEN ACTIVATOR INHIBITOR-1 (PAI-1) AND uPA ACTIVITY
IN THE OVIDUCT AND ASSOCIATION OF PAI-1 WITH THE
PREIMPLANTATION EMBRYO

Introduction............................................................. ............... 115
Materials and Methods...................... .............. ...... .... ......117
Results................... ...................................... 123
Discussion .......... .... ........... ...................... .................. 146

6 EFFECTS OF A PORCINE OVIDUCT-SPECIFIC GLYCOPROTEIN ON
FERTILIZATION, POLYSPERMY, AND EARLY EMBRYONIC
DEVELOPMENT IN VITRO

Introduction................... ............ ............................. ............. 151
Materials and Methods................... ................ .. ............ 153
R esults................... ............. .................. ......... .....162
Discussion........................................ ...... ............... 176

7 SUMMARY AND CONCLUSIONS............................................... 183


REFERENCES............................. .......................................... 193

BIOGRAPHICAL SKETCH...... ........... .................. .................220















LIST OF TABLES


Table Pag

6-1 Effect ofpOSP on fertilization parameters of pig oocytes matured and fertilized in
vitro..................... ................ ................ .. ............. 167

6-2 Effect ofpOSP on zona pellucida solubility of pig oocytes matured in vitro....170















LIST OF FIGURES


Figure ag

2-1 Pericellular activation cascade for plasminogen and matrix metalloproteinases...57

3-1 Representative fluorograph after 2D-SDS-PAGE separation of['H]-labeled
proteins (100,000 cpm) from explant culture medium conditioned by the three
segments of the oviduct .................. .................................72

3-2 Representative Sepharose CL-6B and Sephadex G-100 fractionations of [H]-
labeled proteins in isthmus-conditioned medium.............................74

3-3 Comparison of the identified N-terminal amino acid sequence of the 45,000 Mr
protein from isthmic-conditioned culture media to the N-terminal sequence of
porcine PAI-1 .............. ........ ............... .......... ........ ... ..... 75

3-4 Immunoprecipitation ofplasminogen activator inhibitor-1 from porcine isthmic-
conditioned culture media after fractionation on a heparin-agarose affinity
column and elution with 0.4 M NaCI..............................................76

3-5 Representative immunocytochemical localization of PAI-1 in the isthmus from
crossbred gilts on Day 0 (Panel A), Day 2 (Panel B) or Day 12 (Panel C) of
pregnancy............................... ...... .. .... ....................77

3-6 Representative immunocytochemical localization of PAl-1 in oviductal tissue
from Day 2 cyclic (Panel A and C) and Day 2 pregnant (Panel B and D)
crossbred gilts from the isthmus (Panel C and D) and ampulla (Panel A and
B )........................................................... ..... ...... 78

3-7 Immunogold labeling ofPAI-1 within oviductal tissue from a Day 0 non-pregnant
crossbred gilt ........... ... .. ................ ................. 80

3-8 Immunogold labeling of PAI-1 within oviductal tissue from a Day 9 pregnant
crossbred gilt ................................ ............ ..... ........ ........ 82

4-1 Densitometry ofPAI-l protein in fluorographs from 2D-SDS-PAGE analyses of
isthmic- and ampulla-conditioned medium (100,000 cpm) from Large White
and Meishan gilts during early pregnancy.......................................102








4-2 PAI-1 protein in oviduct flushes from crossbred gilts during early pregnancy... 103

4-3 Densitometry of PAI-I protein in fluorographs from 2D-SDS-PAGE analyses of
isthmic-conditioned medium (100,000 cpm [3H]-leucine) from OVX cross
bred gilts................ .... ..... ............ ..... ... ......... .... ..... 104

4-4 Dot-blot hybridization analyses of oviductal total RNA from cyclic crossbred
gilts................... ................. ................... .............. ........... 105

4-5 Dot-blot hybridization analyses of oviductal total RNA from the infundibulum,
ampulla, and isthmus of crossbred gilts...........................................107

4-6 Dot-blot hybridization analysis of whole oviductal total RNA from OVX steroid-
treated crossbred gilts.......................... .......... ...... .... ..........108

5-1 Representative fluorograph of denaturant-activated or latent PAI-1 incubated in
the presence ofuPA after non-reducing 1D-SDS-PAGE separation..........126

5-2 Oviductal PAI-1 inhibition ofuPA activity over time.............................128

5-3 Dose-dependent inactivation of uPA activity by activated PAI-1.................129

5-4 Inhibition ofuPA activity with activated PAI-1 (40 pg) 30 minutes after initiation
of the enzymatic reaction (cleavage of the chromogenic substrate, 2-44x, by
uPA)............................. .... ...... ................130

5-5 Plasminogen activator activity in oviduct flushes from early pregnant crossbred
gilts....................................................... .............. .............131

5-6 Dose-dependent inhibition ofuPA activity by amiloride.........................132

5-7 Inhibition ofoviductal flush PA activity by amiloride .............................133

5-8 Electron micrograph of a pig preovulatory follicular oocyte with surrounding
cumulus cells....................................... .............. .............. 135

5-9 Immunogold localization ofPAI-1 in association with pig preovulatory follicular
oocytes and cumulus cells......................... ..............................137

5-10 Immunogold localization of PAI-1 in association with oviductal oocytes........139

5-11 Immunogold localization of PAI-I in association with oviductal embryos (2-4
cell) and spermatozoa attached and embedded in the zona pellucida......... 141

5-12 Immunogold localization of PAI-1 in association with early uterine embryos (8-16
cell)................................................... ... ........ .......... ...... .... 143











5-13 Controls for immunogold localization of PAI-1 in association with follicular
oocytes, oviductal oocytes and embryos, and uterine embryos................145

6-1 Representative fluorograph of [3H]-leucine labeled proteins (500 pg) from whole
oviduct explant culture media (Day 0/1) subjected to heparin-agarose affinity
column chromatography and separated by 2D-SDS-PAGE..................166

6-2 Photograph ofpolyspermic porcine oocyte matured and fertilized without
exposure to pOSP ...... ..................................... ...................168

6-3 Effect of in vitro incubation of pOSP and anti-pOSP IgG on the polyspermy rate
of porcine oocytes matured and fertilized in vitro...............................169

6-4 Effect of in vitro incubation ofpOSP on sperm binding to porcine putative
zygotes.................. ..... ..... ............ ...................................... 171

6-5 Photograph of sperm binding to putative zygotes after pre-incubation and
fertilization in the presence of 0 pg/ml pOSP (A) or 100 pg/ml pOSP (B).. 173

6-6 Effect of in vitro incubation ofpOSP on cleavage rate using porcine oocytes
matured, fertilized and cultured in vitro........................................174

6-7 Effect of in vitro incubation of pOSP on blastocyst development using porcine
oocytes matured, fertilized and cultured in vitro. ............................175

7-1 Model for the de novo synthesis and secretion of proteins by the porcine
oviduct........... ............................. ........................... 192














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, CHARACTERIZATION, AND REGULATION OF PORCINE
OVIDUCTAL PLASMINOGEN ACTIVATOR INHIBITOR-1 AND FUNCTIONAL
ANALYSIS OF OVIDUCTAL-SPECIFIC SECRETARY GLYCOPROTEIN

by

Andrew J. Kouba

December 1999

Chairperson: William C. Buhi
Major Department: Animal Science

In mammals the oviduct de novo synthesizes and secretes numerous proteins that

may act to facilitate the processes of fertilization and early cleavage-stage embryonic

development. Several of these proteins have been shown to be steroid-modulated and

have cycle-specific expression. However, the majority of these proteins remain

unidentified and have only been characterized in a limited fashion by molecular weight

and isoelectric point. Functional data relating to these proteins is especially lacking. The

first objective was to identify and characterize the major de novo secretary product of the

isthmus. The second objective was the examine potential functions of the well-

characterized pOSP during in vitro fertilization and embryo culture.

The major de novo secretary product of the isthmus was identified as plasminogen

activator inhibitor-1 (PAI-1). It was demonstrated that this protein is localized to

oviductal epithelium and packaged into putative secretary granules for exocytosis. Upon








release into the oviductal lumen, PAI-I was shown to associate with oviductal oocytes,

embryos, and spermatozoa. Characterization of this protein revealed that (1) estrogen

inhibits mRNA and protein expression, (2) mRNA and protein is greatest during

fertilization and early embryo development, and (3) once activated, PAI-1 can form a

complex with uPA and inhibit its activity. In addition, PAI-1 levels are greatest when PA

activity was highest. These results suggest an important role for the regulation of the

plasminogen/plasmin system during fertilization and early embryonic development.

Functional analysis of pOSP revealed several potential actions of this protein

during fertilization. It was demonstrated that pOSP significantly reduced the incidence of

polyspermy in pig oocytes fertilized in vitro. After fertilization in the presence ofpOSP,

a reduction in the number of sperm cells attached to the zona pellucida was observed.

This may be a potential mechanism whereby pOSP reduces the number of sperm cells

that can attach and penetrate an egg. Pig oocytes fertilized in the presence of pOSP had

increased blastocyst development compared to the control. Therefore, the reduction in

the rate of polyspermic fertilization may have resulted in enhanced development to

blastocyst. The pOSP reduction in polyspermy could be inhibited by anti-pOSP

indicating that the decreased incidence of polyspermic fertilization was specific to pOSP.

The characterization and functional analysis of these two proteins, PAI-I and

pOSP, indicates an important relationship between their expression, association with the

oocyte or embryo, and potential actions during fertilization and development.















CHAPTER 1
INTRODUCTION


This dissertation will attempt to enlighten and convey to the reader the importance

of the mammalian oviduct as not only a conduit for the transport of gametes but also as

an important reproductive organ that contributes to the well-being and development of

the embryo. Numerous studies have been done indicating that a pregnancy can be

established without the presence of features and structures that are located within or are

specific to the oviduct. Perhaps the most pervasive example is that of in vitro fertilization

(IVF), in which an oocyte is matured, fertilized, and developed in vitro with subsequent

transplantation of the embryo directly into the uterus, resulting in a successful pregnancy.

Techniques such as IVF, which remove the oocyte and spermatozoa from contact with

the tubule, have generated much debate on the question of whether the mammalian

oviduct is an essential part of the female reproductive tract.

Over the past decade, numerous advances have been accomplished for embryonic

development in vitro, especially during oocyte maturation and fertilization. These

advances have been primarily due to careful formulation of culture medium and use of

co-culture with oviductal epithelial cells. Interpretation of these studies has led to two

prevalent hypotheses. One view is that the oviduct has a more passive role in terms of

establishing an environment with an optimum pH, temperature, osmotic pressure,

nutrients, and oxygen tension for fertilization and early cleavage-stage development. The

second is the oviduct has a more active role, investing energy into the expression and

I







2

synthesis of molecules, which may facilitate or regulate these events. Evidence to date

suggests that these two views are not mutually exclusive but are cooperative in the

formation of a microenvironment, which sustains fertilization and early cleavage-stage

development. Perhaps the importance of this reproductive organ has to be viewed with

respect to its importance in the establishment of a species population. An insightful

paragraph by Dr. Ronald H. F. Hunter in the book The Fallopian Tubes (1988) touches on

this theory:



For most biologists, successful reproduction needs to be considered in
population terms, and all features that contribute to the stability of a
breeding population, and thereby to perpetuation of the species, need to be
brought to mind. It is in this light that the many specific contributions of
the Fallopian tubes are best interpreted. Whilst individual morphological
and biochemical components may not be of overriding importance in
themselves for the establishment of single pregnancies, together they can
be viewed as contributing to the successful maintenance of a breeding
population.


Although embryonic development in vitro has recently seen significant advances,

scientists are still far from reaching a degree of development and viability similar to that

seen in vivo. In virtually every reported case, cultured embryos exhibit reduced cell

numbers, decreased viability with increasing duration of culture, and reduced

development to later stages when compared with their counterparts growing in vivo

[Bavister, 1988]. These observations indicate that some important characteristics of the

embryonic milieu are missing from the artificial culture environment. Peptide and

macromolecules are the oviductal constituents most suspected of having specific

embryotrophic actions, particularly de novo synthesized and secreted products of the

epithelium. The first objective was to identify an unknown major de novo secretary









product of the porcine isthmus. Sequence analysis revealed this protein to be porcine

plasminogen activator inhibitor-1 (PAI-I). Experiments were designed to characterize

this protein throughout the estrous cycle or early pregnancy and to evaluate its

association with the pre-implantation embryo. The second part of this dissertation

concerns the functional analyses of a well-characterized oviductal protein, porcine

oviductal secretary glycoprotein (pOSP), during fertilization and early embryonic

development. It may be that with the identification and functional analysis of specific

oviductal molecules, conditions for in vitro development of embryos may come to mimic

more closely that of the in vivo situation. This dissertation will present data that indicates

a more active role for this reproductive organ in the successful establishment of

pregnancy.















CHAPTER 2
LITERATURE REVIEW



Biology of the Oviduct

Historical Perspective

The first observations of the oviduct, date back to Aristotle (322-384 BC), who

noted a convoluted structure connecting the uterus to the ovary. However, a common

misinterpretation that perpetuated for several centuries was that the ducts transmitted

female semen to the urinary bladder where it was excreted. Accurate anatomical

descriptions were first made by Galen (AD 130-200), linking the ovaries to the uterus and

describing their termination in the uterine horns of bicornuate domestic animals.

However, his mistake was to perpetuate the concept that this tubule was a conduit for

female semen, which was filtered by the ovary from the bloodstream. This notion was

based upon his observation that the female Fallopian tubes were similar to the male

seminal vessels.

The idea of a tubule which transmitted female semen persisted for over 14

centuries. Descriptive anatomical drawings were not available until the excellent works

of Leonardo da Vinci (1452-1519) were published. Another anatomist, Andreas

Vesalius, extensively characterized the anatomy of the Fallopian tube in 1543 in a seven-

volume series known as De Humani Corporis Fabrica. His analyses of their function

echoed that of the past by supporting the hypothesis that these structures were responsible

4







5

for the transport of female semen. The critical structure of the oviduct and relationship of

this structure to the ovary and uterus were detailed by Fallopius, and due to his detailed

description of the tubule his name was lent to the structure (Fallopian tube). The

collected works of Harvey (1651), Van Home (1668), and de Graaf (1672) established

that female semen was a myth, but that instead an egg originates in the ovary (although

the follicle was misinterpreted as the egg), and that this egg is transported into the uterus

via the Fallopian tubes. It was not until significant advances in the technological

development of the microscope were made near the end of the 17-" century that a more

careful scrutiny of the oviduct and its structure, came about. For an excellent review on

the history of the oviduct see The Fallopian Tubes (1988), edited by Hunter and The

Mammalian Oviduct (1969), edited by Hafez and Blandau.

Prenatal Development and Macroscopic Anatomy of the Oviduct

The oviducts of adult mammals are specialized structures that originate during

organogenesis from the cranial region of the primitive Mullerian ducts [Price et al.,

1969]. Johannes Muller first described the Mullerian ducts in 1830, yet far less attention

has been directed toward the development of the oviduct in relation to the uterus.

Although the development of the oviduct prenatally differs widely among mammalian

species, several general statements can still be made. The oviduct undergoes

morphogenetic and histogenetic changes that distinguish it from the uterus which include

(1) coiling of the oviduct, (2) differences in size (diameter), (3) differentiation of

epithelium and muscle, (4) formation of mucosal folds, (5) development of the fimbriae,

and (6) development of the three segments infundibulum, ampulla, and isthmus [Price

et al., 1969]. For a detailed description of comparative development of the mammalian







6

oviduct, the reader is referred to The Mammalian Oviduct (1969), edited by Hafez and

Blandau.

The paired oviducts connect within the uterus at the uterine tubule junction (UTJ)

and link this structure to the ovary. The oviduct is convoluted in most species (except the

rabbit and man), and is supported by the mesosalpinx, which is a part of the broad

ligament [Nalbandov, 1969]. A specialized portion of the mammalian oviduct is often

referred to as the fimbriae infundibulum, which forms a fringed open funnel. In the pig,

which has open fimbriae, this is an intimate structure which encapsulates the ovary. In

species such as mouse and rat this structure forms a sac around the ovary called the bursa

ovarii. The fimbriae of the infundibulum are extremely motile and direct fluid movement

into the oviduct during the time of ovulation. In the pig, unilateral removal of the

infundibulum does not cause infertility as the contralateral fimbriae can redirect and

capture oocytes from the peritoneal cavity [Nalbandov, 1969]. The segment adjacent to

the infundibulum is referred to as the ampulla and has the largest diameter of the three

segments of the oviduct. The intricate folding of the mucous membrane is also

characteristic of this section. The diameter of the ampulla tapers down as it becomes the

isthmus which, in the pig, connects into the uterus at the tip of the uterine horn. The pig

oviduct averages 26-28 cm in length, with the isthmus comprising approximately 1/3 of

the overall length with the majority being made up of the ampulla.

The portion of the isthmus which enters the uterus is referred to as the uterotubal

junction (UTJ), and in most litter-bearing species such as the pig, it contains a high

density of folds and finger-like processes (villi) [Hafez and Black, 1969]. These

structures are arranged in such a fashion so as to prevent the passage of fluid and uterine







7

contents into the oviduct. Anderson (1927a) showed that in pigs, a greater amount of

pressure is required to force uterine fluid into the oviducts when eggs are present (Day 2-

4 of the estrous cycle) than during the luteal phase. This suggests that the UTJ maintains

the oviductal environment protecting it from contamination by secretions from the uterus.

As reviewed by Hunter (1988), the pig oviduct contains an outer layer of

longitudinal muscle and an inner layer of circular muscle. External to these muscle

layers is a well-vascularized serosa layer of tissue. The isthmus portion of the oviduct

has a greater amount of circular muscle, allowing for a greater contraction potential than

the other two segments. The circular muscle layers of the isthmus and ampullary-isthmic

junction are highly innervated with adrenergic nerve terminals originating in the ovarian

or hypogastric plexus. The innervation of the ampulla and infundibulum is very poor

compared to the isthmus, and is usually restricted to the walls of blood vessels. The high

density of circular muscle and adrenergic innervation of the isthmus indicates an

important role for this segment in the transport of gametes and embryos.

The lymphatic vessels of the pig oviduct follow the ovarian and uterine lymphatic

drainage and show a similar concentration in the isthmus as was described for the

sympathetic innervation of nerve bundles. Anderson's (1927b) study of the domestic pig,

revealed that cyclical variations could be noted in lymphatic development and that these

changes might be due to ovarian steroids. It has been suggested that the lymphatics

might be involved in transmission of embryonic information to either the ovary or uterus,

since the greatest amount of tubule resistance to the embryo is found within the isthmus.

The vascular supply to the oviduct is supported by the uterine and ovarian

arteries, and varies according to the estrous cycle or pregnancy. The capillary bed and







8

larger blood vessels of the infundibulum and ampulla become engorged with blood close

to the time of ovulation and are regulated by estrogen [Hunter, 1988]. This increase in

blood supply may also account for the increased tubule secretion and transudation in the

pig at estrus.

Microscopic Anatomy of the Oviduct

The architecture of the ampulla and isthmus is centered on an epithelium with

numerous tubal mucosal folds that show cyclical changes in morphology. The actual

space within the oviduct lumen is very negligible (unlike the lumen of a gland).

Therefore, gametes often remain in intimate contact with the epithelial lining as opposed

to floating free in an aqueous environment. This epithelial lining is composed of simple

columnar and modified cells of the goblet cell type, which are secretary in nature. There

are two distinct differentiated cell types within the oviduct, referred to as ciliated and

non-ciliated secretaryy), although this is not absolute as sometimes differentiating cells

display both cilia and secretary granules [Jansen and Bajpai, 1982]. Interspersed

throughout the ciliated cells are non-ciliated cells that contain electron-dense secretary

granules that can be identified throughout the cell but are most often found in the apical

region. These secretary granules exude their contents with greater frequency during

ovulation, contributing to the decline in epithelial height shortly thereafter [Hunter,

1988]. In most domestic species, epithelial cell height and secretary activity are greatest

at the time of ovulation.

Estrous Cycle of the Pig and Hormonal Regulation by Ovarian Steroids

The influence of ovarian steroid hormones on the oviductal environment has been

well documented in several species for ciliogenesis, secretion, contractility, and







9

morphology of the tubal mucosa and musculature [Hunter, 1988]. It is difficult to

discuss these processes in relation to the oviduct without first describing some aspects of

the pig estrous cycle and its changing hormonal environment. A prepubertal gilt becomes

sexually mature at 6-8 months of age and is described as polyestrous, which means a lack

of seasonality in its reproductive cycle. The average cycle length for post-pubescent gilts

is 21 days, with Day 0 being the onset of standing estrus. The average length of estrus is

2-3 days, with ovulation occurring 30-40 hours after the onset of estrus. A pig is a

polytoccous, or litter bearing, animal and ovulation occurs over a time period of 1 to 5

hours [Hunter, 1988]. Estrogen begins to rise several days prior to estrus (proestrus) and

returns to basal levels by Day 2 of the estrous cycle [Karlbom et al., 1982]. Progesterone

levels begin to rise by Day 2 of the estrous cycle in conjunction with development of the

corpus luteum, reaching a peak around Days 14-15 and quickly declining thereafter

[Karlbom et al., 1982].

Ovarian steroids are transported to the oviduct through the blood [Giorgi, 1980]

which then bind to specific cytosolic receptors and are transported to the nucleus [Chan

and O'Malley, 1976]. If steroid receptors are absent in the oviduct, the tissue becomes

refractory to the action of that hormone. This provides a unique model in which

oviductal function can be examined in intact and ovariectomized animals. Tissue

concentrations of estradiol and progesterone receptors in the porcine oviduct were first

examined by Stanchev et al. (1985). These researchers looked at both cytoplasmic and

nuclear receptors for estrogen (ER) and progesterone (PR) throughout the estrous cycle.

The ampulla showed cyclic variations in the concentration of cytoplasmic and nuclear

ER, increasing during proestrus, reaching a maximum concentration during standing







10

estrus and declining by Days 3-4 of the cycle. The ampulla also showed changes in

cytoplasmic and nuclear PR levels during the estrous cycle, with increasing

concentrations found from standing estrus through the luteal phase. The isthmus had a

much lower concentration of cytoplasmic and nuclear ER than the ampulla, and isthmic

cytoplasmic ER showed no cyclic variation. The isthmus and ampulla showed similar

levels of PRs and also had a similar expression of nuclear PR during the estrous cycle.

However, no changes were noted for levels of cytoplasmic PR in the isthmus during the

estrous cycle. These studies indicate that the isthmus has a constant low-level expression

of cytosolic receptors, while nuclear receptor concentrations vary according to the

hormonal state of the animal. The ampulla, which is believed to be more sensitized to the

effects of estrogen and progesterone than the infundibulum or isthmus, revealed cyclic

changes in both cytoplasmic and nuclear receptors during the estrous cycle. The higher

proliferative and secretary nature of the ampulla during the preovulatory and ovulatory

period may account for these differences in ER expression [Iritani et al., 1974].

Contractility and Ciliogenesis of the Oviduct

Oviductal muscular activity has been examined throughout the estrous cycle, and

although investigators are not in complete agreement, it appears that the most vigorous

contractions occur at about the time of estrus and ovulation. Observations done nearly 76

years ago in the pig revealed that the strength of oviductal muscular contractions was

related to the dimensions of the preovulatory follicles [Seckinger 1923, Wislocki and

Guttmacher 1924]. The larger the diameter of the follicle the stronger the contractions.

As reviewed by Boling (1969), the predominant movement is peristaltic, and

contractions occur in localized segments or loops rather than traveling long distances.







11

These contractions are primarily abovarian in direction. Kymographic records

radiographicc examination in which the range of involuntary movements are recorded) of

the pig oviduct revealed that the ampulla undergoes regular and mild contractions, while

those of the isthmus are more vigorous [Kok, 1926]. Decreased contractility has been

noted in the castrate rabbit although this activity did not become completely quiescent

[Greenwald, 1963]. Compared to intact animals, a delay in tubal egg transport was

observed in castrate animals. Greenwald (1967) showed that species differences are

observed in egg transport in response to exogenous estrogen. In the rat, single injections

of estrogen on Days 1, 2, and 3 of early pregnancy accelerate oviductal transport of ova

[Ortiz et al., 1979, 1991]. A recent study suggests that protein synthesis is required in rat

oviducts for the estrogen effect on oviductal transport to occur [Rios et al., 1997].

Several reports have indicated that estrogen and progesterone together regulate the

contractility and are summarized by Boling (1969). It was observed that the oviductal

musculature remains relatively quiet under the influence of estrogen, and begins to

contract more vigorously when estrogen is withdrawn. Likewise, tissue under the

influence of estrogen begins to contract vigorously shortly after an injection with

progesterone. These data indicate that perhaps estrogen has an inhibitory effect on

vigorous oviductal contractions. Much of these data appears contradictory, especially

since observations at or near estrus indicate this period as having the strongest

contractions. Contractions of oviductal musculature may be controlled by prostaglandins

of the F series. Prostaglandin F2a has been observed in both the Fallopian walls as well as

in oviduct fluid. Lippes (1979) found that PGF2. in human tubal fluid showed cyclical

variations and had higher pre-ovulatory than post-ovulatory levels. It may be that steroid







12

hormones and prostaglandins program the amplitude and frequency of contractions via

the regulation of calcium transport. However, information regarding control of oviductal

contractions is lacking.

Ciliogenesis refers to the growth and development of oviductal cilia throughout

the estrous cycle and pregnancy. Perhaps the greatest amount of information on this

subject is from investigations on the number and morphology of oviductal cilia

throughout the estrous and menstrual cycles and pregnancy. However, there are also

several studies using experimental procedures such as hypophysectomy and ovariectomy,

followed by steroid hormone replacement. As summarized by Brenner (1969),

overwhelming evidence suggests that estrogen stimulates the growth of ciliated cells and

progesterone antagonizes this effect. In the presence of prolonged progesterone, cells

become undifferentiated and difficult to clarify [Brenner et al., 1983]. Distinct

morphological changes have been noted in relation to a changing steroid hormone

environment in humans [Verhage et al., 1979] and rabbits [Brenner, 1969]. Although the

hormonal regulation of ciliogenesis is widely accepted, the number and character of

ciliated cells varies depending on the species studied, the portion of the oviduct sampled,

and the phase of the sexual cycle [Schaffer, 1908]. For example, in the domestic pig,

active ciliated cells have been seen at all stages of the cycle, without any significant

changes in their distribution, number, or activity [Snyder, 1923]. Only a change in cell

height was observed with a maximum occurring around the time of ovulation (Day 0-3)

and a minimum height about two weeks after ovulation. Observations by Buhi

(unpublished data) indicate that in the pig, gradients in the number of ciliated vs secretary

cells exist between the ampulla and isthmus. The pig isthmus contains about 80%







13

ciliated cells and 20% secretary cells (personal communication). During the luteal phase

of the pig, the distal portion of the ciliated cells become constricted from the rest of the

cell and are cut off in a process called deciliation [Flerko, 1954]. It has been proposed

that deciliated cells fill up with secretary granules and become active secretary cells.

Depleted secretary cells then become peg cells and are eventually lost at the end of the

cycle [Brenner, 1969]. As summarized by Hunter (1988), ciliary beat appears to be

regulated by circulating levels of ovarian hormones and is greatest at or near the time of

ovulation. In most mammals examined thus far, cilia of the ampulla beat in the direction

of the ampullary-isthmic junction [Blandau and Verdugo, 1976], suggesting a role for

cilia in the process of egg transport to the site of fertilization.

Fertilization and Early Embryogenesis in the Oviduct

In most mammals, new life begins in the oviduct through the union of germ cells

during fertilization. Wassarman (1999a) defines fertilization as the process of joining

two germ cells, egg and sperm, whereby the somatic chromosome number is restored and

the development of a new individual exhibiting characteristics of the species is initiated.

This process includes several ordered steps: final maturation of both male and female

gametes, sperm acrosome binding, acrosome reaction and penetration through the zona

pellucida, fusion of sperm and oocyte membranes, the corticol granule reaction, and

activation of the cell. During ovulation, fully-grown oocytes from antral follicles

undergo meiotic maturation to become oocytes that are now capable of interacting with

spermatozoa and being fertilized [Wassarman, 1999a]. Capacitation of spermatozoa

within the oviduct is viewed as the final phase of sperm maturation, conferring upon the

cells the ability to penetrate the egg investments [Hunter, 1995]. One consequence of







14

capacitation is an increased flagellar beat of spermatozoa, termed "hyperactivation,"

which is believed to assist in the penetration of the hamster zona pellucida [Yanagimachi,

1988]. A second consequence is membrane vesiculation due to point fusions on the

anterior portion of the sperm head, releasing proteolytic enzymes such as acrosin [Hunter,

1995]. The acrosome reaction has been suggested to be initiated by the binding of sperm

to the zona pellucida. It is generally accepted that one of the zona pellucida proteins,

ZP3, is the natural agonist for stimulation of the acrosome reaction in acrosome-intact

spermatozoa. Binding of spermatozoa to ZP3 activates G-proteins which stimulate a

change in intracellular pH and Ca++; however, the sperm receptors that activate these G-

proteins remain elusive [Wassarman, 1999b]. In the mouse, the zona matrix is composed

of three ZP glycoproteins, ZP 1-3. ZP3 and ZP2 dimers are located throughout the zona

pellucida and are crosslinked by ZPI to create a three-dimensional matrix [Wassarman,

1999b]. In the pig, rabbit, and non-human primates, ZP1 has been shown to be the

primary sperm receptor [Dunbar et al., 1998]. Although these proteins from different

species are conserved within each family, they exhibit distinct biological properties as the

mouse ZP3 is the primary sperm receptor.

Once sperm has penetrated the egg investments, its plasma membrane must bind

and fuse with the oocyte vitelline membrane. Fertilin-P, a sperm membrane-bound

protein, is thought to be the primary molecule responsible for binding of acrosome-

reacted sperm to the plasma membrane [Frayne and Hall, 1999]. This protein is a

member of the ADAM (disintegrin and metalloprotease domain) family of

transmembrane proteins. The disintegrin domain of fertilin-0 is believed to interact with

integrins on the oocyte plasma membrane, mediating the process of fusion. Recently, it







15

was shown that mice that are homozygous null for fertilin-P (fertilin-1 -/-) have a

significantly reduced fertilization rate [Wassarman, 1999b]. Shortly after spermatozoa-

oocyte membrane fusion, the quiescent egg "awakens" to initiate a series of biochemical

events known as egg activation. This activation leads to an explosive release of Ca"

from intracellular storage sites, a transient increase in intracellular pH, an irreversible

activation of oxidative pathways, lipid metabolism, and protein and DNA synthesis

[Yanagimachi, 1988]. This activation is believed to be initiated by paternal proteins

obtained during fusion. The massive release of intracellular Ca" stimulates a Ca"-

dependent exocytosis of corticol granules (CG) which have an important role in

modification of the egg envelope during fertilization. As reviewed by Yanagimachi

(1988), exocytosis in mouse eggs begins near the point of sperm-egg fusion and

propagates from this point in a wave-like fashion to the opposite side of the egg. The

CG, which contain hydrolytic enzymes and saccharide components, alter the physical and

chemical characteristics of the zona pellucida so that the zona becomes refractory to

sperm attachment and penetration. This occurs due to proteolytic modifications of the

sperm-binding ZP receptors. This process is referred to as the Zonaa reaction" and is a

key event for establishing a block to polyspermy in pigs [Hunter, 1991]. Once the sperm

nucleus is incorporated into the cytoplasm of the mature mouse egg (characterized by

ejection of the second polar body) there is a rapid breakdown of the nuclear envelope,

decondensation ofchromatin, and formation of the male pronucleus.

The formation of the female and male pronucleus and their subsequent migration

through the cytoplasm is due to microtubules and microfilaments of the cytosolic

framework [Yanagimachi, 1988]. Besides fertilization, the oviduct is also the site of







16

early cleavage-stage embryonic development. Compared to other domestic species, the

pig has a relatively short cleavage-stage development within the oviduct, entering the

uterus at about the 4-cell stage [Kopecny, 1989]. Cleavage is invariably asynchronous in

mammalian embryos cultured in vitro and development is retarded by about 2 h each cell

cycle, thus, the development to the blastocyst stage may be an entire day longer than

embryos in vivo [Pedersen, 1988]. This indicates that oviduct components may

contribute to the morphologic and metabolic regulation of the developing embryo,

leading to enhanced rates of development of in vivo compared to in vitro embryos.

Oviductal Fluid

Oviductal fluid which can be found within the tubal lumen throughout the

reproductive cycle varies in volume as well as composition and is most abundant when

gametes or embryos are present [Hunter, 1988]. Because tubal volume was found to be

greatest during estrus and near to or just after ovulation, it was suggested that circulating

ovarian steroids might regulate tubal fluid secretion. Bishop (1956) showed that after

ovariectomy in estrous rabbits, tubal fluid secretion dropped from 1.2 ml to 0.2 ml/24 hr

and that systemic injections of estradiol could restore fluid production to estrus volumes.

Comparable observations in tubal fluid accumulation have also been noted in the ewe,

cow, and pig [Hunter, 1988]. The pig was shown to have a fluid accumulation of 6.3

ml/24 hr during estrus falling to 2.1 ml/24 hr in the luteal phase [Iritani et al., 1974].

Another distinction seen between estrus and the luteal phase, is the direction of fluid

flow. Seminal experiments in tubal ligation revealed significant changes in tubal fluid

retention depending on the stage of the cycle. During estrus, the bulk of the fluid passes

from the ampulla into the peritoneal cavity generating a retrograde flow, however, by the







17

time embryos descend through the uterotubal junction, much of the reduced fluid volume

is being redirected toward the uterus [Hunter, 1988]. The restriction of tubal flow

towards the uterus at estrus, is thought to be primarily due to edema found within the

isthmic mucosa. Constriction of the isthmus is thought to be programmed by circulating

levels of estrogen and wanes as a result of elevating progesterone levels [Hunter,1988].

It is difficult to envision the relationship between tubal fluid flow and movement

of the gametes through the lumen. From the observations of Parker (1931), suggestions

have been made that a microflow system close to the epithelium and within the grooves

or channels of the surface may differ in direction from that of the bulk flow. This

microflow system may be intimately involved with the movement of spermatozoa or

oocytes through the oviductal lumen. Detailed chemical analyses of oviduct fluid has

revealed a complex mixture of constituents derived from plasma and the oviductal

epithelium including the oviduct-specific secretary macromolecules. The passage of

plasma constituents into the lumen has been referred to as transudationn.' Because

certain components in oviduct fluid are present at concentrations different from those in

plasma the term transudation was expanded to 'selective transudation' [Leese, 1988].

Therefore, the term 'secretion' has been applied to the formation and discharge of de novo

proteins from the oviductal epithelium. Protein concentration in oviductal fluid is 10-

15% that found in serum [Gandolfi, 1995]. The molecules and proteins involved are

transported across the epithelium within endocytic vesicles that travel from the

basolateral membrane to the apical surface where their contents are released into the

lumen [Parr and Parr, 1986]. The epithelial cell tight junctional complexes restrict

passage of proteins into the lumen between epithelial cells. During early pregnancy in







18

the pig, total recovered oviductal fluid protein was greatest on days 0-3 indicating a

cyclical variation in total protein [Buhi et al., 1998, Kouba unpublished]. Similar cyclic

changes in total protein have been observed during the human menstrual cycle [Lippes et

al., 1981].

Biosynthetic Activity

The mammalian oviduct is a highly synthetic and secretive tissue. The

biosynthetic capacity has been evaluated by measuring the incorporation of radiolabeled

precursor into secreted nondialyzeable macromolecules. Biosynthetic activity of the

entire oviduct was greatest during proestrus, estrus, and metestrus compared to other days

of the cycle in the pig [Buhi et al., 1989], sheep [Buhi et al., 1991] and cow [Malayer et

al., 1988]. Synthetic differences were not observed between mated/pregnant and cyclic

gilts, indicating that the presence of gametes or embryos does not affect the biosynthetic

activity. These findings are in contrast to that of the rabbit where pregnant animals were

shown to have decreased incorporation rates compared to non-pregnant animals [Roy et

al.,1972]. The greater activity during estrus suggests that ovarian steroid hormones may

regulate this function. In ovariectomized pigs [Buhi et al., 1992] or sheep [Buhi et al.,

1991], biosynthetic activity was greater with estrogen than with other steroid treatments.

Furthermore, pigs unilaterally ovariectomized showed no differences in biosynthetic

activity from either the contra- or ipsilateral oviduct containing an ovary with follicles or

corpus lutea. Therefore, circulating ovarian hormones produced by one ovary were

sufficient to control protein synthesis of both oviducts [Buhi et al., 1997].

A recent study showed that the presence of a persistent dominant follicle

decreased the synthetic activity of oviductal tissue in the cow [Binelli et al., 1999]. These







19

investigators suggest that prolonged exposure to high levels of estrogen may have caused

down-regulation of the estrogen receptor, thus suppressing biosynthetic activity.

Interestingly, fluctuations in the abundance of de novo synthesized proteins in animals

containing a persistent dominant follicle, compared to a fresh dominant follicle, were

observed. It was suggested by Binelli et al. (1999) that these fluctuations might be due to

a release of the inhibitory actions of estrogen on certain proteins. Of the three oviduct

segments, the ampulla had the greatest synthetic activity with estrogen and the lowest

with progesterone. Similar to the biosynthetic capacity of the entire oviduct, the capacity

of the ampulla was greatest on Days 0-4 of the estrous cycle, while the isthmus showed

no differences. Likewise, the ampulla was shown to have greater biosynthetic activity

than the isthmus regardless of day of the estrous cycle. Similar rates of activity were

found for the ampulla and infundibulum.

Proteins. Growth Factors, and Cvtokines of the Oviduct

Fluxes of molecules from the fluid bathing the serosal surface to those bathing the

mucosal surface are generated giving the oviductal epithelium its physiological

characteristics [Leese, 1988]. This flux includes water, electrolytes, metabolic non-

electrolytes, amino acids, steroids and proteins. The secretion of specific oviduct fluid

constituents will be discussed here in light of their possible contributions to gamete-

embryo interactions. The majority of this section will cover the identification of proteins

found within oviductal secretions, and the other constituents will be briefly highlighted.

Water is not thought to be transported into the oviductal lumen through an active process,

but rather in response to osmotic gradients established by the transport of ions [Leese,

1988]. Numerous ions have been identified in oviduct fluid including sodium, chloride,







20

potassium, bicarbonate, magnesium, calcium and inorganic phosphorous [Leese, 1988,

Ellington, 1991, Boatman, 1997]. As summarized in these reviews, electrolyte

concentrations in the oviduct vary considerably from those found in serum. Varying

electrolyte concentrations in culture medias during in vitro fertilization and embryo

culture have shown the importance of these constituents on fertilization and embryonic

development. The volume of literature on this topic is extensive and will not be covered

in this literature review. The presence of non-electrolytes in oviduct fluid has also been

reviewed [see Leese, 1988, Boatman, 1997]. Metabolic precursors found in oviduct fluid

include glucose, lactate, pyruvate and amino acids. Differences in concentrations of

these energy substrates in oviductal fluid and blood plasma of pigs during the peri-

ovulatory period have been observed [Nichol et al., 1992]. Several of these energy

substrates (pyruvate and lactate) are routinely used during in vitro fertilization. Two

amino acids, taurine and hypotaurine, have been shown to be present in oviductal fluid at

concentrations much greater than those found in serum [Guerin et al., 1995]. These

amino acids, as well as the converting enzyme that synthesizes hypotaurine from cysteine

intermediates, such as cysteine sulphinate decarboxylase [Guerin and Menezo, 1995],

have been localized to oviductal epithelial cells. Actions of taurine and hypotaurine

include; osmoregulation, calcium modulation, phospholipid interactions, membrane

protein receptor interactions, and antioxidation [Huxtable, 1992, Boatman, 1997].

Specific functions of these amino acids relative to the oviduct include; motility and

viability of sperm, progression through the cell block, improved development of zygotes

to the morulae or blastocyst stages, and increased cell numbers in vitro [Boatman, 1997].

The oviductal epithelium also produces another potent modulator of oxidative stress, the







21

antioxidant glutathione [McNutt-Scott and Harris, 1998]. Through addition of both

electrolyte and non-electrolyte energy substrates and careful formulation of culture

mediums to mimic oviductal concentrations of these constituents, improvements to in

vitro fertilization and development to blastocyst have been achieved. Therefore, the

relative contribution of these constituents to establishment of an appropriate oviductal

environment for fertilization and early embryonic development should be underscored.

Steroids present in the oviductal environment are not generally assumed to be

produced by the oviduct but arrive through serum transudation or from the ruptured

follicle and cumulus cells. However, one report suggests that the oviductal epithelium

contains the machinery for steroidogenesis [Ogunranti, 1992]. Many of the effects of

steroids on gamete function in the oviduct are mediated indirectly through the

vasculature, serosa, mucosa, and muscularis affecting protein secretion, gamete transport,

and fluid accumulation [Hunter, 1988, Hafez and Blandau, 1969]. Recent evidence

suggests that progesterone may have direct actions on the gametes, inducing

hyperactivation and the acrosome reaction in mammalian spermatozoa [Meizel, 1995,

Boatman, 1997]. Use of an anti-progesterone antibody in vivo showed decreased

cleavage rates and embryonic cell numbers [McRae, 1994]. The actions of progesterone

on spermatozoa are non-genomic in origin, and genomic actions of steroids on oviductal

gametes has not been reported. Similarly, there are no reported non-genomic actions of

steroids on metaphase II-oocytes through to the blastocyst stage [Boatman, 1997].

In the pig, analyses of oviductal fluid proteins reveals the presence of hundreds of

proteins, most of which originate from serum as transudate [Buhi et al., 1999]. The most

abundant serum proteins in oviductal fluid are albumin, heavy and light chains of







22

immunoglobulins and transferring. Similar to measurements of total protein in oviduct

fluid, electrophoretic analyses of serum proteins show cyclical variations in their relative

distribution [Buhi et al., 1999]. Additionally, several other proteins have been localized

to the oviductal epithelium and while these proteins have not been shown to be oviduct-

derived and secreted, deserve some mention. Numerous protease inhibitors and enzymes

have been detected including, but not limited to; alkaline phosphatase, amylase, lactate

dehydrogenase, diesterase, lysozyme, acid phosphatase, esterases, lipase, aminopeptidase,

glucosidase and galactosidase [Hunter, 1988]. These molecules may be important in

detoxifying the oviductal environment of substrates harmful to either gametes or

embryos. Recently, catalase was found in porcine, human, and bovine oviduct fluid

[Lapointe et al., 1998]. Catalase concentration was greatest near estrus, and could bind to

the acrosomal cap of spermatozoa. These investigators suggested that the ability of the

oviduct to maintain sperm viability was partially dependent on the ability of catalase to

protect sperm from oxidative damage.

Local countercurrent transfer of products from the ovary or the uterus to other

parts of the reproductive tract are well known. Molecules originating from these tissues

and found within oviduct epithelium or circulation include PGF2, and PGE2 [Harper,

1988], oxytocin [Schramm et al., 1986], and endothelin 1 [Rosselli et al., 1994]. In

addition, the oxytocin receptor has been identified in the oviduct of the ewe [Ayad et al.,

1990]. These molecules have been shown to be important stimulators of oviductal

muscular activity. Local distributions of prostaglandin, oxytocin, and endothelin-1 in the

bovine oviduct were shown to be cyclic and greater concentrations were detected at

estrus in association with elevated levels of oviductal estradiol [Wijayagunawardane et







23

al., 1997]. In the pig, LH/hCG receptors [Gawronska et al., 1999] and receptor mRNA

[Derecka et al., 1995] have been identified. It was observed that LH/hCG receptors were

greatest at the preovulatory stage of the estrous cycle, and caused relaxation of the

oviduct. It may be that LH in conjunction with oxytocin, endothelin 1 and prostaglandin

regulate gamete/embryo transfer.

Alpha-fetoprotein (AFP) has recently been found in human oviduct fluid and may

be regulated by progesterone [Lippes and Wagh, 1993]. AFP binds estrone and estradiol,

acting as a carrier molecule for this steroid and its metabolites in biological fluids. AFP

has also been shown to have immunosuppressive effects. Wagh and Lippes (1993)

showed that AFP binds to the sperm acrosomal cap and may act as an acrosome-

stabilizing factor, preventing a premature acrosome reaction. Other possible

immunosuppressing agents in the oviduct are mucins. MUC1, a transmembrane mucin

gene product, has been localized in the human Fallopian tube, and similar to mucins in

the cervix may act as a barrier to sperm and pathogens [Gipson et al., 1997]. MUC1 may

also be important for gamete and embryo transfer and prevention of adhesion to oviductal

epithelium.

A large number of peptide growth factors and cytokines in the mammalian

oviduct have been described [Kane et al., 1997, Buhi et al., 1997]. These factors likely

originate from both the oviductal epithelium and serum transudate. Growth factors and

cytokines that have been identified within oviductal fluid or the epithelium during

ovulation, fertilization, and early cleavage-stage development in domestic species

include; epidermal growth factor (EGF), heparin-binding EGF, transforming growth

factor (TGF)-ct,, insulin-like growth factor (IGF)-I and II, IGF binding proteins (1-4),







24

colony stimulating factor (CSF)-1, acidic and basic fibroblast growth factor, platelet

derived growth factor, granulocyte macrophage-CSF-1, interleukin-6 and the IGF-II

receptor [Gandolfi, 1995, Chegini, 1996, Buhi et al., 1997, Kane et al., 1997, Buhi et al.,

1999]. Several growth factor and cytokine mRNAs and proteins in the oviduct appear to

be modulated by steroids and temporally associated with elevated estrogen levels at or

near estrus. For examples of their presence and regulation during the estrous cycle the

reader is referred to the reviews listed above. Another interesting cytokine, leukemia

inhibitory factor (LIF) has been identified by the presence of mRNA and protein in the

bovine [Reinhart et al., 1998] and human [Keltz et al., 1996, Barmat et al., 1997] oviduct.

Reinhart et al. (1998) showed that estrogen but not progesterone, stimulated LIF

synthesis and that this stimulation was not receptor-mediated, suggesting a non-genomic

action for estrogen derivitized products. Retinol-binding protein (RBP) has been

localized in the porcine [Harey et al., 1994], equine [McDowell et al., 1993] and ovine

[Eberhardt et al., 1999] oviduct. Eberhardt et al. (1999a) showed that RBP mRNA and

protein synthesis are stimulated by estrogen and that levels of RBP mRNA are greater on

Day 1 of the estrous cycle than Days 5 or 10. In one experiment, embryos collected at

the 1- to 4-cell stage from retinol-treated superovulated pregnant ewes and cultured in

vitro for 7 days, showed a higher development to blastocyst (79%) than control animals

(5%). Retinoids may have beneficial effects during fertilization and early embryonic

development especially in embryonic morphogenesis, cell growth, and differentiation.

Oviduct-derived proteins have been identified in experiments using radiolabeled

precursors for protein synthesis during oviductal organ explant or epithelial cell cultures.

Electrophoretic characterization of de novo synthesized and secreted proteins by ID- and







25

2D-SDS-PAGE and fluorography have been performed in the pig, cow, ewe, human, and

baboon [Buhi et al.,1997]. However, very few of these proteins have been identified and

have only been characterized in a limited fashion by isoelectric point and molecular

weight. Two of these proteins, the estrogen-dependent oviduct secretary glycoprotein

(OSP) and tissue inhibitor of matrix metalloproteinase-1 (TIMP-1), have been

extensively characterized [Buhi et al., 1997, 1999]. Pig OSP will be discussed in greater

detail later in the literature review. TIMP-1 is a major de novo synthesized and secreted

protein of the isthmic segment of the porcine oviduct and its mRNA and protein

expression are greatest on Day 2 of the estrous cycle or early pregnancy. TIMP-1 protein

production and release in the isthmus portion of the oviduct was shown to be greater with

estrogen treatment indicating potential regulation of this inhibitor by ovarian steroids.

This is supported by radioimmunoassay analyses of oviductal flush TIMP-1 protein,

which showed significantly greater levels at Day 0 than other days of the cycle. TIMP-I,

a serine protease inhibitor, has been shown to improve in vitro development of embryos

in both the cow [Satoh et al., 1994] and pig [Funahashi et al., 1997]. This evidence

suggests that TIMP-1, present in the isthmus at the time of fertilization and early

embryonic development, may have important regulatory actions on these processes.

Recently, serine protease inhibitors of the uterus have been shown to stimulate DNA

synthesis in a glandular endometrial epithelial cell line [Badinga et al., 1999], thereby

suggesting a mitogenic role for TIMP-I stimulation of embryonic development.

Four other de novo synthesized proteins, complement C3b, immunoglobulin A

heavy chain, the carboxy terminal region of preprocollagen, and clusterin, have been

identified in the pig oviduct by N-terminal amino acid microsequence analysis [Buhi et







26

al., 1998, Buhi et al., 1999b]. Similar to pOSP and TIMP-1, these four proteins appear to

show synthetic gradients within the oviduct. Of the four, only clusterin and

preprocollagen are found within the pig isthmus. Preliminary evidence indicates that

these proteins exhibit a cyclic secretion pattern and may be partially controlled by

ovarian steroids [Buhi, personal communication]. Potential functions of these four

proteins include immunoprotection, lipid transport, and tissue remodeling. Of the many

de novo synthesized proteins secreted by the porcine (>14) [Buhi et al., 1997] or bovine

(>30) [Malayer et al., 1988] oviducts, only seven have been positively identified (OSP,

TIMP-1, PAI-1, preprocollagen, immunoglobulin A, clusterin and complement C3b).

Oviduct Secretions and Co-culture on Fertilization and Embryonic Development

A massive amount of literature indicates that the oviduct creates a unique

environment for sustaining embryo development. Prior to modern day embryo culture

systems, embryos removed from the oviduct or produced as a result of in vitro

fertilization tended to slow down or block developmentally when maintained in culture

media [Bavister, 1988]. However, when embryos are transferred to the oviduct in vivo,

development could occur from 1-cell to the blastocyst. This beneficial effect of the

oviduct was shown not to be species-specific as bovine embryos could develop in the

sheep or rabbit oviduct. The ability of the oviduct to overcome the cell block to

development is mimicked in vitro by use of chemically defined media supplemented with

protein additives (example; serum or BSA) and a low oxygen tension. Experiments in

the cow, sheep, and mouse where embryos are transferred to oviducts in vivo or in vitro

indicate an important role for this reproductive organ in sustaining embryonic growth

[Bavister, 1988, Hosoi et al., 1995]. A reduced degree of embryo growth obtained in







27

culture media alone [Bowman and McLaren, 1970, Bavister, 1988] and growth

abnormalities following embryo transfer [Farin and Farin, 1995], have been noted for

embryos produced by in vitro fertilization.

If indeed the oviduct creates a unique environment that sustains embryo

development, supplementation of oviductal fluid or the use of oviductal epithelial cell

(OEC) co-cultures during in vitro embryo development should increase the

developmental capacity. Addition of oviduct secretions/fluid to culture medium has

either had no observable effect or has retarded embryo development when compared to

controls [Sirard et al., 1985, Eberhardt et al., 1994]. However, careful fractionation of

oviductal cell-conditioned medium and subsequent addition of these fractions during in

vitro fertilization or embryo culture has had more promising results [Minami et al., 1992,

Mermillod et al., 1993, Liu et al., 1995, Vansteenbrugge et al., 1996, Hill et al., 1996, Liu

et al., 1998]. Preliminary results by these investigators have characterized, by molecular

weight, the embryotrophic activities of oviductal cell-conditioned media. These activities

have both high (>10 kDa) and low (<10 kDa) molecular weight components. None of the

determined molecular weights correspond to OSP. Perhaps the greatest contribution of

oviductal secretion analysis to date has resulted from careful evaluation of its ionic

composition. The concentrations of oviductal electrolytes differs considerably from that

of blood sera and media formulated to resemble the ionic composition of oviductal fluid

show improved embryonic development [Bavister, 1988, Boatman, 1997]. This media,

referred to as synthetic oviduct fluid (SOF), was commonly used during the late 1980s

and early 1990s. However, its use is uncommon today, as laboratories are moving

towards using protein-free chemically-defined culture media for repeatability.







28

Perhaps the most pervasive example of the contribution of the oviduct to embryo

culture can be seen using OEC co-cultures. Seminal experiments by Gandolfi and Moor

(1987) showed that in vivo fertilized sheep 1-cell embryos had a markedly increased

development (8- to 16-cell stage) when co-cultured with OEC (86%) or fetal fibroblast

cells (93%) than in culture media alone (13%). Interestingly, embryos cultured for 6 days

in the presence of OEC had an increased development to blastocyst (46%) when

compared to the fetal fibroblast cell co-culture (5%). Likewise, embryo transfer

experiments, revealed that embryos cultured in the presence of OEC showed twice the

viability of embryos cultured with fetal fibroblast cells. Similar results have been shown

in cattle [Eyestone et al., 1987, Eyestone and First, 1989, Ellington et al., 1990]. As

suggested by Bavister (1988), morphological criteria alone are not sufficient evidence of

the quality of cultured embryos. Hosoi et al. (1995) designed experiments to identify the

developmental stage at which embryos could be influenced by the oviductal environment.

These investigators found that in both mice and cattle the oviductal influence on

blastocyst development occurs during the switch from maternal to embryonic genome

control. The largest contribution of OEC has been shown in humans. Human embryos

co-cultured with OEC have reduced fragmentation [Bongso et al., 1989, Yeung et al.,

1992, Morgan et al., 1995], better morphological characteristics [Morgan et al., 1995],

higher cleavage [Morgan et al., 1995], blastulation [Bongso et al., 1989, Wiemer et al.,

1993], hatching [Yeung et al., 1992] and pregnancy rates [Bongso et al., 1992]. The

beneficial effect of the OEC may include the removal of toxic components from the

medium and/or production of embryotrophic factors. The benefits of OEC co-cultures

are not restricted to the oocyte alone but may also act on spermatozoa. Human sperm co-







29

cultured with bovine OEC show decreased chromatin structural degeneration [Ellington

et al., 1998] and in the pig, a reduced incidence of polyspermy [Nagai and Moor, 1990,

Dubuc and Sirard, 1995]. In addition to OEC effects on enhancing early embryonic

development, OEC play an important role in sperm-oviduct interactions.

Sperm-Oviduct Interactions

In most domestic species, of the millions of sperm that are ejaculated during a

natural mating, only a few thousand reach the isthmus of the oviduct, and there most are

held in a "reservoir" [Suarez, 1998]. Of these thousand which enter the isthmus only a

few reach the ampulla which is the site of fertilization. This oviductal sperm reservoir

has been suggested to be involved in the prevention of polyspermic fertilization [Hunter,

1991], maintenance of sperm fertility between the onset of estrus and fertilization

[Pollard et al., 1991, Chian and Sirard 1994, Smith, 1998], capacitation [Chian et al.,

1995, Mahmoud and Parrish, 1996], and/or motility hyperactivation [Kervancioglu et al.,

1994]. The oviductal sperm reservoir is established by spermatozoa binding to oviductal

epithelial cells and has been observed in cattle, mice, hamsters, pigs, and horses [Suarez,

1998]. For most species studied thus far, spermatozoa bind to the apical membrane of

isthmic oviductal epithelial cells (OEC) and adhesion is specific to the rostral region of

the sperm head [Smith, 1998]. The narrowness of the isthmic oviductal lumen enhances

sperm entrapment within the isthmus and increases the amount of contact between sperm

and the mucosal surface. The binding moieties between spermatozoa and OEC are still

unknown. Nor is it completely understood how and when sperm are released from these

cells. Some evidence suggests that sperm binding and release may depend on the

physical state and motor functions of the sperm. Smith and Yanagimachi (1991) reported







30

that hamster sperm would not bind to OEC once capacitated or hyperactivated. Likewise,

hyperactivation was shown to be a primary factor for release of bound spermatozoa in the

mouse [DeMott and Suarez, 1992]. Therefore, binding and release of spermatozoa may

be due to changes in surface properties of the sperm head, which might lead to increased

flagellar beat amplitude and asymmetry. Binding of sperm to OEC may be due to a lectin

or carbohydrate present on sperm cells, as sperm attachment was inhibited by fetuin in

the hamster [DeMott et al., 1995] or fucose in the cow [Lefebvre et al., 1997]. Suarez

(1998) and Kervancioglu et al. (1994) suggest that epithelial secretions, initiated by

signals of impending ovulation, could enhance sperm capacitation, binding, and release.

Soluble oviductal factors have been shown to induce capacitation in bull sperm

[Parrish et al., 1989, Chian et al., 1995, Mahmoud and Parrish, 1996]. Heparin has been

shown to induce capacitation similar to oviduct fluid [Parrish et al, 1988] and is widely

used during in vitro fertilization of bovine oocytes. However, the intracellular signaling

conditions which induce capacitation (increased intracellular Ca' and pH) are not

equivalent between heparin and oviduct fluid [Parrish et al., 1994]. Recently, Fazeli et al.

(1999) examined induction of capacitation by porcine OEC. Similar to observations in

other species, capacitated boar spermatozoa do not bind to porcine OEC. These

investigators also found that unbound spermatozoa in co-culture with ampulla or isthmic

OEC were slowly capacitated over time. This observation was not noted for cells of

nonreproductive origin. Kervancioglu et al. (1994) reports similar observations between

human OEC and vero cells (kidney epithelium). This suggests that the induction of

capacitation may be specific to oviductal secretions. With boar spermatozoa, the slow

induction of capacitation in vitro by OEC [Fazeli et al., 1999] is similar to in vivo bovine







31

sperm capacitation which may occur for as long as 24 h [Parrish, 1989]. Besides

suggestions by Parrish (see above references) that a heparin-like glycosaminoglycan is

responsible for capacitation activity in oviduct fluid, other protein components have been

shown to induce capacitation, including the bovine estrus-associated glycoprotein

[McNutt and Killian, 1991, King et al., 1994]. Capacitation can also occur spontaneously

in vitro in a defined medium without the addition of biological fluids. This suggests that

the process is intrinsically modulated by sperm itself However, this does not rule out the

influence of positive/negative regulatory factors within the oviduct. It is generally

accepted that capacitation is induced by the removal of decapacitating factors,

specifically cholesterol from the sperm membrane [as reviewed by Visconti and Kopf,

1998 and Cross, 1998]. Cholesterol-binding proteins are present in follicular and

oviductal fluids [DeLamirande et al., 1997] and may be involved in cholesterol removal.

Structure. Regulation and Biological Actions of Plasminogen Activator Inhibitor (PAI)-I

Controlled and targeted proteolytic activity or inhibition of such activity is an

important process of any biological system integrating tissue remodeling, tissue

destruction, or cell migration. This extracellular proteolysis is tightly regulated by the

plasminogen activator (PA)/plasmin system and inhibitors. One of the regulatory

molecules, plasminogen activator inhibitor-I (PAI-1), regulates the biological activity of

PA by preventing conversion of plasminogen to the broad-spectrum enzyme, plasmin.

Plasmin is primarily responsible for the proteolytic degradation of fibrin as well as other

ECM proteins. Therefore, it is principally involved in fibrinolysis and breakdown of the

ECM and basement membranes. PAI-I belongs to a superfamily of "serpins" serinee

protease inhibitors), which represent about 10% of the total protein in blood plasma [Gils







32

and Declerck, 1998]. Of these serpins, al-proteinase represents about 70%. While PAI-1

has been traditionally thought of as a regulator of fibrinolysis, the plasminogen activator

cascade has been shown to be important in other biological systems requiring

extracellular proteolysis such as ovulation, mammary gland involution, blastocyst

implantation, and tumor metastasis [Ny et al., 1993]. Plasminogen biosynthesis occurs

primarily in the liver and is found in high concentrations within blood plasma (2 mM)

thereby providing an unlimited source of potential proteolytic activity [Ny et al., 1993].

Thus, the rate-limiting step in these reactions are synthesis and activation of PA. The

proteolytic cascade is very dynamic and must be considered a multi-component system.

This multi-level system allows for precise regulation and fine-tuning at discrete focal

areas requiring tissue remodeling, so that systemic proteolysis is controlled.

The primary component of this system, plasmin, is formed by proteolytic

cleavage of plasminogen. One of two specific plasminogen activators, tissue-type (tPA)

or urokinase (uPA), initiates the cleavage of plasminogen. Both belong to the serpin

superfamily, are immunologically distinct and encoded by different genes [Ny et al.,

1993]. PAI-1 binds to both tPA and uPA acting as a suicide substrate and thereby

inactivating their enzymatic activity. In addition to this function, uPA- and tPA-PAI-1

complexes can bind to the multifunctional clearance receptor, low-density lipoprotein

related-receptor protein/a2-macroglobulin receptor (LRP/o2-MR or Megalin) [Ny et al.,

1993]. These complexes are internalized by endocytosis and subsequently degraded.

This is another potential mechanism for PAI-I regulation of uPA and tPA proteolytic

activity. Urokinase PA also has a cell surface receptor (uPAR) along with plasminogen,

and binding of these molecules to cellular surfaces is thought to lead to generation of







33

plasmin, which coordinates extracellular matrix (ECM) remodeling and degradation.

Degradation of receptor-bound uPA is enhanced three- to four-fold following complex

formation with PAI-1 [Ny et al., 1993]. Plasmin is the proteolytic enzyme primarily

responsible for the degradation of fibrin (fibrinolysis), breakdown of ECM and basement

membranes as well as thrombolysis. While tPA generation of plasmin is linked to the

fibrinolytic activity, uPA generation of plasmin is associated with the ECM

remodeling/degradation [Andreasen et al., 1990]. Plasmin can also activate promatrix

metalloproteinases, procollagenase, and progelatinase. A diagram of the pericellular

activation cascade for plasminogen and matrix metalloproteinases can be found in Figure

2-1.

PAI-1, a single chain glycoprotein of 50 kDa, is synthesized in an active form,

which converts spontaneously to an inactive latent form that can be partially reactivated

by denaturing agents such as SDS, guanidinium hydrochloride, and urea [Hekman and

Loskutoff, 1985]. In plasma, PAI-1 occurs in a complex with an integrin, vitronectin,

resulting in a significant stabilization [Declerck et al., 1988, Mimuro and Loskutoff,

1989]. PAl-1 forms a stable equimolar complex with both uPA and tPA, presumably of a

covalent nature [Andreasen et al., 1990], and exists in at least three distinct

conformational forms; active, latent, and substrate-cleaved [Lawrence et al., 1997].

Active PAI-1 decays to the latent form with a half-life of 1-2 h at 370 C, unless stabilized

by vitronectin. Similar to other serpins, PAI-I has a tightly ordered tertiary structure

consisting of three P-pleated sheets, A, B and C, a-helices A through I and a reactive site

loop [Gils and Declerck, 1998]. The reactive site, comprising the "bait" peptide bond, is

located 30-40 amino acid residues from the carboxy-terminal end. Nucleotide







34

sequencing of the cDNA and amino acid sequencing revealed a mature protein of 379-

381 amino acids long with heterogeneity among species primarily in the amino terminus

[Andreasen et al., 1990]. This protein was also found to be a glycoprotein and shares a

high degree of similarity (91%) between species examined, cow, human, rat, rabbit, and

mouse. Recently, pig PAl-1 was cloned and sequenced from a cDNA library prepared

from cultured pig aortic cells [Bijnens et al., 1997]. The cDNA consists of a 131-bp 5'

untranslated sequence, a 1206-bp open reading frame encoding a 402-amino acid protein,

and a 1651-bp untranslated sequence. The preprotein is cleaved at amino acid position

23, resulting in release of the mature 379-amino acid protein, PAI-1.

Hormonal Regulation of PAl-1

PAl-1 is produced by a variety of cell lines and primary cell cultures. These in

vitro culture systems have served as important models for obtaining information about

the intracellular mechanism of action stimulated by various regulatory agents.

Investigations have resulted in the identification of several hormones and steroids that

control the synthesis and secretion of PAI-I. As reviewed by Andreasen et al. (1990),

PAI-1 synthesis and/or activity is stimulated by glucocorticoids, insulin, endotoxin,

interleukin-1, tumor necrosis factor-a, EGF, CSF-1, and phorbol esters. Inhibitors of

PAI-1 synthesis and/or activity are molecules that stimulate intracellular increases in

cAMP production such as follicle stimulating hormone (FSH) and leutinizing hormone

(LH). Other potential regulators of PAl-1 synthesis/activity are prolactin [Liu et al.,

1998] and prostaglandin F2a [Zhang et al., 1996].

In addition to regulation by the molecules described above, PAl-1 expression and

activity is regulated by ovarian steroids. Progesterone has been shown to significantly







35

increase PAl-1 synthesis in several endometrial cell lines [Schatz et al., 1994, Casslen et

al., 1992, Casslen et al., 1995, Miyauchi et al., 1995]. As reviewed by Schatz et al.

(1994), estrogen had either no effect or inhibited PAI-I synthesis. However, the effect of

progesterone on PAl-1 synthesis doubled in response to the presence of both estrogen and

progesterone. Furthermore, exogenous steroids elicit similar changes in expression of

PAI-1 mRNA as seen for PAI-I protein. Progesterone was shown to not interact with the

804-bp promoter region of the human endometrial PAI-1 gene, but rather increased the

stability of PAI-1 mRNA [Sandberg et al., 1997]. After menopause, hormone

replacement therapy may reduce the risk of coronary heart disease that is associated with

fibrinolytic activity. Women receiving estradiol were shown to have significantly lower

plasma levels of PAI-1 [Lindoff et al., 1996, Shahar et al., 1996], indicating that estrogen

has a significant role in inhibition of PA-1 expression. However, some investigators

suggest that estrogen might also increase PAl-1 synthesis [Sobel et al., 1995, Fujimoto et

al., 1996] which is contradictory to the majority of other findings to date.

Plasminogen Activator and its Inhibitor in the Uterus and Ovary

In man and rodents, endometrial stromal cells proliferate and differentiate into

decidual cells as the implanting embryo invades the endometrium in order to establish an

intimate contact with the maternal blood supply. Although the implanting embryo

produces matrix-degrading enzymes for implantation, ECM-degrading enzymes are also

produced in the human endometrium under the control of steroid hormones [Wang et al.,

1996]. In the human, an increase in tissue-factor (TF) and PAl-1, and inhibition oftPA,

uPA and matrix metalloproteinases (MMPs) are associated with progestin-induced

decidualization of estrogen-primed endometrial stromal cells in vivo and in vitro







36

[Lockwood and Schatz, 1996]. These important regulators of fibrinolysis, hemostasis,

and ECM turnover in the decidualized stromal and decidual cells suggest a mechanism to

explain the absence of hemorrhage during invasion of the endometrial vasculature by

trophoblasts. Progesterone withdrawal reduces TF and PAl-1, yet increases uPA, tPA

and MMPs thereby increasing hemorrhage, fibrinolysis, ECM degradation and vascular

injury characterizing menstruation [Lockwood and Schatz, 1996]. Trophoblast cells are

so highly invasive that they are called pseudomalignant and when grafted to ectopic sites,

invade uncontrollably [Axelrod, 1985]. The modified uterine tissue, decidua, is rendered

resistant to invasion by the well-controlled coordinate expression of protease inhibitors

such as PAI-1 [Waterhouse et al., 1993, Teesalu et al., 1996, Lala and Hamilton, 1996].

The PA system has also been shown to be an important regulator of ovulation. As

early as 1916, it was proposed that proteolytic activity was involved in degradation of the

follicle wall at the time of ovulation [Ny et al., 1993]. Levels of plasminogen in

follicular fluid are comparable to levels found in serum. Injection of agents which inhibit

PA and plasmin activity into the ovarian bursa suppress ovulation [Ny et al., 1993].

Combined uPA and tPA gene-deficient mice were also shown to have a 26% reduced

ovulation efficiency [Leonardsson et al., 1995]. The loss of an individual PA is

functionally compensated by the activity of the remaining PA, however the loss of both is

not obligatory for ovulation. As reviewed by Andreasen et al. (1990) and Ny et al.

(1993), ovulation-associated hormones, FSH and LH, are effective stimulators of PA

secretion from granulosa cells. In the preovulatory follicle, the LH surge stimulates a

cascade of proteolytic enzymes, including PA, plasmin, and MMPs. These enzymes

bring about the degradation of the perifollicular matrix and the decomposition of the







37

meshwork of collagen fibers which provide strength to the follicle wall [Tsafriri and

Reich, 1999]. Increased ovarian proteolytic activity is controlled by high levels of PAI-1

and TIMP-1 expression in theca cells from growing follicles, ensuring their development

by protection from enzymes diffusing from ovulatory follicles [Tsafriri and Reich, 1999].

FSH, LH, and activators of cAMP suppress PAl-1 synthesis and activity in granulosa

cells. However, one report in the rat indicates cell-specific expression of PAI-1 and

TIMP-1 mRNAs in the LH/hCG-stimulated ovary suggesting a co-expression of both

enzymes and their inhibitors during ovulation [Chun et al., 1992]. PAI-1 was also shown

to be a major secretary product of the corpus luteum and its expression is stimulated by

PGF2a [Smith et al., 1997]. These investigators suggest an important role for PAI-1

during ECM remodeling of the ovary following ovulation.

PA in the Oviduct During Fertilization and Embryo Development

A significant amount of literature has been compiled on PA activity during

fertilization and its association with the preimplantation embryo. However, there are no

data on its physiological inhibitor, PAl-1, in the oviductal lumen or its association with

the early embryo. The importance of PAl-1 in the reproductive system is not fully

understood. PAI-1 gene-deficient mice are viable, fertile, and without abnormalities in

organogenesis and development [Carmeliet et al., 1993]. Other factors (PAI-2, 2-

antiplasmin, a2-macroglobulin, Cl-esterase inhibitor and ca-antitrypsin) known to reduce

plasminogen activation and/or plasmin activity, may compensate for the loss of PAl-1.

However, to the authors knowledge, these proteins have not been identified in the

oviduct. This portion of the discussion will concern data available on PA that may lead

to insights on PAI-1 function within the oviduct.







38

As early as 1968, PA was shown to be associated with the oviductal mucosa

[Tympanidis and Astrup, 1968]. PA activity has also been shown to be associated with

rat [Liedholm and Astedt, 1975] and mouse [Sherman, 1980] preimplantation embryos

from the 2-cell to blastocyst stage. Sherman (1980) showed that this activity was

associated with the zona pellucida, as denuded embryos have no PA activity. These

earlier studies concluded that zona-associated PA did not originate from the embryo but

from genital tract secretions. The observation that PA activity steadily decreased as

preimplantation development proceeded both in vivo and in vitro raises the possibility

that the zona pellucida contains residual activity from follicular or oviductal fluid

[Sherman, 1980].

PA activity has been associated with several important reproductive processes

including oocyte maturation, fertilization, and early embryogenesis. Pig cumulus-cell

oocyte complexes (COCs) have been shown to produce two PAs during in vitro

maturation [Kim and Menino, 1995]. These PAs were shown to be tPA and tPA-inhibitor

complex. Stimulators of PA activity in pig COCs, like cAMP and okadaic acid, were

found to inhibit oocyte maturation. Kim and Menino (1995) suggest that although PA

production is temporally associated with oocyte maturation, coordination of these two

processes is differentially regulated. The production of PA in rat COCs has also been

reported [Liu et al., 1986, 1987]. PA activity in rat [Salustri et al., 1985, Ny et al., 1987]

and pig [Kim and Menino, 1995] COCs is through a stimulation of protein kinase A and

C. Recently, cow COCs matured in vitro, were shown to contain mRNA for all elements

of a proteolytic cascade including uPA, PAl-1, MMP-1 and TIMP-1 [Bieser et al., 1997].

These investigators suggest a potential role for this extracellular proteolysis in cumulus







39

expansion during oocyte maturation. In the human, transcription of PAI-1 and PAI-2

genes have been described for cumulus cells and granulosa-luteal cells [Piquette et al.,

1993]. Similarly, cow COCs show production of PA during in vitro maturation and this

production was stimulated by EGF [Park et al., 1999]. This activity was shown to be

associated with uPA and was not present in denuded oocytes, indicating the importance

that cumulus cells play in its synthesis. For a current discussion on the production of PA

in COCs relative to the time of germinal vesicle breakdown and resumption of meiosis,

the reader is referred to Park et al. (1999).

As reviewed by Huarte et al. (1993) the PA/plasmin proteolytic cascade may have

a significant role during fertilization. Proteolytic enzymes participate in multiple phases

of mammalian fertilization including the acrosome reaction, sperm-binding to the zona

pellucida, zona pellucida penetration, and the zona reaction. Mouse gametes express

plasminogen-dependent proteolytic activities; ovulated eggs synthesize and secrete tPA

while ejaculated spermatozoa have uPA activity. Likewise, plasminogen was shown to

bind to both mouse spermatozoa and eggs, and the presence of plasminogen increased the

in vitro fertilization rate [Huarte et al., 1993] presumably due to increased plasmin

generation. In addition, antibodies, which inhibit the catalytic activity of plasmin, were

shown to decrease the fertilization rate. Exogenously added plasminogen was likely

converted to plasmin by zona pellucida-associated uPA or tPA activity. These findings

are supported by previous work indicating that a positive correlation existed between the

ability of oocytes fertilized in vitro and the PA activity of the corresponding follicular

fluid and granulosa cells [Deutinger et al., 1988, Milwidsky et al., 1989]. Zhang et al.

(1992) showed that tPA was released from rat oocytes as a result of oocyte activation and







40

suggested that tPA may act in the perivitelline space on the zona pellucida during

fertilization and/or activation. As described by Zhang et al. (1992), tPA may be a

component of corticol granules that are involved in the zona block to polyspermy. In

their study, activation-induced zona hardening (limited proteolysis of zona protein ZP2)

was prevented by a anti-tPA monoclonal antibody. Inadvertent activation under in vitro

conditions in mouse and human IVF systems often times leads to premature loss of

fertility in respective oocytes. Addition of leupeptin, a serine protease inhibitor, during

mouse oocyte calcium-ionophore activation reversed the decrease in the capacity of

oocytes to fertilize and develop in vitro [Tawia and Lopata, 1992]. The data support

evidence for PA activity in zona hardening and suggests a coordinate expression between

activator and inhibitor in regulating this process. Additionally, supplementation of

culture medium with proteases increased hatching rate of mouse embryos [Lee et al.,

1997], while addition of protease inhibitors to culture medium inhibited hatching in vitro

[Dabich, 1981, Yamazaki et al., 1985]. This suggests an important role for oviductal-

derived protease inhibitors in the prevention of premature hatching, prior to the correct

developmental stage within the uterus. The effect may be directed towards proteolysis of

the zona matrix. Zona of porcine oviductal oocytes and embryos were shown to be more

resistant to proteolytic digestion than either follicular oocytes or embryos recovered from

the uterus [Broermann et al., 1989], indicating an important role for protease inhibitors in

protecting the zona pellucida from degradative proteases.

Preimplantation mouse [Harvey et al., 1995, Zhang et al., 1996], rat [Zhang et al.,

1994], sheep [Menino et al., 1989, Bartlett and Menino, 1993] and cow [Dyk and

Menino, 1991, Berg and Menino, 1992] embryos were shown to have PA activity.







41

Kaaekuahiwi and Menino (1990), showed that as embryonic size and cell number

increase and development progresses, bovine embryos liberate more PA. Recently,

bovine embryonic PA was shown to induce changes in the electrophoretic protein profile

of the zona pellucida when embryos were incubated with exogenous plasminogen

[Cannon and Menino, 1998].

In addition to PA's association with maturation, fertilization and embryogenesis, it

might also have an important role in spermatozoa physiology. Urokinase PA is

synthesized in epithelial cells of the caudal epididymis, vas deferens, and seminal

vesicles [Huarte et al., 1987], while tPA was found in the prostate gland [Reese et al.,

1988]. In addition, Sertoli cells have been shown to secrete PA and PAI-1 [reviewed by

Gilabert et al., 1995]. The presence of uPA and tPA has been shown in ejaculated

spermatozoa of man and various animal species [Smokovitis et al., 1987]. These

investigators report a direct influence on sperm motility and suggest that the PA are

membrane bound. Recently, boar spermatozoa were shown to have uPA and tPA

associated with spermatozoal membranes and differences were detected in their

localization [Smokovitis et al., 1992]. The outer acrosomal membranes contained tPA

while the inner acrosomal membrane contained both tPA and uPA.

Structure. Regulation, and Biological Actions of Oviductal Secretory Glycoprotein (OSP)

In vitro Synthesis and Hormonal Regulation of OSP

Experiments by Oliphant and Ross (1982) in the rabbit describe the identification

and purification of three sulfated glycoproteins from oviduct fluid. This investigation

was one of the first studies utilizing a radiolabeled precursor, [35S], to evaluate

incorporation into macromolecular components of oviductal epithelial cells and oviductal







42

explant tissue in culture. Utilizing this technique, Oliphant and Ross (1982) reported the

de novo synthesis and secretion ofglycoproteins by oviductal epithelium. Subsequently,

numerous studies have been done in various species. Evaluations of stage-specific

proteins in the pig oviduct were first reported by Buhi et al. (1989). The de novo

synthesis of porcine oviductal proteins by cyclic and early pregnant tissues in explant

culture revealed that incorporation of [3H]-leucine into nondialyzeable macromolecules

was greatest on Days 0 and 2 and secretary activity was lowest on Days 10 to 15. The

ampulla was shown to have greater incorporation than the isthmus and increased rates of

incorporation were temporally associated with elevated estrogen at proestrus, estrus, and

metestrus [Buhi et al., 1989]. 1D-SDS-PAGE and fluorographic analysis of oviductal

culture medium revealed three proteins of 335 k, 115 k, and 85 k Mr, associated with

proestrus, estrus, and metestrus. High resolution 2D-SDS-PAGE further resolved the

isoelectric and molecular weight variants of these radiolabeled proteins initially detected

by 1D-SDS-PAGE [Buhi et al., 1990]. The major 115 k Mr band was resolved into two

major glycoproteins, one basic (100 k M,, pi > 8) and one acidic (100 k Mr, pl 4.5-5.5).

The 85 k Mr protein was resolved into a very acidic (pl< 4) protein of 75 to 85 k Mr.

These three proteins have been designated pOSP El (85 k Mr), pOSP E2 (100 k Mr;

acidic) and pOSP E3 (100 k Mr; basic). It was suggested by Buhi et al. (1990) that these

denatured and reduced subunits may be part of a higher molecular weight complex and

were related (pOSP E2 had the same N-terminal sequence as El and the N-terminal

sequence for E3 was identical to an internal sequence of E2). Incorporation of [35S]-

glucosamine indicates that these three estrus-associated proteins are glycosylated [Buhi et

al., 1990].







43

Oviduct-specific secretary proteins have been identified in human [Verhage,

1988, Buhi et al., 1989], baboon [Fazleabas and Verhage, 1986, Verhage and Fazleabas,

1988], cow [ Malayer et al., 1988, Boice et al., 1990], sheep [ Sutton et al., 1984, Sutton

et al., 1986, Gandolfi et al., 1989, Buhi et al., 1991, Murray, 1992], mouse [Kapur and

Johnson, 1985] rhesus monkey [Verhage et al., 1997], and hamster [Robitaille et al.,

1988, Abe et al., 1998]. Although isoelectric points are similar for the subunits examined

in other species, differences have been noted in molecular weights for these

glycoproteins. Studies on these glycoproteins have led to several key observations. First,

these glycoproteins are primarily secretary products of the ampulla, although their

synthesis can be seen in other segments at a much lower rate [Verhage and Fazleabas,

1988, Buhi et al., 1990, Boice et al., 1990, Gandolfi et al., 1991, Murray, 1992, and

O'Day-Bowman et al., 1995]. Second, the presence of these proteins only during estrus,

temporally elevated at estrus or the follicular stage in humans indicates that this protein is

regulated by estrogen. Therefore, the pattern (but not presence) of distribution for these

proteins is dependent on the presence or absence of specific ovarian steroids. In the pig,

treatment of ovariectomized (OVX) gilts with estrogen increased incorporation rate in the

ampulla and revealed that the oviduct-specific glycoproteins were estrogen-dependent

[Buhi et al., 1992]. This is supported by the observation that progesterone

antagonizes/abrogates the stimulatory effect of estrogen on their synthesis. Similar

observations on OSP steroid regulation using OVX animals have been made in sheep

[Buhi et al., 1991, Murray and DeSouza, 1995] and baboon [Verhage and Fazleabas,

1988]. Newborn golden hamsters (1.5 days) injected daily with estrogen or progesterone

showed that estrogen could induce hamster OSP synthesis in the undifferentiated







44

epithelial cells of neonates, while progesterone could not [Abe et al., 1998]. This family

of estrogen-dependent and oviduct-specific glycoproteins is the most abundant

radiolabeled secretary product of the oviduct, yet assigning specific functions to these

proteins has been elusive. Investigations on the localization of OSPs in the oviduct and

their association with the ovulated oocyte and early embryo, as detailed below, suggest

possible involvement during fertilization and early cleavage-stage embryonic

development.

Immunolocalization ofpOSP in the Oviduct. Oocvte. Spermatozoa and Early Embryo

Immunogold localization of pOSP in the pig has been evaluated at the cellular

level utilizing electron microscopy [Buhi et al., 1993]. These investigators found that

pOSP was localized to putative secretary granules in non-ciliated secretary cells of the

ampulla from both cyclic and OVX estrogen-treated gilts. This study was not extended

to evaluate distribution of pOSP in either the infundibulum or isthmus of the pig. Similar

observations of OSP localizing to secretary granules in the ampulla have been observed

in sheep [Murray, 1992, Gandolfi et al., 1991], cow [Boice et al., 1990], baboon [Verhage

et al., 1990, Verhage et al., 1989], rhesus monkey [Verhage et al., 1997], and human

[O'Day-Bowman et al., 1995, Rapisarda et al., 1993]. Studies done in the human and

baboon identified OSP in secretary granules of the isthmus as well as the ampulla,

indicating that this segment also synthesizes OSP. Evaluations of colloidal gold densities

in segments treated with estrogen, progesterone, or estrogen + progesterone have not

been performed. Therefore, it is unknown whether there are dynamic changes of OSP in

secretary granules relative to hormonal status, except for the observation with estrogen

alone, as detailed above. Hamster OSP of oviductal origin is associated with uterine







45

epithelial cells during the first 3 days of pregnancy, but is reduced by Day 4, and is

absent by Days 5 and 6 [Roux et al., 1997]. Because Days 5 and 6 correspond to the time

of implantation in the hamster, and levels of OSP remain constant in the oviduct, these

investigators suggest that OSP may be involved in uterine receptivity.

The macromolecular composition of the zona pellucida in the mammalian

oocyte/embryo is quite complex and three major glycoproteins have been identified each

containing about 15 isoelectric species [Dunbar, 1983]. Several studies have revealed

changes in the composition of the porcine zona pellucida during development from the

oocyte to the 4-cell embryo. Brown and Cheng (1986) observed that the zona pellucida

of the follicular oocyte to the early embryo acquired three glycoproteins from the oviduct

at estrus. These investigators observed two proteins of 250 k and 90 k Mr when the zona

pellucida was analyzed by 1D- and 2D-SDS-PAGE under non-reducing conditions.

When these proteins were electrophoresed under reducing conditions, subunits of 90 k,

79 k, and 69 k M, were detected. Similar differences in the macromolecular composition

of zona pellucida from pig oocytes, eggs, and zygotes have been observed by Hedrick et

al. (1987), although there is some discrepancy in the molecular weight of these proteins

when compared to those of Brown and Cheng. These are the first studies indicating that

proteins of oviductal origin associate with the zona pellucida and that their molecular

weight (80-90 k) corresponds to that of pOSP. The study by Buhi et al. (1993) supports

these earlier results and specifically indicates that oviductal pOSP associates with

unfertilized oviductal oocytes and early embryos. Pig OSP immunoreactivity was

observed through early development to Day 7 hatched porcine embryos but had

disappeared by Day 9 of pregnancy [Buhi et al., 1993]. In addition to association with the







46

zona pellucida, pOSP also associates with flocculent material in the perivitelline space

and with vitelline and blastomere membranes. Limited data supports the possibility that

the embryo itself endocytoses OSP during its early development. Hamster OSP has been

localized in coated pits, apical vesicles, mutivesicular bodies, and lysozome-like

structures [Kan and Roux, 1995]. The association of pOSP with the early developing

embryo suggests this protein may have an important role during fertilization and early

cleavage-stage embryonic development.

Localization of OSP to the zona pellucida of oviductal oocytes or early embryos

has been shown in the hamster [Kan et al., 1989, Kan and Roux, 1995, Abe and Oikawa,

1990, Leveille et al., 1987], cow [Wegner and Killian, 1991], sheep [Gandolfi et al.,

1991], rhesus monkey [Verhage et al., 1997], and baboon [Boice et al., 1990]. In the

mouse, OSP was not associated with the zona pellucida and was only localized within the

perivitelline space [Kapur and Johnson, 1985]. An interesting observation by Kan et al.

(1995) was that hamster OSP localized to the flocculent material in the perivitelline space

in oviductal oocytes only after fertilization. These investigators suggest that OSP

interacts with secretions of corticol granules after fertilization and may be involved in the

block to polyspermy. This hypothesis is supported by our work on polyspermy described

in Chapter 4.

While observations for an association of OSP with the oviductal oocyte or early

embryo are consistent across species studied thus far, a similar consensus does not exist

for localization of OSP to spermatozoa. In the hamster [Kimura et al., 1994, Boatman

and Magnoni, 1995] and cow [King and Killian, 1994, Abe et al., 1995], OSP has been

shown to associate with the spermatozoal membrane. However, a report in the human







47

[Reuter et al., 1994] failed to find an association of human OSP with human

spermatozoa.

Molecular Biology ofpOSP

Cloning and characterization of a full length cDNA for pOSP has been reported

by Buhi et al. (1996). A full length clone (2022-bp) was identified and sequenced,

yielding an open reading frame of 1581 bp that coded for a protein of 527 amino acids.

A putative signal sequence was identified (a.a. 1-21) and is highly conserved, having

100% homology among bovine, human, and ovine signal sequences. Sequence analysis

of pOSP revealed 3 consensus N-glycosylation sites as well as several potential O-

glycosylation sites, suggesting large amounts of carbohydrate addition. This would also

explain the differences between the predicted native molecular weight of 55,600 and

those observed by 2D-SDS-PAGE (El; 75 k, E2,3; 100 k M,). The pOSP cDNA

sequence revealed significant identities and similarity to oviductal secretary

glycoproteins (OSP) from a number of other species, especially in the 5' end of the open

reading frame. Sequence analysis of pOSP cDNA also revealed two potential

phosphorylation sites, a consensus heparin-binding sequence, a region similar to the

chitinase catalytic site, and a potential C-terminal region similar to a chitinase binding

domain. However, this protein does not have chitinase activity and is missing an

essential amino acid required for such activity. Observations have been made

establishing a structural similarity between OSPs and the chitinase protein family [Arias

et al., 1994, Sendai et al., 1995, Suzuki et al., 1995, Buhi et al., 1996]. Sendai et al.

(1995) has suggested that the chitinase-like structure in OSPs may be involved with

carbohydrate moieties of the oocyte or surface of the spermatozoa. Since chitinase binds







48

to chitin (poly-pl,4-N-GlcNAc) and hydrolyzes it, Suzuki et al. (1995) suggests that

OSPs may be a GlcNAc binding protein, which interacts with carbohydrate components,

located on the zona pellucida. OSPs have been cloned in several other species including

the baboon [Donnelly et al., 1991], human [Arias et al., 1994], hamster [Suzuki et al.,

1995], mouse [Sendai et al., 1995], cow [Sendai et al., 1994], sheep [DeSouza and

Murray, 1995] and rhesus monkey [Verhage et al., 1997]. These researchers also found

that a significant degree of homology exists among oviduct-specific secretary

glycoproteins of various mammalian species studied thus far. However, OSP cDNA

appears to lack homology to the frog oviduct-specific protein-1 gene found in Xenopus

oviduct cells [Donnelly et al., 1991]. Sequence analysis of the various species listed

above revealed several N- and O-glycosylation sites, which may reflect why differences

are observed between species in isoelectric point and molecular weight when evaluated

by 2D-SDS-PAGE. Identification of the polypeptide precursors and characterization of

their biosynthetic maturation within oviductal cells for hamster OSP has been examined

in detail and hypothesis on the polymorphism of these variants due to translational and

posttranslational events has been reviewed by Malette et al. (1995). Southern blot

analysis ofgenomic DNA has shown that there is a single copy of the hamster OSP gene

in the hamster genome and PCR amplification revealed that it is contained within a single

exon, excluding the possibility of alternative splicing [Paquette et al., 1995]. Although

these studies have not been done in other species to date, immunological and northern

blot analysis suggest that the family of OSPs detected by 1- and 2D-SDS-PAGE all

derive from a single gene. Hamster OSP has recently been referred to as a secretary

mucin, due to extensive O-glycosylation (> 50%) not seen in other species, tandemly







49

repeated amino acid motifs, and the presence of multiple alleles (Paquette et al., 1995].

These tandem repeats are present in rodents and humans but are not seen in ungulates,

raising interesting questions from an evolutionary viewpoint. A review of the various

sequences cloned revealed significant differences in the C-terminal region of the OSPs

and may reflect species differences in the OSP molecule. For instance, Sendai et al.

(1995) identified a unique seven-residue repeat sequence (21 repeats) in the C-terminal

side of mouse OSP which are not seen in cow, sheep, pig or baboon OSP.

Probes for pOSP mRNA detected only a single 2.25 kb message in oviductal

tissue and mRNA was present in all 3 segments of the oviduct. Levels of mRNA were

greatest in the ampulla indicating this segment as the primary site of its expression.

Northern blot analysis of numerous other tissues (heart, intestine, cervix, kidney,

endometrium, myometrium, spleen, lung, aorta, liver, or stomach) confirmed that

expression of this protein is specific to the porcine oviduct. Steady-state levels of pOSP

mRNA in the oviduct were greatest at estrus (Days 0,1) consistent with elevated estrogen,

decreased rapidly by Day 2, and remained low throughout diestrus. These data correlates

well with previous reports on pOSP protein synthesis [Buhi et al., 1990, Buhi et al.,

1992]. Levels of pOSP mRNA in oviducts of steroid-treated OVX gilts confirmed that

this protein's expression is up-regulated by estrogen and that progesterone antagonizes

this effect. Data from other species including human [Arias et al., 1994], baboon

[Donnelly et al., 1991], and bovine [Sendai et al., 1994] have also shown that the

message for this protein is greatest during the late follicular phase and in estradiol-

dominated oviducts.







50

A study in the hamster, utilizing in situ hybridization for detection of OSP

message, found evidence for OSP in both the ampulla and isthmus, although the signal

intensity was greatest in the ampulla [Komiya et al., 1996]. The distribution of OSP

message was different within the two sections, with the ampulla having signal detected in

both the perinuclear and basal regions, while the isthmus only had signal in the basal

compartment of epithelial cells. Komiya et al. (1996) also found that the message for

OSP is greatest when serum estradiol/progesterone levels are higher and that the mRNA

for OSP wanes with age and correlates with decreasing serum estradiol levels. However,

these investigators also showed that progesterone is required for hamster OSP gene

expression, unlike data from the baboon [Donnelly KM et al, 1991] and pig [Buhi et al.,

1996], which shows OSP strongly suppressed by progesterone. Another interesting

difference is that the hamster [Paquette et al., 1995] rabbit [Donnelly et al., 1991], and

mouse [Donnelly et al., 1991] showed a constant level of mRNA expression for OSP

throughout the estrous cycle which differs from that of other species observed thus far

where a strong estrogen-dependence is required for its expression. Contrary to findings

by Komiya et al. (1996), Murray and DeSouza (1995) found that transcripts encoding

sheep OSP were in the basal compartment and at the apical tips in fimbria and ampulla

epithelial cells at the free margins of mucosal folds. Because mRNA was localized to the

apical tips, these investigators suggested that mRNA encoding OSP was translated at a

unique cytoplasmic foci. Electron microscopy reveals that all the cytoplasmic machinery

required for protein synthesis is also located within the apical tips of oviductal epithelial

cells, which suggests a quick response time from the point of translation to release of

OSP into the lumen [Murray and DeSouza, 1995].







51

Cyclical, structural, and functional changes in the oviduct are thought to be

brought about primarily by the actions of estradiol and progesterone [Verhage and Jaffe,

1986], and numerous reports have been cited above showing regulation of OSPs and

pOSP by estrogen/progesterone levels. Recent studies have demonstrated that human

[Lie et al, 1993] and pig [Gawronska et al., 1999] oviducts contain luteinizing hormone

(LH)/human chorionic gonadotropin (hCG) receptors, suggesting that factors other than

circulating ovarian steroids may regulate oviductal function. One study by Sun et al.

(1997), revealed that culturing bovine oviductal epithelial cells with hCG resulted in a

time and dose-dependent increase of OSP protein and transcript. A nuclear transcription

run-on assay was used to determine that hCG did not increase the transcription rate of the

gene but rather "stabilized" its transcript [Sun et al., 1997]. These investigators suggested

that higher LH levels present during the periovulatory periods may act to stabilize the

message for OSP in the oviduct, thus leading to an increased secretion rate. Recently,

heterozygous mutant mice have been generated, which lack a large portion of the OSP

coding region [Sendai et al., 1999]. In the future, homozygous offspring may clarify the

role of this protein in vivo.

Biological Actions of OSPs

The observations that in all species examined so far, with the notable exception of

the mouse, OSPs associate with the zona pellucida, suggest an intimate role for these

estrogen-dependent proteins during fertilization and early cleavage-stage embryonic

development. However, the abundant information detailing a morphological association

of OSPs with spermatozoa, oocytes, or embryos does not correspond to a suitable amount

of functional information. Functional information regarding OSPs have been hindered by







52

difficulties in the protein purification process. In fact, no direct in vivo data on precise

functions of these proteins are available to date. However, several in vitro observations

of activities associated with the inclusion of OSP have led to some attractive hypotheses.

Macromolecules secreted by the hamster oviduct in the ampulla region have been

shown to facilitate penetration of sperm through the egg investments [Boatman et al.,

1994] and follicular eggs collected from the ampulla were shown to be more penetrable

and fertilizeable than follicular eggs collected from the ovarian bursa or unovulated

follicles [Boatman and Magnoni, 1995]. Similar increases in penetration rates have been

observed in the bovine [Martus et al., 1998] and were linked to the addition of

concentrated bovine OSP during in vitro fertilization (IVF). One conflicting report in the

hamster [Kimura et al., 1994] indicates that penetration rates were decreased in the

presence of OSP. These data, although partially conflicting, indicate that association of

OSP with the oocyte may have a functional role during fertilization. It is unknown

whether these observed effects were due to actions on spermatozoa, oocyte or both.

Martus et al. (1997) showed that exposure of bovine oocytes to bovine OSP during any

phase of the IVF process increased fertilization rates, while exposure of spermatozoa to

bovine OSP had no effect.

Several investigators suggest that OSPs induce capacitation [McNutt et al., 1992,

Anderson and Killian 1994] and/or the acrosome reaction. Therefore, an increase in the

number of capacitated or acrosome-reacted spermatozoa could possibly lead to the above

observations on increased penetration. However, direct evidence for the estrogen-

dependent OSP on these activities is lacking.







53

Another potential modulator for this increase in penetration may be due to actions

on spermatozoa movement characteristics. Abe et al. (1995) found that bovine OSP

effectively maintained the viability and motility of bovine spermatozoa relative to control

medium and that this activity was dose-dependent. The ability of bovine OSP to

maintain motility and viability may allow a greater proportion of spermatozoa to

penetrate the egg investments and subsequently fertilize the oocyte. The OSPs may also

have a role in gamete recognition and binding. In the hamster [Schmidt et al., 1997a,

1997b] and human [O'Day-Bowman et al., 1996], OSP was found to increase the number

of tightly bound sperm attached to the zona pellucida. This increase might also lead to

the increased penetration rates described above. These data are supported by the

observation that a redistribution in the localization of hamster OSP occurs depending on

the state of capacitation [Boatman and Magnoni, 1995].

Functional data on embryonic development are especially lacking, and

investigations to date are less than encouraging. No effects of bovine OSP

[Vansteenbrugge et al., 1997] or ovine OSP [Hill et al., 1997] were observed on bovine

embryonic development. Numerous investigations have shown the need for homologous

systems when studying this protein, therefore observations by Hill et al. (1997) may be

due to the heterologous system employed. Previous studies by Hill et al. (1996 a, b)

examining ovine OSP effects on ovine embryonic development demonstrated several

subtle but consistent results. The most significant of these being that a reduced

proportion of one-cell embryos underwent first cleavage, but no overall decrease on the

proportion of blastocysts that formed. These investigators suggest that this may reflect a

selection mechanism occurring in vivo. Bovine OSP has been found to increase the







54

number of bovine zygotes that cleaved compared to controls, but by Day 7 of embryonic

development, no difference could be detected in the number of blastocysts that formed

[Martus et al., 1997]. However these investigators did observe an increased number of

blastocysts on Day 6 of in vitro culture. This may suggest that bovine OSP, while not

increasing the number of blastocysts, increases their rate of development. However, a

slight loss in the number of blastocysts between Day 6 and Day 7 was shown and might

indicate that the concentration of bovine OSP tested or length of exposure (>6 days) is

detrimental to subsequent development. This increased rate of development may be due

to an unknown stimulation on the expression of maternal or embryonic genes and

increased protein synthesis. One study indicates that in vivo fertilized pig embryos (one-

or two-cell stage), cultured in vitro in the presence of one subunit (97 k Mr) of semi-

purified pOSP, showed increased rates of incorporation of methionine into protein at the

four-cell stage [Wallenhaupt et al., 1996]. The data described above indicate that pOSP

may facilitate or enhance fertilization and early cleavage-stage embryonic development.

This review thus leaves a number of unanswered questions which need to be

addressed. The first is; what are the unidentified de novo synthesized and secreted

macromolecules of the oviduct? The second is; for those proteins which have been

identified and characterized, what is the specific functions) in relation to reproductive

processes in the oviduct; union of gametes, fertilization, and early cleavage-stage

embryonic development. Chapters 3, 4, and 5 describe the identification,

characterization, and regulation of PAI- in the oviduct. The objectives of these three

chapters include; 1) localization of PAI-I within the oviduct, 2) hormonal regulation of

oviductal PAI-1 by ovarian steroids, 3) expression of PA-1 mRNA and protein during







55

early pregnancy, 4) evaluation of PAI-1 and uPA activity in the oviduct during early

pregnancy, and 5) localization of PAI-1 on oviductal oocytes and embryos. Chapter 6

will examine the functionality of pOSP during in vitro fertilization and embryo culture.

Effects of pOSP on the fertilization rate, polyspermy rate, and development to blastocyst

will be examined. Results of these studies will begin to define the roles of oviductal PAI-

1 and pOSP during fertilization and embryonic development.

















Figure 2-1. Pericellular activation cascade for plasminogen and matrix metalloproteinase.

uPA, urokinase plasminogen activator; uPA-R, urokinase plasminogen activator receptor;
PAI-1, plasminogen activator inhibitor-1; a2 AP, alpha-two anti-plasmin inhibitor; TIMP,
tissue inhibitor of matrix metalloproteinase; MMP, matrix metalloproteinase. Proteolytic
enzymes and inhibitors (outlined in blue) involved in the regulation of fibrinolysis and
extracellular matrix remodeling/degradation.















cell
tPAl-1
TIMPns


pla II noTel s IIi















CHAPTER
IDENTIFICATION AND LOCALIZATION OF PLASMINOGEN ACTIVATOR
INHIBITOR-1 (PAI-1) IN THE PORCINE OVIDUCT


Introduction

The porcine oviduct provides an important microenvironment for final maturation

of gametes, fertilization, and early cleavage-stage embryonic development. In part to

provide an effective environment for these reproductive processes, numerous proteins

derived from serum as a transudate or from oviductal epithelium contribute to the

composition of oviductal luminal fluid [Feigelson and Kay, 1972, Sutton et al., 1984,

Buhi et al., 1997]. Oviductal luminal fluid and its protein composition vary during the

estrous cycle and early pregnancy, and maximal levels of oviductal fluid appear to

coincide with elevated estrogen [Hunter, 1988]. When estrogen is maximal, Days 0 and 1

of the estrous cycle or early pregnancy, biosynthetic activity in the oviduct is greatest

[Buhi et al., 1989]. The infundibulum and ampulla, regardless of day of the estrous cycle

or early pregnancy, have a biosynthetic activity significantly greater (2-3 times) than that

of the isthmus [Buhi et al., 1997]. This activity, measured in explant culture-conditioned

media, reflects the de novo synthesis and secretion of secretary proteins from these three

segments.

The relative importance of oviductal-derived proteins to fertilization and early

embryonic development is not known, nor have many of these proteins been identified.

Several studies have shown the importance of utilizing the oviduct or its constituents for

58







59

enhancing embryo development in vitro [Archibong et al., 1989, White et al., 1989, Liu

et al., 1998]. In the pig, glycoproteins derived from oviductal fluid have been shown to

associate with the zona pellucida of oocytes after ovulation, at fertilization and during

early embryonic development [Brown and Cheng, 1986, Hedrick et al., 1987, Buhi et al.,

1993]. It has been suggested that these proteins may be facilitating the beneficial effects

of oviductal epithelial cell co-cultures on embryo development in vitro [Buhi et al.,

1997].

The pig oviduct has been shown to synthesize and secrete de novo 14 major

proteins into explant culture medium [Buhi et al., 1990]. The majority of these proteins

have been described electrophoretically by isoelectric point and relative molecular mass.

Two of these proteins have been identified and characterized, the porcine oviduct-specific

secretary glycoprotein (pOSP) [Buhi et al., 1992] and tissue inhibitor of matrix

metalloproteinase-1 (TIMP-1) [Buhi et al., 1996]. Of the numerous proteins yet to be

identified, one protein with an apparent molecular weight of 45,000, appeared to be

synthesized and distributed similar to TIMP-1. Like TIMP-1, this protein was shown to

incorporate 3H-glucosamine, be composed of 5-6 isoelectric species, and be synthesized

primarily by the isthmus [Buhi et al., 1990]. Thus, with characteristics similar to

oviductal TIMP-1, it was suggested that the 45,000 Mr protein may also be a protease

inhibitor.

This study was designed to further identify and characterize the 45,000 M, de

novo synthesized protein of the porcine oviduct and to better understand specific protein

contributions of each oviductal segment relative to ovulation, gamete transport,

fertilization and early cleavage-stage embryonic development. With identification of the







60

unknown 45,000 M, protein as porcine plasminogen activator inhibitor (PAI)-1

(described here in Chapter 3), studies were designed to examine its synthesis within the

infundibulum, ampulla, and isthmus, and to evaluate its distribution throughout the

oviduct. While PAI-1 has been localized in the uterus and ovary from a variety of

species and potential mechanisms examined, no studies to our knowledge have been done

to assess PAI-1 within the oviduct.

Materials and Methods

Materials

Acrylamide, N,N' diallyltartardiamide, urea, Nonidet P-40, and sodium dodecyl

sulfate were purchased from Gallard-Schlesinger (Carle Place, NY); X-Omat AR film

and photography reagents were a product of Eastman Kodak Co. (Rochester, NY); amino

acids and protein standards were purchased from Sigma-Aldrich (St. Louis, MO);

ampholines were from Pharmacia-Biotech (Piscataway, NJ); all other supplies and

reagents for gel electrophoresis were purchased from Bio-Rad Laboratories (Richmond,

CA) or Fisher Scientific (Orlando, FL). All medium and culture supplies were purchased

from Life Technologies (Grand Island, NY). L-[4,5-3H]leucine (specific activity, 120

Ci/mmol) was obtained from Amersham (Arlington Heights, IL). Immobilin-P (PVDF)

membranes were purchased from Millipore Corporation (Bedford, MA); Vectastain ABC

Elite kit was obtained from Vector Laboratories (Burlingame, CA). Affinity-purified

goat anti-human PAI-1 was purchased from American Diagnostica (Greenwich, CT);

goat IgG and normal rabbit serum (NRS) were purchased from Sigma-Aldrich. All other

reagents, including column chromatography supplies, were products from Sigma, Fisher,

Life Technologies or Pharmacia-Biotech.









Tissue Collection. Explant Culture, and Electrophoresis

Florida crossbred (Yorkshire x Duroc x Hampshire) or European Large White

gilts were observed daily for behavioral estrus in the presence of an intact boar. After the

completion of at least two estrous cycles, animals for pregnancy studies were bred on the

first day of standing estrus, designated as Day 0, and 24 h later (except for gilts assigned

to Day 0). Pregnant or cyclic gilts were taken to the abattoir on Days 0, 2, and 12 for

slaughter. Oviducts were collected aseptically and separated by gross dissection into the

three functional segments, which were then cultured as previously described [Buhi et al.,

1989]. For culture, 0.5 g of tissue was incubated in leucine-deficient (modified) Eagle's

minimum essential medium (MEM) containing 100 PCi [3H]-leucine for 24 h at 390 C in

a defined atmosphere. Conditioned culture media was then separated from tissue and

frozen at -200 C until analyzed by 2D-SDS-PAGE and fluorography or utilized for

purification of the 45,000 Mr protein.

In order to evaluate synthesis of the 45,000 Mr protein within the three segments,

oviductal tissue was collected from Day 2 pregnant Large White gilts and cultured as

described above. Culture media was dialyzed (12,000 molecular weight cut-off) against

10 mM Tris-HC1 buffer (pH 7.6), containing 0.15 M NaCI and 0.02% (w/v) NaN3,

followed by dialysis against dH20 (2 changes, 4L each, 24 h each) at 40 C. Total protein

content of dialyzed culture media was measured by the Bio-Rad microassay (according to

manufacturer's instructions) and radiolabeled proteins measured by liquid scintillation

spectrophotometry. Samples representing each oviductal segment, containing 100,000

cpm, were lyophilized, solubilized in Laemmli's buffer [Laemmli, 1970] and analyzed







62

by 2D-SDS-PAGE and subjected to fluorography as described previously [Buhi et al.,

1995]. All X-ray films were exposed for 14 days at -800 C and developed.

Protein Fractionation

Explant culture media conditioned by isthmic tissue on Days 0, 2, and 12 of early

pregnancy, were pooled and subjected to gel filtration chromatography on a Sepharose

CL-6B column as described previously [Buhi et al., 1990] with some modifications. The

Sepharose CL-6B column (1.8 x 92 cm) was equilibrated in column buffer [10 mM Tris-

HCI, 0.4 M NaCI, 0.02% (w/v) NaN3; pH 7.5] at 40 C. Culture media (7 ml) centrifuged

at 2,200 x g for 10 minutes at 40 C to remove particulate material, was added slowly to

the column. After collection of the void volume, elution profiles were generated by

collecting 2 ml fractions, measuring protein (absorption at 280 nm) and determining

radioactivity in each fraction. Chromatographic peaks corresponding to the elution of

ovalbumin (45,000 Mr standard) were pooled from individual column runs.

Immunoglobulins were removed from pooled fractions by incubation with Protein A-

Sepharose beads in 10 mM phosphate-buffered saline (PBS), pH 8.0, overnight at 40 C.

Beads were then separated from the supernatant by centrifugation at 1,700 X g for 10

minutes at 40 C, washed 3x in PBS, and pooled supernatants dialyzed against dH20 (2

changes, 4 L each, 24 h each, 40 C). The dialyzed sample was analyzed for protein

content and radioactivity, as indicated above, and lyophilized. Lyophilized samples were

then resuspended in column buffer (lml/15 mg protein) and further fractionated on a

Sephadex G-100 column (1.5 x 75 cm) at 40 C. The Sephadex G-100 column was

calibrated previously with Blue Dextran, apotransferrin, ovalbumin and cytochrome C as

molecular weight markers. Elution profiles were generated using methodology







63

established for the Sepharose CL-6B column. The pooled chromatographic peaks from

individual column runs corresponding to the 45,000 Mr standard were utilized for

subsequent Western blotting and N-terminal amino acid sequence analysis. Purification

at each step was examined by 2D-SDS-PAGE and fluorography [Buhi et al., 1990].

Western Blotting and Sequence Analysis

The pooled chromatographic peaks containing the 45,000 Mr protein were

separated by Tris-tricine 2D-SDS-PAGE [Schagger and von Jagow, 1987]. Proteins were

then transferred by semi-dry electrophoresis (Milli-Blot SDE system, Millipore Corp,

Bedford, CA) to a PVDF membrane as described previously [Buhi et al., 1995] with

some modifications. Following separation of proteins by Tris-tricine 2D-SDS-PAGE, the

gel was rinsed in three changes of dH20 for 5 min each, then equilibrated in 25 mM Tris-

HCI buffer (pH 9.4) containing 2% (w/v) SDS, for 30 min. The gel was washed in three

changes of dH20 for 5 min, and incubated in two changes of cathode C buffer (25 mM

Tris-HCl, 5.25 g/L norleucine, 10% (v/v) methanol; pH 9.4) for 10 min at room

temperature. After assembly of the TransUnit sandwich, proteins were transferred for 1 h

under constant current (2.5 mA per cm2). Subsequently, the membrane was washed

(dH20, 5 min), and stained in 0.1% (w/v) Coomassie blue R-250 in 50% (v/v) methanol

for 1 min. The membrane was destined (50% methanol, 5 min) with constant rocking

and rinsed in three changes of dH20. The PVDF membrane was allowed to air dry,

proteins corresponding to the 45,000 Mr protein were excised, and subjected to N-

terminal amino acid microsequencing at the Interdisciplinary Center for Biotechnology

Research (ICBR) facility using a 470A gas phase protein sequencer (Applied

Biosystems) with an on-line analytical HPLC system. The peptide sequence was







64

analyzed with the National Center for Biotechnology Information (NCBI) BLAST

program [Altschul et al., 1990].

After identification of the 45,000 M, protein as porcine PAI-1 (see Results), a

one-step partial-purification method was employed using heparin-agarose affinity column

chromatography. PAI-1 has been shown previously to quantitatively bind to heparin-

Sepharose and elute with increasing concentrations of NaCI [Ehrlich et al., 1991].

Isthmic-conditioned culture media (Day 12 pregnant and Day 1 cyclic) were pooled,

centrifuged as described above, diluted (1:3) in 20 mM Tris-HCI (pH 7.6, 40 C)

containing 0.02% (w/v) NaN3, and slowly loaded onto a heparin-agarose column (2.5 x

8.2 cm) at 40 C. PAI-1 was eluted utilizing stepwise increments of NaCl (0.1-3.0 M), and

the protein pooled and dialyzed against dH20 (two changes, 24 h each, 4L each, 40 C).

Protein content was determined as above, and aliquots containing 0.5 mg of protein

lyophilized and used for immunoprecipitation. Fractions were analyzed for the presence

of PAI- by 2D-SDS-PAGE and fluorographic analysis.

Immunoprecipitation

PAI-1, semi-purified by heparin-agarose affinity column chromatography, was

immunoprecipitated with a polyclonal rabbit anti-hPAI-1 antiserum (kindly provided by

Dr. Schleef, Scripps Institute, La Jolla, CA). Lyophilized protein samples (0.5 mg) were

solubilized in 900 Il of NET buffer [50 mM Tris-HC1, 0.15 M NaCI, 0.1% (v/v) Nonidet

P-40, 1 mM EDTA, 0.25% (w/v) gelatin, and 0.02% (w/v) NaN3; pH 7.5 at 250 C].

Protein A-Sepharose beads were equilibrated in NET buffer and 100 Al of swollen beads

were incubated with 100 pl of either undiluted PAI-1 antiserum or NRS and 300 ul of

NET buffer for 1 h at 250 C. After incubation, Protein A-Sepharose beads were pulse-







65

centrifuged (Beckman, microcentrifuge) for 30 sec and washed in three changes of NET

buffer (0.5 ml). Semi-purified PAI-I protein (100-200 pl) was then incubated with NRS

or rabbit anti-hPAI-1 antibody-coated beads (200 il), respectively, for 2 h at 250 C with

constant rotation. Complexes were pelleted by centrifugation and washed as above.

Proteins conjugated to the Protein A-Sepharose beads were solubilized in Laemmeli

buffer [Laemmeli, 1970], boiled for 3 min, separated on a 10% (w/v) 1D-SDS-PAGE gel

and subjected to fluorography [Buhi et al., 1995]. Heparin-agrarose fractionated proteins,

which were not used for immunoprecipitation, were removed from solution using

standard acetone precipitation procedures [Harlow and Lane, 1988] and solubilized in

Laemmeli buffer as above for positive radiolabeled PAI-I identification.

Immunocvtochemistry

A polyclonal affinity-purified goat anti-hPAI-1 was used to immunolocalize PAI-

1 in porcine oviductal tissues from cyclic and early pregnant animals. To compare

distribution of PAI-1 in cyclic and early pregnant animals, infundibulum, ampulla and

isthmic tissues were collected on Days 0, 2, and 12, (n=3 animals/day) and

immunocytochemistry performed as described [Chegini et al., 1992]. Tissues were cut

into 5 mm portions, fixed in Bouin's solution, embedded in paraffin, sectioned (0.5 pm),

and mounted on precoated glass slides. A goat Vectastain ABC Elite kit (Vector

Laboratories, Burlingame, CA) was used according to the manufacturer's instructions.

Controls included use of an affinity-purified goat IgG and the absence of primary

antibody. Goat anti-hPAI-1 and goat IgG were used at a dilution of 1:10 in PBS (pH

7.4).









Immunogold Electron Microscopy (EM)

Oviductal tissue from the three segments of Day 0 non-pregnant and Day 9

pregnant crossbred gilts, were fixed for 1 h in PBS, pH 7.4, containing 0.5% (v/v)

glutaraldehyde, 4% (v/v) paraformaldehyde at 40 C. The two days selected were times of

elevated estrogen (Day 0) or progesterone (Day 9) production. After fixation and rinsing

in PBS, tissues were dehydrated in graded ethanol series and embedded in Unicryl

(British BioCell International, UK ) under UV light at -100 C for 2 days. Thin sections

(0.5 mm) were cut and collected on Formvar-coated 100 mesh nickel grids, and PAI-I

antigen detected by immunogold labeling. The polyclonal rabbit anti-human PAI-I and

preimmune rabbit sera, diluted 1:1000 in a high salt Tween buffer (0.02 M Tris-HCI, 0.5

M NaCI, 1% [v/v] Tween 20, pH 7.2) supplemented with 1% (w/v) ovalbumin, were

incubated overnight with grids in a humid chamber at 40 C. Sections were then

incubated with a secondary antibody (goat anti-rabbit IgG, 1:30 dilution in PBS)

conjugated to 18 nm colloidal gold (Jackson ImmunoResearch Laboratories, Inc, West

Grove, PA) for 1 h at room temperature. Sections were then post-stained with 2% (w/v)

uranyl acetate and Reynolds lead citrate. Grids were examined on a Hitachi H-7000

transmission electron microscope (Hitachi Scientific Instruments, Danbury, CT). Digital

micrographs were taken on a Gatan BioScan/Digital Micrograph 2.5 (Gatan Inc,

Pleasanton, CA).

Results

Electrophoretic Analysis

Representative 2D-SDS-PAGE and fluorographic analyses of explant culture

media from the infundibulum, ampulla, and isthmus containing radiolabeled de novo







67

synthesized proteins are shown in Figure 3-1. The 45,000 Mr protein, found in all three

segments, is seen as the major radiolabeled protein from the isthmic portion of the

oviduct (Figure 3-1C). The level of protein expression appears to be reduced in the

ampulla (Figure 3-1B) and nearly absent in the infundibulum (Figure 3-1A). Thus, the

45,000 Mr protein appears to have a specific spatial synthesis and release in the oviduct.

This specific spatial expression is similar to that reported for the matrix metalloproteinase

inhibitor, TIMP-1, in the oviduct [Buhi et al., 1996] (Figure 3-1).

Purification, Blotting, and N-terminal Sequencing

With identification of the 45,000 Mr protein as the major radiolabeled protein of

the isthmus, the next objective was to purify this protein for N-terminal sequencing. A

two-step gel-filtration chromatographic procedure was developed in order to separate and

enrich this protein. Elution profiles for protein and radioactivity from Sepharose CL-6B

and Sephadex G-100 columns are shown in Figure 3-2A and 3-2B, respectively.

Fractionation of explant culture media on these two columns, each showed two peaks of

radioactivity. The broad second peak in both graphs, contained the greatest amount of

radioactivity and included the 45,000 Mr protein. A representative fluorograph of the

45,000 Mr protein after fractionation on a Sephadex G-100 column is shown in Figure 3-

2C. This protein appears to contain at least 5 isoelectric species, of which 3 species,

including an acidic and basic species, were submitted for N-terminal amino acid

sequence analysis. A search of protein, RNA and DNA data banks indicated that the

derived N-terminal amino acid sequence of 26 amino acids, identical for all three

isoelectric species, was 96% identical (100% similar) to porcine PAI-1 (Figure 3-3). This

sequence (1-26) corresponded to amino acid positions 20-46 of mature PAI-1 protein,







68

indicating removal of the hydrophobic leader peptide prior to release from the cell

[Bijnens et al., 1997].

Immunoprecipitation

To confirm that the 45,000 Mr protein identified by N-terminal amino acid

microsequencing was PAl-1, an anti-hPAI-1 serum was used for immunoprecipitation of

this protein from isthmic culture media after fractionation by heparin-agarose affinity

column chromatography. Fractionated culture media containing de novo synthesized

radiolabeled proteins were treated with either Protein A-complexed rabbit anti-hPAI-1

serum or Protein A-complexed NRS. Bound proteins were solubilized and examined by

1D-SDS-PAGE and fluorography (Figure 3-4A and 3-4B). As shown in Figure 3-4B,

anti-hPAI-1 serum specifically recognized and precipitated radiolabeled PAl-1, while

NRS did not. These results indicate that this antibody does not cross-react with other

radiolabeled proteins present within the fractionated culture media. Western Immunoblot

analysis showed that the anti-hPAI-1 antisera cross-reacted with only 2-3 minor

unlabeled proteins, possibly oftransudate origin. The primary protein recognized by this

antibody on the Western Immunoblot was the PAI-1 family (5-6 isoelectric species) (data

not shown).

Immunocytochemistry

With identification of PA-1, the next objective was to examine its distribution

throughout the oviduct on Days 0, 2, and 12 of the estrous cycle or early pregnancy.

Immunoreactive PAl-1 was detected in all three segments of the oviduct regardless of

day of cycle examined and no differences in staining intensity could be detected between

days (infundibulum not shown). Representative data of the immunocytochemical







69

localization in the isthmus on these days is shown in Figure 3-5. No difference in

staining intensity could be detected (subjective visualization of three representative

animals) between pregnant and cyclic tissues within the three segments (Figure 3-6).

PAI-1 was localized primarily within the oviductal epithelium, while only background

staining could be identified within muscle and stroma tissue (Data not shown for muscle

and stroma). Here, PAI-1 appeared to be heavily concentrated at the apical region of the

epithelium (Figure 3-5C, arrow) with little staining found in the basal region. However,

PAI-I immunoreactivity was also associated within cells lining blood vessels in the

stroma. Immunocytochemistry pictures shown are representative of 3 animals/day.

Because staining intensity varied between each animal examined, subjective

comparisons on the level of staining intensity between pregnant and cyclic gilts were not

made.

Electron Microscopy

Since immunocytochemistry was not able to resolve PAI-1 association with

specific cellular structures or organelles within the epithelium, an examination of PAI-1

distribution within the epithelium was performed using electron microscopic

immunogold localization. PAI-I was found primarily in epithelium of isthmic non-

ciliated secretary cells and appeared to be concentrated in putative secretary granules at

or near the apical border of cells in both Day 0 non-pregnant and Day 9 pregnant animals

(Figures 3-7A and 3-8A). In addition, this protein was also found to be associated with

the isthmic luminal epithelial border and cilia (Figures 3-7B and 3-8B). Localization of

PAI-1 within ampullary epithelium (both ciliated and non-ciliated) was negligible (Figure

3-8C). Isthmic tissue incubated with preimmune sera (control) showed no labeling







70

within secretary granules or at the epithelial border (Figure 3-8D). As detailed in the

immunoprecipitation results, the antibody employed showed no potential cross-reactions

with other proteins found within oviduct-coniditioned culture medium.

















Figure 3-1. Representative fluorograph after 2D-SDS-PAGE separation of [HI-labeled
proteins (100,000 cpm) from explant culture media conditioned by the three segments of
the oviduct.

Tissue was collected from the A) infundibulum, B) ampulla or C) isthmus of Large
White X Landrace gilts on Day 2 of pregnancy. The 45,000 Mr protein is indicated by an
arrow in each panel. TIMP-1 is marked by an arrowhead only in panel C. Molecular
weight markers (x 103) are indicated and the pH gradient runs from left (pH 8) to right
(pH 4).







72




















s. ? r *i
iT -

U -


4* .

















Figure 3-2. Representative Sepharose CL-6B and Sephadex G-100 fractionations of
[3H]-labeled proteins in isthmus-conditioned medium.

Fractions (2 ml) collected from the CL-6B column elution (Panel A) were counted for
radioactivity and measured for protein (see Methods). In Panel A, tubes 58-75 (peak 2,
bracket) containing the 45,000 M, protein, were pooled, treated with Protein A beads to
remove immunoglobulins and further factionated on a G-100 column (Panel B). In panel
B, tubes 48-60 (peak 2, bracket), containing the 45,000 Mr protein were pooled, and
separated by Tris-tricine 2D-SDS-PAGE for electroblotting. Arrows in panel A and B
reflect elution point of 45,000 Mr ovalbumin molecular weight standard. Panel C, shows
a representative fluorograph of the 45,000 Mr protein after chromatographic separation by
Tris-tricine 2D-SDS-PAGE. Three isoelectric species (marked with arrows) were
excised and subjected to N-terminal amino acid sequence analysis. Molecular weight
markers (x103) are indicated and the pH gradient runs from left (pH 8) to right (pH 4).












1 20 00 ,n


.0W






o 10 s 30 40 5a so 70 0









oI -ieo
a 10 20 0 40 so 1 7




o tio a so
vms--------------- --, R







Fraion anuir

C
97
66





29' *r 0
~ ~ 10 M M W C t *
















20.... 25 .... 30 ... 35 .. 40 .... 45 .
Porcine
EGSAS SHHQS LAARL ATDFG VKVFR QV PAI-1

EGSAS SHHQS LAARL ATDFG VKXFR QV 45,000 M,
isthmic protein
1.. 5 .. .. 10 .. 15 .. .. 20 .. 25 .



Figure 3-3. Comparison of the identified N-terminal amino acid sequence of the 45,000
Mr protein from isthmic-conditioned culture media to the N-terminal sequence of porcine
PAI-1.

Proteins were fractionated by size-exclusion chromatography using a Sepharose CL-6B
and Sephadex G-100 column, separated by Tris-tricine 2D-SDS-PAGE, transferred by
electroblotting to PVDF membranes and subjected to N-terminal amino acid micro-
sequence analysis. X (residue 22) indicates an undetermined amino acid residue. A 96%
sequence identity was found to porcine plasminogen activator inhibitor-i (Accession #
1870170).






-----.- 76











2 3

I
















1 2 3




Figure 3-4. Immunoprecipitation of plasminogen activator inhibitor-1 from porcine
isthmic-conditioned culture media after fractionation on a heparin-agarose affinity
column and elution with 0.4 M NaCI.

A representative one-dimensional SDS-polyacrylamide gel (Panel A) containing
Coomasie blue-stained proteins from either a non-precipitated sample (lane 1), proteins
precipitated using a rabbit anti-human PAI-1 (lane 2) or normal rabbit sera (lane 3).
Although not visible on the gel, the arrow indicates the apparent location of PAI-I
protein. Panel B shows a fluorograph generated from the same gel in Panel A. Lanes are
as indicated above. PAI-1 found in semi-purified culture media, was specifically
precipitated by antibody (lane 2), but not normal rabbit serum (lane 3).










rr


Figure 3-5. Representative immunocytochemical localization of PAI-1 in the isthmus
from crossbred gilts on Day 0 (Panel A), Day 2 (Panel B) or Day 12 (Panel C) of
pregnancy.

The control is shown in Panel D. Arrow indicates intense staining of PA-1 within the
apical region ofisthmic epithelium. X 20.







































Figure 3-6. Representative immunocytochemical localization of PAI-1 in oviductal
tissue from Day 2 cyclic (Panel A and C) and Day 2 pregnant (Panel B and D) crossbred
gilts from the isthmus (Panel C and D) and ampulla (Panel A and B).

The infundibulum is not shown. Arrow indicates intense staining of PAI-1 within the
apical region of the epithelium. X 20.

















Figure 3-7. Immunogold labeling of PAI-1 within oviductal tissue from a Day 0 non-
pregnant crossbred gilt.

Gold particles are seen in secretary granules of isthmic non-ciliated cells (Panel A,
arrow) as well as associated with the epithelial luminal border and cilia (Panel B, arrow).
X 20,000.























Figure 3-8. Immunogold labeling of PAl-I within oviductal tissue from a Day 9 pregnant
crossbred gilt.

Gold particles can be seen clustered within putative secretary granules of isthmic non-
ciliated cells (Panel A, arrow) and associated with the epithelial luminal border and cilia
(Panel B, arrow). Gold particles were not detected in secretary granules of the ampulla
although some scattered particles were associated with cilia (Panel C). Immunogold
staining was not detected in the isthmus treated with preimmune serum (Panel D, arrow).
X 20,000.





82









; .,, f-.



P o. ,


^' ,, ;. \.y. %' % ,,
'* j, ." '.^"" .' t-" "

.. -,- -.,- :'.-,',.

,A, ". "' .'." '.,'IL ''Yd










Discussion

The mammalian oviduct secretes numerous de novo synthesized proteins into the

oviductal lumen, some of which are present before ovulation, or during fertilization and

early cleavage-stage embryonic development. In order to characterize the 14 major de

novo proteins secreted by the porcine oviduct, explant tissue (infundibulum, ampulla, and

isthmus), was incubated with [3H]-leucine, and synthesized and secreted proteins

separated by 2D-SDS-PAGE, detected by fluorography, and characterized by molecular

weight and isoelectric points [Buhi et al., 1990]. Two of these proteins have been

extensively examined, while others remain unidentified [Buhi et al., 1997], including a

45,000 M, protein synthesized primarily by the isthmic portion of the oviduct. Sequence

analyses revealed that this isthmic protein was porcine plasminogen activator inhibitor-1.

Recently, an unidentified 45,000 M, protein found in bovine oviductal fluid was shown to

associate with the zona pellucida of bovine oocytes [Staros and Killian, 1998] and may

be the protein under investigation.

PA-1, a member of the serpin family of serine protease inhibitors, is the primary

inhibitor ofurokinase plasminogen activator (uPA) and tissue-type plasminogen activator

(tPA). It is a glycoprotein consisting of 379 amino acids with an apparent molecular

mass of 45,000 [Andreasen et al., 1990]. Among the serpins, PAI-1 is secreted in an

active form that rapidly converts to an inactive latent form that can be reactivated by

phospholipids and a variety of denaturants [Hekman and Loskutoff, 1985]. The

inhibitory activity of PAl-I, however, is stabilized by vitronectin in either serum or

extracellular matrix (ECM), thus preventing transformation to the inactive latent form

[Declerck et al., 1988]. Both tPA and uPA initiate proteolysis by converting plasminogen







84

to the broad-specificity enzyme plasmin. This extracellular protease is reported to be

involved in the remodeling of ECM, fibrinolysis, cell migration, and tumor metastasis

[Andreasen et al., 1990]. Plasmin can also activate pro-matrix metalloproteinases

(MMPs) thereby regulating the pericellular activation cascade leading to ECM

degradation [Murphy et al., 1992]. Complex control of this activation cascade is

regulated by PAI-1 and tissue inhibitors of matrix metalloproteinases (TIMPs).

Immunoprecipitation confirmed the presence of PAI-1 which in the oviduct,

appeared to consist of at least five or more isoelectric species as shown by fluorographic

analysis, N-terminal amino acid sequencing and immunoblot analysis (data not shown for

immunoblot). The presence of these isoforms suggests that newly synthesized and

secreted PAI-1 was post-translationaly modified. One modification was characterized as

glycosylation with incorporation of ['H]-glucosamine into PAI-1 during tissue explant

culture [Buhi et al., 1990]. Localization and synthesis of PAI-1 appears to mimic a

matrix metalloproteinase inhibitor from the oviduct, TIMP-1 [Buhi et al., 1996]. Both

TIMP-1 and PAI-1 show a similar spatial expression by the oviduct segments, with a

greater expression in the isthmus relative to either the ampulla or infundibulum. This

would suggest an important function relative to its spatial expression. The isthmus

portion of the oviduct has been described as a spermatozoa reservoir where sperm

undergo capacitation and hyperactivation [Suarez, 1998]. In the pig, the ampulla-isthmic

junction is the location of fertilization and early cleavage-stage embryonic development

[Buhi et al., 1997]. This protein may therefore facilitate or regulate these important

reproductive events that occur within or near the isthmus. However, PAI-1 may act in

other segments of the oviduct as well, due to retrograde movement of oviductal fluid into







85

the peritoneal cavity during estrus [Hunter, 1988]. Localization of PAI-I in the

infundibulum and ampulla epithelium and protein secretion by explant tissues, suggests

that these segments may also contribute to luminal PAl-1, although synthesis of PAI-1 in

these segments appears to be very low.

Fibrin deposits have been located on the tubal mucosa of the oviduct, which could

possibly interact with the ECM components of the zona pellucida and prevent tubal

transport [Liedholm and Astedt, 1975]. Liedholm and Astedt (1975), observed

fibrinolytic activity associated with the unfertilized ovum in rats and suggested that this

activity may be involved with the prevention of adhesion to fibrin deposits that may

hinder gamete transport. This activity was also shown to be associated with spermatozoa.

Prevention of cellular adhesion to the oviductal mucosa and fibrin deposits might also be

regulated by uPA and tPA. The developing embryo may produce uPA/tPA in response to

signals from the surrounding oviductal environment. Fibrin has been shown to increase

expression of uPA mRNA and protein in 3.5 day old uterine embryos of the mouse

[Zhang et al., 1996]. Both mouse and rat preimplantation embryos have been shown to

have tPA activity [Zhang et al., 1992, Carroll et al., 1993] and uPA activity [Zhang et al.,

1994, Harvey et al., 1995]. Expression of this proteolytic activity, or the respective

mRNA, was found to be developmental- and stage-specific. While the fibrinolytic

activity of the ovum and proteinase expression by the embryo may facilitate their

transport through the oviduct, these molecules are potent modulators of their immediate

environment with respect to ECM remodeling. PAl-1, an important regulator of both

fibrinolysis and plasminogen activators, may act as a stabilizing or counter-regulatory

factor for maintaining ECM integrity of the oviductal epithelium. Therefore, production







86

of PAI-1 by the oviduct might act to prevent premature nidation of the preimplantation

embryo. While the pig has noninvasive placentation, the trophoblast of the pig has

invasive potential as shown by its transfer to ectopic sites [Samuel, 1971]. Part of this

invasive potential may be due to fibrinolytic and plasminogen activator activity of the

embryo. Liedholm and Astedt (1975) suggested that the fibrinolytic activity of the ovum

may be depressed by an inhibitor of plasminogen activation. Plasmin-induced proteolysis

has been shown to be important for implantation in invasive species such as the mouse

[Sappino et al., 1989, Zhang et al., 1996]. However, in the pig, an endometrial inhibitor

of plasmin production has been suggested to protect the endometrium from the

blastocyst-induced proteolysis during placentation [Fazleabas et al., 1982]. PAI-1

secretion within the oviduct, may therefore have a similar function to that of the

endometrial plasmin inhibitor.

Immunolocalization results indicated that PAI-I is localized within the oviductal

epithelium and is heavily concentrated near the apical membrane, suggesting secretion

into the lumen. Immunogold EM of the isthmus, revealed that PAI-1 was located in

putative secretary granules within the lumen and associated with cilia. While PAI-1 was

localized to the ampulla using immunocytochemistry, immunogold electron microscopy

was unable to detect its presence. This may reflect a difference in specificity and

dilutions of two different antibodies used, difference in tissue sections examined between

the two different procedures and 2D-SDS-PAGE and fluorographic analyses, which

suggest that PAI-1 protein secretion is low in the ampulla. Both immunocytochemistry

and fluorographic analysis have shown that PAI-1 is localized in the ampulla and

infundibulum, however, the inability to locate immunoreactivity utilizing electron







87

microscopy may be that protein synthesis of this molecule is very low in these segments

and is present in secretary granules in very small amounts. PAI-1 was found in secretary

granules from isthmic tissue exposed to either a high estrogen (Day 0 cyclic) or high

progesterone (Day 9 pregnant) environment indicating that this protein is synthesized

throughout the estrous cycle or early pregnancy. Evidence from our laboratory

(unpublished) indicates that PAI-1 secretion may vary during the estrous cycle or early

pregnancy and its synthesis might be controlled by ovarian steroids. Reports on PAI-1

regulation in endometrial stromal and decidual cell cultures suggest that this protein is

up-regulated by progesterone, while estrogen antagonizes this effect [Schatz and

Lockwood, 1993]. The oviduct along with the uterus, is a major target for ovarian

steroids and steroid-modulated and cycle-specific changes have been noted for two other

de novo synthesized products of the epithelium, pOSP and TIMP-1 [Buhi et al., 1997].

Because the zona pellucida can be subject to proteolytic degradation, oviductal

PAI-1 may also protect the integrity of the zona pellucida from embryonic or oviductal

plasminogen activator activity. To our knowledge, PAI-I mRNA or activity has not been

examined in the early cleavage-stage embryo or in the oviduct. The porcine zona

pellucida of oviductal oocytes or embryos was found to be more resistant to proteolytic

degradation than that of either follicular oocytes or embryos collected from the uterine

environment [Broermann et al., 1989], suggesting a potential interaction of oviductal

protease inhibitors with the oocyte/embryo. Changes in resistance of the zona pellucida

to proteases may be dependent upon addition of glycoproteins/inhibitors obtained during

transit through the oviduct. Thus, an oviduct-specific factor may act to protect the zona

pellucida and embryo from degradation by proteolytic enzymes. Proteases are present in







88

oviductal flush and include the plasminogen activators and matrix metalloproteinases

(Kouba AJ, unpublished). PAI-1 working together with TIMP-I, may act to tightly

regulate this proteolytic activity. Plasminogen, the natural substrate for uPA and tPA, is

found in many extracellular fluids including seminal plasma [Kobayashi et al., 1992] and

follicular fluid [Beers, 1975], and may be enriched in oviductal fluid at or near the time

of fertilization. Estrogens have been shown to stimulate the uptake of plasminogen from

plasma by the mouse uterus [Finlay et al., 1983] and a similar function may occur in the

oviduct during estrus. Plasminogen has been shown to bind to mouse spermatozoa,

oocytes, and cumulus cells, which enhanced the local generation of plasmin [Huarte et

al., 1993] and this binding allows for PA-induced proteolysis to discrete focal areas.

Huarte et al. (1993) also showed that addition of plasminogen or antibodies to plasmin

during in vitro fertilization could increase and decrease, respectively, the fertilization

rate. Therefore, PAI-1 within the oviduct may inhibit oocyte or embryonic generation of

plasmin from plasminogen, due to their inherent tPA and uPA activity, thus maintaining

integrity of the zona pellucida while not affecting fertilization.

Further objectives will be to evaluate the biological role of PAI-1 within the

oviduct. PAl-1 may act in a autocrine/paracrine fashion to prevent proteolytic

degradation of ovulated oocytes or early embryos in the oviduct or uterus, prevent

premature hatching, regulate ECM remodeling of the oviduct or early embryo, inhibit

embryonic invasion of the oviductal lining, and promote embryonic development.




Full Text
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 EK0R6QVH9_75X4WK INGEST_TIME 2014-04-14T16:26:40Z PACKAGE AA00014255_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES