Maternal-embryonic interactions during early pregnancy in cattle

MATERNAL-EMBRYONIC INTERACTIONS DURING
EARLY PREGNANCY IN CATTLE

By

MARIO BINELLI

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

Dedicated to Mario Rodolpho Giovanni Binelli (late) and Arnaldo Monteiro

de Oliveira (Grandfathers), Guilherme Jose Binelli (Father), Ricardo Binelli

(Brother), Luiz Alberto de Oliveira (Uncle), Cicero Spiritus, Paul Campbell, Zilmar

Ziller Marcos, H. Allen Tucker and William W. Thatcher (Mentors)

To my family and friends

and

To an enlightened humankind

ACKNOWLEDGMENTS

I would like to express my greatest appreciation to Dr. William Thatcher,

my supervisory committee chair, for giving me the honor of being his student and

friend. His dedication, generosity, enthusiasm, sincerity, intelligence and ability

to excel under a lot of pressure are remarkable and a great source of inspiration.'

I am proud of being a member of one of the most respected reproductive

physiology laboratories in the world. I also want to thank him for his patience

when I was away from the lab pursuing extracurricular activities. My gratitude is

extended to the other members of my committee: Dr. William Buhi, for changing

my abstract conception of what was a protein to something more real like a spot

in a 2-D gel; Dr. Peter Hansen, for being my "second advisor", sharing his

knowledge and laboratory; Dr. Howard Johnson, for asking me "so, what is

novel?" when I first described my dissertation project to him, and for letting me

work in his laboratory during the beginning of my program; and to Dr. Frank

Simmen, for his advice on molecular matters and for the being such an example

of a humble personality behind a powerful mind.

I am also indebted to Dr. Prem Subramaniam, Dr. Lokenga Badinga, Dr.

Rosalia Simmen, Dr. Naser Chegini, Dr. Joel Yelich, Dr. Maarten Drost, Dr.

Herbert Head, Dr. Michael Fields and Dr. Nancy Denslow for sharing their

laboratory and/or scientific expertise.

I wish to thank Dr. Michael Roberts, Dr. Thomas Hansen and Dr. Douglas

Leaman for providing important reagents used for the research described in this

dissertation.

I want to express my eternal gratitude to Dr. Thais Diaz, Dr. Eric Schmitt,

Dr. Alice de Moraes, Fabiola Paula-Lopes, Dan Arnold and Ricardo Mattos for

closely supporting me through the struggles of graduate school and for providing

their sincere friendship and scientific support.

I am grateful to fellow graduate students, post-docs, visitors and friends in

the departments of Dairy and Poultry and Animal Science, including Dr. Luzbel

de la Sota, Dr. Divakar Ambrose, Dr. Joan Burke, Dr. Maria de Fatima Pires, Dr.

Sandra Coelho, Ellen Van de Leemput, Jim Hampton, Nina Nusbaum, Monte

Meyer, Frederico Moreira, Aydin Guzeloglu, Metin Pancarci, Flavia Lopes,

Cassia Orlandi, Jan Vonk, Arthur Araujo, Dr. Carlos Arechiga, Dr. Lannett

Edwards, Morgan Peltier, Yaser AI-Katanani, Rocio Rivera, Saban Tekin,

Andrew Majewski, Inseok Kwak, Jason Blum, Max Huidsen, Tomas Belloso, Dr.

Alfredo Garcia, Dr. Rafael Roman, John Fike, Dr. Maria Cadario, Andres

Kowalski, Dr. Karen Reed, Dr. Michael Green, Michael Porter and Andy Kouba.

From helping me to solve some unearthly statistics to sharing a smile their

contributions will not be forgotten.

I want to thank, for their technical, scientific and professional expertise,

Marie-Joelle Thatcher, Idania Alvarez, Frank Michel, Jesse Johnson, Jennifer

Williams, Lauren Knickerbocker, Mary Ellen Hissen, Susan Gottshall, Larry

Eubanks, Eddie Fredriksson, Mary Russell, Dale Hissen, James Lindsey, Chris

Wilson, Joyce Hayen, Werner Collante, Kathy Austin, Debbie Akin, Stephany

Suggs, Peggy Briggs, Melissa Thomas, Patricia Hancock, the Dairy Research

Unit crew, the Large Animal Clinic crew, students and residents at the School of

Veterinary Medicine and the Meat's Laboratory crew.

As a important part of my University of Florida experience I wish to

express thanks for the generosity of Ms. Louise Curtelis, the hard work and

comradery of fellow officers and friends at the Graduate Student Council and the

Brazilian Student Association and the financial support from the Animal

Molecular and Cell Biology program during the first year of my graduate studies.

One true friendship is worth one thousands PhDs, so I want to show my

sincere gratefulness to all friends in Gainesville, the ones mentioned above and

also Mauricio, Ilka, Carlos da Costa, Patricia, Dirceu and Ana Clara, Daniella and

Fabiano, Ronaldo and Valeria, Cleisa and Cartaxo, Lawry, Claudio and

Cristiane, Tom e Jose Melvin, Ricardo Harakava, Cristina and Warley, Annie e

Wigberto, Debbie and Ken, Joe and Kirsten, Karina e Uilson, Liana and Silvano,

Deise and Alfredo, Raul and Helena, Michelle, Mary Duryea, Paul and Joan and

my friends at the School of Tai Chi Chuan, Marcus Harvey, Dorota Porazinska,

Stacey Chestain, Joe Sadek and Jackie Wilson .

Distance makes one appreciate the value of a family. I would have never

been able to complete my graduate studies without the endless support of Nice,

Arnaldo, Aurora, Lola, Dinda, Dindo, Carmen, Maga, Guilherme, Paula, Neco,

Ana, Rico, Beto, Aldo, Ines, Janete, Maria Helena, Farjala (late), Lenita, Lucia,

Elisa, Leandro, Lucio, Elisa Kampf, Lolo Kampf, Claudia, Gabi e Yara. I Thank

them for their dedication, faith, love, patience, inspiration, support and

encouragement.

Most specially of all, with my whole heart I wish to thank my wife Nana, for

helping me to externalize the best of my being and for being side-by-side with

me on this long road.

TABLE OF CONTENTS

ACKNOWLEDGMENTS ........................................... iii

ABSTRACT ............................................... xi

CHAPTERS

1 INTRODUCTION ................... ..... .. ............ 1*

2 LITERATURE REVIEW ..................... ............. 3

Maternal-embryonic Communication as a Requirement for Successful
Pregnancy ........................................... 3
Communications Between Gametes and Maternal Units ....... 4
Communications Between Conceptus and Maternal Units ...... 6
Problems Associated with Fertilization Failure and Embryonic Mortality in
Cattle .......................... ... ............. 9
Susceptible Periods During Pregnancy ..................... 9
Causes of Fertilization Failure and Embryonic Mortality ....... 10
Oviductal Function and Reproductive Failure in Cattle ............. 16
The Oviduct Environment ................ .......... 16
Steroid Regulation and Protein Synthesis .................. 20
Regulation of Reproductive Processes Occurring in the Oviduct
.............. .............................. 23
Oviductal Function and Reproductive Failure in Cattle ........ 25
Uterine Function and Reproductive Failure in Cattle ............... 27
The Uterine Environment .............................. 27
Regulation of Reproductive Processes Occurring in the Uterus
....... .................................... 31
The JAK-STAT Pathway .............................. 47
Uterine-Conceptus Interactions and Reproductive Failure in Cattle
.. ................................ 70
Manipulating Uterine Function to Minimize Embryo Mortality ... 73
Objectives of This Dissertation ............................. 75

3 PERSISTENT DOMINANT FOLLICLE ALTERS PATTERN OF
OVIDUCTAL SECRETARY PROTEINS FROM COWS AT ESTRUS .. 76

Introduction ............................ .. ............ 76
Materials and Methods ................... ........... 79
Materials .......... ........... .... ............. 79
Preparation of Medium ............................ 80
Animals and Treatments ............................. 80
Tissue Culture ................ .................. 82
Two-Dimensional Electrophoresis ..................... 83
Densitometry .................... ... ............. 83
Hormone Assays .................................. 84
Statistical Analysis ............................ 85
Results ............................................. 85
Ultrasonography and Hormone Measurements ............. 85
Incorporation Rate ................................. 88
Fluorography and Densitometry ....................... 88
Discussion .................. .. ...... ............ 95

4 EFFECTS OF BOVINE INTERFERON-TAU ON THE JAK-STAT SIGNAL
TRANSDUCTION PATHWAY AND SYNTHESIS OF PROTEINS IN
BOVINE ENDOMETRIUM AND ON THE MECHANISM OF
GENERATION OF PROSTAGLANDIN F2a IN ENDOMETRIAL
EPITHELIAL CELLS ..................... .............. 103

Introduction ..................... ................... 103'
Materials and Methods ............................ .. 105
Materials ........... ........ .. .... ............ 105
Experiment 1 .......... ............. ............ 107
Experiment 2 ........................ ............ 114
Statistical Analysis ........................... 117
Results ............ .................. ............. 117
Experiment 1 ........................... ...... 117
Experiment 2 .......... ............ ............. 121
Discussion .......... ......... .. .... .............. 123

5 BOVINE INTERFERON-TAU STIMULATES THE JAK-STAT PATHWAY
IN BOVINE ENDOMETRIAL EPITHELIAL CELLS ............... 133

Introduction ..................... ................... 133
Materials and Methods ....................... .... 135
Materials ................... ... .. ............ 135

viii

Cell Culture and Cell Extracts .......................... 136
Immunoprecipitation ............................... 142
Immunoblots .................. ..... ........... 142
Nature of BEND Cells ................................ 143
Dose Response to bIFN-r ........................... 144
Time Response to blFN-c ............................. 145
Validation of Immunoprecipitation and Immunoblots Procedures
............................................ 145
Nuclear Translocation of STATs ........................ 146
Coimmunoprecipitation of STATs ....................... 147
Densitometric Analysis .................. ......... 147
Statistical Analysis ............................... 148
Results .................................... ...........148
Nature of BEND Cells ................................ 148
Validation of Immunoprecipitation and Immunoblots ......... 151
Dose Response to blFN-r ........................... 151
Time Response to bIFN-r ............................. 156
Nuclear Translocation of STATs ........................ 156
Validation of Time Responses to bIFN-T ................. 167
Coimmunoprecipitation of STATs ....................... 167
Discussion ................ ........................... 173

6 BOVINE INTERFERON-TAU STIMULATES BINDING OF STAT
PROTEIN COMPLEXES TO DNA AND STIMULATES SYNTHESIS OF
INTERFERON RESPONSE FACTOR-1 (IRF-1) PROTEIN IN BOVINE
ENDOMETRIAL (BEND) CELLS ............................ 187

Introduction .............. ............ ...... ... .... 187
Materials and Methods .................................. 188
Materials ................ ...................... 188
Probes .......................... ................. 190
Electrophoretic Mobility Shift Assays .................... 192
Immunoblotting for IRF-1 ............................. 195
Statistical Analysis ................................ 196
Results ...................................... ........... 196
Electrophoretic Mobility Shift Assays .................... 196.
Immunoblotting for IRF-1 ............................. 201
Discussion ................... ........... ............ 201

ix

7 INTERFERON-TAU MODULATES PHORBOL ESTER-INDUCED
SECRETION OF PROSTAGLANDIN AND PROTEIN EXPRESSION OF
PHOSPHOLIPASE-A2 AND CYCLOOXYGENASE-2 FROM BOVINE
ENDOMETRIAL (BEND) CELLS ...........................208

Introduction .............................. ............ 208
Materials and Methods ................ ..................... 210
Materials ........................................ 210
Cell Culture and Sample Collection ................... 211-
Radioimmunoassay ............... ............... 212
Preparation of Extracts ............................ 214
Immunoblotting ................................. 214
Experimental Designs ............................. 215
Statistical Analysis ............................... 216
Results ............. ..................... .......... 219
Experiment 1 .................................... 219
Experiment 2 ..................................... 223
Experiment 3 ........................ .............. 227
Experiment 4 ........................ .............. 232
Discussion .............................. ..... ........ 234

8 GENERAL DISCUSSION .................................. 246

LIST OF REFERENCES ........................................ 264

BIOGRAPHICAL SKETCH ................ .................... 291

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

MATERNAL-EMBRYONIC INTERACTIONS
DURING EARLY PREGNANCY IN CATTLE

By

Mario Binelli

August 1999

Chairperson: William W. Thatcher
Major Department: Animal Science

Maternal-embryonic physiological communications are an important

feature of processes in the reproductive cycle. Communications occurring during

oviductal transit of gametes/embryos and during maternal recognition of

pregnancy for maintenance of the corpus luteum (CL) were studied in cattle.

Objectives were 1) to study the distribution pattern of oviductal secretary proteins

secreted by cows bearing persistent or fresh dominant follicles (PDF or FDF,

respectively); 2) to examine the signal transduction system stimulated by bovine

interferon-r (blFN-t) in endometrium; and 3) to characterize the effects of blFN-T

on prostaglandin F,(PGF,) production by bovine endometrial (BEND) cells.

Presence of PDF reduces fertility in cattle. Proteins synthesized from

infundibulum, ampulla and isthmus from oviducts ipsilateral and contralateral to

CL of cows bearing PDF or FDF were examined by two-dimensional

fluorography. Presence of PDF altered distribution of secretary proteins in a

side- and region-specific manner. Changes in the oviductal environment may

contribute to decreased fertility of cows bearing a PDF. Conceptus-produced

blFN-t suppresses endometrial PGF2, pulses in vivo, and is required for

maintenance of pregnancy. The hypothesis was that blFN-r stimulated

synthesis of endometrial proteins through the jak kinases (JAK)-signal

transducer and activator of transcription (STAT) pathway of signal transduction.

Presence of and bIFN-r-induced tyrosine phosphorylation of STAT proteins were

demonstrated via immunoprecipitation (IP) and immunoblotting (IB) techniques,

while blFN-r-induced secretary proteins were measured by fluorography in

endometrial explants obtained from day 15 cyclic cows. BEND cells were used

for remaining experiments. Presence of STATs, tyrosine-phosphorylation, dimer

complex formation and nuclear translocation were measured through IP, co-IP

and IB. Binding of activated STAT complexes to cis-acting elements present in

the regulatory region of interferon-inducible genes was determined using

electrophoretic mobility shift assays. Bovine IFN-T stimulated synthesis of

interferon-regulatory factor-1 (IRF-1) in BEND cells as determined by IB. The

blFN-t regulates synthesis of phorbol 12,13 dibutyrate (PDBu)-induced PGF2,,

as measured through radioimmunoassays (RIA). Regulation was associated

with blFN-t -induced suppression of phospholipase-A2 and cyclooxygenase-2

protein expression and enzymatic activity measured through IB and RIA.

Collectively, experiments elucidated mechanisms that are involved in maternal-

conceptus crosstalk required for successful reproductive outcome.

CHAPTER 1
INTRODUCTION

The world's human population is increasing at a fast rate, and

consequently the need for basic nutrients, including carbohydrates, lipids,

proteins and minerals, is also increasing. Animal agriculture historically has

been one of the most important sources of nutrients for humans. Dairy cows

efficiently metabolize feed nutrients and synthesize milk, which provides protein

and energy in a suitable form for human consumption. Lactation is a final step in

the reproductive cycle which is dependent on successful production of viable

gametes, conception, pregnancy and parturition. First insemination conception

rates for dairy cattle is 52 to 57% (Mawhinney and Roche, 1978). Improvements

on such rates will be required for dairy products to continue being a viable

source of nutrients for humans.

Classically, study of animal performance has focused on genetic and

environmental effects on a given production trait, such as meat or milk

production. Reproduction poses an interesting scenario, in which reproductive

processes are modulated by the interactions between both the maternal and

embryonic genomes. Moreover, reproductive processes are influenced by

external environment, but more importantly, by the internal environment (i.e., the

reproductive tract). Both maternal and embryonic units influence such internal

environment, and physiological crosstalk is a hallmark of the process. As a

result, a complex set of coordinated interactions takes place and it is these

interactions that will dictate a successful reproductive outcome. My thesis is that

failure in maternal-embryonic communications leads to decreased reproductive

rates.

This dissertation examines the role of maternal-embryonic

communications during two physiological windows of the reproductive cycle and

their importance on the overall reproductive process. Chapter 3 describes

steroid hormone-modulation of protein synthesis and secretion in the oviduct,

where final stages of gamete maturation, fertilization and early embryo

development occur. Chapters 4, 5, 6 and 7 characterize changes in intracellular

and secretary processes of the maternal endometrium in response to a

conceptus-secreted factor, interferon-r, that is essential for maintenance of

pregnancy.

A greater understanding of the basic mechanisms regulating reproductive

processes, specially mechanisms involving maternal-embryonic interactions, is

necessary to improve conception rates in cattle.

CHAPTER 2
LITERATURE REVIEW

Maternal-embryonic Communication as a Requirement for Successful Pregnancy

Communications of a physiological nature are very common between

embryonic (and pre-embryonic) and maternal units from the development of the

oocyte (i.e., before fertilization; Eppig et al., 1997a) to parturition (Bazer and

First, 1983; Fuchs and Fields, 1999) and lactation (Thatcher et al., 1980). The

maternal unit constitutes all tissues in the female reproductive tract that directly

or indirectly interacts with gametes or concepts conceptss = embryo and

associated embryonic membranes). Appropriate exchange of hormonal signals

between the two units is required for successful establishment and completion of

several windows of the reproductive cycle. Moreover, each window requires

unique signals that have been studied in detail and that have unique implications

on the outcome of the reproductive process (i.e., live, viable offspring). A

general review of the literature on several critical windows on the reproductive

cycle will be presented. A more detailed review will be offered on embryonic-

maternal interactions during the time of embryo transit through the oviduct and

during the window of maternal recognition of pregnancy occurring at around day

3

4

17 of pregnancy. The focus of this review will be on the bovine species, but data

from other species will be presented whenever appropriate.

Communications Between Gametes and Maternal Units

Gametes are differentiated cells with the specific function of conveying

genetic information from each paternal and maternal unit to a zygote during the

process of fertilization. To ultimately undergo fertilization, both paternal and

maternal gametes interact with somatic cells in maternal reproductive tissues.

Eqq-oocvte. During development of follicles, oocytes change the program

of granulosa cells in the follicle, so that the default program of mural granulosa

cell differentiation is suppressed. As a consequence, cells surrounding the

oocyte become more specialized in functions that favor development of the

oocyte. In mature Graafian follicles, this layer of differentiated granulosa cells

forms a small pedicle of cells, the cumulus oophurus, which contains the oocyte

and protrudes towards the interior of the antrum. Eppig and coauthors (1997a)

hypothesized that oocytes control their own microenvironment by regulating

differentiation of the supporting cells that are in direct communication with them.

For example, expression of luteinizing hormone (LH)-receptors in cumulus cells

is abrogated by presence of the oocyte (Eppig et al., 1997b).

Oocyte-oviduct. In cattle, the oviduct is divided in three functional regions,'

namely infundibulum (INF), ampulla (AMP) and isthmus (IST; Hafez, 1993a).

The INF is opened to the peritoneal cavity of the body, the IST connects the

oviduct with the uterus, and the AMP is localized between INF and IST. The

initial interaction of the oocyte with the oviduct takes place at the INF. The INF

"picks up" the ovulated oocyte and initiates its transport towards the uterus.

Overall flow of oviductal fluid is towards the body cavity (Hafez, 1993a).

Therefore, in order for the oocyte to be transported to the site of fertilization

(AMP), it must interact with ciliated epithelial cells lining the oviductal lumen.

Balance between oviductal fluid flow and ciliary beating towards the uterus yields

a net movement of the oocyte towards the AMP and ultimately, uterus. During

this trajectory towards the site of fertilization, the oocyte is under the influence of

products secreted by the oviduct which could modulate its development and the

process of fertilization (Buhi et al., 1997a). For example, Kouba and coworkers

(1999) determined that a major secretary protein of the AMP, named POSP

(porcine oviductal secretary protein), has a role in decreasing occurrence of

polyspermy during fertilization in pigs. This supports the findings of Nancarrow

and Hill (1995) that an estrus-associated glycoprotein, a protein homologous to

POSP in sheep, increased blastocyst formation after in vitro fertilization.

Furthermore, Staros and Killian (1998) identified six proteins in oviductal fluid,

including a POSP-like protein that are associated with bovine oocytes in vitro.

Sperm-oviduct. Austin (1951) and Chang (1951) independently reported

that freshly ejaculated rat and rabbit spermatozoa were incapable of penetrating

an oocyte. The ability of fertilization was only acquired after the sperm spent a

period of time in the female reproductive tract, a process called sperm

capacitation. These early findings supported the idea of the necessity of

interaction between male gametes and products of the oviduct prior to

fertilization. In fact, incubation of sperm cells with oviductal fluid capacitated and

sustained sperm mobility in vitro (Parrish et al., 1989; McNutt and Killian, 1991).

Moreover, Boatman and Magnoni (1995) identified and purified an oviductal

factor (oviductin) that acts to enhance sperm penetration in follicular oocytes.

Communications Between Conceptus and Maternal Units

Conceptus-oviduct. Following fertilization, the concepts continues to

interact with the AMP and IST before it reaches the uterus. During this period,

the concepts undergoes initial cell divisions, and there is a possibility of a

continual influence of oviductal products on the concepts. For example, Buhi

and coworkers (1993) showed that gold particles immunoreactive with porcine

oviductal secretary protein are associated with flocculent material in the

perivitelline space surrounding the blastomeres and in the zona pelucida of

embryos from the four-cell stage to blastocysts. Moreover, semi-purified

oviductal specific protein improved cleavage rates of embryos fertilized and

developed in vitro (Hill et al., 1996). In contrast, at least in pigs, presence of

gametes or embryos did not affect production of specific oviductal proteins,

suggesting that a regulation of oviductal function by gametes or embryos is

probably not important during this stage of the reproductive cycle (Buhi et al.,

1990).

A series of growth factors have been identified in the oviduct during the

estrous cycle and early pregnancy in different species (summarized in Buhi et

al., 1997a). For example, Paula-Lopes and coworkers (1999) reported synthesis

and secretion of interleukin-1 p both in oviduct and uterus of cyclic cows.

Moreover, interleukin-1 p stimulated in vitro development of embryos when added

before day 5, which suggests a oviductal effect (Paula-Lopes et al., 1998).

Although limited research has been performed to date to elucidate specific roles

of growth factors in fertilization and early embryo development, it is possible that

growth factors act in an autocrine and paracrine fashion to influence these

processes (Chegini, 1996).

Conceptus-uterus. In cattle, it takes 72 to 84 h from the time of ovulation

to the time embryos enter the uterus (Betteridge and Flechon, 1988). From entry

into the uterus to parturition there are multiple examples of interactions between

the concepts and the maternal unit, which will be discussed elsewhere in this

review. A striking example of such interactions is the process of maternal

recognition of pregnancy associated with maintenance of the CL, which will also '

be presented in detail afterwards. Briefly, maintenance of pregnancy requires a

steady supply of progesterone (P4) from the corpus luteum (CL). In contrast, in

cycling animals, it is necessary that P4 concentrations decrease so that animals

can return to estrus. This decrease is accomplished in response to pulses of

uterine prostaglandin-F2,(PGF2,,) that have lytic actions on the CL. Uterine

physiology must be changed to suppress this pulsatile release of PGF2afor the

concepts to survive. The conceptus-secreted interferon-r (IFN-t) interacts with

the uterine tissue to decrease production of PGF2,, thereby allowing for

maintenance of CL and consequent sustained elevated P4 concentrations.

The process of luteolysis includes the action of follicular estrogen on a P4-

primed uterus, which is capable of secreting PGF2,. Driancourt and coworkers

(1991) formulated the hypothesis that one possible aspect of the antiluteolytic

mechanism induced by concepts could involve attenuation in development of

follicles on the ovary adjacent to the pregnant uterine horn. They determined

that the number of follicles greater that 7 mm was reduced in the ovary

containing the CL after day 22 of pregnancy. Moreover, size of the largest

follicle was greater on the ovary contralateral to the pregnant uterine horn. An

additional study comparing follicular development in pregnant versus

hysterectomized cows indicated that products of pregnancy, either secreted

directly or induced by the concepts, decreased intraovarian follicular

development in a local manner (Thatcher et al., 1991). This could enhance

embryonic survival by attenuating luteolytic mechanisms (Thatcher et al., 1994b).

The examples above illustrate the common theme of maternal-gametic

and maternal-embryonic interactions, and their occurrences throughout the

reproductive cycle. Failure of appropriate communication between maternal and

embryonic units can lead to disruption of the reproductive cycle and termination

of pregnancy. Next I will examine the issue of embryonic mortality in cattle. In

later sections I will discuss the implications of a failure in maternal-embryonic

cross-talk on embryonic mortality.

Problems Associated with Fertilization Failure and Embryonic Mortality in Cattle

Susceptible Periods During Pregnancy

Calving rates to a single insemination are reported to be 52 to 57% for

dairy cows (Mawhinney and Roche, 1978) despite fertilization rates of about 89%

(Henricks et al., 1971). Diskin and Sreenan (1980) utilized beef heifers to

determine embryo survival during discrete periods within pregnancy. They

reported up to 93% survival rates to day 8, 66% to day 16 and 58% to day 42.

These data indicate that minor losses are due to fertilization and embryonic

death before day 8, which encompasses the period of embryo permanence in

the oviduct and development to the blastocyst stage in the uterus. However, in a

group of infertile cows, there are appreciably greater losses (-40%) due to failure

of fertilization (Tanabe and Cassida, 1949) and additional losses by day 35

(40%; Ayalon 1978). In normal cows, a large percentage of embryos is lost

between days 8 and 16 of pregnancy, which is the period of embryonic

elongation and maternal recognition of pregnancy associated with CL

maintenance. The substantial losses of pregnancies during the first 16 days of

pregnancy has obvious economic impacts in the livestock industry, and

represent an opportunity for animal scientists and reproductive physiologists to

improve calf crops. This prompted a great quantity of research to understand

mechanisms involved in the processes of fertilization failure and embryonic

mortality. Moreover, understanding the mechanisms underlying these alterations

in reproductive development will provide the basis for creation of technologies

aimed to attenuate fertility problems in the field.

Causes of Fertilization Failure and Embryonic Mortality

Inadequate embryo unit. Both genetic and environmental effects can

account for early embryonic deaths. There is an estimated 7.5% death of early

bovine embryos because of occurrence of chromosomal abnormalities (Wilmut et

al., 1986). Such abnormalities may be inherited or arise during meiosis,

fertilization or early cleavage stages (King, 1985).

Environmental effects such as heat stress also decrease the ability of

embryos to develop properly. For example, Putney and coworkers (1989)

exposed superovulated heifers to hyperthermic conditions for 10 hours after the

onset of estrus. There was no difference in the rate of fertilization between heat-

stressed and non-heat stressed heifers (control). However, only 12% of

embryos recovered from stressed heifers were normal vs. 69% of embryos

recovered from control heifers. Since the period of heat stress was administered

prior to ovulation and fertilization, it was hypothesized that the detrimental effect

of the heat stress was exerted on the oocyte within the follicular environment. In

an attempt to investigate the effects of heat stress on embryonic survival from

the time of ovulation to 7 days after estrus, Putney and coworkers (1988a) kept

heifers either in thermoneutral or hyperthermic conditions starting 30 hours after

onset of estrus. They found similar results to the ones described above, where

heat stress increased the proportion of abnormal embryos compared to controls.

To pinpoint critical stages of susceptibility of embryos to elevated maternal

temperatures, Ealy and coworkers (1993) submitted cows to heat stress on days

1, 3, 5 or 7 of pregnancy. Embryos recovered on day 8 were compared to

embryos from cows not heat stressed. Only heat stress at day 1 caused

decreased development. Collectively, data from these three heat stress

experiments suggest that environmental factors can affect embryo development

in multiple stages. Alternatively, it is possible that the toxic effects of heat stress

are exerted in the oviductal and uterine environments, which could become sub-

optimal for fertilization and embryo development, resulting in abnormal embryos.

As a final example of environmental effects on embryo development, Putney and

coworkers (1988b) incubated day 17 conceptuses at normal (39 C, 24 hours) or

high (39 C, 6 hours; 43 C, 18 hours) temperature regimens and measured de

novo protein synthesis by these conceptuses. They found that heat stress not

only decreased overall protein synthesis, but more importantly, decreased

secretion of IFN-T. This indicates that under the influence of heat stress,

embryos are less capable to sending appropriate signals to the uterus, which are

required for maintenance of an environment conducive to pregnancy.

12

Inadequate maternal unit. As mentioned above, results from the work of

Putney and coworkers (1988a; 1989) suggest that the elevated number of

abnormal embryos in heifers that underwent heat stress could be a result of the

effect of high temperatures on the maternal unit, and not a direct effect on the

embryo. In that regard, embryos produced by in vitro maturation/fertilization

techniques that are exposed to elevated temperatures (40.5 C) for 12 hours had

development comparable to that of controls (56% blastocyst formation; Rivera,

Lopes and Hansen, personal communication), supporting the concept that heat

stress effects on maternal units may create a toxic environment that is conducive

to development of abnormal embryos. Moreover, in the experiment of Putney

and others (1998b), they incubated endometrium explants removed from cows at

day 17 of the estrous cycle at 39 or 43 C as described above, and measured

secretion of PGF2 in the medium. Heat stress caused a pronounced increase in

PGF2 production over time compared to controls. This finding suggests that

heat stress favors luteolysis and consequent loss of pregnancies.

Measurements of P4 concentrations in milk following insemination of dairy

cows revealed that inseminated-pregnant cows had slightly higher P4 compared

to inseminated-non-pregnant cows (Lamming et al., 1989). This finding

prompted the hypothesis that luteal insufficiency could be a cause of increased

embryonic mortality in lactating dairy cows. Possible causes of decreased luteal

function include (1) poor development of the ovulatory follicle, resulting in a low

quality CL (i.e., low weight and consequent low P4 secretion) and (2) insufficient .

13

luteinizing hormone (LH) support of continuous luteal P4 secretion. Strategies to

provide supplemental P4 through administration of exogenous P4 (Van Cleef et

al., 1996), use of human chorionic gonadotropin (hCG; Schmitt et al., 1996a),

gonadotropin releasing hormone (GnRH) injections (Schmitt et al., 1996b), and

GnRH implants (Ambrose et al., 1998) have increased circulating concentrations

of P4 but yielded mixed results on pregnancy rates.

Asynchronv between embryonic and maternal units. Embryo survival may

be impaired because of failure in some aspects of the relationship between the

embryonic and maternal units, despite the fact that both are normal (Wilmut et

al., 1986; Thatcher et al., 1994b). An example of such failures is lack of

synchrony between uterus and embryo. During early pregnancy, embryo

development depends upon a sequence of changes in the uterine secretions,

which in turn is dependent on progressive changes in the maternal hormonal

milieu. This phenomenon became established when it was observed that

embryos transferred between animals that were not in estrus at the same time

caused abnormal development and death of the embryo (Wilmut and Sales,

1981). Moreover, when cows were treated with P4 from days 1 to 5 of the

estrous cycle and received a day-8 embryo on day 5, pregnancy was

maintained, indicating that uterine development had been advanced as a result

of the exogenous P4 (Geisert et al., 1991). A condition that can cause

asynchrony is exposure of cows to heat stress. Biggers and others (1987)

determined that high environmental temperatures between days 8 and 16 of

14

pregnancy caused a 50% reduction in weight of concepts compared to control

cows. These retarded embryos may not be able to send the appropriate

antiluteolytic signals to the maternal endometrium, thereby allowing luteolysis

and consequent loss of pregnancy to occur.

Inadequate manipulations of the system. As a means to improve

conception rates in livestock operations, researchers have developed several

management practices that often include manipulations of the endocrine system

of animals. Such practices may sometimes yield unexpected results, including

decreased fertility due to disruption of appropriate maternal-embryonic

communications. One widespread practice is synchronization of estrous cycles.

Estrus synchronization systems are used for artificial insemination, timed

insemination and embryo transfer. Most commonly, synchronization is achieved

with combinations of treatments with PGF2a, progestins and gonadotropin

releasing hormone (GnRH; Thatcher et al., 1996). Synchronization with

progestins is based on the principle that exogenous progestins, such as

progesterone delivered by a controlled internal drug release (CIDR) device, can

maintain a sub-luteal concentration of progestin in blood during a period which

permits CL regression. In the absence of a CL, removal of the progestin source

will result in a synchronized estrus (Macmillan and Peterson, 1993). However,

sub-luteal concentrations of progesterone increase LH pulse frequency, which

stimulates continuous growth of a dominant follicle (Cooperative Regional

Research Project, NE-161, 1996; Savio et al., 1993a; Savio et al., 1993b). This

"persistent" dominant follicle (PDF) is estrogenic, and subsequent fertility, as

measured by conception rate at first service (number of pregnancies / number of

animals inseminated), is lower compared to animals bearing normal DFs [37.1%

vs. 64.8% in heifers, (Savio et al., 1993b); 23.6% vs. 58.2% for cows and heifers

(Cooperative Regional Research Project, NE-161, 1996). Possible explanations

for reduced fertility include alterations in the oocyte and /or in the oviductal

environment. In a study by Ahmad et al. (Ahmad et al., 1995), embryos

recovered at Day 6 of pregnancy from cows bearing PDF were less developed

(i.e., were less able to reach the 16-cell stage) than embryos from cows

ovulating a fresh dominant follicle (FDF). In addition, Revah and Butler (1996)

showed that oocytes recovered from PDF showed expanded cumulus cells and

condensed chromatin dispersed in their ooplasm. In contrast, compact cumulus

cells and intact germinal vesicles were found in oocytes from FDF. Thus, the

PDF may induce premature oocyte maturation and/or alter oviduct function,

which could affect early embryonic development and decrease fertility.

Processes of sperm capacitation, fertilization and early embryonic development

in this altered oviductal environment can contribute to decreased fertility

experienced by cows developing persistent follicles.

Oviductal Function and Reproductive Failure in Cattle

The Oviduct Environment

The oviduct environment can be simplistically described as presenting

physical and chemical characteristics which are conducive to the reproductive

processes occurring within the oviduct. Physical and chemical characteristics

are described below in the sub-sections "Functional anatomy and morphology"

and "The oviduct fluid", respectively. To exemplify the functions of these

characteristics, Rieger and others (1995) examined development of embryos in

vitro, either in coculture with oviductal cells (both physical and chemical

influences) or in serum-free medium pre-conditioned by oviductal cells (chemical

influences, only). In both systems, embryos reached the 4-cell stage in 48

hours. However, embryos developing in the coculture system reached the

blastocyst stage 24 h before the others and also had significantly more cells.

There was no treatment where only physical factors were present, but the

conclusion from their data is that probably both physical and chemical

characteristics are necessary for best embryo development.

The ovarian cycle. There is a close association between oviduct function

and concentrations of circulating ovarian steroid hormones. Therefore, it is

appropriate to describe the changes occurring in such hormones during the

1T

estrous cycle. Other aspects of the estrous cycle will be discussed elsewhere in

this chapter.

The ovarian cycle consists of cyclic growth and demise of two ovarian

structures, the follicle and the CL. Considering one estrous cycle the period

comprised between two ovulations, two to three follicular waves of dominant

follicle growth occur (Savio et al., 1988; Sirois and Fortune, 1988). Each

follicular wave is comprised of periods of recruitment, selection, dominance and

turnover or atresia. The ovulatory follicle generated in the last wave does not

turn over, but ovulates. The main steroid secretary product from follicles are

estrogens, such as E2. There is a positive relationship between size of follicles

and E2 concentration in the circulation. Since maximum growth of follicles occurs

during the dominance phase, the peri-ovulatory period is characterized by

highest concentrations of circulating E2 during the estrous cycle. The ovulated

follicle undergoes functional and structural changes to form a CL. The CL grows

at a rapid rate to reach a maximum size in about 11 days, remains at its

maximum size until about day 16 of the estrous cycle and then regresses (the

process of luteolysis). Parallel to changes in CL size are changes in secretion of

luteal P4. Turnover of the dominant follicle is associated with high concentration

of P4, typical of mid-cycle. In contrast, final differentiation and growth of the

ovulatory follicle prior to ovulation only occurs in a low P4 environment.

Functional anatomy and morphology. The oviducts are suspended in the

mesosalpynx, a peritoneal fold of the broad ligament. As mentioned earlier, the

oviduct can be divided into three functional regions: the funnel shaped

abdominal opening near the ovary INF, which terminates in the fringe-like

fimbriae; the more distal dilated AMP and the IST, the narrow proximal portion of

the oviduct, connecting to the uterus (Hafez, 1993a). The oviduct can be simply

described as a muscular tube with a mucosal lining. There are two muscle

coats: an external longitudinal and an internal circular coat (Leese, 1988).

Thickness of the musculature increases from the ovarian to the uterine end of

the oviduct. Muscular contractions function to mix oviductal contents, aid sperm

transport, help denudate the egg, promote fertilization and regulate egg

transport. Patterns of oviduct muscular contractions vary with the stage of the

estrous cycle, indicating hormonal regulation of this process. Before ovulation,

contractions are gentle, but become more vigorous at ovulation. Muscular

contractions in the ovarian direction are more common than in the uterine

direction (Hafez, 1993a). The oviductal mucosa possesses characteristic folds,

with high, branched folds in the AMP and decreasing heights towards the IST to

become low ridges. The mucosa consists of one layer of columnar epithelial

cells, underlined by a submucosa containing smooth muscle fibers and

connective tissue. The oviductal epithelium contains both ciliated and non-

ciliated, secretary cells. Ciliated cells are most abundant in the INF and least in

the IST. Rate of cilia beating is affected by levels of ovarian hormones, with

maximal activity occurring at the periovulatory period. Cilia beating is

synchronized and toward the uterus. The opposite direction of coordinated cilia

19

beating (towards the uterus) and oviduct muscular contractions (towards ovary)

maintain eggs in constant rotation, which is essential for fertilization and to

prevent oviduct implantation (Hafez, 1993a). Non-ciliated epithelial cells have

primarily a secretary function. They contain secretary granules at their apical

aspect, and these accumulate during the follicular phase of the estrous cycle,

and are released into the lumen after ovulation (Murray, 1992). Treatment of

ovariectomized sheep with E2 stimulates hypertrophy of secretary organelles and

accumulation of granules in non-ciliated cells of the AMP (Murray, 1995).

Oviductal secretions contribute to the formation of the oviductal fluid, discussed

next.

The oviduct fluid. Reproductive processes occurring in the oviduct are

exposed and subjected to regulatory influences of ingredients in the oviductal

fluid. Chemical analyses of the oviductal fluid indicated that it is a mixture of

constituents derived from the plasma, through selective transudation, plus

specific proteins synthesized and secreted by the oviductal epithelium (Leese,

1988). The major classes of components are water, gases (02), electrolytes (Ca,

Na, K, Cl), non-electrolytes (glucose, fructose, complex carbohydrates) and

proteins. Some proteins originate from serum (albumin, immunoglobulins) while

others are synthesized de novo in the oviduct [plasminogen activator inhibitor

(Kouba et al., 1997) and bovine oviductal glycoprotein (Boyce et al., 1990)].

Functions of oviductal fluid electrolytes and non-electrolytes are reviewed in

Leese (1988). De novo synthesized oviductal proteins may affect reproductive

processes such as fertilization, and early embryonic development (Buhi et al.,

1997a). More importantly, changes in the optimal milieu of de novo synthesized,

secretary oviductal proteins may lead to sub-optimal micro-environments

conducive to reproductive failure (Binelli et al., 1999, Chapter 3).

Steroid Regulation and Protein Synthesis

Macromolecules present in oviductal fluid have been suggested to serve

important roles in sperm capacitation (Anderson and Killian, 1994), fertilization

(Boatman and Magnoni, 1995), and early embryo development (Gandolfi et al.,

1989). Therefore, alterations in oviductal biosynthetic activity and protein

synthesis and secretion may affect conception rate. Steroid modulation of

oviductal synthesis and secretion of proteins has been characterized in sheep

(Buhi et al., 1991, Murray, 1993), baboon (Verhage and Fazleabas, 1988) and

swine (Buhi et al., 1989; Buhi et al., 1990). Buhi and others (1989) measured

the biosynthetic capacity of the oviduct (i.e., rate of incorporation of radiolabeled

amino acid precursor into newly synthesized protein) of pigs during the estrous

cycle, early pregnancy and in ovariectomized animals following steroid

replacement (Buhi et al., 1992). These studies indicated that bioactivity of the

oviduct is related to the hormonal status of animals. For example, incorporation

rate of radiolabeled leucine was greater when ovariectomized animals were

treated with E2 compared to P4. These findings were consistent with what was

found with intact animals, where a greater incorporation rate was found in the

periestrus stage of the estrous cycle.

It is important to keep in mind that functional regions within the oviduct

have specific roles probably associated with particular arrays of secretary

products. Thus, it is expected that different steroid environments (e.g., estrous

cycle vs. pregnancy) have distinct effects on each oviductal region,

characterizing a biosynthetic gradient of proteins across regions. For example,

in studies with bulls (Anderson and Killian, 1994), it has been demonstrated that

culture medium conditioned by IST tissue at estrus capacitated more sperm than

did medium conditioned by AMP. This increase was abolished by heating the

conditioned medium and inactivating proteins before incubation with sperm.

Staros and Killian (1998) showed that four unidentified oviductal proteins and a

P1-like protein (Boice et al., 1990; Binelli et al., 1999; Chapter 3) from non-luteal

oviductal fluid would associate with the zona pellucida, suggesting a modulation

of sperm/egg binding or embryonic development by oviduct-derived proteins.

Biosynthetic protein gradients have been reported in the pig and sheep (Buhi et

al., 1992; Buhi et al., 1996; 28, DeSouza and Murray, 1995; Murray, 1993).

Moreover, DeSouza and Murray (1995) reported differential secretion of a

chitinase-like protein, similar to P1 in response to steroid treatments in sheep,

while Buhi et al. (1996) showed differential expression POSP mRNA among

oviductal regions in pigs.

While some evidence has accumulated for roles of oviductal secretary

proteins on reproductive products, much less is known about roles of embryonic

secretary products. To the best of my knowledge, no reports have focused on

effects of presence of embryo on the pattern of secretary proteins from the

bovine oviduct. Buhi and coworkers (1989), working with porcine oviductal

secretary proteins, failed to demonstrate differences in rate of incorporation of

non-dialyzable, 3H-leucine labeled molecules between pregnant and cyclic

oviducts. Moreover, one-dimensional SDS-PAGE analysis of secretary proteins

did not indicate changes in patterns of de novo synthesized, secretary proteins.

This indicates that presence of the embryo had little effect on modulating

secretion of macromolecules from the oviductal epithelium. In contrast, a recent

report (Wakuda et al., 1999) showed that presence of embryos in mice which

had their uterotubal junction ligated on day 1 of pregnancy, enhanced

implantation rate of embryos transferred to the uterus. This was in comparison

with pseudopregnant mice, which had uterotubal junction ligated before or after

mating with vasectomized males, and, also mated females which had uterotubal

junction ligated before mating (all mice had blastocysts transferred to uterus on

day 4). Embryo-dependent factors have not been identified, but clearly

influenced embryo development in that species.

Regulation of Reproductive Processes Occurring in the Oviduct

As summarized by Hafez (1993b) and discussed by Harper (1982) and

Anderson (1991), transport of unfertilized and fertilized eggs and sperm in the

oviduct is regulated by four primary forces: (1) frequency and force of

contractions of the oviductal musculature, influenced by endocrine and neural

mechanisms; (2) direction and intensity of beating of cilia, which conditions

movement of oviductal fluids; (3) secretary activity of non-ciliated cells, which is

dependent on the E2/P4 ratio; and (4) hydrodynamic properties of luminal fluids.

Changes in these factors are modulated by concentrations of ovarian steroids.

The outcome of these activities is efficient transport of gametes and embryos

and fertilization. Next, I will emphasize the concerted actions of these factors for

the mechanisms of egg pick up and fertilization.

Ega Dick up. At the time of ovulation, there is a noticeable increase in

frequency and amplitude of contractions in the smooth musculature supporting

the oviduct. Contractions of the mesotubarium superior and mesosalpinx draw

the oviduct in a crescent shape and slide the fimbriae over the surface of the

ovary. The fringe-like folds in the INF contract rhythmically to repeatedly touch

the ovarian surface (Hafez, 1993a). This pattern of movements constitutes an

efficient mechanism to pick up ovulated oocytes. Moreover, maximum density of

ciliated cells in the oviduct occur in the INF. During ovulation, the strokes of cilia

in the fimbriated portion of the oviduct are synchronized to propel the oocyte

towards the oviductal lumen. Furthermore, volume of oviductal fluid sharply

increases 2 days before estrus to reach maximum rate one day after estrus,

which coincides with the period of ovulation and reception of oocyte by the INF

(Perkins et al., 1965). Muscular contractions, cilia beating and fluid secretion are

controlled by ovarian steroids, being stimulated by high periovulatory E2

concentrations.

Fertilization. In the cow, oocytes are transported rapidly to the site of

fertilization, above the isthmoampullar junction, where they spend most of their

time in the oviduct (Aref and Hafez, 1973), and then are transported rapidly

through the IST into the uterus (Anderson, 1991). A balance among the effects

of cilia beating, muscular contractions modulated by catecholamines and fluid

flow rate cause this egg "lock up" at the site of fertilization. It has been proposed

that the IST of the cow is contracted throughout estrus, and that norepinephrine

release after estrus causes relaxation of the IST to allow embryo transit into the

uterus (Isla et al., 1989). This is supported by data from EI-Banna and Hafez

(1970), who showed a dramatic change in the surface area of the IST lumen

from estrus to 3 days after estrus (0.06 mm2 to 1.89 mm2). Meanwhile, if the

animal had been inseminated, sperm are migrating up the reproductive tract and

arrive at the IST portion of the oviduct, where their movement is slowed (Hunter

and Wilmut, 1982). It is hypothesized that biochemical and biophysical

properties of the IST may work to impede upward migration of spermatozoa,

including narrow isthmic lumen, viscous isthmic mucous and oviductal

25

musculature contractions (Ellington, 1991; Hafez 1993b). Within the IST, sperm

undergo hyperactivation, which is required for final sperm transport, completion -

of sperm capacitation and the acrosome reaction. Eventually, spermatozoa

become exposed to ampullary fluid, detach from the IST epithelium and continue

migration towards the site of fertilization. The control of concerted, opposite

direction-movement of sperm and eggs at similar times in the oviduct is

intriguing. Perhaps the isthmoampullary junction acts to retain oocytes in the

ampulla, while spermatozoa are allowed to enter the IST (Anderson, 1991). Low

doses of estrogen cause "tube locking", retaining ova at the isthmoampullary

junction, while larger doses promote quick movement through the isthmus and to

the uterus (Hawk, 1988).

Oviductal Function and Reproductive Failure in Cattle

The fact that oviductal function is regulated in the multi-factorial,

integrated fashion described above could lead one to hypothesize that

perturbations in the system could easily lead to reproductive failure. However, in

normal cattle, embryonic losses occurring during the time when the embryo is in

the oviduct are small, relative to other phases, as described above. This could

be interpreted at least in two ways. First, one could say that the oviduct plays

only a passive role on the processes of gamete transport, fertilization and early

embryonic development. In this view, the oviduct would keep default modes of

function (i.e., similar in presence or absence of an embryo), modulated by

26

patterns of steroids. Gametes and embryos would tolerate mild perturbations in

the system and the reproductive processes would be carried out by internal, pre-

determined programs, modestly influenced by the oviductal environment. The

oviduct would basically provide a physical substratum for events to occur. Leese

(1988) suggested that this possibility could be appropriately tested by trying to

culture embryos on an epithelium anatomically related to the oviduct, such as the

trachea (i.e., ciliated, secretary, containing active chloride ion pump). To support-

this first possibility, there is the fact that embryos can be matured and fertilized in

vitro in the absence oviductal cells, tissue or conditioned medium. Alternatively,

one could say that oviduct-gametes/embryos relationships have been optimized

in the course of evolution, to become a robust system, with little chance for

failure. Specific interactions would be required for success of reproductive

processes, including synthesis and secretion of oviductal proteins in a regional

and timely fashion. Moreover, such unique set of proteins would interact with

gametes/embryos to maximize reproductive output. To test this last possibility,

secretary proteins in the oviduct would first need to be identified. Then,

removing specific proteins from the system with use of immunoneutralization,

knockouts, transgenics and anti-sense models for example, should provide

evidence for their importance. For example, addition of specific antibodies for a

hamster E2 -dependent oviduct protein prevents in vitro fertilization (Sakai et al,

1988).

Uterine Function and Reproductive Failure in Cattle

As mentioned before, there is approximately a 30% rate of embryonic

losses occurring from days 8 to 16 of early pregnancy, which represents a period

of uterine localization of the embryo. In this section I will attempt to describe key

aspects of uterine physiology that can be considered when trying to solve the

problem of embryonic mortality during this period.

The Uterine Environment

The uterus is considered to be an extension of the oviduct (Bartol, 1999),

therefore, several of the principles regarding biophysical and biochemical

properties discussed for the oviduct also will apply here. Compared to the

oviduct, bovine embryos will spend a much longer interval of time in the uterus

(280 days vs 4 days, on average; Catchpole, 1991) which permits a much

broader set of communications between the maternal and embryonic units. To

support this notion is the fact that although there are not remarkable embryo-

induced changes between the cyclic and the pregnant oviduct, this is the

opposite for the uterus. The thesis of this section is that presence of the embryo

conditions the uterine environment, to support embryonic development.

The uterine cycle. Similar to the oviduct, the uterus also undergoes

changes dependent on the stage of the estrous cycle in response to changes in

concentrations of ovarian steroids. However, the uterus has the unique role of

controlling length of ovarian cycles (described above) and, as a consequence,

its own uterine cycle. The uterine cycle can be divided into a long progestational

phase and a short, estrogen-dominated phase (Hansel and Convey, 1983). At

the end of progestational phase, the uterus gains the ability to produce and

secrete PGF2,, which acts to cause structural and functional demise of the CL

(McCracken et al., 1971). An immediate consequence of PGF2,actions is a

decrease in circulating concentrations of P4 (Nett et al., 1976). This initiates the

estrogen dominated period that lasts until the next ovulation and formation of

new CL. Controlling CL life span, the uterus controls the ovarian cycle. During

pregnancy, presence of the concepts blocks luteolytic mechanisms so that the

CL remains functional and the uterus remains in a progestational stage until

parturition (McCracken et al., 1984). Similar to oviducts, uterine morphology and

secretary activity are modulated by ovarian steroids, as discussed below.

Functional anatomy and morphology. The uterus is suspended in the

pelvis by the mesometrium, a caudal division of the broad ligament. In cows, the

uterus can be described anatomically in two continuous portions, the gestational

part of the uterus (consisting of uterine horns and uterine body), and the cervix.

Similar to the oviduct, the uterus is a tube-shaped organ, which contains a lumen

(Bartol, 1999). Histologically, a cross section of the uterus reveals an inner

mucosal layer, the endometrium, an adluminal layer of smooth musculature, the

myometrium and an outer, serious peritoneal coat of the uterus, the perimetrium

(Bartol, 1999). For the remaining of this discussion, I will focus on characteristics

and processes occurring in the endometrium. The endometrium is lined with a

single layer of epithelial cells and contains simple, coiled, tubular glands. Glands

are relatively straight at estrus, but become more coiled and complex as

progesterone levels rise as the estrous cycle progresses (Hafez, 1993a).

Glandular secretions, the hystotroph, constitute a nutrient-rich mixture required

for development of the concepts (Bazer and First, 1983), and will be discussed

next. Underneath the luminal epithelium and around the glands is the

endometrial stroma, composed of stromal cells distributed in greater or lower

density patterns, depending on the location. In the cow, between 100 to 150

aglandular ridges are present, the caruncles (Flood, 1991). During pregnancy,

caruncles become attached to specialized areas of the allantochorion of the

concepts, the cotyledons, to form placentomes. Placentomes are units for

exchange of gas and nutrients between maternal and embryonic units. A better

description on the process of attachment and formation of placentomes is given

afterwards in this review. A highly dynamic and organized microvasculature

supplies myometrial and endometrial tissues. They originate from uterine

branches of the ovarian arteries (supply uterine body and uterine horns), uterine

arteries (supply uterine body and uterine horns) and urogenital artery (supplies

caudal uterus and cervix). In the cow, uteroovarian relationships exist in that

demise of the CL (luteolysis) is regulated by the uterine horn adjacent to the

ovary containing CL. Luteolysis is accomplished by countercurrent exchange of

the uterine produced luteolysin, PGF2,, between the uteroovarian vein and the

ovarian artery. The later is coiled about the surface of the uteroovarian vein

(Bartol, 1999). Countercurrent exchange of PGF2,was demonstrated elegantly

by Knickerbocker and coworkers (1996). The authors sampled blood originating

from uterine branch of ovarian artery(UBOA) and facial artery (FA), and

measured changes in PGF2,concentrations in response to a challenge with E2.

There was a greater concentration of PGF2a in UBOA compared to FA, indicating

existence of local countercurrent exchange between uterine venous drainage

and ovarian artery.

The hystotroph. Hystotroph is the secretions present within the uterine

lumen for nourishment of the developing concepts (Roberts and First, 1983). A

broader definition should also include functions such as paracrine regulation of

concepts physiology and development and protection of the concepts from the

maternal immune system. Solymosi and Horn (1994) measured protein content

in uterine milk (i.e., hystotroph) of cows and determined that 73% of the dry

matter content was composed proteinaceous material. Information on nature of

proteins contained in uterine milk is limited in cattle. Electrophoretic analysis

reveled at least nine proteins, seven minor and 2 major, which were identified as

lactoferrin and acid phosphatase (Bazer and First, 1983). Lactoferrin may have

a bacteriostatic function in the uterine luminal environment. The acid

phosphatase has basic PI, which is similar to uteroferrin in pigs. Uteroferrin is

involved in iron transport to the concepts (Roberts and Bazer, 1988), but

whether bovine acid phosphatase has the same role in cattle is unknown.

31

MacKenzie and coauthors (1997) reported expression of retinol-binding protein

(RBP) in bovine uterus, and monitored steroid modulation of expression of this

protein during estrous cycle and early pregnancy. The hypothesized role of RBP

is to regulate transport of vitamin A to the concepts. Finally, expression of

growth factors involved in the growth hormone (GH)-insulin-like growth factor

(IGF) axis were examined by Kirby and others (1996). They reported expression

of IGF-1, IGF-binding protein (IGFBP)-2 and 3 and GH receptor in the uterus.

Collectively, uterine milk has functions similar to those of the oviductal fluid, to

provide an adequate microenvironment for concepts development. As the

embryo develops, it starts to contribute with its own secretions to the pool of

molecules composing the uterine fluid, which becomes more complex and may

exert regulatory functions that influence both the concepts and the uterus.

Regulation of Reproductive Processes Occurring in the Uterus

As mentioned previously, the uterus plays specific roles both during the

estrous cycle and during pregnancy. During the estrous cycle, the uterus

regulates ovarian function and the ovarian cycle and the uterus in turn is

regulated by actions of ovarian steroids. During pregnancy, functions of the

uterus include transport, storage and maturation of spermatozoa, recognition

and reception of embryos, provision of an embryotrophic environment for

concepts development during gestation, and expulsion of fetus and placenta at

parturition (Bartol, 1999). At this physiological state, uterine function is regulated

32

both by ovarian steroids and embryonic bioactive molecules. During pregnancy,

misregulation of uterine function may lead to embryonic mortality. For the

remaining of this section I will focus on the mechanisms regulating PGF2,

production from the cyclic uterus, and on mechanisms of maternal recognition of

pregnancy related to CL maintenance occurring in the pregnant uterus.

Cyclic uterus. The reason for the uterine cycle is to provide repeated

opportunities for pregnancy at relatively short intervals. In practice, the turning

point in the uterine cycle is the commitment to either luteolysis or pregnancy. In

the absence of pregnancy, the progestational stage is finished by the uterus-

induced demise of the P4 source. Demise of the CL is accomplished by pulsatile

secretion of uterine PGF,(Nancarrow et al., 1973; Kindahl et al., 1976). Then,

the estrogen-dominated uterus prepares for reception of sperm, initially, and

reception of the early developing embryo. This cycle repeats until successful

establishment of pregnancy. However, early pregnancies will be terminated if

the uterine cycle is not interrupted.

Secretion of PGF,,. During the estrous cycle, the presence of two distinct

patterns of PGF, release is easily distinguishable: a basal release and a

stimulated, pulsatile release. Initial measurements of PGF2were performed in

samples collected from the venous drainage of the uterus, which required

surgical cannulation (Nancarrow et al., 1973). Measuring peripheral

concentrations of 15-keto-13, 14-dihidroprostaglandin F2,, (PGFM) the main

metabolite of PGF2,found in the circulation, facilitated study of such patterns

33

(Kindahl et al., 1976). It was determined that for most of the estrous cycle, basal

secretion of PGFM ranged from 25 to 70 pg/ml in one heifer and from 60-100

pg/ml in a second heifer. However, around the time of luteolysis, four peaks of

about 500 pg/ml and four peaks of about 250 pg/ml were observed for the first

and second heifers, respectively (Kindahl et al., 1976). Despite the clear among

animal variability in this small experiment, there was an evident decrease in P4

concentrations, from ~5 ng/ml to less than 1 ng/ml, within 24 hours after the first

PGFM peak for both heifers.

Pulsatile release of PGF,,. Generation of PGF2, pulses requires presence

of a stimulatory signal and a responsive uterus. A responsive uterus contains

receptors for the stimulatory signal, functional intracellular pathways to transduce

the stimulus into a secretary pulse and adequate amounts of substrate for PGF2,

synthesis.

In cattle, nature of the stimulatory signal for production of luteolytic pulses

of PGF2, remains unclear. It has been accepted generally that oxytocin is the

major stimulator of PGF, secretion in cattle. Armstrong and Hansel (1959)

demonstrated that exogenous oxytocin caused luteolysis in heifers. Moreover,

injections of oxytocin increased concentrations of PGF2. in the uterine vein

(Milvae and Hansel, 1980) and increased concentrations of PGFM in peripheral

circulation (Lafrance and Goff, 1985) in cows. However, while oxytocin is able to

stimulate PGF2 secretion in these experiments, it remains unclear whether

oxytocin is in fact required for the process of luteolysis. In a recent report,

Kotwica and others (1997) demonstrated that administration of an efficacious

oxytocin receptor antagonist (CAP-527) failed to block normal luteolysis in cows.

Moreover, treatment of endometrial explants from days 16-17 cyclic cows with

oxytocin failed to stimulate secretion of PGF2,, both in static (Arnold et al., 1999)

and in perifusion (Del Vecchio et al., 1990) culture systems. This raises the

possibility that perhaps ligands other than oxytocin are required for luteolysis in

cattle. Alternatively to oxytocin, possible stimulators of pulsatile PGF2.secretion

include E2 and LH.

Irradiation of ovarian follicles, which reversibly eliminates production of

follicular E2, delays luteolysis and extends the length of the estrous cycle

(Hughes et al., 1987). Thus, follicular E2 plays a major role in these events. In

heifers, Thatcher and coworkers (1986) demonstrated that injections of E2 on

day 13 of the estrous cycle stimulated release of PGFM starting 3 hours after the

injection, peaking at 6 hours and returning to basal levels by 10 hours. In

addition, heifers injected with E2 underwent luteolysis 96 hours after injections,

while it took 125 hours for control heifers. To confirm that the PGFM increases

measured in the peripheral circulation represent PGF2 of uterine origin,

Knickerbocker and others (1986) measured a sharp increase in PGF2, in the

uterine vein of cows treated with E2. Collectively, these findings indicate that

both endogenous and exogenous E2 are able to stimulate secretion of PGF2,

and to cause luteolysis.

35

A novel concept regarding control of luteolysis involves the actions of LH

in the endometrium. Friedman et al. (1995) reported the presence of LH binding

sites in bovine endometrium that were maximal in endometrium from days 15 to

17 which corresponds to the time of luteolysis. In addition, production of PGF2,,

was stimulated when endometrial cells from days 15 to17 of the estrous cycle

were treated with LH in vitro. In a series of preliminary experiments (Fields,

personal communication) ovariectomized, P4-treated cows were injected with

either saline or E2 and 4 hours later injected with either saline or human chorionic

gonadotropin (hCG, a long half life LH analog). Concentrations of PGFM in

plasma were elevated only for groups pre-treated with E2. In addition, hCG

injection elicited a pronounced release of PGFM compared to saline. It can be

concluded that exposure to E2 is required for the endometrium to secrete PGFM

in response to LH. Physiologically, it could be hypothesized that E2 acts at the

endometrium to enhance responsiveness to circulating LH, thereby evoking

PGF2, secretion during luteolysis. Indeed, a decline in P4 will elevate plasma LH

that may contribute to a continued secretion of PGF2,to re-enforce the luteolytic

process. Mechanistically, this could be accomplished by increasing

concentration of LH receptors in the endometrium or by connecting intracellular

pathways stimulated by LH with the PGF2, secretary machinery.

In summary, it is doubtful that oxytocin is the sole stimulator of pulsatile

secretion of PGF,,. It is more probable that other effectors such as E2 and LH

act in concert with oxytocin to stimulate luteolysis.

Since the dominant paradigm in the field of luteolysis has been that

oxytocin is the major stimulator of PGF2, pulsatile secretion, experiments to test

presence of a responsive uterus had as an endpoint development of oxytocin

receptors and secretion of PGFa in response to oxytocin. Based on this

paradigm, it has been well established that exposure of the uterus to

progesterone is required for the uterus to acquire responsiveness to oxytocin.

McCracken (1980) proposed that P4 has the ability to inhibit synthesis of E2

receptors, and synthesis of oxytocin receptors is an E2 -dependent process. As

long as P4 inhibits synthesis of E2 receptors, E2 is unable to stimulate synthesis

of oxytocin receptors. Moreover, McCracken and others (1984) suggested that

the uterus eventually becomes refractory to inhibitory effects of P4, allowing

oxytocin receptors to be expressed, which leads to pulses of PGF,. Lafrance

and Goff (1988) treated long-term ovariectomized heifers with P4 for 0, 7, 14 or

21 days then measured PGFM in response to an oxytocin challenge. After 7, 14

or 21 days of P4 -priming there was a significant increase in plasma PGFM after

oxytocin injection, but no increase was noticed in animals that did not receive P4.

Silvia and coauthors (1991) put forth the question of whether requirement for

long term exposure to P4 is due to stimulatory effects that take at least 10 days

to build up (7 days in Lafrance and Goff, 1988) or to slow development of the

condition whereby the uterus is desensitized to P4 inhibitory effects. Collectively,

the concept of a responsive uterus means that this organ has been primed by P4,

and as a result, the uterus becomes responsive to the luteolytic stimulus.

I will next review intracellular pathways for generation of the pulsatile

secretion of PGF2a. Little experimentation has been done to uncover potential

intracellular pathways activated by estradiol or LH, so focus is on the well studied

and established pathway of oxytocin stimulation (Flint et al., 1986; Burns et al.,

1997; Thatcher et al., 1997). Oxytocin receptors start to increase in the P4 -

primed, responsive uterus. Oxytocin originating from the neurohypophyseal lobe

of the pituitary gland binds to the seven transmembrane-domain, G protein-

coupled receptors and activates phospholipase C (PLC). The PLC cleaves

membrane phosphotydilinositol bisphosphate, yielding inositol trisphosphate (IP3)

and diacylglycerol (DAG). The IP3 binds to specific receptors in the endoplasmic

reticulum resulting in release of calcium from internal stores into the cytosolic

compartment. The DAG activates protein kinase C (PKC), leading to serine

phosphorylation of cytosolic, calcium-dependent phospholipase A2 (PLA),

probably through a MAP-kinase dependent pathway (Lin et al., 1993). The IP3-

stimulated increase in cytosolic calcium acts to further stimulate PLA, activity

(Clark et al., 1991). Stimulated PLA2 translocates to the membrane where

phospholipid substrates are located (Clark et al., 1991). Activated, membrane-

bound PLA2 cleaves arachidonic acid (AA) from phospholipids. Free AA is

converted to prostaglandin H2 (PGH2) by the enzyme cyclooxygenase-2 (COX-

2). Prostaglandin F2,synthase converts PGH2into PGF2a, which is then released

into the uterine circulation. In the endometrium, this process occurs

preferentially in epithelial cells compared to stromal cells (Danet-Desnoyers et

38

al., 1994). As discussed beforehand, PGF2,gains access to the ovary through a

counter-current mechanism. Binding of PGF2,to receptors in the CL stimulates

release of luteal oxytocin that in turn binds oxytocin receptors in the

endometrium to elicit further release of PGF,,, characterizing a positive feedback

loop.

Arnold and coauthors (1999) demonstrated that a responsive uterus is not

necessarily a oxytocin-responsive uterus. They incubated endometrial explants

obtained from day 17 cyclic cows with oxytocin or with intra-cellular stimulators of

the PGF2 -generating cascade described above. They showed that despite

oxytocin failure to stimulate PGF2, secretion, the stimulator of PKC activity,

phorbol 12, 13 dibutyrate (PDBu), and the stimulators of PLA2 activity, calcium

ionophore and melittin, were able to induce PGF2, release acutely. This supports

the notion that alternative ligands to oxytocin can play a role on pulsatile

secretion of PGF2,.

A final comment on the role of P4 -priming of the uterus for pulsatile PGF2,

secretion, relates to P4 ability to induce accumulation of lipid droplets in bovine

uterine epithelial cells (Brinsfield and Hawk, 1973). In mice, such lipid droplets

contain phospholipids (Silvia et al., 1991), which are substrate for PLA2 and

source of AA, as mentioned above. Progesterone also induces synthesis of

COX-2 (Raw et al., 1988).

Pregnant uterus. In cattle, the vast majority of embryos are found in the

uterine horn ipsilateral to the ovary where ovulation occurred, indicating a readily

"attachable" embryo, which only migrates minimally within the uterus (Flood,

1991). Embryos undergo rapid morphological changes in the first 3 weeks of

pregnancy. After a series of cellular divisions, formation and hatching of

blastocysts (day 9-10 after ovulation; Betteridge and Flechon, 1988),

conceptuses start to elongate on day 12 (Betteridge et al., 1980), to occupy the

whole length of the uterine horn ipsilateral to the CL by day 17 and to reach the

tip of the contralateral horn on day 21 (Kastelic et al., 1988; Flood, 1991). The

first intimate connection between the concepts and the uterus occurs between

days 18 and 20 of pregnancy, when numerous papillae penetrate the openings

of uterine glands (Guillomot et al., 1981). There is intimate contact of matemal

and embryonic tissues starting with apposition of apical cell membranes of

aligned epithelia from both units. Actual adhesion begins around day 22 and is

completed on day 27 after insemination. Adhesion is characterized by

interdigitation of embryonic and maternal microvilli (Flood, 1991). The next

series of events include development of placentomes and growth of placental

tissues.

Maternal recognition of pregnancy associated with CL maintenance

As mentioned previously, the turning point in the uterine cycle is the

commitment to either luteolysis or pregnancy. In cattle, commitment to

pregnancy is only accomplished if adequate signaling exists between maternal

and embryonic units. Maternal recognition of pregnancy has been defined as

the process by which the periattachment concepts signals its presence to the

40

maternal unit, as reflected by maintenance of the CL (Short, 1969; reviewed in

Hansen, 1991). More specifically, the process of maternal recognition of

pregnancy requires that embryonic molecules interact with the uterine

endometrium and change its program, so that pulsatile secretion of PGF2'is

blocked and thereby luteolysis is impeded. The net result is continuous

secretion of P4 by the CL, which is required for continuation of pregnancy. Roles

of P4 include continuous stimulation of uterine secretions and inhibition of

smooth muscle contractions (Hafez, 1993a). In cattle, the critical period for

maintenance of pregnancy is around day 17 of the estrous cycle. Betteridge and

others (1980) transferred embryos to synchronized recipients and demonstrated

that pregnancy was only maintained if embryos were transferred prior to day 17.

Moreover, inter-estrus interval increased from 20 to 25 days when conceptuses

were removed on day 17 vs. day 15 of pregnancy (Northey and French, 1980).

Based on the model proposed above, pregnancy effects on suppression

of pulsatile PGF2, could be exerted at several levels: (1) suppression of the

PGF2,-releasing stimulus (i.e., oxytocin, LH, E2), (2) alterations of the P4-primed

uterus (i.e., PGF,,-synthesizing machinery), and (3) decrease in substrate

required for PGF,2synthesis (i.e., AA). Another possibility is presence of an

conceptus-induced luteoprotective action, where CL would become less

susceptible to luteolytic effects of PGF2,,. However, since PGF2,,pulses are

effectively blocked during early pregnancy, this possibility will not be considered

in this discussion.

There is good evidence for pregnancy-induced suppression of luteolytic

stimulus in cattle relative to the attenuation of E2 effects. Pregnant cows have

reduced circulating concentrations of E2 (Pritchard et al., 1994), probably as a

result of reduced folliculogenesis (total production of follicles) and decreased

production of E2 per follicle (decreased aromatase activity; Thatcher et al., 1991).

Moreover, in day 18 pregnant cows, administration of E2 stimulates only a

modest increase in PGF2a secretion, indicating that presence of the concepts

attenuates E2 effects (Thatcher et al., 1984).

Regarding alterations on the P4 -primed uterus, Arnold and others (1999)

incubated endometrial explants obtained from day 17 cyclic or day 17 pregnant

cows with intracellular stimulators of PGF2, synthesis, and measured

concentrations of PGF2 secreted into the culture medium. Melittin, PDBu and

calcium ionophore each stimulated release of PGF,2 from explants of cyclic cows

compared to control treatment (medium alone). In contrast, all stimulators

mentioned above failed to induce release of PGF2, in explants originated from

pregnant cows. This indicated that pregnancy affected the intracellular PGF,,-

generating machinery to suppress its ability to stimulate PGF2,,. Interpretation of

these data suggests that pregnancy may have inhibitory effects at each of the

steps stimulated by treatments, which include PKC (PDBu) and PLA2 (melittin,

ionophore). Alternatively, pregnancy may affect a distal, convergence point in

the pathway, for example, at the COX-2 level. Effects on the enzymatic

machinery can be to decrease expression and/or activity of PKC, PLA2and COX-

42

2. In fact, existence of a pregnancy-induced inhibitor of COX-2 activity has been

found in the endometrium of cows (Basu and Kindahl, 1987; Gross et al, 1988).

Danet-Desnoyers and others (1993) identified linoleic acid as the active molecule

in bovine endometrium which acted to decrease COX-2 activity. Moreover,

linoleic acid acted as a competitive inhibitor of AA on a PGF2, generator assay

(Thatcher et al. 1994b). It is possible that altered lipid metabolism in the

pregnant endometrium increases availability of linoleic acid to inhibit COX-2

activity and thereby decrease PGF2, production.

Finally, pregnancy could change lipid composition and metabolism in the

endometrium to inhibit PGF2,synthesis. Thatcher and others (1995) compared

concentrations of free linoleic and free AA in endometrial microsomes from day

17 cyclic and pregnant cows. They found that pregnancy decreased

concentrations of AA and increased concentrations of linoleic acid compared to

estrous cycle, to result in a change of the ratio of linoleic to AA of 0.6 to 2.4 in

endometrium between cyclic and pregnant cows.

Effectors of maternal recognition of pregnancy: IFN-T

A considerable amount of research focused on identification and

purification of concepts products with the PGF2,-secretion inhibitory activity

required for maintenance of pregnancy. A family of molecules has been

identified as the embryonic antiluteolytic factor in ruminants, named IFN-r

(Thatcher, 1999). For a historical prospective, see Martal et al. (1979), Godkin et

al. (1982), Bartol et al. (1985), Helmer et al. (1987), Imakawa et al. (1989) and

Roberts et al. (1992). Isoforms of blFN-T are glycosylated, have molecular

weights between 22 and 24 kD and vary in isoelectric forms between PI 6.3 and

6.8 (Helmer et al., 1987; Anthony et al., 1988).

Antiluteolytic effects of bovine bIFN-t (blFN-r) have been examined both

in vivo and in vitro. Intrauterine infusions of highly enriched blFN-t complex

(Helmer et al., 1989b) and recombinant blFN-r (Meyer et al., 1995) extended

lifespan of CL in cows, compared to control infusions. Moreover, PGF,, release

in response to an oxytocin injection was suppressed in day 17 cyclic cows

infused with recombinant blFN-r compared to controls (Meyer et al., 1995).

Danet-Denoyers and others (1994) tested the ability of blFN-r to suppress

basal and oxytocin-stimulated secretion of PGF,, from primary cultures of

endometrial epithelial cells obtained from day 15 cyclic cows (Danet-Desnoyers

et al., 1994). Twenty four hour-incubation with bIFN-t reduced both basal and

oxytocin-stimulated secretion of PGF2,,. This agrees with data from Meyer and

coworkers (1996) and Xiao et al. (1999). Meyer and others (1996) reported that

endometrial epithelial cells obtained from cows which received intrauterine

infusions of blFN-T secreted less basal and oxytocin-stimulated PGF2, compared

with cows infused with a control protein. Moreover, Xiao and coworkers (1999)

cultured endometrial epithelial cells obtained from cows of days 1 to 4 of the

estrous cycle in presence of oxytocin or a combination of oxytocin and bIFN-r.

Similar to the data described above, blFN-T effectively reduced both oxytocin-

and phorbol ester-stimulated PGF2, secretion. In contrast, Asselin and others

(1998) showed that blFN-T increased secretion of PGF2,from endometrial

epithelial cells from days 1 to 5 of the estrous cycle. However this effect was

only significant when extremely high doses of blFN-r (20 ~gg/ml) were used.

Collectively, these data support the concept that blFN-T interacts with

endometrial epithelium and affects the PGF2,-generating machinery to decrease

PGF2. production. To further test this possibility, Arnold and others (1999)

infused either blFN-r or a control protein (bovine serum albumin) in the uterus of

cows from days 14 to 17 of the estrous cycle. Secretion of PGF2,was measured

in medium conditioned by endometrial explants cultured in presence of specific

intracellular stimulators of PGF2,synthesis. Incubations with calcium ionophore

and PDBu stimulated PGF2,secretion compared to medium alone in

endometrium from control cows, but not from blFN-T -treated cows. In contrast,

melittin stimulated secretion of PGF2from explants originating from blFN-r -

infused-cows. Overall, their data indicated that in vivo treatment with blFN-T

attenuated PGF2, production probably at the level of PKC, since PDBu

stimulation of PGF2,was reduced by blFN-r, whereas melittin stimulated PGF2,

secretion. This is in variance with the ubiquitous inhibitory effects of pregnancy

on stimulated PGF2 secretion (mentioned above), suggesting that other products

of pregnancy, and not blFN-t alone, probably also operate to inhibit PGF2,

production.

In an effort to pinpoint specific enzymes that blFN-r altered in the PGF2,-

generating cascade, Xiao and others (1999) measured messenger ribonucleic

acid (mRNA) and protein expression for COX-2 in endometrial epithelial cells

treated with oxytocin and with oxytocin in combination with blFN-r. Oxytocin

maximally stimulated COX-2 mRNA and protein from 3 to 24 hours compared to "

controls. Treatment with blFN-T reduced this effect of oxytocin, and this was

consistent with a reduction in PGF2,secretion in medium. In contrast, Asselin

and coworkers (Asselin et al., 1997) found that blFN-T actually stimulated

expression of COX-2, which would contradict the antiluteolytic role of blFN-r.

However, they also reported that blFN-T stimulated expression of an endometrial

prostaglandin E2-9-ketoreductase, which catalyzes the conversion of PGF, into

PGE2 (Asselin and Fortier, 1998). Since PGE2 has been shown to have luteo-

protective actions (Pratt et al., 1977), they proposed a model whereby blFN-T

actually-stimulates the PGF, -generating machinery, but a conversion of PGF2,

to PGE2 at the end of the cascade would support an antiluteolytic effect of blFN-

T.

It is expected that in order to stimulate intracellular changes resulting in

decreased PGF2,production, blFN-T needs to stimulate a receptor-mediated

mechanism of signal transduction. Such a mechanism should evoke intracellular

46

second messengers to ultimately regulate molecules involved in the generation

of PGF.,, Such regulation could involve synthesis of proteins inhibitory to the

PGF2 production cycle, or, could acutely activate molecules already present in

the cell to suppress PGF2,stimulatory actions. There is limited information on

the nature of IFN-r receptors. Knickerbocker and Niswender (1989) measured

numbers of unoccupied binding sites for IFN-r in endometrium of cyclic and

pregnant sheep. Number of unoccupied binding sites decreased for both cyclic

and pregnant ewes from day 4 to day 12. Then it increased for cyclic animals,

but was still decreased for pregnant ewes, indicating that blFN-T binding sites

were possibly being occupied by conceptus-secreted IFN-r. Interestingly, affinity

for binding sites increased after day 12 for pregnant ewes but decreased for

cyclic ewes.

Hansen and coauthors (1989) reported use of cross-linking experiments

to characterize association of iodinated ovine IFN-r to membrane peptides.

They identified binding of IFN-T to both 100 and 70 kD membrane polypeptides.

Comparison of binding kinetics of IFN-T with IFN-a in this experiment suggested

existence of different receptors for these two ligands. However, Li and Roberts

(1994) showed a reciprocal displacement of IFN-T and IFN-a from bovine

endometrial cell membranes, suggesting that binding sites for these two

molecules were the same. Recently, Han and Roberts (1998) reported cloning

and characterization of receptors for IFN-T in cattle endometrium. Sequences of

47

receptor subunits IFNaR1 and IFNaR2 are similar to ones utilized by other type I

interferons such as IFN-a. However, these receptors were not linked with

functional data, to demonstrate that such subunits are necessary and sufficient

to suppress PGF,,synthesis.

The JAK-STAT Pathway

The observations above lead to the assumption that blFN-t stimulated a

signal transduction system, the JAK-STAT pathway, similar to other type I

interferons (Schindler et al., 1992; Darnell et al., 1994; Darnell, 1997; Figure 2-

1). In this paradigm, interferon receptors do not contain intrinsic kinase activity,

but they are physically associated with protein tyrosine kinases from the Janus

family (JAK kinases). Binding of interferon to its receptor causes

phosphorylation of tyrosine residues in the JAK kinases and in the cytoplasmic

tail of the receptor. The tyrosine phosphorylated receptor attracts signal

transducer and activation of transcription, or STAT, proteins to close contact.

Members of the STAT family of proteins then become phosphorylated on

tyrosine residues and form homo- and hetero-dimers. Dimerized STATs migrate

to the nucleus where they bind to the specific regulatory elements located in the

promoter region of interferon-regulated genes. In this manner, STAT proteins

5 6
R1 R2

Nucleus

ISGF-3 A

IISRE
IFN-lnducble gene
STAT-I STAT-2

Figure 2-1. The JAK-STAT pathway of signal transduction and gene activation.
1) Binding of type I interferon (IFN) to the interferon-a-receptor (R)-2 chain causes
recruitment of R1; 2) dimerization of IFN receptor complex causes reciprocal tyrosine
phosphorylation of associated JAK kinases (tyk-2 and jak-1); 3) JAK kinases phosphorylate
receptor subunits in tyrosine residues; 4) unphosphorylated, cytosolic STAT proteins bind
receptor complex through SH2 domains present in STAT proteins; 5) JAK kinases
phosphorylate tyrosine residues in bound STATs, STATs dissociate from receptor complex
and associate in a heterodimer (STATs 1 and 2); 6) dimerized STATs translocate to the
nucleus, bind to the DNA binding protein p48, forming the ISGF-3 transcription complex,
which stimulate synthesis of IFN-inducible genes.

stimulate transcription of genes and synthesis of interferon-specific proteins. I

will next examine characteristics of molecules involved in this pathway, and then

describe evidence for existence of this pathway in the bovine endometrium.

Type I interferon receptors

Type I interferon receptor consists of two chains, IFNaR1 and IFNaR2,

which can be presented in different forms. The IFNaR1 is present as a full chain

(IFNaRla) and as a shorter splice variant (IFNaRls). The IFNaR2 chain exists

in soluble, short and long forms, designated IFNaR2a, IFNaR2b and IFNaR2c

respectively. Probably IFNaRla and IFNaR2c are the predominant forms

(Petska, 1997). Petska (1997) reviewed a series of experiments where the

different IFNaR chains were expressed in Chinese hamster ovary cells, and

ability of different type I interferons to signal through the different chain

combinations was evaluated. There is a remarkable diversity of such

interactions, in which specific interferons can only signal through specific

combinations of chains, but not others. Petska (1997) proposes that differential

expression of individual chains and ability of individual interferons to signal

through specific chain arrangements confers tissue-specific responsiveness to

interferons. For example, Platanias and coworkers (1996a) reported that IFN-p

signaling requires association of IFNaR1 with p100, a tyrosil phosphoprotein,

which was later identified as a particular chain of the interferon receptor complex.

To the best of my knowledge, these types of experiments have not been

conducted in bovine reproductive tissue, to test signaling ability of blFN-r.

Instead of the antiviral assays used in the experiments mentioned above

(Pestka, 1997), functional assays measuring suppression in synthesis of PGF2,

from endometrial cells would be in order. Moreover, existence of a blFN-r-

specific receptor chain remains elusive.

Colamonici and coworkers (1994a; 1994b) demonstrated that the tyrosine

kinase p135t*2, or tyk-2 is associated physically with the IFNaR1 chain of the

interferon receptor. Immunoblots revealed the ability of monoclonal antibodies to

IFNaR1 and to tyk-2 to reciprocally coimmunoprecipitate both proteins.

Association of tyk-2 was mapped to a 46-amino acid juxtamembrane region of

the IFNaR1 chain. Furthermore, they demonstrated that tyk-2 could directly

phosphorylate tyrosine residues in the IFNaR1 chain after stimulation with IFN-a

(Colamonici et al., 1994b). Besides binding to extracellular interferons and

associating with JAK kinases, the interferon receptor complex also has other

functions in the JAK-STAT pathway. The unstimulated IFNaR2 chain may

contain-associated unphosphorylated STAT proteins (Li et al., 1997). Binding of

interferon brings IFNaR2 and IFNaRI, which contains tyk-2, together.

Dimerization of receptor chains elicits transfer of STATs to the IFNaR1, where

STATs become tyrosine phosphorylated. This confirms the previous finding that

tyrosine 466 in the chain of IFNaR1 acts as a docking site for association of the

SH2 domain of STAT-2, and such binding is required for tyrosine

51

phosphorylation of STAT-2 by tyk-2 (Yan et al., 1996). Similar to STAT-2, STAT-

3 activation also requires binding to IFNaR1 (Yang et al., 1996). Furthermore, it

has been demonstrated that phosphotyrosine modules (i.e., sequence of amino

acids surrounding the tyrosine residue in the receptor chain) play a major role in

selecting which STAT binds (Gerhartz et al., 1996). The authors demonstrated

that a two point mutation in the phosphotyrosine module changed the specificity

of interferon-gamma receptor from STAT-1 to STAT-3.

Chains of the interferon receptor may also play roles independent of the

JAK-STAT pathway. For example, Abramovich and others (1997) reported

binding of a protein-arginine methyltransferase to the IFNaR1 chain. This finding

suggests that methylation of proteins may be a signaling mechanism

complementary to tyrosine phosphorylation, and methylation may be required for

full stimulation by interferons. In fact, cells deficient in this methylase activity by

antisense become less sensitive to the antiproliferative effect of interferons.

Finally, Platanias and coauthors (1996b) reported that the interferon receptor

mediates tyrosine phosphorylation of insulin receptor substrate 2 (IRS-2). The

IRS-2 molecules associate with IFNaR1 and become phosphorylated by tyk-2.

Moreover, phosphorylated IRS-2 associates with the p85 regulatory subunit of

the phosphatydilinositol 3'-kinase, suggesting that this kinase participates in the

interferon signaling cascade downstream from IRS-2. Collectively, the examples

presented above illustrate actions of the multifunctional interferon receptor. It is

tempting to speculate that some of such actions may be required for the

antiluteolytic roles of bIFN-r in the endometrium.

JAK kinases

Janus kinases or JAKs tyk-2 and jak-1 are associated respectively with

IFNaR1 and IFNaR2 and are involved in tyrosine phosphorylation of STAT

proteins. The carboxy-terminal domains of the jak kinases share considerable

sequence homology with the catalytic domains of other protein tyrosine kinases.

The amino-terminal half of the jaks contains regions of sequence homology to

other members of the jak family and the extreme amino-terminal domain

probably is involved in association with interferon receptor chains (Williams and

Haque, 1997). Ligand-mediated dimerization of interferon receptor chains is

required for interferon-stimulated signal transduction. Dimerization evokes

reciprocal tyrosine phosphorylation and consequent activation of JAKs

associated with interferon receptor chains (Ihle et al., 1995). Phosphorylation of

the kinase is the first of three tyrosine phosphorylations culminating in STAT

activation. Activated JAKs phosphorylate tyrosine residues on the interferon

receptor chains, which serve as docking sites for STATs, as mentioned above.

Lastly, STATs are phosphorylated by the JAKs (Darnell, 1997). Activated JAKs

are not specific for particular STATs. Different receptors can activate the same

STATs through different JAKs. Moreover, STAT docking sites can be

interchanged between different cytokine receptors, and the STAT specific for the

docking site present will be activated by binding of the ligand specific for the

extracellular domain of the receptor (Stahl et al., 1995). Therefore, STAT

activation is determined more by specific interactions between STATs and their

receptors than by specific JAKs associated with receptor chains.

STATs

Unlike other common intracellular second messengers, STATs not only

convey the extracellular signal to the interior of the cell, but they themselves

carry such a signal to the nucleus, acting as transcription factors to activate

transcription of genes induced by particular ligands. I will focus this discussion

on STATs 1, 2 and 3, although STATs 4 to 6 have been described (Darnell,

1997). STAT-la and 1p are encoded by alternative splicing of a single mRNA

transcript. Human STAT-la consists of 750 amino acids, while the extreme

carboxy-terminal 38 amino acids are missing for STAT-11. STAT-2 is composed

of 851 amino acids. STATs 1, 2 and 3 have significant sequence homology (Fu

et al., 1992; Zhong et al., 1994). The domain distribution in the STAT molecule

includes a centrally-located DNA-binding domain, a carboxy-terminal

transcription activation domain, and SH2 and SH3 domains located in between

them (Fu, 1992; Figure 2-2). The SH2 domain allows docking to tyrosine

phosphorylated sites in the IFN and cytokine receptors, as discussed above, and

also STAT dimerization. SH2 domain sequences are specific for each STAT, but

mutant STATs 1 and 3, in which SH2 domains were swapped, completely

400 500 600 700

DNA
binding

Y

Figure 2-2. Domain structure of the STAT-1 protein. The diagram represents the linear
structure of STAT-1 oriented in a amino- (leftmost) to carboxy- terminus sequence. DNA
binding domain, SRC homology 2 (SH2) domain, SH3 domain and transcription activation
domain (TAD) are represented in the sequence they occur in the STAT-1 molecule. The site of
tyrosine (Y) phosphorylation is also represented.

reversed their specificity for interaction with specific phosphotyrosine motifs

(Hemmann et al., 1996). This indicates that the SH2 domain is the sole

determinant of specific STAT factor recruitment to receptors. STATs contain a

unique tyrosine residue in the carboxy-terminal region (Y701, Y690 and Y705 for

STATs 1, 2 and 3 respectively). A recently developed model for STAT activation

(Li et al., 1997; Figure 2-1) proposes that unphosphorylated STATs 1 and 2 are

associated with the IFNaR2 chain. Binding of interferon causes dimerization of

this chain with IFNaR1. Tyrosine phosphorylated residue 466 of IFNaR1 binds

the SH2 domain of STAT-2, which is then phosphorylated on tyrosine 690,

providing a docking site for the SH2 domain in STAT-1. STAT-1 is

phosphorylated on tyrosine 701, and then dimerizes with STAT-2 through

reciprocal binding of tyrosine phosphorylated residues with SH2 domains.

However, an unsolved question is what drives SH2 domains of STATs to

dissociate from a higher affinity interaction with receptor phosphotyrosine to form

dimers which association is mediated by a lower-affinity phosphotyrosil

interaction (Greenlund et al., 1995). In light of this question, Gupta and

coauthors (1996) proposed an alternative model for STAT binding and dimer

formation. After binding to the receptor phosphotyrosine motif, the STAT shifts

its target to the tyrosine motif in the tyrosine kinase. Tyrosine phosphorylation of

STAT would cause a conformational change to destabilize this interaction with

the kinase, and STATs would then be driven to form more energy-stable

interactions with other STATs and form dimers. They based this model on the

finding that SH2 domains from STATs 1 and 2 bind with high affinity to

phosphotyrosine motifs on JAK kinases.

STAT dimers are competent to bind DNA. Known DNA binding

heterodimers are STAT 1:2 and STAT 1:3 (strong binding) and STAT 2:3 (weak

binding). Homodimers are STAT 1:1 and STAT 3:3 (strong) and STAT 2:2 (form

seldom in absence of STAT-1; Darnell, 1997). In variance with the notion that

tyrosine phosphorylation is required for STAT dimerization, Stancato and

coworkers (1996) demonstrated that STAT complexes exist in the cytosol of

unstimulated cells. Moreover, such association was independent of tyrosine

phosphorylation, since the Y701F STAT-1 mutant still bound to STAT-2 in

reticulocyte lysates. Such an interaction was weak, since it was not observed in *

extracts obtained with high-salt, detergent-containing buffers.

Current models for the mechanism of STAT activation of gene transcription

propose that following dimerization, STAT complexes translocate to the nucleus.

However, mechanism of transport to the nucleus remains unclear, since STAT

proteins lack the nuclear localization signal (NLS; Johnson et al., 1998b), which

are required for nuclear transport mediated through the importin mechanism

(Gorlich and Mattaj, 1996). Johnson and coauthors (1998b) proposed an

intriguing model for nuclear translocation of STATs after activation by interferon

gamma (IFNy). Since the carboxy-terminal domain of the IFNy molecule contain a

NLS, they propose that following binding to IFNy a complex containing the IFNy-

receptor, jak kinases, STATs and the bound ligand become internalized by

endocytosis. Upon cytoplasmic localization, the NLS sequence in the IFNy

molecule could associate with the importin protein complex, which would then

catalyze the transport of this complex to the nucleus, where STAT-mediated

transcription activation would ensue. They provided evidence for actual nuclear

translocation of a peptide containing the carboxy-domain of the IFNy molecule.

Although seemingly unique, they provide evidence of over 30 cytokines and/or

their receptors, which utilize STATs as signal transducers that contain NLS in

their sequence, indicating that this ligand-receptor-assisted nuclear translocation

is a viable, and intriguing mechanism. Among such cytokines and receptors are

the human IFNa and the human IFNaR1 molecules. Data in a recent paper is in

variance with this concept (Milloco et al., 1999). Those authors engineered a

STAT-1-estrogen receptor chimera, in which the estrogen receptor ligand binding

domain was fused to the carboxy-terminus of STAT-1 molecules. After

transfection to STAT-deficient U3a cells, this "conditionally active STAT"

underwent dimerization following estrogen/tamoxifen treatment. Moreover, these

chimeras were able to undergo nuclear translocation and activated transcription

of interferon-induced genes such as IRF-1. The authors concluded that tyrosine

phosphorylation of STAT is probably only a trigger for dimerization, since

dimerized, non-phosphorylated STAT chimeras also were able to stimulate

interferon-specific gene activation. Furthermore, since the estrogen receptor

domain used in the chimera did not contain any NLS, dimerization alone was

sufficient to promote nuclear translocation, sequence-specific DNA binding and

transcription activation functions of the chimeric STATs. A study conducted by

Strehlow and Schindler (1998) indicated that the amino-terminal 100 amino acids

of particular STATs mediated their nuclear translocation activity. Chimeric

constructs in which those amino acids in STAT-1 were substituted by those of

STAT-2 abolished nuclear translocation of STAT-1, while other functions were

maintained, such as activation by receptor, dimerization and DNA binding.

Collectively, it is fair to say that the mechanism of STAT nuclear translocation

remains unclear. Although the work of Johnson et al. (1998b) puts forth an

exciting proposition for such a mechanism, data from Milocco and others (1999)

argues against the requirement of a ligand-receptor-assisted transport

mechanism. However, existence of both mechanisms is feasible in vivo.

Interferon-directed gene activation

After translocation to the nuclear compartment, STAT complexes can act

as transcription factors, to direct expression of interferon-induced genes. The

best studied transcription activation complex containing STAT dimers is called

interferon-stimulated gene factor 3 (ISGF-3), which is composed of a STAT 1:2

dimer and a nuclear DNA binding protein, p48 (Darnell et al., 1994; Bluyssen et

al., 1996). ISGF-3 was first identified in electrophoretic mobility shift assays as a

complex induced by interferon treatment. It was formed independent of protein

synthesis, and was found to bind to consensus sequences on the regulatory

region of interferon-stimulated genes (Kessler et al., 1988). Consensus

sequences are known as interferon-stimulus response elements (ISREs).

Williams and Haque (1997) present a summary of sequences of ISREs of known

interferon-induced genes. A second interferon-induced transcription-activation

complex also was identified and named ISGF-2 (Kessler et al., 1988). Such a

complex is formed contingent on protein synthesis, presents different pattern of

migration in mobility shift assays and was later identified as the transcription

factor interferon regulatory factor 1 (IRF-1; Parrington et al., 1993). Interestingly,

IRF-1 and p48 are from the same family of proteins and can bind to the same

promoter elements (i.e., ISREs) in the regulatory region of interferon-stimulated

genes (Kessler et al., 1988; Parrington et al., 1993). The p48 and STAT 1:2

dimer do not associate in a stable manner to form the ISGF-3 complex in the

absence of DNA. However, contacts of amino acids 150 to 250 in the STAT-1

59

molecule with the carboxy-terminal portion of DNA-bound p48 stabilizes ISGF-3

(Horvath et al., 1996). Vickenmeier and coworkers (1996) reported direct binding

of recombinant, tyrosine phosphorylated STAT-1:1 dimers to tandem DNA

sequences. STAT-2 also forms homodimers, but requires p48 for strong

transactivation of transcription (Bluyssen and Levy, 1997). However, interactions

with DNA were not stable. Addition of STAT-1 increased the affinity and altered

sequence selectivity of p48-DNA interactions. In this scenario, ISGF-3 assembly

involves p48 functioning as an adaptor protein to recruit STAT-1 and STAT-2 to

an ISRE, STAT-2 contributes with potent transactivation but is unable to directly

contact DNA, while STAT-1 stabilizes the complex by contacting DNA directly.

Alternatively to transcription-induction through ISRE binding, interferons also

induce genes like IRF-1 which lack ISREs. Such genes are induced through

sequences named Inverted Repeats, present in their promoters (Haque and

Williams, 1994).

JAK-STAT pathway regulation

As in other tyrosine-phosphorylation-induced signaling systems, biological

responses resulting from activation of the JAK-STAT pathway are transient (Shuai

et al., 1992). Although the pathway of activation via the JAK-STAT pathway is

well established, few molecules have been identified that switch the signal off

(Starr and Hilton, 1999). Intuitively, one would predict that regulation of a tyrosine

phosphorylation pathway could occur through the actions of phosphatases, to

inactivate phosphotyrosil groups on receptors, JAKS and STATs, and proteases,

to degrade activated complexes. There is evidence for occurrence of both

mechanisms of regulation in the JAK-STAT pathway (i.e., phosphatases and

proteases), but more recent data indicate presence of novel regulatory molecules

also playing a role. Callus and Mathey-Prevot (1998) showed that treatment of

Ba/F3 cells with a specific proteasome inhibitor led to stable tyrosine

phosphorylation of the interleukin-3 (IL-3) receptor and STAT-5, after stimulation.

with IL-3. Further investigation revealed that stable phosphorylation events were

due to prolonged activation of JAKs. Moreover, Kim and Maniatis (1996)

demonstrated that after activation with interferon-y, STATs became ubiquitinated

and quickly degraded. In contrast with data from Kim and Maniatis (1996), but in

agreement with data from Callus and Mathey-Prevot (1998), Haspel and others

(1996) reported that proteasome inhibitors increased time of activation of STAT-1

by prolonging signals from the receptor (i.e., preventing degradation of receptor-

JAKs complexes), but not by blocking removal of phosphorylated STATs. This

was based on the finding that 35S-labeled STAT-1 translocated to the nucleus

upon tyrosine phosphorylation and later returned to the cytoplasm in non-

phosphorylated configuration. Data from Strehlow and Schindler (1998) agrees

and expands these findings, in that chimeric STATs with mutated amino-terminal

domains exhibited defects in nuclear translocation and deactivation, indicating

that these two events might be linked (i.e., deactivation may be dependent on

previous nuclear localization). Indeed, David and others (1993) demonstrated

that a nuclear tyrosine-phosphatase is responsible for deactivation of

61

phosphorylated STATs. To support the existence of a mechanism for regulating

activity of STATs based on phosphatases, Haque and coauthors (1995) reported

that treatment of cells with orthovanadate, molybdenate and tungstate, which are

effective inhibitors of protein-tyrosine phosphatases, resulted in accumulation of

interferon-y-induced phosphorylated STATs. Involvement of novel molecules in

the regulation of the JAK-STAT pathway was reviewed by Starr and Hilton (1999).

They propose a model in which suppressers of cytokine signaling (SOCS)

proteins such as SOCS1 bind directly to JAKs to inhibit their catalytic activities.

Another protein, CIS, binds to activated receptors to prevent docking of STATs.

SH2-domain phosphatase-1 (SHP-1) dephosphorylates JAKs or activated

receptors. Finally, a protein inhibitor of activated STAT (PIAS) inactivates STAT -

dimers. Song and Shuai (1998) demonstrated that SOCS 1 and SOCS3 inhibited

interferon-mediated antiviral and antiproliferative activities in HeLa cells. This

was linked with abolished tyrosine phosphorylation and nuclear translocation of

STAT-1 in response to interferon-a. Chung and others (1997a) reported that

PIAS3 directly interacted with STAT-3 and inhibited DNA binding of both STAT-

3:3, STAT-1:3 dimers. Binding of STAT-1 homodimers was not affected.

Moreover, cotransfections of both STAT-3 and PIAS3 showed a decrease in

luciferase activity from an IRF-1 reporter gene with increasing amounts of PIAS3.

Specificity of interferon signaling

Taken together, information presented in previous sections offers several

opportunities for occurrence of specific cellular responses to interferons. Such

opportunities include: (1) milieu of subtypes of interferons present at the receptor,

in which for example, different iso-forms of ovine IFN-T have different abilities to

extend estrous cycle length in ewes (Ealy et al., 1998); (2) composition of the

receptor complex, where recruitment of particular subunits may affect which

STATs are recruited; (3) amino acid context of the phosphotyrosine module on

the receptor chain, and amino acid context of the SH2 domain on STATs will also

determine which STAT will dock to which receptor chain; (4) which STATs are

present and which dimers will form upon ligand binding; (5) mechanism of nuclear

translocation of STATs, since whether STATs translocate as dimers alone or in

combination with ligand-receptor complexes may influence the configuration and

specificity of the transcriptional activation complex; (6) formation of single or

multiple transcription activation complexes, which will depend on nature of dimers

and interacting nuclear proteins; (7) dynamics of downregulation of JAK-STAT

pathway, in which specific branches of the pathway may be inhibited while others

may remain active to elicit specific responses; finally (8) interactions with other

cellular pathways, which will be discussed next.

Cross talk with other intracellular pathways

The best known cross-talk between JAK-STAT and other signaling

pathways is that represented by serine and threonine phosphorylation of STAT

63
residues, both constitutively and in response to ligands (see Leaman et al., 1996

for review). Such phosphorylation events are important, since treatment of cells

with kinase inhibitors disrupts STAT-3:3 DNA complexes. A mitogen-activated

protein kinase (MAPK) may be involved in phosphorylation of serine residues of

STAT-1, because the serine 727 lies in a consensus sequence for MAPK

phosphorylation. In fact, Stancato and coworkers (1997) proposed a model in

which activation of MAPK was dependent on activated JAK kinases. Binding of

interferon-a/p induced tyrosine phosphorylation of JAK-1, which stimulated

activity of membrane bound Raf-1. Activated Raf-1 phosphorylates MEK and

activates MAPK. MAPK in turn phosphorylates serine residues on STAT-1,

contributing to modulation of activity for this signal transducer. However,

modulation of STAT activities by MAPK may be stimulatory or inhibitory. For

example, Chung and others (1997b) reported serine phosphorylation of STAT-3

by growth factors, while STAT-1 was poor substrate for several MAPK tested.

Interestingly, serine phosphorylation of STAT-3 negatively modulated tyrosine

phosphorylation of this protein, and consequently inhibited dimerization, nuclear

translocation and gene activation.

Signal transducers such as IRS-1 and IRS-2 that are activated in response

to insulin, IL-2, IL-4 etc, are tyrosine phosphorylated by JAK-1. Epidermal growth

factor(EGF) is able to activate tyrosine phosphorylation of STATs 1 and 3 (David

et al., 1996). Interestingly, this does not require presence of JAKs. Moreover,

truncated receptor constructs containing the intrinsic kinase activity but lacking

the autophosphorylation domains were also effective in phosphorylating STATs.

This indicates that an alternative mechanism, where docking through SH2 domain

of STATs is not required for phosphorylation, is in place for EGF-induced STAT

phosphorylation.

The obligatory intracellular bacterium of macrophages, Ehrlichia

chaffeensis, blocked tyrosine phosphorylation of STAT-1, JAK-1 and JAK-2 in

response to IFN-y within 30 minutes of infection (Lee and Rikihisa, 1998). Also, .

PKA activity was increased 25 fold after infection. Inhibitors of PKA activity

partially abrogated the E chaffeensis-induced inhibition of STAT-1 tyrosine

phosphorylation, suggesting negative regulation of the JAK-STAT pathway by the

PKA-dependent mechanisms.

Another interesting theme is the occurrence of synergistic effects as a

result of coactivation of cellular pathways involving the JAK-STAT system. For

example, cooperation of interferon-y and tumor necrosis factor (TNF) during

inflammatory responses is a result of cooperation between STAT-1 and the

transcription factor NF-Kp. Synergistic expression of several genes involved in

the inflammatory process was contingent on presence of both transcription

factors (Ohmori et al., 1997). Stimulation by oncostatin M (OSM) induces

expression of matrix metalloproteinases (MMPs). Analysis of the regulatory

region of MMP-1 gene revealed presence of an AP-1 site as well as a STAT

binding element. Korzus and coworkers (1997) reported enhancement of MMP

expression due to synergistic actions of AP-1 and STAT-1. Such an effect was

Ras-dependent, which implies crosstalk between the MAPK and the JAK-STAT

pathways of signal transduction.

Yet another example of crosstalk is between the JAK-STAT pathway and

the PI 3' kinase, which has both lipid and serine kinase activities. Pfeffer and

others (1997) reported that PI 3' kinase is tyrosine phosphorylated through the

JAK-STAT pathway. Tyrosine phosphorylated STAT-3 proteins, bound to the

IFNaR1 chain of the interferon receptor, serve as a docking site for PI 3' kinase,

which couples its SH2 domain to tyrosine phosphorylated residues in the STAT

molecule. Upon docking, the PI 3' kinase is activated by JAKs, which then

promotes serine phosphorylation of STAT-3 to increase STAT-3 activity. In

another study (Uddin et al., 1997) interferon-a stimulated serine kinase activity of-

PI 3' kinase, which in turn activated the signal transducer IRS-1. Moreover,

stimulation with interferon p caused activation of MAPK, and such stimulation was

inhibited by Wortmannin, an inhibitor of PI 3' kinase activity. This suggests

involvement of the PI 3' kinase on MAPK activation. In contrast with data from

Pfeffer and others (1997), wortmannin failed to inhibit formation of the ISGF3

complex and interferon-mediated induction of ISG-15, indicating that the PI 3'

kinase probably is not required for interferon effects.

Finally, the work of Flati and others (1996) indicates that stimulation of

cells with interferon-a causes activation of PLA2, as measured by release of AA in

66

culture medium. PLA2 was associated with JAK-1, and inhibitors of PLA, activity

prevented formation of active ISGF3 transcription complexes. However, such

inhibition did not block binding of activated STAT-1 to inverted repeat sequences,

such as present in the regulatory region of IRF-1. Moreover, treatment of cells

with interferon-a stimulates tyrosine phosphorylation of PLA2. The authors argue

for a structural role of PLA2, which may be required for correct assembly of the

ISGF3 transcription complex.

The JAK-STAT pathway in bovine endometrium

In addition to the blFN-T receptor data mentioned above, very little has

been done on elucidation of the signal transduction system activated by blFN-T in

the endometrium. In fact, one of the main objectives of this dissertation was to

provide evidence of existence, as well as details on the blFN-r -activated,

endometrial JAK-STAT pathway.

Spencer and others (1998) conducted two studies to detect induction of

interferon-stimulated transcription factors, IRF-1 and IRF-2. Both factors were

absent in cyclic ewes and present in pregnant ewes (days 11 and 13, cyclic and

days 13, 15 and 17, pregnant). In cyclic ewes with ligated uterine homes,

unilateral infusion ovine IFN-T induced expression of IRF-1 and IRF-2 but not the

uterine horn receiving a BSA infusion. Since expression of these factors is

contingent on a functional JAK-STAT pathway, these data support existence of

such a pathway in the endometrium. Bathgate and coworkers (1998) also

reported existence of IRFs in endometrium of pregnant cows. Perry and

coauthors (1999) reported presence of STATs 1 and 2 and IRF-1 in the nucleus

of BEND cells stimulated with blFN-T. More detailed evidence for the JAK-STAT

pathway (i.e., existence, tyrosine phosphorylation, nuclear translocation, dimer

formation of and gene activation via STAT proteins) in bovine endometrium is

presented in Chapters 4, 5 and 6.

Bovine blFN-T -simulated protein synthesis in the endometrium

A main proposition of this dissertation is that proteins synthesized or

suppressed as a result of activation of the JAK-STAT pathway interact with the

PGF, synthesizing machinery to inhibit PGF2, secretion in the endometrium. In

this section, I will describe the data available on proteins synthesized in the

endometrium in response to blFN-T and their possible influences in the PGF2,

system.

Rueda and coworkers (1993) reported secretion of 12 and 28 kD proteins

both from pregnant endometrial explants and cyclic endometrial explants

stimulated with blFN-r in vitro. In a subsequent paper, Naivar and others (1995)

further characterized those proteins and discovered a novel, 16 kDa secretary

protein (P16). Endometrium explants were obtained from day 18 pregnant cows

and incubated in presence or absence of blFN-r. Both basal and stimulated

secretion of all three proteins increased in culture medium in a time-dependent

manner. More importantly, the 12 kD protein (Rueda et al., 1993), now renamed

as P8, was induced only in response to blFN-t, but not in response to IFN-a,

suggesting the possibility of blFN-r eliciting specific signal transduction and

protein synthesis. Moreover, P8 but not P16 secretion could be stimulated by

phorbol ester (Staggs et al., 1998). Amino acid analysis of the P8 revealed

identity with the alpha chemokine family: 92-100% identity with bovine bGCP-2

(Teixeira et al., 1997). Functions of bGCP-2 remain elusive, but it has been

suggested (Hansen et al., 1999) that being a chemokine, bGCP-2 may attract

concepts cells to attachment sites in the endometrium. Also, bGCP-2 may

attract cells from the immune system, to release cytokines beneficial to embryonic

development. P16 was identified as a bovine ubiquitin-cross reactive protein

(Austin et al., 1996a,b). The bUCRP mRNA (Hansen et al., 1997) and protein

(Austin et al., 1996b) are induced by blFN-r, and sequence analysis of the

bUCRP gene revealed presence of a conserved ISRE in the promoter region,

indicating putative activation by blFN-r (Perry et al., 1997). Analysis of the

primary structure of bUCRP revealed presence of critical amino acids and

domains implicated in functions of ubiquitin, such as conjugating with other

proteins. However, bUCRP lacked residues required for targeting proteins to

proteasomal degradation (Austin et al., 1996a). Therefore, it was proposed that a

possible role for bUCRP was to modify uterine proteins during early pregnancy

(Hansen et al., 1999). In fact, Johnson and others (1998a) reported that specific

conjugates of bUCRP and endometrial cytosolic proteins were formed in

69

response to treatment with blFN-r. Moreover, such complexes were distinct from

complexes containing ubiquitin, indicating a blFN-T-induced, specific action.

Although proteins present in the bUCRP conjugates have not yet been identified,

an attractive hypothesis is that blFN-, induces conjugation of bUCRP to proteins

involved in the cascade of PGF2,production in the endometrium. Such targeting

could modify function of such proteins to make them less able to stimulate PGF2,

production.

Research from Spencer and coworkers (1998) also showed that

endometrial estrogen receptors and oxytocin receptors were reduced in the

uterine horns infused with ovine IFN-r, and this was negatively correlated with

observed increase in IRF-1 and IRF-2 expression. Since IRF-2 has been

implicated as an inhibitor of gene transcription (Harada et al., 1994), the authors

hypothesized that perhaps interferon-induced IRFs were involved in inhibition of

gene transcription for estrogen and oxytocin receptors. In fact, Fleming and

coworkers (1998) cloned the ovine estrogen receptor gene and discovered IRF

response element (IRE) consensus sequences in the promoter region, further

supporting the hypothesis of interferon modulation of estrogen receptor

expression. Deletion constructs of the estrogen receptor promoter linked to

luciferase reporter gene were transfected into endometrial cells. Treatment of

these cells with ovine IFN-r caused reduction in luciferase expression only in

constructs containing the IREs. Using the same rationale, Bathgate and others

70

(1998) sequenced the bovine oxytocin receptor gene and also found IREs in the

regulatory region, and such sites bound bovine IRF-1 and -2. Again, the

suggestion is that perhaps blFN-T-induced transcription repressors may

downregulate expression of oxytocin receptors, to ultimately decrease PGF2,

secretion in the pregnant uterus.

Hypothetical model for blFN-T -mediated suppression of PGF,, secretion in the

endometrium

The hypothetical model shown in Figure 2-3 depicts some of the

possibilities discussed thus far.

Uterine-Conceptus Interactions and Reproductive Failure in Cattle

Thus far, this review has illustrated the enormous amount and intricacy of

interactions that need to occur between embryonic and maternal uterine tissues

in order for a successful pregnancy to be established. Given the high

percentage of embryonic mortality occurring during early pregnancies, it

becomes apparent that a precise program of interactions must be followed, and

that deviations from such a program may lead to pregnancy termination. Such a

program includes both embryonic and maternal components. For example, the

embryonic unit must be able to effectively interact with maternal endometrium,

undergo elongation and send antiluteolytic signals to the maternal unit in order to

survive. The maternal unit should provide a quiescent and nutritive environment,.

Uterine Lumen
INF-r

Figure 2-3. Hypothetical model of interferon-t (IFN-r)-stimulated gene activation and effects on
molecules involved in the PGF2a synthetic pathway. Estrogen receptors (ER) are up-regulated
before luteolysis and stimulate synthesis of oxytocin (OT) receptors (OTR). Oxytocin binding to
OTR stimulates phospholipase C (PLC), which cleaves phsphatydilinositol (PI) yielding inositol
trisphosphate (IP3) and diacylglycerol (DAG). The IP3 stimulates release of Ca" from
intracellular stores, and DAG activates protein kinase C (PKC). The PKC activates
phospholipase A2 (PLA,) which, in the presence of Ca+, cleaves arachidonic acid (AA) from
membrane phospholipids. Molecules of AA and linoleic acid (LA) regulate the enzyme
prostaglandin synthase (PGS) to produce PGF2a. In the pregnant cow, embryonic trophoblastic
cells secrete IFN-t into the uterine lumen. Receptors on endometrial epithelial cells bind IFN-T,
and dimerize. Dimerization of receptors promote phosphorylation (represented by a circled "p") of
associated tyrosine kinases from the JAK family, such as tyk-2 and jak-1. Phosphorylated
receptors attract signal transducer and activators of transcription (STAT) proteins. The STAT
proteins are phosphorylated in tyrosine residues by the JAK kinases and form a complex that
migrates to the nucleus. In the nucleus, the complex associates with a 48 kD DNA-binding
protein, and this new complex binds to interferon-stimulated response elements (ISRE) in the
regulatory region of interferon-induced genes, activating transcription of such genes and
synthesis of proteins. Synthesized proteins may act to specifically block one or more steps on the
PGF,2 synthetic pathway (arrows with [-] signs; see text for details and abbreviations).

72

conducive for embryonic attachment, and should bear intracellular mechanisms

to receive and transduce antiluteolytic signals from the concepts that ultimately

inhibit the default, PGF2,-secretory pathway of the uterus. Thatcher and Hansen.

(1992) reported that day 17 conceptuses varied in size from 15 to 250 mm.

Since inhibition of PGF2.is probably dependent on total amount of blFN-r

secreted and on area of endometrium occupied by the concepts, smaller

conceptuses would have already a smaller chance of survival. Environmental

effects such as heat stress (discussed above) decrease concepts development

and apparently compromises ability of the concepts to secrete blFN-r, leading

to failure in pregnancy recognition. There is also evidence for a role of the

uterus to stimulate secretion of blFN-r by concepts. Hernandez-Ledezma and

coworkers (1992) cultured IVF (in vitro fertilization)-produced embryos to

blastocyst stage and either continued in vitro culture or transferred conceptuses

to synchronized recipient cows. Embryos were recovered 4 days later, placed in

culture dishes and secretion of blFN-r was quantified. Secretion of blFN-e was

highly stimulated by exposure to the uterine environment, indicating that optimal

production of the antiluteolytic signal by the concepts is not solely determined

by the concepts. Stojkovic and coworkers (1999) reported that bovine embryos

derived by embryo flushing and in vitro production produced more blFN-r in long

term culture than embryos derived from nuclear transfer or embryo splitting Such

differences may contribute to lower pregnancy rates following embryo transfer to

recipients. Failure of cows to extend CL lifespan in response to blFN-T have

been reported (Helmer et al., 1989b; Meyer et al., 1995). This indicates failure in

the interferon receptor system, JAK-STAT-mediated signal transduction, post-

signaling mechanisms within the endometrium or a combination of these factors.

These responses have not been examined in a population of cows and warrant

further investigation.

Manipulating Uterine Function to Minimize Embryo Mortality

Bovine IFN-r administration

Based on the variation of concepts size and consequent capacity to

secrete at the critical time of maternal recognition of pregnancy for CL

maintenance, it is reasonable to propose that supplementing blFN-- at that

critical period may decrease embryonic losses. The rationale is that a slightly

underdeveloped concepts that may be unable to deliver the appropriate

antiluteolytic signal may be rescued by exogenous blFN-r administered at

around day 17 after insemination. Lack of availability of recombinant blFN-r and

structural similarity with blFN-a prompted Newton and others (1990) to test

fertility effects of blFN-a. Interferon-a extended CL lifespan but caused side

effects such as increased body temperature. Barros and coworkers (1992)

conducted a field experiment where blFN-a was administered daily from days 14

to 17 of pregnancy or as a single injection on day 13. Conception rates were

actually decreased by about 10% compared to control animals. This was

attributed to bIFN-a-induced side effects such as hyperthermia and acute drops

in P4 concentration. Alternative delivery systems and use of actual blFN-T may

still make this technology useful in the field (see Thatcher et al., 1994a for

discussion).

Fat feeding

Another possible manipulation of this system consists of attempting to

make the uterus less luteolytic, by changing the proportion of luteolytic AA to

antiluteolytic linoleic acid. For example, Thatcher and others (1997) described

an experiment in which Menhaden fish meal was fed to cows for 25 days. Fish

meal contains both eicosapentanoic and docosahexaenoic fatty acids, which had

been shown to be able to decrease PGF2 secretion. Indeed, cows fed fish oil

had a much attenuated secretion of PGFM in response to an oxytocin challenge.

Prevention of heat stress

One single environmental challenge that has negative effects both in the

embryonic and maternal units during the period of maternal recognition of

pregnancy is heat stress, as discussed previously. Therefore, strategies to

reduce effects of high environmental temperatures, such as observed in tropical

and subtropical latitudes, warrant investigation and application.

75

Objectives of This Dissertation

1) To study the distribution pattern of oviductal secretary proteins secreted

by cows bearing persistent or fresh dominant follicle;

2) To examine the signal transduction system stimulated by blFN-r in

endometrium;

3) To characterize the effects of blFN-T on PGF, production by BEND

cells.

CHAPTER 3
PERSISTENT DOMINANT FOLLICLE ALTERS PATTERN OF OVIDUCTAL
SECRETARY PROTEINS FROM COWS AT ESTRUS

Introduction

Synchronization of the estrous cycle in cattle is a very important tool for

reproductive management. For example, synchronization systems are used

widely for artificial insemination, timed insemination and embryo transfer. Most

commonly, synchronization is achieved with combinations of treatments with

prostaglandin F2a (PGF2,), progestins and GnRH (Thatcher et al., 1996).

Synchronization with progestins is based on the principle that exogenous

progestins, such as progesterone delivered by a Controlled Internal Drug

Release (CIDR) device, can maintain a sub-luteal concentration of progestin in

blood during a period which permits CL regression. In the absence of a CL,

removal of the progestin source will result in a synchronized estrus (Macmillan

and Peterson, 1993).

During the estrous cycle in cattle, two to three follicular waves of dominant

follicle development occur (Savio et al., 1988; Sirois and Fortune, 1988). Each

follicular wave is comprised of periods of recruitment, selection, dominance and

turnover or atresia. The ovulatory follicle generated in the last wave does not

77

turn over, but ovulates in a low progesterone (P4) environment. Turnover of the

dominant follicle (DF) is associated with high concentrations of P4, typical of mid-

cycle, which lowers LH pulse frequency (Kinder et al., 1996). Turnover of the

first wave DF can be blocked by exogenous progestins and injection of PGF2a

(Cooperative Regional Research Project, NE-161, 1996; Savio et al., 1993a;

Savio et al., 1993b). The resulting sub-luteal concentration of progestin in

plasma permits an increase in LH pulse frequency which sustains growth of the

DF. This "persistent" DF (PDF) is estrogenic, and subsequent fertility, as

measured by conception rate at first service (number of pregnancies / number of .

animals inseminated), is lower compared to animals bearing normal DFs [37.1%

vs. 64.8% in heifers, (Savio et al., 1993b); 23.6% vs. 58.2% for cows and heifers,

(Cooperative Regional Research Project, NE-161, 1996). Fertility after Al,

however, is restored to levels comparable to controls if the PDF is turned over

and a freshly recruited follicle is allowed to ovulate. Possible explanations for

reduced fertility include alterations in the oocyte and /or in the oviductal

environment. In a study by Ahmad et al. (1995), cows ovulating a PDF had

embryos that at day 6 of pregnancy were less developed (i.e., were less able to

reach the 16-cell stage) than embryos from cows ovulating a fresh (F) DF. In

addition, Revah and Butler (1996) showed that oocytes recovered from the PDF

showed expanded cumulus cells and condensed chromatin dispersed in their

ooplasm. In contrast, compact cumulus cells and intact germinal vesicles were

found in oocytes from FDF. Thus, the PDF may affect oocyte maturation,

oviduct and uterine function, which could affect early embryonic development

and decrease fertility.

Macromolecules present in oviductal fluid have been suggested to serve

an important role in sperm capacitation (Anderson and Killian, 1994), fertilization

(Boatman and Magnoni, 1995) and early embryo development (Gandolfi et al.,

1989). Therefore, alterations in oviductal biosynthetic activity including protein

synthesis and secretion may affect conception rate.

Steroid modulation of oviductal synthesis and secretion of proteins has

been characterized in sheep (Buhi et al., 1991, Murray, 1993), baboon (Verhage

and Fazleabas, 1988) and swine (Buhi et al., 1989; Buhi et al., 1990). An altered

steroid environment, associated with development of a PDF, may alter oviductal

protein synthesis and secretion. In turn, the altered pattern of protein synthesis

and secretion could affect optimal oviductal function, fertilization and early

embryo development that contributed to reduced embryonic survival in

synchronized cows. The present experiment tested the hypothesis that the

presence of a PDF alters protein synthesis and secretion of oviductal explants

from cows at estrus.

Specific objectives were: 1) to induce a PDF or a FDF with the strategic

use of PGF2a, progesterone-containing CIDR and GnRH; 2) to compare the

biosynthetic activity and the array of secretary proteins synthesized in the

infundibulum (INF), ampulla (AMP) and isthmus (IST) at estrus in oviducts

ipsilateral (IPSI) and contralateral (CONTRA) to the DF of cows bearing a PDF

versus a FDF.

Materials and Methods

Materials

Impervo paint was from Benjamin Moore and Co. (Jacksonville, FL) and

All-weather Paintstick was from LA-CO Industries, Inc./Markal Company

(Chicago, IL). Donations of Lutalyse were made by Pharmacia-Upjohn Co.

(Kalamazoo, MI), Buserelin from Hoescht-Roussel Agri-Vet (Somerville, NJ) and

CIDR-B devices were donated by EAZI-BREEDT, InterAg (Hamilton, New

Zealand). Eagles' minimum essential medium (MEM, catalog number 86-5007),

non-essential aminoacids (100x), anti-mycotic/antibiotic solution (100x) and

MEM vitamin solution (100x) were from Life Technologies (Gibco Laboratories,

Grand Island, NY). L-[4,5-3H] leucine (leu; 159 Ci/nmol) was from Amersham

Life Sciences, Inc. (Arlington Heights, IL) and L-leu, L-methionine, L-glutamine,

D(+) glucose, bovine pancreatic insulin, riboflavin and molecular weight

standards were purchased from Sigma Chemical Co. (St. Louis, MO).

Spectra/por 3 dialysis membrane was from Spectrum Medical Industries Inc.

(Houston, TX). Acrylamide, NN'-methylenebisacrylamide, sodium dodecyl

sulphate, Nonidet-P40, urea, agarose, diallyltartardiamide were from BDH

Laboratory Supplies (Poole, England). Ampholines were from Pharmacia

(Uppsala, Sweden), TEMED and ammonium persulphate were from Bio-Rad

(Hercules, CA). Glycine was from ICN Pharmaceuticals, Inc. (Costa Mesa, CA).

Coomassie brilliant blue, fast green, bromophenol blue, P-mercaptoethanol,

hydrochloric acid, sodium hydroxide, tris hydroxymethyll) aminomethane, sodium

salicylate, acetic acid and chromatography paper were from Fisher Scientific

(Fairlawn, NJ) and X-OMAT x-ray film was from Eastman Kodak Co. (Rochester,

NY).

Preparation of Medium

Leucine-deficient modified minimum essential medium (MEM; 10% normal

concentration of leu) was prepared as described by Buhi and coworkers (Buhi et

al., 1990). Briefly, MEM was supplemented with glucose (3g/1), methionine (1.5

mg/I), leu (5.2 mg/I), sodium bicarbonate (2.2 g/l), MEM vitamins (10 ml/l), non-

essential amino acids (10 ml/I), insulin (7.41 mg/I), sterile filtered and adjusted to

pH 7.4. Before use, medium was supplemented further with glutamine (292

mg/l), methionine (13.5 mg/I) and antimycotic-antibiotic solution (10 mill).

Animals and Treatments

During the pre-treatment period, estrous cycles of six mature non-lactating

cows were synchronized (Figure 3-1). A used CIDR device containing

approximately 1.2 g (Van Cleef et al., 1992) of P4 was placed into the vagina of

81

each cow for 7 days. One day prior to CIDR removal, cows received an injection

Pre-Treatment Period

PGF2a

CIDR

0 6 7 0
Estrus

Treatment Period

GnRH
(+1-)
PGF2a PGF2a

CIDR

7 9 16 18
Estrus
Slaughter

Ultrasonography, Blood Collection

Figure 3-1. Experimental protocol (see text).

of prostaglandin-F2o (PGF2a, Lutalyse, 25 mg) to regress the CL. To aid with

estrus detection, tail heads were painted (Impervo) and chalked (All-weather

Paintstick). Cows were observed twice daily for signs of estrus, and paint scores

were assigned (Macmillan et al., 1988). The day of standing estrus was

designated experimental day 0. During the treatment period, ovaries were

examined by transrectal ultrasonography using an Aloka echo camera model

SSD 500 linear array ultrasound scanner equipped with a 7.5MHz transducer

(Aloka Co., Japan). From Days 5 to 18, follicles and CL were measured daily

and sizes recorded. In addition, blood samples were collected in heparinized

evacuated tubes (Vacutainers, Becton Dickson Vacutainer System USA,

Rutherford, NJ) by tail venipuncture and stored in an ice bath. Plasma was

harvested by centrifugation (1800 x g for 30 minutes) and stored at -200C until

assayed for estradiol-17p (E2) and P4. On day 7, all cows were injected with

PGF2a and received one used CIDR device (Savio et al., 1993b). On day 9 cows

were assigned randomly to one of two treatment groups. Cows of the FDF group

(n=3) received an injection of GnRH agonist (Buserelin, 8 mg), to induce turnover

of any large size follicles present at that time, and allow recruitment of fresh

follicles (Schmitt et al., 1996c). Cows of the PDF group (n=3) did not receive the

GnRH agonist. On day 16 CIDR devices were removed, and cows received an

injection of PGF2, (25 mg). Cows were checked for signs of estrus twice daily

and slaughtered when observed in standing estrus (day 18 or 19). The

experimental models for persistent and fresh follicles resulted in a greater

pregnancy rate for heifers inseminated at estrus induced by FDF (Savio et al.,

1993b; Schmitt et al., 1996c).

Tissue Culture

On the day of slaughter, reproductive tracts were removed aseptically,

and oviducts were identified as IPSI or CONTRA to DF, dissected, trimmed free

of mesosalpynx and divided into INF, AMP and IST regions based on gross

anatomical characteristics. Segments of tissue between IST and AMP were

discarded. Tissue from each region was cut longitudinally to expose the lumen,

and then minced into fragments of -50mm3. Tissue fragments from each

functional region were cultured (Buhi et al., 1990) in LEU-deficient minimal

essential medium supplemented with 3H-LEU in the ratio of 100 mg tissue/3 mL

medium/20 mCi 3H-LEU for 24 hours at 370C in a controlled atmosphere of

N2:02:CO, (50%:47.5%:2.5% by volume). For AMP and INF, 500 mg of tissue

were cultured per dish, while for IST variable amounts of tissue (between 140

and 290 mg) were used.

Two-Dimensional Electrophoresis

After 24 hours incubation, conditioned media were dialyzed extensively

(MW cut-off 3500) against Tris buffered saline (10 mM Tris, 150 mM NaCI) pH

7.6 (two changes of 4 liters each/24 hours) and then dialyzed against deionized

water (two changes of four liters each/24 hours). Radioactivity in the retentate

was determined by liquid scintillation spectrometry, and incorporation rate was

defined as dpm non-dialyzable macromolecules/mg wet tissue. For each

sample, a volume of dialyzed conditioned medium containing 4 x 105 DPM was

lyophilized and submitted for two-dimensional SDS-PAGE as previously

described(Buhi et al., 1991). Gels were stained with Coomassie blue, soaked in

1 M Na salicylate solution, dried and exposed to x-ray film for 35 days at -800C.

Densitometry

Fluorographs were developed, and after qualitative analysis 20 protein

spots were selected and analyzed quantitatively by densitometry (Alphalmager

84

2000, Alpha Innotech Corporation, San Leandro, CA). Since a constant amount

of DPM was loaded for all samples, the capacity of tissues to synthesize and

secrete macromolecules (DPM/mg of tissue) was not accounted for and,

therefore, unadjusted densitometric measurements were biased. Different

secretary capacities were corrected by expressing the densitometric

measurements per unit secretary tissue. In this way, densitometric

measurements from tissues with greater secretary capacity were adjusted

upwards and vice versa for tissues with lower secretary capacities. Adjustments

were calculated by the equation: adjusted Arbitrary Density Units (ADU) =

ADU/mass of tissue equivalents, where one tissue equivalent is the mass of

tissue needed to synthesize and secrete 4 x 105 DPM of labeled

macromolecules. Mass of tissue equivalents was obtained by dividing 4 x 105

DPM by incorporation rate (DPM non-dialyzable macromolecules/mg of tissue)

for individual tissue samples.

Hormone Assays

Concentrations of E2 and P4 in plasma were measured by

radioimmunoassays previously validated in our laboratory [E2:(Badinga et al.,

1992); P4:(Knickerbocker, 1986)]. Intra- and inter-assay coefficients of variation

were 15.5 and 12.4%, respectively, for E2and, 6.8 and 8.1%, respectively, for P4

Statistical Analysis

Data were analyzed by least squares analysis of variance using the

General Linear Models of Statistical Analysis System (SAS, 1988).

Concentrations of E2 and P4 in plasma and diameter of DF were analyzed by split

plot ANOVA. The mathematical model used treatment (FDF or PDF), cow

(treatment), day, treatment by day and error. Rate of incorporation of

radioactivity into oviductal tissue and natural Log of adjusted ADU measurement

of proteins were calculated and analyzed by least squares ANOVA. The

mathematical model was: treatment (FDF or PDF), cow (treatment), side (IPSI or

CONTRA to the DF), region (INF, AMP and IST), all higher order interactions

and error. Orthogonal contrasts for treatment (PDF vs. FDF), region (INF and

AMP vs. IST and INF vs. AMP), and treatment by region interactions were used

to compare means.

Results

Ultrasonographv and Hormone Measurements

Size of DF was analyzed in two phases during the treatment period: from

day 5 to day 9 (period prior to injection of GnRH) and from day 10 until day 16

(Figure 3-2). Both FDF and PDF cows had similar sizes of DF from day 5 to day *

9. However, a significant (p<0.01) treatment by experimental day interaction was

20-

010-
U-
GnRH
5. PGF2a (+/-) PGF2a

r CIDR
5 6 7 8 9 10 11 12 13 14 15 16
Day of Treatment Period

Figure 3-2. Least squares means ( SEM) of diameter of the dominant follicle (DF) of cows
bearing a fresh DF (FDF, treated with GnRH on d 9) or persistent DF (PDF, not treated with
GnRH on d 9) during the Treatment Period. Treatments with PGF,, CIDR and GnRH are
indicated. Day 0 represents day of estrus at the beginning of Treatment Period.

detected from day 10 to day 16. All cows with FDF ovulated the first wave DF

and a newly recruited DF was detected on day 11 which reached 12 mm by day

16. In contrast, the first wave DF of PDF group was sustained and reached a

size of 22 mm by day 16.

Concentrations of E2 (Figure 3-3, panel a) and P4 (Figure 3-3, panel b)

were analyzed between experimental day 7 (day of PGF2a injections) and day 18

or 19. There was a significant (p<0.01) treatment by experimental day

interaction for both E2 and P4 concentrations in plasma. After GnRH injection on

day 9, E2 concentrations decreased in plasma of FDF cows and remained

between 5 and 10 pg/ml until day 16 and increased to 22 pg/ml at day 18

-I- FDF -- PDF

I CIDR I
5 6 7 8 9 10 11 12 13 14 15 16 17 18
Day of Treatment Period
b T

6019 CIDR Y5 I
5 6 7 8 9 10 11 12 13 14 15 16 17 18 19
Day of Treatment Period
Figure 3-3. Least squares means (t SEM) of concentrations of ovarian steroids in plasma
of cows bearing a fresh dominant follicle (FDF, treated with GnRH on d 9) or persistent
dominant follicle (PDF, not treated with GnRH on d 9) during the Treatment Period.
Treatments with PGF2,, CIDR and GnRH are indicated. Day 0 represents day of estrus at the
beginning of Treatment Period.
a) estradiol-17p (E2); b) progesterone (P4).

	Title Page
	Dedication
	Acknowledgement
	Table of Contents
	Abstract
	Introduction
	Literature review
	Persistent dominant follicle alters...
	Effects of bovine interferon-tau...
	Bovine interferon-tau stimulates...
	Bovine interferon-tau stimulates...
	Interferon-tau modulates phorbol...
	General discussion
	List of references
	Biographical sketch

MILLISECOND	CLASS.METHOD	MESSAGE
0	sobekcm_page_globals.constructor
0	sobekcm_page_globals.constructor	Application State validated or built
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.constructor	Navigation Object created from URI query string
0	sobekcm_database.verify_item_lookup_object
0	sobekcm_page_globals.display_item	Retrieving item or group information
0	sobekcm_page_globals.get_entire_collection_hierarchy	Retrieving hierarchy information
0	sobekcm_assistant.get_entire_collection_hierarchy
0	cached_data_manager.retrieve_item_aggregation
0	cached_data_manager.retrieve_item_aggregation	Found item aggregation on local cache
0	item_aggregation_builder.get_item_aggregation	Found 'all' item aggregation in cache
0	system.web.ui.page.page_load (ufdc.page_load)
0	sobekcm_page_globals.constructor.on_page_load
0	html_echo_mainwriter.add_style_references	Adding style references to HTML
0	html_echo_mainwriter.add_text_to_page	Reading the text from the file and echoing back to the output stream
34	html_echo_mainwriter.add_text_to_page	Finished reading and writing the file