Citation
Cyclic nature of bovine uterine luminal proteins and their relationship to peripheral plasma progesterone and estrogen levels /

Material Information

Title:
Cyclic nature of bovine uterine luminal proteins and their relationship to peripheral plasma progesterone and estrogen levels /
Creator:
Mills, Albert Carter, 1943-
Publication Date:
Copyright Date:
1975
Language:
English
Physical Description:
ix, 108 leaves : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Blastocyst ( jstor )
Cattle ( jstor )
Embryos ( jstor )
Estrus cycle ( jstor )
Gels ( jstor )
Plasmas ( jstor )
Rabbits ( jstor )
Rats ( jstor )
Secretion ( jstor )
Uterus ( jstor )
Animal Science thesis Ph. D
Cattle -- Physiology ( lcsh )
Dissertations, Academic -- Animal Science -- UF
City of Gainesville ( local )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis--University of Florida.
Bibliography:
Bibliography: leaves 91-106.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Albert Carter Mills, III.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
029132895 ( AlephBibNum )
02274300 ( OCLC )
ABZ3809 ( NOTIS )

Downloads

This item has the following downloads:


Full Text















CYCLIC NATURE OF EOV1NE ULERINE LUMINAL PROTEINS
AND THEIR RELATIONSHIP TO
PERIPHERAL PLASMA PROGESTERONE AND ESTROGEN LEVELS







By



ALBERT CARTER MILLS, III


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL
OF TIHE Ul;IVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE IJQUIRE;MENTS FOR THE
DEGREE OF 1DOCIOR OF PHILOSOPIIY



UNIVERSITY OF FLORIDA


1975













ACKNOWLEDGMENTS


The author wishes to express sincere gratitude to the

members of his supervisory committee: Dr. A. C. Warnick,

chairman; Dr. Fuller W. Bazer; Dr. D. E. Franke; Dr. D. H.

Barron and Dr. P. T. Cardeilhac. Sincere appreciation is

extended to Dr. A. C. Warnick for his constant support, guid-

ance and assistance during the author's graduate program.

Special thanks are expressed to Dr. Fuller W. Bazer for

his continual willingness to be of assistance in the planning

and conduct of this research.

Deep appreciation is expressed to Dr. W. W. Thatcher for

his constant help and encouragement with the hormonal and

statistical analyses conducted during the course of this

research. Thanks are also due Dr. C. J. Wilcox, Dr. P. S.

Kalra, Dr. F. C. Gwazdauskas and Susan Acree for their help

with these analyses.

The author wishes to thank his fellow graduate students,

Dr. James V. Knight, Thomas T. Chen and Eddy Muljono for

their willing assistance in the collection and processing of

samples for this research. The valuable technical assistance

of Mary Bates Smith during the many hours of electrophoresis

and protein determination is also appreciated.








The handling of slaughter of experimental animals by

Dr. A. Z. Palmer and Jerry S. Scott is greatly appreciated.

Thanks are due Mr. Dean E. Pogue for care and feeding of the

animals.

The author wishes to thank his wife, "Dotty," for her

constant encouragement and help during the writing of this

dissertation.














TABLE OF CONTENTS


ACKNOWLEDGMENTS . . . . . . .

LIST OF TABLES . . . . . . .

LIST OF FIGURES . . . . . . .

ABSTRACT . . . . . . . .

INTRODUCTION . . . . . . .

REVIEW OF LITERATURE . . . . .

CHAPTER

1 NONSURGICAL COLLECTION AND STUDY 0
BOVINE UTERINE LUMINAL PROTEINS .

Materials and Methods .. ..
Results and Discussion ..

2 CHANGES IN BOVINE UTERINE LUMINAL
PROTEINS DURING THE ESTROUS CYCLE
AND THEIR RELATIONSHIP TO PLASMA
PROGESTERONE AND ESTRADIOL LEVELS

Materials and M\ethods .. ..
Results and Discussion .

GENERAL DISCUSSION . . . . .

SUMMARY . . . . . . . . .

LIST OF REFERENCES . . . . .

BIOGRAPHICAL SKETCH . . . . . .


Page

* ii

v

vi

vii

1

4













LIST OF TABLES


Table Page

1 Average total protein recovered from the
bovine uterus during the estrous cycle ... .40

2 Average total uterine protein recovered
and plasma progesterone and estradiol
concentration from cattle on each day of
the estrous cycle . . . . . . .. 46

3 Least-squares analysis of variance: data
obtained between days 0 and 20 of the
estrous cycle . . . . . . . ... 52

4 Least-squares analysis of variance: data
obtained between days 4 and 18 of the
estrous cycle . . . . . . . . 53

5 Least-squares analysis of variance: data
obtained between days 0 and 3 of the
estrous cycle . . . ... . . . . . 54

6 Correlation of plasma progesterone and
estradiol concentrations with total uterine
protein and presence of Rf 0.35 uterine pro-
tein band following electrophoresis . . . 58

7 Number and percent of polyacrylamide gels
of uterine protein containing electro-
phoretic bands not present in corresponding
plasma gels . . . . . . . . . 65














LIST OF FIGURES


Figure Page

1 Averages for total uterine protein
recovered on days 0 to 20 of the
estrous cycle . . . . . . . ... 56

2 Averages for plasma progesterone
concentration on days 0 to 20 of
the estrous cycle . . . . . . ... 60

3 Averages for plasma estradiol concen-
tration on days 0 to 20 of the estrous
cycle . . . . . . . . ... .. . 63

4 Polyacrylamide gel electrophoresis
toward the anode of bovine uterine
protein and plasma collected on days
0 to 20 of the estrous cycle . . . ... 67

5 Typical Sephadex G-200 gel filtration
protein profiles of bovine uterine
protein collected on days 0, 5 and 15
of the estrous cycle and bovine plasma . .. 72








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


CYCLIC NATURE OF BOVINE UTERINE LUMINAL PROTEINS
AND THEIR RELATIONSHIP TO
PERIPHERAL PLASMA PROGESTERONE AND ESTROGEN LEVELS


by

Albert Carter Mills, III

June, 1975

Chairman: A. C. Warnick
Major Department: Animal Science


Bovine uterine protein secretions were examined during

the estrous cycle for changes in total recoverable protein,

Scphadex G-200 gel filtration protein profiles and electro-

phoretic protein patterns. Qualitative and quantitative

changes were correlated with plasma progesterone and estra-

diol concentrations.

Preslaughter plasma and immediate postslaughter 0.33 MI

saline uterine flushings were collected from 67 cattle of

mixed breeding (Angus, Hereford, Holstein and Brahman X

British crosses). Three samples were collected for each day

of the estrous cycle and 4 samples each were obtained on days

3, 5, 13 and 15. Plasma samples were assayed for progester-

one (competitive protein binding procedure) and estradiol

(radioimmunoassay). Total uterine protein was determined by

Lowry's method. Qualitative polyacrylamide gel electropho-

resis and Sephadex G-200 gel filtration protein profiles were

obtained for each sample.








There was a significant (P<.025) difference in total

uterine protein recovered on days 0 to 20 of the estrous

cycle. From days 4 to 18,cows yielded significantly (P<.05)

greater amounts of total uterine protein than heifers. Plasma

progesterone concentration was significantly (P<.05) corre-

lated with total uterine protein (r=.23) from days 0 to 20

after correction for breed, parity and day effects by least-

squares analysis. This correlation coefficient was .44

(P<.05) when data collected during the luteal phase (days 4

to 18) of the estrous cycle were considered separately. The

correlation coefficient between estradiol and total uterine

protein was .26 (P<.05) when all data were considered (days 0

to 20) and .73 (P<.01) when data collected during estrus and

metestrus (days 0 to 3) were considered separately. However,

after correction for breed, parity and day effects by least-

squares analysis, the correlations between estradiol and

total uterine protein were not significant (P>.10).

A protein band,not present in plasma samples,was found

in 16 '(62%) of the polyacrylamide gels following electropho-

resis of uterine samples obtained between days 13 and 20 of

the estrous cycle. There was a highly significant (P<.01)

difference in the frequency with which this protein band was

present from days 0 to 20. This band (Rf=0.35) was present

in all samples on days 15 and 16. Based on its Sephadex

G-200 elution volume elutedd just prior to albumin), molecular

weight of the protein represented by this band was estimated

to be approximately 100,000. A highly significant (P<.01)


viii








correlation (r=.46) was found between plasma concentration of

progesterone and the presence of this protein. A second pro-

tein band (Rf=0.72) which migrated just ahead of the trans-

ferrins appeared in all uterine protein samples analyzed from

the entire estrous cycle, but it was absent in plasma. The

electrophoretic mobility of this protein was similar to that

of the rabbit uterine specific protein "blastokinin." One or

two prealbumin protein bands (Rf=1.09 and 1.12) were present

in 39 (58%) of the uterine protein gels between days 0 and 20

but not in plasma gels.

Sephadex G-200 gel filtration uterine protein profiles

did not reveal any significant differences in the size classes

of proteins present between days 0 and 20 of the estrous cycle.

However, there was some slight trailing in uterine samples of

two low molecular weight protein fractions which were never

present in plasma samples. These two"low-profile"fractions

were in the 10,000 to 15,000 and <10,000 molecular weight

range.

This study indicated that both the qualitative and quan-

titative aspects of bovine uterine protein secretion changed

during the luteal phase of the estrous cycle. Correlation

of these changes with plasma progesterone concentration sug-

gested that these changes were associated with plasma proges-

terone concentration.













INTRODUCTION


In recent years there have been several studies on the

protein composition of uterine fluids in mice (Homburger

et al., 1963; Mintz, 1970), hamsters (Noske and Daniel, 1974),

rats (Junge and Blandau, 1958; Howard and DeFeo, 1959; Albers

and Castro, 196]; Ringler, 1961; Heap and Lamming, 1962;

Kunitake et al., 1965), rabbits (Stevens, Hafs and Hunter,

1964; Krishnan and Daniel, 1967; Beier, 1968; Daniel, 1971a;

Murray and Daniel, 1973; Beier, 1974a), swine (Murray, 1971;

Murray et al., 1972a; Squire, Bazer and Murray, 1972), sheep

(Heap and Lamming, 1960,1963; Heap, 1962; Perkins et al.,

1965; Iritani, Gomes and VanDemark, 1969; F. W. Bazer, unpub-

lished data) and cattle (Fahning, Schultz and Graham, 1967;

Schultz, Fahning and Graham, 1971; Roberts and Parker, 1974a,b).

These reports support the concept that uterine fluid proteins

are a product of active secretion and not simple diffusion

from blood. Synchronization of uterine secretions and

stage of embryonic development has been shown to be criti-

cal to normal growth and implantation of the embryo (Dick-

mann and Noyes, 1960; Rowson and Moor, 1966; Rowson, Moor

and Lawson, 1969; Adams, 1969,1971,1973; Rowson et al.,

1972a; Beier, Mootz and Kihnel, 1972). It has been estab-

lished that the embryo must reach the uterine environment








at the proper time for continued and normal development to

occur beyond the early blastocyst stage. This suggests that

uterine protein secretions play an active role in early em-

bryonic growth and development.

Work with mice (Mintz, 1970; Pinsker and Mintz, 1973)

has indicated that estrogen controls the secretion of a uter-

ine factor which affects implantation. In the rabbit (Krish-

nan and Daniel, 1967; Beier, 1968; Urzua et al., 1970; Arthur

and Daniel, 1972; Johnson, 1972; Whitson and Murray, 1974;

Goswami and Feigelson, 1974) and pig (Murray, 1971; Knight,

Bazer and Wallace, 1973a,b, 1974b; Knight et al., 1974c;

Chen, 1973; Chen et al., 1973a; Schlosnagle et al., 1974)

recent work has established that uterine protein secretions

are under the regulation of progesterone and that these pro-

teins are associated with early development of the embryo and

trophoblast. Several workers have also suggested that the

composition of bovine uterine fluid is under hormonal control

(Olds and VanDemark, 1957; Heap and Lamming, 1960; Heap, 1962;

Fahnin'g et al., 1967; Schultz et al., 1971).

The working hypothesis which served as the basis for

this study was that the uterus of the cow secretes specific

proteins in response to changing plasma progesterone and/or

estradiol concentrations, which are required by the embryo for

complete development. Therefore, this study was designed to

examine bovine uterine protein secretions during the estrous

cycle for changes in total recoverable protein, Sephadex





3


G-200 gel filtration protein profiles and electrophoretic

protein patterns, and to correlate these changes with plasma

progesterone and estradiol levels.














REVIEW OF LITERATURE


Recent research has made it clear that uterine factors

play a very important part in mammalian reproduction. The

evidence indicates that these factors influence embryonic

development in very specific ways. Some components) of the

uterus appears to be important for embryonic development to

the early blastocyst stage and absolutely essential for con-

tinued normal development past the early blastocyst stage.

Evidence supporting the presence and importance of uterine

protein secretions and their hormonal regulation will be con-

sidered herein.

Embryos restricted to the oviductal environment develop

to the early blastocyst stage in mice (Kirby, 1962; Orsini

and McLaren, 1967; Whittingham, 1968), rats (Alden, 1942),

rabbits (Pincus and Kirsch, 1936; Adams, 1958) and swine

(Murray et al., 1971; Pope and Day, 1972). However, con-

tinued development beyond the early blastocyst stage in the

oviduct is retarded (Pope and Day, 1972) or the embryos

degenerate (Adams, 1958; Kirby, 1962; Orsini and McLaren,

1967; Murray et al., 1971). Kirby (1962) suggested the

requirement of a "uterine factor" for development of mouse

embryos beyond the blastocyst stage. He observed that

mouse embryos recovered from the uterus continued to









develop when transplanted beneath the kidney capsule as

opposed to only trophoblast development when tube-locked

embryos were treated similarly. Thus, the conclusion was

reached that the uterus must provide some factors) which is

absolutely essential for continued normal development to occur

past the early blastocyst stage of embryonic development.

Early research with in vitro culture of embryos demon-

strated the importance of the uterine environment and its

secretions. Lewis and Gregory (1929) found that rabbit ova

developed no further than the blastocyst stage when cultured

in rabbit blood plasma. It did not appear from earlier stud-

ies that simple nutrients constituted the uterine factors

which enabled uterine embryos to develop continuously. Fail-

ure of development of the embryo past the early blastocyst

stage in vitro was the general rule regardless of the culture

media or the species studied (Chang, 1949; Hammond, 1949;

Whitten, 1956,1957; McLaren and Biggers, 1958; Tarkowski, 1961;

Brinster, 1963; Adams, 1965; Cole and Paul, 1965; Maurer,

Whitener and Foote, 1969; Rundell, 1969). According to

Krishnan and Daniel (1967) and Beier (1968), a specific uter-

ine protein secretion (termed "blastokinin" and "uteroglobin"

by the respective authors) is involved in blastocyst formation

by rabbit morulae during culture in vitro in a chemically

defined medium.

More recent data have provided evidence that successful

in vitro embryo culture can be achieved in chemically defined

culture media up to the expanding blastocyst stage in mice








(Whitten and Biggers, 1968; Whitten, 1970), rats (Folstad,

Bennett and Dorfman, 1969), rabbits (Onuma, Maurer and Foote,

1968; Maurer, Onuma and Foote, 1970; Kane and Foote, 1970a,b,c)

and sheep and cattle (Tervit, Whittingham and Rowson, 1972).

When transferred to recipients many of these cultured blasto-

cysts developed to form normal young (Whitten, 1970; Maurer

et al., 1970; Tervit et al., 1972). Tubal mouse embryos cul-

tured in vitro to the blastocyst stage and transferred to an

ectopic site were also reported to be capable of forming well-

differentiated embryos (Billington, Graham and McLaren, 1968).

Therefore, it would seem that blastokinin is not essential

for blastulation to occur as originally proposed by Krishnan

and Daniel (1967) and Beier (1968). It was later indicated

that blastokinin may have a more general function in embryo-

genesis after the free-blastocyst stage (El-Banna and Daniel,

1972a,b; Daniel, 1972c). Development of embryos past the

blastocyst stage in vitro has not been reported. This sup-

ports the conclusion that uterine secretions, the protein

milieu in particular, are necessary for continued embryonic

development.

A possible explanation of differences found in the extent

of blastocyst development in vitro in the absence of the uter-

ine environment is found in the work of Nicholas (1942),

Fawcett, Wislocki and Waldo (1947), Runner (1947), Fawcett

(1950), and Kirby (1962). Their data indicated that tropho-

blast of transplanted oviducal embryos expand in a morpho-

logically normal manner, but the embryonic disc cells fail








to develop. Also, Daniel (1971a) acknowledged that blasto-

cyst growth, presumably promoted by blastokinin, might be only

volume change due to fluid accumulation in the blastocele

rather than an increase in the number of cells or rate of

mitosis by trophoblast cells. Fluid accumulation in the

blastocyst is an energy-coupled process and not a simple pro-

cess of osmosis (Tuft and Boving, 1970; Gamow and Daniel,

1970). Fluid accumulation could, therefore, be an important

function associated with the presence of blastokinin.

Synchrony between stage of embryonic and uterine devel-

opment has been demonstrated to be essential for normal embry-

onic development in mice (Kirby, 1962), rats (Dickmann and

Noyes, 1960), rabbits (Adams, 1969,1971,1973; Beier et al.,

1972), sheep (Rowson and Moor, 1966), swine (Hunter, Polge

and Rowson, 1967; Bazer et al., 1969) and cattle (Rowson

et al., 1969,1972a). The work of Bazer et al. (1969) indi-

cated that embryos in synchrony with the uterus were at an

advantage in competing for a possible uterine factors) as

compared to an asynchronous embryo. This, along with Kirby's

(1962) work with mice, suggests that embryos must reach a

precise stage of development before they can utilize the uter-

ine factors) apparently necessary for their continued devel-

opment and that the uterine factors) is secreted only at a

specific time. Therefore, the embryo must be present in the

uterus at the proper time in order to utilize the secretionss.

Additional information on the relationship between the

uterus and embryonic development is found in experiments with









animals having delayed implantation (Dickmann and DeFeo, 1967;

Chang, 1968; Daniel, 1970). The work of Dickmann and DeFeo

(1967) and Chang (1968) demonstrated that uterine rather than

embryonic factors determine whether or not implantation will

be delayed. Therefore, it is obvious that the state of the

uterus determines the state of the blastocysts it contains.

This is also additional support for the view that uterine

secretions are necessary for continuous embryonic development

and that secretion of the required factors) is not constant

but occurs at a specific time.

In recent years, uterine luminal fluids from several

species have been examined and found to consist primarily of

proteins. In the rat, in contrast to early work by Warren

(1938) which describes uterine fluids only as the necessary

medium for movement of spermatozoa and studies by others on

the chemical composition of uterine secretions (Shih, Kennedy

and Huggins, 1940; Howard and DeFeo, 1959; Heap and Lamming,

1960,1962,1963; Heap, 1962), the first studies of protein

components of uterine luminal fluid were those of Bredeck and

Mayer (1955), Junge and Blandau (1958), Ringler (1961), Albers

and Castro (1961) and Kunitake et al. (1965). Marked differ-

ences between potassium content of ucerine fluid and plasma

indicated to Howard and DeFeo (1959) that the fluid in the

uterine lumen was a secretion rather than a transudate from

plasma. Other chemical composition studies (Shih et al.,

1940; Heap and Lamming, 1960,1962,1963; Heap, 1962) revealed

that certain chemical constituents of the rat uterine lumen








were greater at estrus than at diestrus and that they were

under the influence of estrogen. Junge and Blandau (1958)

found low levels of four major electrophoretic protein com-

ponents in rat uterine fluid. Ringler (1961) reported that

rat uterine fluid contained proteins with five different

electrophoretic mobilities and that-only one, a prealbumin

fraction, was specific to uterine fluid. Based on this find-

ing he suggested that uterine luminal fluid was essentially

an ultrafiltrate of plasma supplemented by specific uterine

secretions. Albers and Castro (1961) also found that rat

uterine fluid contained at least five protein components when

they used the techniques of Ouchterlony gel-diffusion and

immunoelectrophoresis. Only one of their five proteins was

specific to uterine fluid, and it migrated similar to serum

B-globulins. Kunitake et al. (1965) demonstrated nine protein

components of rat uterine fluid by disc electrophoresis.

Using Ouchterlony gel-diffusion analysis with polyacrylamide

discs and rabbit anti-rat sera, they showed that at least four

of these proteins were common to serum. Five of the nine pro-

tein components were thought to be specific to uterine fluid.

These results added further evidence to the concept that

uterine fluid is, in part, a secretary product of the uterine

endometrium. Using immunofluorescence analysis, Joshi and

Murray (1974) recently reported data which indicated that

rat uterine fluid contained a peptidase unique to uterine

secretion, not found in blood.








Studies on the chemical composition of rabbit uterine

secretions revealed greater levels of certain constituents

during the luteal phase (pseudopregnancy) than at estrus

which was contrary to data from the rat (Shih et al., 1940;

Heap and Lamming, 1960,1962,1963; Heap, 1962). Also, these

chemical constituents appeared to be under control of pro-

gesterone rather than estrogen. The first study on proteins

of rabbit uterine fluid was conducted by Stevens et al. (1964)

using diffusion in agar gel, moving boundary electrophoresis

and immunoelectrophoresis to characterize the proteins in

fluid from the ligated uteri of estrous rabbits. Eight elec-

trophoretic components were separated using moving boundary

electrophoresis. Two of these were specific to uterine fluid

and were not found in serum, one migrating as a prealbumin

and the other as an a-globulin. Diffusion of uterine fluid

in agar gel revealed 13 antigenic components, three of which

appeared to be specific to uterine fluid. At least five anti-

gens, not found in serum, were identified by means of immuno-

electrophoresis. Two of these had electrophoretic mobilities

similar to prealbumins and three similar to B-globulins.

Kirchner, Hirschhauser and Kionke (1971) demonstrated pro-

tease activity in the B-glycoprotein fraction of rabbit uter-

ine secretion. They assumed a relation between uterine pro-

tease and implantation of the blastocyst.

Using Sephadex G-200 gel filtration, Krishnan and Daniel

(1967) studied uterine fluid protein constituents of rabbits

in early pregnancy. They were able to demonstrate the








presence of five major protein fractions. One of these frac-

tions, Fraction IV, was not present until day 3 post coitum;

it reached a maximum concentration on day S and then declined

until day 9. It was also found in uterine fluid of pseudo-

pregnant rabbits on day 7 post coitum and in blastocelic fluid,

but was not observed in maternal serum, fetal serum, fetal

amniotic fluid or in uterine fluid accumulated by ligation of

the rabbit uterus during days 3 to 10 of gestation. When

used to supplement Ham's F10 culture medium in rabbit embryo

culture, it promoted embryonic development to the expanding

blastocyst stage and was thus termed "blastokinin." Beier

(1968), at approximately the same time as Krishnan and Daniel

(1967), demonstrated that flushed rabbit uterine fluid and

blastocelic fluid obtained on day 6 of pregnancy contained

both plasma and uterine specific proteins. He described one

of the uterine specific proteins present in blastocelic fluid

as "uteroglobin" and suggested that it was involved in blasto-

cyst development. There is no doubt that "uteroglobin" and

"blastokinin" are the same protein (Kirchner, 1972; Beier,

1970; Hamana and Hafez, 1970; Beier, Kihnel and Petry, 1971;

Daniel, 1971a). Hamana and Hafez (1970) demonstrated the

presence of this protein in rabbit blastocelic fluid between

days 5 and 8 of pregnancy with a maximum intensity at 6.5 to

7 days post coitum. Petzoldt (1974) recently reported that

micro-disc electrophoretic protein patterns of day 4, 5, 6

and 7 blastocyst fluids were similar to corresponding









patterns of uterine secretion samples.

Using immunofluorescent techniques, Kirchner (1972) was

able to show that uteroglobin was a product of the uterine

endometrium and that it diffused across the blastocyst cover-

ings into the blastocele. Labeled amino acid work by Murray

and Daniel (1973) showed that blastokinin, as well as other

macroglobulin fractions, was produced by the uterine endo-

metrium. Bullock and Connell (1973) recently demonstrated

the presence of a protein, similar to blastokinin, in uterine

flushings from nonpregnant rabbits. They also observed an

electrophoretic band with mobility similar to blastokinin in

flushings from nonpregnant rabbits and pregnant rabbits on

days 1 and 2. Thus, it appears that blastokinin is present

in minute quantities in the nonpregnant state and that its

secretion is greatly accelerated during early pregnancy.

Blastokinin is a glycoprotein with amino acids constitut-

ing approximately 74% of its weight and carbohydrates 6%

(Krishnan and Daniel, 1968). Beier (1968) also found it to

be a glycoprotein. Beier (1968) showed that estradiol and

progesterone had a profound effect on the level of this uter-

ine glycoprotein in day 6 post coitum uterine fluid. He

suggested that uteroglobin secretion was under the control

of progesterone and that excessive amounts of estradiol

caused a reduction in the amount of uteroglobin produced.

The molecular weight of blastokinin was originally esti-

mated by Sephadex G-200 gel filtration to be 27,000 (Krishnan









and Daniel, 1967,1968) and by ultracentrifugation (Beier,

1968,1970) to be about 30,000. However, Murray, McGaughey

and Yarus (1972b) found the above figures to be overestimated.

Using gel filtration, SDS polyacrylamide gel electrophoresis

and equilibrium sedimentation centrifugation, they reported

the molecular weight to be approximately 15,000. Amino acid

and spectrophotometric analyses of blastokinin suggested that

the protein had a minimal molecular weight of approximately

14,200 (McGaughey and Murray, 1972). Bullock and Connell

(1973) confirmed the latter reports and estimated the molecu-

lar weight of blastokinin to be 14,525 by gel filtration and

10,045 by polyacrylamide gel electrophoresis in the presence

of sodium dodecyl sulfate. The possibility of subunits for

blastokinin was suggested by McGaughey and Murray (1972).

The existence of subunits was given as a possible explanation

for the difference between the two methods of molecular weight

determination used by Bullock and Connell (1973). An immuno-

assay for blastokinin using radial immunodiffusion with goat

anti-blastokinin antiserum was developed by Johnson, Cowan

and Daniel (1972). They presented evidence that blastokinin

possesses at least two antigenic determinants which may occur

on separate subunits. This was added support for the possible

subunit structure for blastokinin.

Recent data have verified most of the earlier observa-

tions that blastokinin is the predominant uterine specific

glycoprotein in rabbits (Hamner, 1970; Daniel, 1971a; Beier

et al., 1971; Beier and Beier-Hellwig, 1973; Beier, 1974a,b).








Blastokinin was presumed to be the most important because of

its progesterone dependent accumulation and its possible

influence on development of the embryo. Fluid in the uterine

lumen of rabbits contains considerable protein, and the volume

of this fluid, total concentration of its macromolecular com-

ponents and its protein patterns change continuously from

ovulation to implantation. These proteins (predominantly

glycoproteins of 25,000 to 50,000 MW) are a product of selec-

tive filtration from plasma proteins and of biosynthesis by

epithelial cells of uterine endometrial glands and endometrial

surface epithelium. Data also indicate that ovarian hormones

regulate secretion and synthesis of the proteins. The most

recent paper by Beier (1974b) included a summary of recent

knowledge on iinunologically identical proteins of uterine

secretion and blastocyst fluid in the rabbit during the late

preimplantation period (days 6 and 7 post coitum). He listed

serum identical proteins as albumin, transferring, immunoglobu-

lin and a-macroglobulin. Uterine "secreto-proteins" were

listed as uterine prealbumin, uterine postalbumin, uteroglo-

bin, uterine B-glycoprotein and S-uterus-macroglobulin. The

serum identical transferring and n-macroglobulin were not

present in blastocyst fluid until day 8 post coitum.

Added support for the importance of uterine specific

proteins in embryonic development, specifically blastokinin,

is found in studies on the relationship between uterine fluid

proteins and the diapausing state of blastocysts from mammals

having delayed implantation (Daniel, 1968,1970,1971b,1972a;








Daniel and Krishnan, 1969). Comparison of uterine fluid

during the free-blastocyst period revealed a much lower con-

centration of protein from mammals with delayed implantation.

The rabbit had much higher levels of total uterine protein

than did the lactating rat during facultative delayed implan-

tation, or the mink, fur seal and armadillo during obligate

delayed implantation. Polyacrylamide gel disc electrophore-

sis revealed no protein with similar mobility to rabbit

blastokinin in the lactating rat and armadillo; however, a

similar band was found in low concentration in uterine fluid

of mink (Daniel, 1968) and fur seal (Daniel, 1971b,1972a).

In fur seal this component of uterine fluid differed in

immunologic properties from rabbit blastokinin. Similarity

between an electrophoretic band of any of these species and

blastokinin must be taken in light of the discussion by Beier

and Beier-Hellwig (1973) which draws attention to the electro-

phoretic difficulty of differentiating between a specific

postalbumin band and hemoglobin artifacts. It was obvious to

Daniel' (1971b) that the uterine fluid proteins of the fur seal

differed both qualitatively and quantitatively between the

time when the embryo was dormant and the time it became

reactivated to implantation. Daniel and Krishnan (1969) also

observed gross qualitative and quantitative differences in

the protein content in uterine secretions of certain mammals

depending on whether or not they exhibited delayed implanta-

tion. Diapause embryos of animals with.obligate delayed

implantation (mink, fur seal and armadillo) were stimulated









to grow (as measured by total blastocyst expansion and in-

creased mitotic index) in vitro in the presence of rabbit

blastokinin, but not in media containing serum. Blastocysts

from heavy lactating rats, where a facultative delayed implan-

tation had been produced, expanded in F10 culture medium

alone or media containing serum or macromolecular components

of uterine secretions from day S pregnant rats. However,

supplementation of the medium with rabbit blastokinin had no

effect. The mitotic index was higher than in uncultured con-

trols in all conditions. Growth of active day 5 rat blasto-

cysts, as evidenced by increased mitotic activity, was stimu-

lated by culture medium supplemented with uterine macronolecu-

lar components. Thus, there was evidence that in the case of

facultative delayed implantation there apparently was some

other uterine condition inhibiting growth of blastocysts, in

addition to a protein deficiency. It was concluded that

delayed implantation resulted from failure of the m-other to

provide sufficient protein and/or certain specific proteins

as needed for active growth of the blastocysts.

Quantitative and qualitative variation in protein con-

stituents of swine uterine fluid were studied on days 2 to 18

and day 20 of the estrous cycle (Murray, 1971; Murray et al.,

1972a; Squire et al., 1972). The maximum average total uter-

ine protein level was found on day 15 with levels increasing

from day 10 to day 15. The uterine protein level decreased

sharply after day 15 to levels on days 17, 18 and 20 which

were similar to those observed before day 10 of the estrous








cycle. Sephadex G-200 gel filtration revealed two protein

fractions (IV and V) which were present only during the luteal

phase of the estrous cycle. Three additional fractions (I to

III) were present throughout the estrous cycle. Estimated

molecular weights of Fractions I to V were 400,000 and

greater; 200,000; 90,000; 45,000; and 20,000, respectively.

Fraction IV contained a lavender-colored protein and was

present only between days 12 and 16 of the estrous cycle.

Polyacrylamide gel disc electrophoresis demonstrated that the

lavender protein was basic, as it migrated toward the cathode

at pH 8.0. Fraction V appeared between days 9 and 16 of the

estrous cycle with peak concentration of total uterine pro-

tein occurring on day 15. This fraction constituted greater

than 20% of the total recoverable protein on days 9 to 16.

Polyacrylamide gel disc electrophoresis showed that this frac-

tion was made up of six acidic proteins migrating toward the

anode at pH 8.0 with Rf values of 0.91, 0.84, 0.80, 0.74,

0.19 and 0.18 (albumin Rf=1.00). Murray (1971) suggested

that these two fractions appearing during the luteal phase

of the estrous cycle might be related to rapid growth and

expansion of the trophoblast of the more advanced embryo.

These data on the cyclic nature of porcine uterine protein

secretion indicated quantitative and qualitative changes

during the estrous cycle.

In subsequent work with the purple Fraction IV discussed

above, it was demonstrated that the Fraction IV basic proteins

(one to three) were present in allantoic fluid after day 30 of









pregnancy, and evidence was presented which suggested that

these proteins were involved in some aspect of placental

development which might, in some way, have affected fetal

development (Chen, 1973; Chen and Bazer, 1973; Chen, Bazer

and Roberts, 1973b). Intravenous administration of sheep

anti-Fraction IV antisera on days 7, 9, 11, 13 and 15 follow-

ing mating decreased placental length and allantoic fluid

protein concentration (P<.01). Administration of the anti-

sera on days 34, 36, 38, 40 and 42 of pregnancy significantly

(P<.01) reduced placental weight and length, and fetal wet

weight and crown-rump length. Krishnan (1971) administered

chicken anti-blastokinin antisera to rabbits during early

pregnancy (days 2, 4 and 6 post coitum) and reported either

abnormal embr-yonic development or complete cessation of preg-

nancy. Daniel (1972b), in some preliminary work with rabbit

antiserum to swine luteal phase uterine protein, substanti-

ated, in part, Krishnan's observations. Of three sows treated

with antiserum, two did not give birth to any offspring and

did not return to estrus until 173 and more than 237 days

after breeding.

More recent work demonstrated that Fraction IV contained

only one protein component, represented by a single electro-

phoretic band which moved toward the cathode at pH 7.0, with

a molecular weight of 32,000 and an isoelectric point at

approximately pH 9.7 (Chen et al., 1973a). The protein was

specific to the uterus as antiserum prepared against it did

not cross-react with extracts from other tissues. It was








made up of 12.5% carbohydrate by weight and large amounts of

basic amino acids. It was pointed out that the purple protein

evidently self-associated to form dimers, under certain condi-

tions, which easily dissociated in the presence of high salt

concentration. Schlosnagle et al. (1974) reported that the

purple protein contained one atom of iron per 32,000 molecular

weight polypeptide and showed acid phosphatase activity.

Using fluorescent antibody techniques, Chen et al. (1975)

showed that the uterine purple acid phosphatase was formed in

the epithelial cells of the endometrial surface and glands,

and was detectable within the areolae by about day 30 of ges-

tation. It was suggested that the purple protein was synthe-

sized and secreted by the uterine surface and glandular epi-

thelial cells and that the placental areolae served as sites

of absorption and transport into the chorio-allantoic mem-

branes and allantoic fluid during pregnancy.

As in the rabbit, Heap (1962) and Heap and Lamming (1960,

1963) reported a significant increase in certain chemical

constituents of the sheep uterus during the luteal phase of

the estrous cycle. The work with cannulated sheep uterine

fluid by Perkins et al. (1965) and Iritani et al. (1969)

showed an increase in volume around estrus, but no luteal

phase increase was noted. More recent work with sheep uter-

ine flushings (F. W. Bazer, unpublished data) indicated a

quantitative and qualitative change in uterine protein secre-

tion during the estrous cycle. There was an increase in

total recoverable protein (also, some low molecular weight








proteins were present) during the luteal phase of the estrous

cycle.

Early work on uterine fluid of cattle (Olds and VanDemark,

1957; Heap,.1962; Heap and Lamming, 1962; Fahning et al.,

1967) indicated that its chemical composition varied with

stage of estrous cycle, with highest levels being reported

during the luteal phase. This indicated hormonal control of

bovine uterine fluid composition. Schultz et al. (1971) fur-

ther illustrated this when they found that concentrations of

reducing substances, total protein, potassium, chloride,

inorganic phosphate, and alkaline and acid phosphatase activi-

ties all varied significantly with stage of the estrous cycle.

These above reports all support the concept that bovine uter-

ine fluid is the result of active secretion and not merely a

product of diffusion from the blood. Until the present study

was conducted no attempt had been made to characterize the

proteins present in bovine uterine fluid nor to relate them

to actual plasma steroid concentrations. After the present

study had been completed, Roberts and Parker (1974a) reported

that the protein components of bovine luminal fluid were

mainly serum proteins, but that small amounts of uterine-

specific proteins had been detected by disc electrophoresis

at pH 4.5. They had examined the luminal fluid from uterine

horns of cows at different stages of the estrous cycle or

early pregnancy by polyacrylamide gel electrophoresis at pH

4.5 and 8.9, isoelectric focusing, immunoelectrophoresis,

and gel filtration.








In preliminary studies of human uterine secretions,

Daniel (1971a) was not able to find a protein comparable to

blastokinin. Beier et al. (1971) state that the intra-tubal

and intra-uterine protein pattern apparently develops not

only in the rabbit but in the human through selection of

individual plasma proteins on the one hand and through bio-

synthesis of uterus-specific proteins on the other. Shirai,

lizuka and Notake (1972) presented disc-electrophoretic evi-

dence for a postalbumin "blastokinin-like" fraction which

exhibited a prominent peak during the mid-secretory phase of

the human menstrual cycle. Noske and Daniel (1974) also

reported the appearance of a postalbumin protein band in the

hamster on day 3 post coitum, which had an electrophoretic

mobility similar to that of blastokinin. Both of these

reports must be taken in light of the electrophoretic dif-

ficulty of differentiating between a specific postalbumin

band and hemoglobin artifacts (Beier and Beier-Hellwig, 1973).

Composition of bovine uterine fluid was suggested to be

under hormonal control (Olds and VanDemark, 1957; Heap and

Lamming, 1960; Heap, 1962; Fahning et al., 1967; Schultz

et al., 1971) because of observed cyclical changes in certain

chemical constituents of bovine uterine fluid during the

estrous cycle. Schultz et al. (1969) reported that proges-

terone administered to ovariectomized cows resulted in a

greater increase than did estradiol in the size of nuclei of

epithelial cells lining the mucous glands of the uterine

endometrium. They suggested that the increase in nuclear








size might have been associated with an increased rate of

cell secretion. In the ewe (Murdoch, 1972), progesterone

stimulated activity of acid and alkaline phosphatases in the

intercotyledonary endometrium.

Roblero (1973) found that progesterone administered to

ovariectomized pregnant nice resulted in embryos which had

significantly more cells than embryos from ovariectomized

females not receiving progesterone. Other work with the

mouse (Mintz, 1970; Pinsker and Mintz, 1973) has indicated

that estrogen controls the secretion of a uterine factor

which affects implantation. Yasukawa and Meyer (1966), in a

study on the effect of progesterone and estrone on preimplan-

tation and implantation stages of embryo development in the

rat, reported that changes necessary for and indicative of

impending implantation had been induced by the synergistic

action of estrone and progesterone on the blastocyst. Several

chemical composition studies with rat uterine fluid (Shih

et al., 1940; Heap and Lamming, 1960,1962,1963; Heap, 1962)

revealed that certain chemical constituents were present in

greater amounts at estrus than diestrus and were therefore

considered to be influenced by estrogen. In the rabbit, con-

centration of chemical constituents was greater during the

luteal phase (pseudopregnancy) than at estrus and was con-

sidered to be under the control of progesterone rather than

estrogen. Wu and Allen (1959) reported graded effects of

progesterone on pregnancy maintenance in castrated rabbits.

Beier (1968,1970) reported that estradiol and progesterone








had a profound effect on uteroglobin on day 6 of pregnancy

in rabbits. He reported that excess estradiol caused a reduc-

tion in the amount of uteroglobin produced, which suggested

to him that uteroglobin might be under the control of proges-

terone and that estrogen antagonized the effect of progester-

one. Pincus and Kirsch (1936) reported a detrimental effect

of estrogen on blastocyst growth and development. Adminis-

tration of estrogen postcoitally in the rabbit has been

reported to lower uterine carbonic anhydrase after 24 hours

(Makler and Morris, 1971). El-Banna and Daniel (1972a) found

that progesterone stimulated rabbit blastocysts in vitro when

in combination with uterine proteins. Urzua et al. (1970)

found that blastokinin was present in uterine fluid of ovari-

ectomized rabbits that had received progesterone, or proges-

terone and estradiol in combination, but it was absent in

rabbits treated with estradiol alone. Arthur and Daniel

(1972) found blastokinin in the uterine fluids of ovariectom-

ized rabbits following administration of progesterone but not

estrogen, and the kinetics of the dose- and time-responses

obtained indicated that this was the normal relationship

during pregnancy. Blastocysts transferred to uteri of cas-

trate rabbits given progesterone grew and differentiated up

to the time of implantation. Whitson and Murray (1974) found

that endometrial cells from mature female estrous rabbits

were capable of synthesizing blastokinin in vitro following

treatment with progesterone for 48 hours. Urzua et al. (1970)

and Arthur, Cowan and Daniel (1972) reported that blastokinin








bound progesterone and estradiol, but binding of progesterone

to blastokinin was inhibited by estradiol and vice versa.

Recently, Goswami and Feigelson (1974) reported on the

differential regulation of a low molecular weight protein in

rabbit oviductal and uterine fluids by progesterone and estra-

diol-176. Based on similar molecular weight and electropho-

retic mobility, this protein was thought to be the same as

blastokinin. It was present in uterine and oviductal fluids

of intact estrous rabbits and assumed a unique cone-shaped

profile, upon polyacrylamide gel electrophoresis followed by

a modified Amido Black staining and destaining procedure.

The protein was under hormonal control and was absent from

both uterine and oviductal fluid following ovariectomy.

Exogenously administered progesterone strongly induced it in

uterine fluid (very slightly in oviductal fluid) of ovari-

ectomized rabbits. Estradiol had a much greater effect than

progesterone on the induction of this cone-forming protein

in oviductal fluid, although of lesser absolute magnitude

than the effect of progesterone on its induction in uterine

fluid. Contrary to the work of Urzua et al. (1970) and

Arthur et al. (1972), neither the uterine nor the oviductal

fluid cone-forming protein was demonstrated to bind proges-

terone in vitro to any discernable extent. More extensive

Sephadex G-30 gel column chromatography was used by Goswami

and Feigelson (1974), which clearly indicated that the cone-

forming protein was eluted in a fraction widely separated

from the single peak of free progesterone. Thus, it was








concluded that rabbit genital tract fluid "cone protein" does

not detectably bind progesterone. Additional immunological

work will have to establish whether blastokinin and this cone-

forming protein are the same.

Data in the pig (Murray, 1971) indicated that maximum

uterine secretary activity occurred in gilts during the

luteal phase of the estrous cycle when the uterus was under

the influence of progesterone. He reported an abrupt decrease

in protein content of uterine flushings which coincided with

regression of the corpora lutea. Knight et al. (1973a)

reported a positive relationship between quantity of recover-

able uterine protein and quantity of luteal tissue in super-

ovulated and unilaterally ovariectomized-hysterectomized

gilts. It was suggested that progesterone, possibly in syner-

gism with estrogen, was primarily responsible for inducing

quantitative and qualitative changes in the protein milieu

of uterine flushings. In a subsequent study, Knight et al.

(1973b) indicated that progesterone was the essential hormone

which'regulated the quantitative and qualitative aspects of

porcine uterine protein secretion during the estrous cycle

and presumably during pregnancy. Work with the porcine purple

uterine protein (Chen, 1973; Chen et al., 1973a; Schlosnagle

et al., 1974) also established that its secretion was regu-

lated by progesterone. Uterine protein secretions of ovari-

ectomized gilts treated with progesterone until either day 7,

9, 11, 13, 15, 17 or 19 after onset of estrus were quanti-

tatively and qualitatively similar to uterine protein









secretions of intact nonmated gilts up to day 15 of the

estrous cycle (Knight et al., 1974c). However, on days 17

and 19 total protein was greater in the treated gilts and the

qualitative aspects were maintained. Knight et al. (1974b)

reported that increased uterine protein secretions, resulting

from high progesterone levels, enhanced placental development

and allantoic fluid volume. This led to an increase in the

placental surface area which was in contact with the maternal

endometrium. It was suggested that the establishment of maxi-

mum placental surface area early in gestation might be of

critical importance with respect to fetal growth and survival

as pregnancy progressed towards term. In more recent work,

Knight (1975) presented data which indicated that the develop-

ment of adequate placental mass was apparently the key factor

necessary for adequate and sustained fetal growth and develop-

ment.

The role of the uterus is not limited solely to control

of embryonic development. It is also involved in control of

luteal function. Wiltbank and Casida (1956) reported that

hysterectomy during the luteal phase of the estrous cycle

prolonged the lifespan of corpora lutea in cows and ewes.

Since this report, it has been clearly demonstrated that the

uterus is responsible for regression of corpora lutea of

estrous cycles in cattle, sheep, swine, horses and other

species (reviews by Anderson, Bowerman and Melampy, 1963;

Bland and Donovan, 1966; Anderson, Bland and Melampy, 1969;

Schomberg, 1969; Rowson, 1970). These reviews point out that








hysterectomy results in maintenance of the corpora lutea for

approximately the length of normal gestation. Cyclical

luteal regression appears to be under local control of the

uterine horn adjacent to the corpora lutea of cattle, sheep

and swine (Hansel and Echternkamp, 1972). When one uterine

horn was removed in sheep, swine (Anderson, 1966) or cows

(Ginther et al., 1967) corpora lutea on the ovary adjacent to

the retained horn regressed at the expected time, but corpora

lutea on the opposite ovary were maintained beyond the normal

estrous cycle length.

Luteolytic effects of endometrial extracts have been

demonstrated. Williams et al. (1967) reported that acetone-

dried extracts of bovine uteri caused luteal regression when

injected into pseudopregnant rabbits. Schonberg (1967) found

that swine uterine flushings obtained on days 1 to 10 and day

20 of the estrous cycle had no detrimental effect on in vitro

growth of granulosa cells or their progesterone production,

but day 12 uterine flushings sometimes showed a luteolytic

effect. Flushings obtained between days 14 and 18 of the

estrous cycle were markedly luteolytic and destroyed the

granulosa cells within 6 to 8 hours. The uterine fluid com-

ponent with luteolytic activity was later reported to be

thermolabile and nondialyzable with an estimated molecular

weight of about 200,000 (Schomberg, 1969). Barber (1972)

demonstrated that the component of swine uterine flushings

showing the greatest in vitro luteolytic effect on granulosa

cells was the Sephadex G-200 gel filtration Fraction II








(Murray, 1971), which also had a molecular weight of approxi-

mately 200,000. Mazer and Wright (1968) reported that a non-

dialyzable luteolytic factor was present in the uterus of the

hamster on days 6 and 7 of pseudopregnancy. Lukaszewska and

Hansel (1970) found that aqueous extracts (precipitable in

55% ammonium sulfate) of day 10 to 13 bovine endometrium were

luteolytic when injected intraperitoneally into pseudopregnant

hysterectomized hamsters. The active factor was thought to

be either a large molecular weight protein or a smaller mole-

cule bound to protein. Further studies showed that lipid

extraction of the active protein fraction removed the luteo-

lytic activity. When the lipid extract was subjected to thin

layer chromatography, the fraction containing prostaglandins

(discussed below) did not have luteolytic activity. Hansel,

Concannon and Lukaszewska (1973) showed clearly that luteoly-

tic activity was separated from prostaglandin activity and

that the luteolytic factor studied had an electrophoretic Rf

value similar to that of the arachidonic acid standard. It

was suggested that the bovine endometrium exerts its local

luteolytic effect by providing the corpus luteum with one

or more precursors (arachidonic acid) which could be converted

into prostaglandin or other luteolysin in situ by the corpus

luteum.

Much evidence has accumulated on the luteolytic proper-

ties of PGF20, a prominent uterine prostaglandin. It has been

identified by Goding et al. (1972) as "the" luteolytic hor-

mone and by McCracken et al. (1972) as a luteolytic hormone








in sheep. Prostaglandin F2a has also been shown to cause

luteal regression in cattle (Lauderdale, 1972; Rowson, Teruit

and Brand, 1972b, Hansel et al., 1973; Inskeep, 1973; Chenault,

1973; Lauderdale et al., 1974), horses (Noden, Hafs and Oxen-

der, 1973), swine (Muljono et al., 1974), rats (Pharriss and

Wyngarden, 1969), hamsters (Gutknecht, Wyngarden and Pharriss,

1971) and guinea pigs (Blatchley and Donovan, 1969).

Several facts have prevented acceptance of PGF2a as

"the" luteolytic hormone by all investigators. Hansel et al.

(1973) argued that measurements of PGF2a in uterine venous

plasma on any given day of the estrous cycle were highly vari-

able, and that agreement between different laboratories was

very poor. They stated that it was not clear whether the

rise in plasma PGF2a preceded the fall in progesterone in

every case studied. Some in vitro studies indicated that

PGF,2 was luteotrophic rather than luteolytic (Hansel et al.,

1973; Speroff and Ramwell, 1970; Sellner and Wickersham, 1970).

Hansel et al. (1973) also found prostaglandins E2, El, A1 and

A2 to be luteotrophic when incubated with bovine luteal

tissue. Wilson, Butcher and Inskeep (1972) reported levels

of PGF2, in the endometrium and uterine venous blood of preg-

nant ewes to be higher than in nonpregnant ewes on days 13,

14, 15, 16 and 18 after onset of estrus.

In maintenance of luteal tissue during pregnancy, it is

possible that some luteotrophin may be produced by the embryo

which overrides the uterine luteolytic factor (probably PGF,2

or a precursor such as arachidonic acid), or the embryo may









exert a direct antiluteolytic effect (i.e., absorption or

inactivation of the luteolytic factor). Evidence supporting

the latter mechanism was presented by Warren, Hawk and

Williams (1971). They found that infusion of homogenized

embryos into the uterus of the ewe prevented IUD-induced

luteal regression.

That the embryo or trophoblast could be producing a

luteotrophin or overriding the effect of the uterine luteo-

lytic factor is suggested by much recent work with the blasto-

cyst. The steroids, estrogen and progesterone, are present

in rabbit blastocysts and uterine fluid (Seamark and Lutwak-

Mann, 1972; Beier, 1974b); and estradiol bound to rabbit

blastocysts has been implicated in the development or implan-

tation of the blastocysts, consistent with the hypothesis that

estradiol may act as a local signal from the blastocyst to the

uterus (Bhatt and Bullock, 1974). In the rat (Dickman and

Dey, 1973,1974a,b; Dey and Dickman, 1974a) and mouse (Dey and

Dickman, 1974b) data were presented which suggested that

blastocysts can synthesize steroid hormones which are critical

for morula to blastocyst transformation and implantation. The

results suggested that one of the hormones synthesized was

estrogen. Perry, Heap and Amoroso (1973) reported production

of both progesterone and estrogen by pig blastocysts (days 14

to 16). Extremely high estrone and estradiol concentrations

were reported in allantoic fluid on various days of pregnancy

in the gilt (Knight et al., 1974a; Knight, 1975), suggesting

a placental source for the estrogens. Research on the








influence of the marsupial embryo on the uterus (Renfree,

1972) provided evidence that the embryo or placenta exerts

a twofold effect on the uterus due to a hormone produced by

the embryo or its membranes or to an immunological response.

The suggested effects of the embryo were: to exceed the

influence of the corpus luteum by increasing the weight of

the endometrium in the uterus which carries it, and to stimu-

late production of uterine-specific proteins. Of consider-

able interest are the recent reports that a substance similar

to HCG or LH was found in human plasma within 6 days of

fertilization (Saxena et al., 1974), and in rabbit blasto-

cysts on days 5 and 6 post coitum (prior to implantation) at

concentrations ten times higher than in pregnant rabbit plasma

(Haour and Saxena, 1974). Experiments involving the surgical

transfer of blastocysts in sheep (Moor and Rowson, 1966a,b)

and pigs (Dhindsa and Dziuk, 1968) have shown that embryos

must be present in the uterus before day 13 for pregnancy to

be established and for the corpora lutea of the estrous cycle

to be converted into corpora lutea of pregnancy. This appar-

ent ability of the blastocyst to signal its presence and to

modify corpus luteum function prior to attachment to the

uterine endometrium implies production of some hormonal fac-

tor. Such an antiluteolytic factor must be suppressing endo-

metrial synthesis of the uterine luteolysin, blocking its

direct transfer from the uterus to the site of action in the

corpus luteum, or having a luteotrophic effect on the corpus

luteum in the presence of secretion of the uterine luteolysin.








Heap and Perry (1974) speak of estrogens being luteotrophic

in the pig, and postulate that the uterine endometrium is

able to conjugate estrogens produced by the blastocyst which

might provide a source of circulating estrogens of low bio-

logical potency, yet readily metabolizable to an active form,

in other tissues (such as the corpus luteun) where their

stimulatory (luteotrophic) effect might be expressed.

Sufficient data have been reviewed on restriction of

embryos to the oviduct, in vitro culture, synchrony of embry-

onic and uterine development and delayed implantation to sup-

port the concept that some uterine components) is extremely

important for continued normal development of the embryo past

the early blastocyst stage. Uterine secretions in many

species have been reviewed and it has been established that

these secretions (predominantly proteins) are the product of

active secretion by the uterine endometrium supplemented by

an ultrafiltrate of plasma. It has been established that

these uterine protein secretions change quantitatively and

qualitatively during the estrous cycle in most animals studied,

and that their secretion is regulated by steroids. They have

also been associated with early development of the embryo and

trophoblast. The relationship between the uterus and main-

tenance of corpus luteum function has been established, and

much data presented and discussed which implicates the embryo

in a definite role, in some cases, in prevention of luteal

regression.













CHAPTER 1

NONSURGICAL COLLECTION AND STUDY
OF BOVINE UTERINE LUMINAL PROTEINS



Bovine uterine fluids obtained following slaughter and

from the live cow have been examined by several researchers.

Olds and VanDemark (1957) obtained fluid by individually

passing each uterine horn through a hand-operated clothes

wringer under moderate pressure. Roberts and Parker (1974a)

obtained uterine saline washings within a few minutes of stun-

ning at slaughter. Gupta (1962) used a return flow uterine

catheter attached to a 250 rml Erlenmcyer flask and a vacuum

pump to collect uterine fluid samples during estrus. A tech-

nique described by Fahning, Schultz and Graham (1966) was

used to aspirate uterine fluid from cattle during the estrous

cycle by Fahning et al. (1967) and Schultz et al. (1971). Loe

(1970) modified the procedure of Fahning et al. (1966) and

aspirated fluids on days 0 and 2. Heap and Lamming (1960)

and Heap (1962) obtained uterine washings from cows by using

an endotracheal tube with an inflatable cuff placed into the

uterus via the cervix. They injected an isotonic solution

into the uterus and recovered it by massaging the uterus per

rectum.

The procedure described in this chapter for nonsurgically

flushing the uteri of cows and heifers was developed using









information gained by C. K. Vincent, J. W. Rundell and A. C.

Mills (unpublished data) and A. C. Mills (unpublished data)

in attempts to nonsurgically recover embryos from cattle,

and from information in the procedures described by Rowson

and Dowling (1949), Fahning et al. (1966) and Heap (1962).

Initially, it was determined from nonsurgical embryo recovery

that most of the fluid injected into the uterus could be

recovered (A. C. Mills, unpublished data). The study described

below was undertaken to determine if there were proteins spe-

cific to the bovine uterus and if any variation in these pro-

teins might be related to plasma steroid concentration and

to the estrous cycle.



Materials and Methods


In this preliminary study of bovine uterine protein

secretions approximately seven uterine flushing samples per

day of the estrous cycle (total of 144) were collected from

live cows and heifers. Approximately three samples per day

were collected from a group of crossbred heifers, three per

day from dairy cows (greater than 35 days postpartum) and one

per day from a set of triplet Angus X Holstein heifers. More

than one sample was collected from the triplets and from

several of the crossbred heifers. In most cases the animals

had been in estrus at least twice with a normal estrous cycle

preceding each collection.









The cattle were checked visually twice daily for estrus

with the aid of Kamar heat detector patches and/or two penec-

tomized Angus bulls fitted with marking collars. Day of

standing estrus was designated as day 0. The cattle were

penned on the morning of the randomly selected day of the

estrous cycle on which a sample was to be collected and then

restrained in a standing position immediately before collec-

tion.

Uterine fluid samples were collected by inserting an

18 Fr urethral catheter through the cervix (rectal manipula-

tion) and inflating the 5 cc cuff just inside the body of the

uterus. Tension was kept on the catheter so that the cuff

was kept firmly against the cervix. A 50 ml glass syringe

filled with 0.33 M saline was connected to the catheter and

the saline was then carefully injected into the uterus. Care

was taken to make sure that the fluid filled the entire uterus

and not just one horn, and that excess pressure was not put

on the uterus by the injection of too much fluid. In some

cases the entire 50 ml was not injected into the uterus.

After momentarily massaging the uterus (less than 1 minute)

the fluid was allowed to flow back into the syringe under

slight negative pressure with rectal manipulation of the

uterus as needed. When no more fluid could be recovered a

pinch clamp was placed on the catheter, the syringe contain-

ing the uterine flush was disconnected and a second 50 ml

glass syringe filled with 0.33 M saline was connected and the








procedure was repeated. The uterine flushings were placed

into stoppered bottles and put into ice water until processing.

Immediately after a uterine flush was completed in the

group of crossbred heifers a 10 ml jugular blood sample was

taken (heparinized vacuum tubes) and placed in ice water. In

the group of dairy cows a heparinized Frlenmeyer flask was

used. No blood was collected from the triplets. The blood

samples were centrifuged for approximately 10 minutes after

which the plasma was removed and stored at -200C for possible

steroid analysis and for comparison of the plasma and uterine

proteins.

Uterine flushings were taken from the ice water and cen-

trifuged at 10,000 rpm for 10 minutes at 53C. They were

sterilized by filtration of the supernatant through a 0.45 p

Millipore filter and concentrated to 1 to 2 ml by vacuum

dialysis. Next the samples were dialyzed against 0.05 M

phosphate-citrate buffer, pH 7.4, for 36 to 48 hours at 30C

and then stored at -200C until further analysis.

Protein concentration of each stored uterine sample was

determined by the method of Lowry et al. (1951) and the total

quantity of protein in each sample calculated by multiplying

that value by sample volume. Total recovered protein was

adjusted (divided by percent recovery) if the saline flush

recovery was less than 100%.

Sephadex G-200 gel filtration was used to fractionate a

portion of each uterine protein sample to obtain a protein

profile. The columns used were 1.5 x 90 cm with a bed height









of approximately 85 cm. They were calibrated using Blue Dex-

tran (MW=2,000,000), to determine the void volume, and aldo-

lase (MW=158,000), ovalbumin (MW=45,000) and ribonuclease A

(MW=13,700) as standard globular proteins of known molecular

weight. Gel filtration was carried out at 30C. The eluent

was 0.05 M phosphate-citrate buffer, pH 7.4, with flow rates

of approximately 6 ml per hour. An aliquot of each fraction

(approximately 2.5 ml) was used for protein determination and

the resulting optical density of each fraction was plotted

against the elution volume, at which the fraction occurred,

to construct a protein profile.

Sephadex G-75 gel filtration was used to construct a

uterine protein profile for each day of the estrous cycle from

a pooled aliquot of the crossbred heifer samples (approxi-

mately three per day). The procedures were the same as for

the Sephadex G-200 protein profiles above. Bovine plasma

Sephadex G-200 and G-75 protein profiles were also obtained

for comparison with the uterine protein Sephadex G-200 and

G-75 profiles.

Polyacrylamide gel disc electrophoresis, using the basic

procedure of Clarke (1964), was used to further study the pro-

tein components of the uterine flushings. A 7% polyacrylamide

gel, 5 mm in diameter and 70 mm long, was used without spacer

or sample gels. Samples containing about 0.3 mg protein were

applied in 0.1 ml of 1 M sucrose and two drops of bromophenol

blue directly on the separating gel. The samples were run

in 0.05 M Tris 0.38 M glycine buffer, pH 8.0, at a constant








current of 2.5 mA per tube and a nominal voltage of 90 volts

for about 1.5 hours or until the marker dye had migrated about

65 mm. The gels were stained with amido black solution (1 gm

per 100 ml of 7% acetic acid) for 1 hour and destined elec-

trophoretically in 3% acetic acid. The electrophoretic

mobility of uterine proteins was expressed in terms of their

mobility relative to that of albumin (Rf) when albumin was

assigned an Rf value of 1.0. Electrophoresis of bovine plasma

was also conducted, using the same procedure, for comparison

with uterine protein polyacrylamide gels.

Prior to making the conclusions discussed below concern-

ing the quality of the uterine protein flushings obtained

nonsurgically, plasma progestin concentrations were determined

for the group of crossbied heifers (approximately three per

day). For analysis, approximately 2,000 CPM of progesterone-

1,2-3H (Amersham/Searle, 31.7 Ci/mM) were added to 2 ml of

plasma. This isotopic steroid served as an internal standard

for correction for procedural losses. The mixture was ex-

tracted vigorously twice with 10 ml of isooctane and aliquots

of the extracted progestins were quantified by competitive

protein binding assay (Murphy, 1967). Details of the pro-

cedures used were described by Gwazdauskas (1972) and Chow

(1972).








Results and Discussion


An average of 84.3 19.7% of the 0.33 M saline put into

the uterus was recovered and the samples appeared to be rela-

tively free of blood contamination during most of the estrous

cycle. Total uterine protein recovered averaged 46.93 41.31

mg, but there were no apparent differences associated with day

or stage of the estrous cycle. Data presented in Table 1 com-

pare average total uterine protein recovered during the

estrous cycle in this study with that recovered by various

methods in other studies. Average total uterine protein col-

lected with the nonsurgical flushing procedure of this study

is comparable to that collected by Heap (1962) who used a

similar type procedure. It is also similar to that found by

Olds and VanDemark (1957), using a postslaughter stripping

technique, and by Gupta (1962) and Schultz et al. (1971) who

nonsurgically aspirated fluid from the uterus. Schultz et al.

(1971) reported protein concentration of plasma and their

aspirated uterine fluid to be practically the same. The

"surgical-like" postslaughter flushing that will be discussed

in Chapter 2 of this study and the postslaughter flushing

reported by Roberts and Parker (1974a), published after the

present study was completed, yielded much lower quantities

of average total uterine protein. The high levels of uterine

protein recovered in nonsurgical flushings in this study and

in those of Olds and VanDemark (1957), Heap (1962), Gupta








Table 1. Average total protein recovered from the
bovine uterus during the estrous cycle.


Av. Total
Method of No. of Recovery Uterine
Study Collection Records (Av.) Protein (mg)

Present Nonsurgical 144 84.319.7%a 46.9341.31a
(Chap. 1) flushings of 100 ml

Present "Surgical-like" 67 97.2+9.7%a 4.644.35a
(Chap. 2) postslaughter of 50 ml
flushings

Olds & Postslaughter 4 0.75 ml 35b
VanDemark stripping
(1957)

Gupta Nonsurgical 21c 2.5 ml 78b
(1962) aspiration

Schultz Nonsurgical 88 2.9 ml 232b
et al. aspiration
Tr971)

Heap Nonsurgical 11 58% of 76b
(1962) flushing 60 ml

Roberts Postslaughter 135 --f 7.5b,d
& Parker flushing 20b,e
(1974a)


aStandard deviation.
Estimates calculated from reported data.
cIncludes only animals in estrus.
During the first 2 to 3 weeks of pregnancy.
eDuring the 3rd week of pregnancy.
Not reported.








(1962) and Schultz et al. (1971) indicate varying degrees of

blood contamination due to techniques employed.

Sephadex G-200 and G-75 gel filtration protein profiles

and electrophoretic polyacrylamide gels of uterine flushings

were similar to those of bovine plasma collected at the same

stage of the estrous cycle. However, uterine flushings con-

tained varying amounts of hemoglobin throughout the estrous

cycle as evidenced by a hemoglobin bands) on the uterine

polyacrylamide gels and a hemoglobin fraction on many of the

uterine Sephadex gel filtration protein profiles. Practically

all of the uterine electrophoretic gels had a distinct hemo-

globin band not present on plasma gels. Sephadex G-200 pro-

tein profiles of uterine flushings revealed varying propor-

tions of hemoglobin which resulted in either a fourth frac-

tion beyond the albumin peak (third fraction) or a distinct

extended shoulder on the albumin peak. This variation may be

explained by relative differences in the proportion of albumin

and hemoglobin present in each sample. Plasma Sephadex G-200

protein profiles contained only three distinct fractions

(albumin being the third).

There was an obvious increase in the proportion of hemo-

globin relative to the albumin peak in Sephadex G-200 protein

profiles of uterine flushings recovered on days -1, 0, 1 and

2 as compared to the rest of the estrous cycle. This could

possibly be explained by a combination of metestrus bleeding

(Schultz et al., 1971, reported samples aspirated from the

uterus during metestrus to be consistently red and brownish-








red with blood elements present) and increased blood flow to

the reproductive tract under the influence of estrogens

(Abrams, Caton and Bazer, 1972; Abrams et al., 1973; Gwazdaus-

kas et al., 1974) around estrus. With increased blood flow to

the reproductive tract it would be expected that the quantity

of hemoglobin present due to collection procedure might

increase along with blood flow.

Progestin analyses were discontinued on plasma samples

collected immediately after successful uterine flushing

because of the possibility that stress of the uterine fluid

collection procedure might have altered plasma progestin con-

centration (Gwazdauskas, 1972). Also, after it was decided

that the uterine flushing technique had resulted in varying

degrees of blood contamination, it was evident that any

attempt to correlate progestin concentration with total uter-

ine protein recovered would be meaningless.

The levels of uterine protein in nonsurgical flushings,

the presence of hemoglobin in uterine flushings, and the

similarity between Sephadex G-200 gel filtration and poly-

acrylamide gel electrophoretic protein profiles of plasma and

uterine flushings collected in this study led to the conclu-

sion that the nonsurgical collection technique resulted in

varying degrees of blood contamination. Thus, the attempt to

characterize uterine protein secretions of the bovine using

this technique precluded any attempt to study bovine uterine

protein secretions. Any uterine protein secretions present

in the uterine flushings were obviously masked by the presence





43


of much greater quantities of plasma proteins as a result

of blood contamination.













CHAPTER 2

CHANGES IN BOVINE UTERINE LUMINAL PROTEINS
DURING THE ESTROUS CYCLE AND THEIR RELATIONSHIP
TO PLASLMA PROGESTERONE AND ESTRADIOL LEVELS



In Chapter 1, data were presented which indicate blood

contamination of uterine flushings collected nonsurgically.

These data and similar average total uterine protein recovered

by previous workers (Gupta, 1962; Heap, 1962; Schultz et al.,

1971) indicated that uterine protein samples collected by non-

surgical methods through cervical penetration do not represent

the actual intraluminal uterine protein milieu. That this is

the case was further evidenced by much less total uterine pro-

tein being recovered by the "surgical-like" postslaughter

flushing technique, described in this section, from several

heifers slaughtered on various days of the estrous cycle dur-

ing a preliminary investigation.

The following study was initiated in an attempt to accu-

rately determine quantitative and qualitative changes in the

intraluminal protein milieu of the bovine and to correlate

these changes with plasma progesterone and estradiol concen-

trations.








Materials and Methods


Sixty-seven cycling heifers and cows of mixed breeding

(Angus, Hereford, Holstein and Brahman X British crosses)

served as the experimental animals. Heifers were approximately

three years of age and cows were greater than 90 days postpar-

tum. Animals were checked visually twice daily for estrus

with the aid of Kamar heat detector patches and two penectom-

ized Angus bulls fitted with marking collars. Day of standing

estrus was designated as day 0. Cattle were randomly slaught-

ered within each breed and parity (heifers or cows) group on

days 0 to 20 of the estrous cycle. Each animal had a normal

estrous cycle prior to slaughter. Feed was withheld for 24

hours prior to slaughter. The number of animals slaughtered

on each day of the estrous cycle is presented in Table 2.

Prior to slaughter, blood samples (100 ml) were collected via

jugular venipuncture into heparinized flasks. The flasks were

immediately placed into an ice.bath, centrifuged at 10,000 rpm

for 10 minutes at 30C and plasma was removed and stored at

-20C until analyzed for progesterone and estradiol.

Immediately postslaughter, after removal of the repro-

ductive tract and recording of ovarian data, each uterine horn

was flushed with 25 ml of 0.33 M NaCl. This was completed in

less than 30 minutes post-stunning. Each uterine horn was

clamped as near as possible to the bifurcation. A small

incision was made in the oviduct approximately 1 cm above

the tubo-uterine junction and a poly-vinyl catheter (I.D.=








Table 2. Average total uterine protein recovered
and plasma progesterone and estradiol
concentration from cattle on each day of
the estrous cycle.


Av. Plasma Av. Plasma
Day of Av. Total Progesterone Estradiol
Estrous No. of Uterine Protein Concentration Concentration
Cyclea Cattle Recovered (mg)b (ng/ml)b (pg/ml)b

0 3 14.53413.02 0.53 0.41 5.63 2.32
1 3 1.48+ 0.21 1.21+ 0.95 3.30 2.00
2 3 5.02- 3.57 1.18 1.15 3.00 1.63
3 4 0.98 0.53 1.20 0.75 2.50 0.98
4 3 1.51' 1.51 2.30- 1.78 2.00+ 0.14C
5 4 2.211 2.10 8.7211.72 2.32 1.11
6 3 7.14 3.01 3.50- 1.73 7.87+ 9.48
7 3 1.62 1.23 5.27 3.15 2.77 0.95
8 3 7.23 3.74 7.00i 3.57 3.03+ 0.49
9 3 2.73 0.58 5.45 1.06c 2.05 0.07c
10 3 5.94 2.20 5.40- 3.07 4.16 3.25
11 3 4.13 0.66 6.20- 3.33 5.13 6.23
12 3 4.04+ 1.36 9.28- 5.59 2.20+ 0.44
13 4 2.88 0.75 9.85+ 3.04 2.52+ 0.40
14 3 5.28 0.68 9.60 2.31 4.05+ 2.51
15 4 3.38 0.97 8.50 3.24 3.25 2.43
16 3 6.70 2.82 9.70+ 4.69 12.30+10.77
17 3 5.79 1.53 9.10 3.91 2.60 0.61
18 3 4.01 2.99 14.70 9.79 2.30 1.73
19 3 2.68 1.12 2.30 0.60 5.31- 4.47
20 3 11.23 5.57 6.701 6.80 5.40 2.11


aDay O = estrus.
bAverage standard deviation.
Represents 2 values instead of 3.









1.25 mm) was inserted through this incision into the oviduct,

through the tubo-uterine junction and into the uterine lumen.

A glass syringe was connected to the catheter and 25 ml of

0.33 M NaCI was carefully injected into the lumen of the uter-

ine horn. The uterine horn containing the hypertonic saline

was massaged prior to the saline being withdrawn into the

syringe under slight negative pressure. Flushings from both

uterine horns were pooled and stored in an ice bath until

centrifugation at 10,000 rpm for 10 minutes at 3C. The

flushings were sterilized by filtration through a 0.20 p

Millipore filter, concentrated to 1 to 2 ml by vacuum dialysis

and dialyzed against 0.05 M phosphate-citrate buffer, pH 7.4,

for 36 to 48 hours at 30C. The samples were then stored at

-200C until further analysis.

Protein concentration of each sample was determined by

the method of Lowry et al. (1951), and the total quantity of

protein in each sample calculated. Total recovered protein

for each sample was adjusted if percent recovery of the saline

flush was less than 100% (total recovered protein divided by

percent recovery).

Sephadcx G-200 gel filtration was used to fractionate a

portion of each uterine protein sample to obtain a protein

profile. Sephadex G-200 protein profiles were also obtained

from plasma samples for comparison with the Sephadex G-200

uterine protein profiles. The columns used (1.5 x 90 cm with

a bed height of approximately 85 cm) were calibrated using

Blue Dextran (MW=2,000,000), to determine the void volume,







and aldolase (MW=158,000), ovalbumin (MW=45,000) and ribo-

nuclease A (MW=13,700) as standard globular proteins of known

molecular weight. Gel filtration was carried out at 3C.

The eluent was 0.05 M phosphate-citrate buffer, pH 7.4, with

flow rates of approximately 6 ml per hour. An aliquot of

each fraction (approximately 2.5 ml) was used for protein

determination, and the resulting optical density of each

fraction was plotted against the elution volume, at which the

fraction occurred, to construct a protein profile.

Polyacrylamide gel disc electrophoresis, using the basic

procedure of Clarke (1964), was used to further study the pro-

tein components of the uterine flushings. A 7% polyacrylamide

gel, 5 mm in diameter and 70 mm long, was used without spacer

or sample gels. Samples containing about 0.3 mg protein were

applied in 0.1 ml of 1 M sucrose and two drops of bromophenol

blue directly on the separating gel. The samples were run

in 0.05 M Tris 0.38 M glycine buffer, pH 8.0, at a constant

current of 2.5 mA per tube and a nominal voltage of 90 volts

for about 1.5 hours or until the marker dye had migrated

about 65 mm. The gels were stained with amido black solution

(1 gm per 100 ml of 7% acetic acid) for 1 hour and destined

electrophoretically in 3% acetic acid. The electrophoretic

mobility of uterine proteins was expressed in terms of their

mobility relative to that of albumin (Rf) when albumin was

assigned an Rf value of 1.0. Electrophoresis of correspond-

ing bovine plasma samples from each female was also conducted,

using the same procedure, for direct comparison with each

uterine protein polyacrylamide gel.








For analysis of steroids, approximately 2,000 cpm each

of progesterone-1, 2-3H (Amersham/Searle, 31.7 Ci/mM) and

estradiol-179-6, 7- H (New England Nuclear, 40 Ci/mM) were

added to 10 ml of plasma. These isotopic steroids served as

internal standards for correction for procedural losses. The

mixture was extracted vigorously three times with two volumes

of freshly distilled diethyl ether. Progesterone and estra-

diol in the ether extract were isolated by chromatography on

1 x 40 cm Sephadex LH 20 columns, as described by Chenault

(1973), and stored at 4C until assayed. The solvent elution

system was chloroform:ethanol (96:4). Aliquots of isolated

progesterone were quantified by competitive protein binding

assay (Murphy, 1967). Details of the procedure were described

by Gwazdauskas (1972) and Chow (1972). The isolated estradiol

was quantified by radioimmunoassay (Hotchkiss, Atkinson and

Knobil, 1971) using an antiserum1 produced by immunization of

sheep with the conjugate 1, 3, 5 (10) estratriene-3, 170-diol,

178-succinyl-bovine serum albumin. Details of the procedure

were described by Chenault (1973).

Least-squares procedures (Harvey, 1960) were used to

evaluate the effects of day of estrous cycle, breed and

parity on the variables measured (total uterine protein and

plasma progesterone and estradiol concentration) for days 0

to 20, 0 to 3 and 4 to 18 of the estrous cycle. Regression

curves were calculated and plotted using the individual


The estradiol antiserum was generously supplied by Drs.
T. Nett and L. Estergreen of Washington State University,
Pullman.









observations for total uterine protein recovered and for plasma

progesterone concentration on days 0 to 20 of the estrous cycle.

Correlations were determined for progesterone and estradiol

concentration with total uterine protein and with the presence

of a Rf 0.35 uterine protein band following electrophoresis.



Results and Discussion


Volumes of uterine flushings recovered did not vary sig-

nificantly due to day of the estrous cycle. An average of

97.2 9.7% of the volume of saline introduced into the uter-

ine lumen was recovered (Table 1). This was comparable to

the surgical recovery of uterine flushings in gilts by Murray

et al. (1972a) who used a similar flushing technique in anes-

thetized pigs.

Concentrated uterine flushings obtained during proestrus,

estrus and early metestrus had a faint yellow color, but

samples collected during diestrus were clear. The yellow

color was thought to be due to albumin and/or a low concen-

tration of hemoglobin in the sample as would be expected dur-

ing metestrous bleeding. Schultz et al. (1971) reported uter-

ine samples during metestrus as being consistently red and

brownish-red with blood elements present.

Average total uterine protein recovered with the "surgi-

cal-like" postslaughter flushing technique used in this study

(Table 1) was 4.64 4.35 mg. This compares favorably with

the 7.5 mg average total uterine protein recovered by Roberts









and Parker (1974a) with a similar postslaughter flushing

technique during the first 2 to 3 weeks of pregnancy.

Average total uterine protein recovered and average

plasma progesterone and estradiol concentrations from the

cattle on each day of the estrous cycle are presented in

Table 2. During days 0 to 20 of the estrous cycle a signifi-

cant (P<.025) difference was found in total uterine protein

recovered (Table 3). The differences in plasma progesterone

concentration during days 0 to 20 approached significance

(P<.10), but there were no significant differences in plasma

estradiol concentration during any period of the estrous

cycle (Tables 3, 4 and 5). Flushings from cows yielded sig-

nificantly (P<.05) more total uterine protein than those from

heifers on days 4 to 18 of the estrous cycle as shown by the

significant parity effect in Table 4. During days 0 to 3 of

the estrous cycle there was a significant (P<.025) breed

effect on plasma progesterone concentration.

Figure 1 shows the total uterine protein averages on

days 0 to 20 of the estrous cycle. Also, the fifth-order

regression curve best describing the individual observations

is shown. Total uterine protein tended to be higher during

the middle and late luteal phase than during the early luteal

phase of the estrous cycle. Least-squares analysis of data

obtained between days 4 and 18 of the estrous cycle (Table 4)

revealed that differences in total uterine protein recovered

during this period approached significance (P<.10).




















4:

l 4 K:


(D 0 C: t r-<

a3 34 a






d- -o o o-
a a >








C)CO .

CO



C) r4 I c o

0 0 m -
0 C C) o No )



4 -14,1 0 mo -I
"CJ P < 4 4I
'o 11 c_
ti W O











00
0) 4 -4 4:
0 l



C) U C iU N- N-








4c cK




*U c





cJ 0 0. ) v v
-' 0 V



U > 0 n

0 1 0 IQ
(1) n .

a, 0, C )




LT) 3 !- rt
Q m a cr



























4a
-d





0



+1 U
a >,
Su



tn

U ^
4-


00






0r V)


*lH
4-C)
0 +








-d 4
a)
00
CU)






1- 0









o
0.









C) -a




1C1
44




CCl
H


V V V
C ). i..



v v v



*K *K


Ln oC

C) 'H
*- O









0 -
ss
4 41


0 a44
CT -4












o o







bC U
OCC





00






DaC.
4-1 f +-


J- 44 31
C-]








Table 5. Least-squares analysis of variance:
data obtained between days 0 and 3
of the estrous cycle.


Mean Squares
Total Plasma Plasma
Uterine Progesterone Estradiol
Source D.F. Protein Concentration Concentration

Days 3 94.7924 0.3010 3.7530

Breed 3 6.5862 1.7986** 1.2614

Parity 1 3.3065 0.4474 0.6401

Remainder 5 69.0571 0.1931 4.3072


**P < .025.














V)
CC




Ca).



.CC) -1 -

CD .,i
-4 >


o C)

Hn 0
Or0C



-1 c






C)u tr,
LO

o rl






Z ) 0
(1C) 4C
O' (OH




4) r4








4- 4-J
C0 C

)4-' C)
qo U
















L) r_
CC)O









Q L4
L H



C)
H 0)
4-'u
o, F-
4-' C
ar Ur







00 0
Ci C 44






C)h
CC


t~co






Cs-.













o- ,


.01


- -


-,O
-0









OD







-U)

0



w
LL.


CD
(0g


(0 c- N a, fj

(OVY) N13108d -l~i0i


vf C\J









Total uterine protein recovered was highest on days 0

and 20 of the estrous cycle. The Sephadex G-200 protein pro-

files and polyacrylamide electrophoretic gels of uterine

samples with low corresponding plasma progesterone concen-

trations on these two days closely resembled those of corre-

sponding plasma. Two cows had apparent functional luteal

tissue on day 20 as evidenced by gross appearance of corpus

luteum and plasma progesterone concentration. The uterine

fluid Sephadex G-200 protein profiles and polyacrylamide

electrophoretic gels of these two cows did not resemble those

of corresponding plasma. These day 0 and 20 quantitative and

qualitative observations could possibly be explained by in-

creased blood flow to the reproductive tract under the influ-

ence of elevated estradiol concentration (Abrams et al., 1972;

Abrams et al., 1973; Gwazdauskas et al. 1974) if there was

any movement of plasma proteins by diapedesis into the uterus.

The .73 (P<.01) correlation found between plasma estradiol

concentration and total uterine protein recovered on days 0

to 3 (Table 6) also supports this explanation of the similar-

ity between uterine fluid and plasma proteins on day 0 of the

estrous cycle.

Figure 2 shows the plasma progesterone concentration

averages on days 0 to 20 of the estrous cycle. In spite of

variable progesterone concentrations among females within

days, the fourth-order regression curve shown in this figure

which best describes the individual observations is similar

to those reported by other workers









Table 6. Correlation of plasma progesterone and
estradiol concentrations with total
uterine protein and presence of Rf 0.35
uterine protein band following electro-
phoresis.


Correlation Correlation with
Days of with Total Presence of
Estrous Uterine Rf 0.35 Uterine
Hormone Cycle D.F. Protein (r)a Protein Band (r)


Progesterone


Estradiol


0-20
0-3
4-18


0-20
0-3
4-18


64 .08 (.23*)
12 -.26 (-.53)
45 .21 (.44")


64 .26* (.00)
12 .73** (.55)
45 .26 (-.16)


.46** (.27*)
_b

.35* (.20)


.17 (.00)
-_b

.17 (.03)


aFigures in parentheses are the residual correlation
coefficients after correction by least-squares analysis
for breed, parity and day effects.
Rf 0.35 uterine protein band not present on days 0 to 3.

*P < .05.
**P < .01.
























0 -H


0 H i


- It rH
003
In C








01*C

C) 4) C)



o NU
00
crC 0
u 0



COO



400


Oct
04-









0 +*.
Ca) 4 C
oos-

oD .c ,U
04 0
- 'l U)



















NC 4-)
0
C C) 0,



4-i 4 C o

W n n

ccM
^ !<






















































(o N OJ


C_


0











I,
-<






--



-C,

\ 0
V\H



\ a
0 o LL






(9-/N NOI - -*




\ N




o c6 (D It


(1CV/9N) 3NO831S390dd









(Stabenfeldt, Ewing and McDonald, 1969; Shemesh, Lindner and

Ayalon, 1971; Wettemann et al., 1972). The high mean plasma

progesterone concentration on day 20 was due to inclusion of

samples from two cows having apparently functional luteal

tissue (based on ovarian examination at slaughter and corre-

sponding low plasma estradiol and high plasma progesterone

concentrations).

The plasma estradiol curve of average concentration on

days 0 to 20 of the estrous cycle is shown in Figure 3. As

previously mentioned, there were no significant day effects

on estradiol concentration during any period of the estrous

cycle. Least-squares analysis of variance did not indicate

significant breed or parity effects on estradiol concentration.

Variation in estradiol concentration between females (as evi-

denced by large standard deviations) could account for the

dissimilarity between the estradiol curve in Figure 3 and

that reported by Wettemann et al. (1972). Their cattle were

sampled each day over the entire estrous cycle, whereas,

because of slaughter, in the present study each observation

was from a different animal.

After correction for breed, parity and day effects by

least-squares analysis on days 0 to 20 of the estrous cycle,

a significant (P<.05) correlation (r=.23) was found between

plasma progesterone concentration and total uterine protein

(Table 6). On days 4 to 18 the correlation was .44 (P<.05),

indicating a greater effect of progesterone on total uterine

protein recovered during the luteal phase of the estrous




















C)




0
oi





0
4N CS






U)0







4-J
.H






S"-
OC


















U
0 L0
a>



























0
CJ












r) U



o
vi














o m
rd
c> e
0ct








ci













C

C

-'-4
* *



























0







LL
(D
LO


O0 (C D o C


1- 0 D oj


(7-1A/9d) 7101GV1lS3








cycle. These correlations between progesterone concentration

and total uterine protein suggest that progesterone may be

causing the quantitative changes in bovine uterine protein

secretion during the luteal phase of the estrous cycle as

reported for pigs (Knight, 1972; Knight et al., 1973b). A sig-

nificant correlation was not found between estradiol concen-

tration and total uterine protein after correction for breed,

parity and day effects by least-squares analysis. Before

correction, however, the correlation between estradiol con-

centration and total uterine protein was .26 (P<.05) for

days 0 to 20 and .73 (P<.01) for days 0 to 3 of the estrous

cycle. As previously mentioned, uterine protein samples

qualitatively resembled those of corresponding plasma on day

0. Therefore, this .73 correlation between plasma estradiol

concentration and total uterine protein on days 0 to 3 would

be expected if, in fact, the uterine samples collected around

estrus consisted primarily of plasma proteins moving into the

uterine lumen by diapedesis under the influence of an estrogen-

induc6d increased blood flow at estrus.

Electrophoretic data in this study indicated qualitative

changes in the proteins of the uterus during the estrous cycle.

Table 7 gives the number and percent of polyacrylamide gels

of uterine protein containing electrophoretic bands not pres-

ent in corresponding plasma gels. Figure 4 gives a repre-

sentative sample of polyacrylamide gel electrophoresis toward

the anode of uterine protein collected on days 0 to 20 of the

estrous cycle and of plasma from each third of the estrous








Table 7. Number and percent of polyacrylamide gels
of uterine protein containing electro-
phoretic bands not present in correspond-
ing plasma gels.


Day of
Estrous No. of
Cycle Samples


Rfa 0.35 Rfa 0.72 Rfa 1.09 Rfa 1.12
() (%) ( ) (0%)


(25)
(67)
(ioo)
(100)
(67)
(67)

(67)


(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)
(100)


0
1 (33)
0
0
1 (33)
0
2 (67)
0
2 (67)
1 (33)
1 (33)
2 (67)
1 (33)
0
2 (67)
3 (75)
3 (100)
1 (33)
0
3 (100)
1 (33)


2 (67)
2 (67)
0
0
3 (100)
1 (25)
1 (33)
1 (33)
2 (67)
2 (67)
3 (100)
2 (67)
2 (67)
2 (50)
2 (67)
0
1 (33)
3 (100)
0
2 (67)
0


aRf of albumin = 1.0.


~~I ~C_~_




























Polyacrylamide gel electrophoresis toward the
anode (+) of bovine uterine protein and plasma
(P) collected on days 0 to 20 of the estrous
cycle. Two uterine proteins not appearing in
plasma are indicated by the arrowts (i-) Albumin
(A) and transferring (T) are also indicated.


Figure 4.




1 2 3 4 5 6 P


I *
|i .^ >-e


9 10 11 1' 13



I ; *
tlMPjI


p1


14 15 16 17 18 19 20 P


L I 'now I


e b



4,


0
4






Ilil8


1
I


L









cycle. Sixteen (62%) polyacrylamide gels of uterine protein

samples obtained between days 13 and 20 of the estrous cycle

had an electrophorctic protein band (Rf=0.35) which migrated

just behind the transferring (Figure 4). A similar protein

was not present in corresponding plasma gels. This band was

present in all gels of samples collected on days 15 and 16.

Two uterine samples (both from cows having functional luteal

tissue) on day 20 yielded this Rf 0.35 electrophoretic pro-

tein band. The corresponding plasma progesterone concentra-

tions from both cows were still elevated. Neither the day 19

samples nor the remaining day 20 uterine sample yielded this

band; however, they had low corresponding progesterone concen-

trations. There was a highly significant (P<.01) difference

in the presence of this protein band from days 0 to 20 of the

estrous cycle (Table 3), and a correlation of .46 (P<.01) was

found between plasma progesterone concentration and its pres-

ence (Table 6). Based on its Sephadex G-200 gel filtration

elution volume elutedd just prior to albumin), the molecular

weight of the uterine protein having an Rf of 0.35 was esti-

mated to be approximately 100,000. The above correlation

between progesterone concentration and the presence of the

Rf 0.35 protein suggests that progesterone may also be caus-

ing the qualitative changes in bovine uterine protein secre-

tion during the luteal phase of the estrous cycle as reported

for pigs (Knight, 1972; Knight et al., 1973b).

A second uterine protein electrophoretic band (Rf=0.72)

migrated between the transferring and albumin (Figure 4) and









did not appear in corresponding plasma gels. This protein

band was found in all uterine protein samples collected from

days 0 to 20 of the estrous cycle. The electrophoretic

mobility (Rf) of the protein represented by this band was

similar to that reported for the rabbit uterine specific pro-

tein "blastokinin" (Bullock and Connell, 1973). Also, the

electrophorotic band(s) representing hemoglobin found in the

first study (Chapter 1) migrated to approximately the same

position, i.e., between transferring and albumin, as did this

Rf 0.72 band.

One or two prealbumin protein bands (Rf=1.09 and 1.12)

were present in 39 (58%) of the uterine protein gels between

days 0 and 20 but not in plasma gels. An Rf 1.12 band is

clearly visible on the day 11 gel just in front of the albumin

band at the bottom edge of the photograph in Figure 4. Most

of the prealbumin electrophoretic bands are not visible in

the photographs or else appeared below the photographic field.

There was also a prealbumin band (Rf=1.27) present in all

uterine protein gels and some plasma gels (very faint) that

migrated with the marker dye used in the electrophoretic pro-

cedure. It was just as prominent in polyacrylamide gels of

nonsurgical uterine flushings in the first study. This Rf

1.27 band appeared at the bottom of the gels and, therefore,

is not visible in Figure 4.

For each major protein band on polyacrylamide gels of

day 70 bovine allantoic fluid (F. W. Bazer and W. W. Thatcher,

unpublished data) there was a similar protein band present in








gels of uterine protein samples obtained in this study. The

bands included were the Rf 0.35 protein, transferring, Rf 0.72,

albumin (Rf=1.0) and a band that migrated between the Rf 0.72

band and albumin. In some uterine protein and plasma gels

there were two bands instead of one visible in the area

between the Rf 0.72 band and the albumin band. The poly-

acrylamide gels of bovine day 70 allantoic fluid were also

similar to day 35 porcine allantoic fluid (F. W. Bazer and

W. W. Thatcher, unpublished data). There were no prealbumin

bands present on polyacrylamide gels of bovine or porcine

allantoic fluid which had not been processed, i.e., lyophilized

or vacuum dialyzed.

No significant differences in the size classes of pro-

teins present between days 0 and 20 of the estrous cycle were

found in Sephadex G-200 gel filtration uterine protein pro-

files. There was, however, some slight trailing of two low

molecular weight protein fractions which were never present

in plasma samples. These two "low-profile" fractions were

estimated to be in the 10,000 to 15,000 and <10,000 molecular

weight range. Figure 5 shows typical protein profiles follow-

ing Sephadex G-200 gel filtration of uterine samples collected

on days 0, 5 and 15 of the estrous cycle and of bovine plasma.

The Rf 0.35 protein was eluted by Sephadex G-200 gel filtra-

tion just ahead of Fraction III (predominantly albumin, 75 to

95 ml) and probably contributed to the indistinct separation

between Fractions II and III (75 to 85 ml) on the day 15 pro-

tein profile in Figure 5. Based on polyacrylamide gels of






























Figure 5. Typical Sephadex G-200 gel filtration protein
profiles of bovine uterine protein collected
on days 0, 5 and 15 of the estrous cycle and
bovine plasma.







10






50 60 70 80 90 100 110 120 130 140 150
z



- 5- DAY-5

1
0 '
I- 50 60 70 80 90 100 110 120 130 140 150
LL
I-
o DAY- 0
wI 5-



10-



5- r PLASMA


ELUTION VOLUME (ML)


50 60 70 80 90 100 110 120 130 140 150









concentrated (vacuum dialysis) pooled Sephadex G-200 fractions

of the uterine protein samples, the Rf 0.72 protein apparently

contributed predominantly to Fraction IV (95 to 125 ml portion

of the uterine protein profiles in Figure 5), and the pre-

albumin electrophoretic proteins (Rf=1.09, 1.12 and 1.27)

apparently contributed predominantly to Fraction V (125 to

150 ml portion of the uterine protein profiles in Figure 5).

On some polyacrylamide gels of these pooled Sephadex G-200

fractions there iere two electrophoretic bands in the Rf 0.72

area. In the first study (Chapter 1) hemoglobin occurred on

the polyacrylamide gels as either one or two bands in this

same area (some cattle in this study apparently had two

molecular species of hemoglobin). This observation could be

an indication that the Rf 0.72 protein is hemoglobin or, at

least, that in some samples in this study there was some

hemoglobin contamination. It was noted that if the Rf 0.72

protein was not hemoglobin it would be very difficult to

separate the two by either Sephadex G-200 gel filtration or

by polyacrylamide gel electrophoresis. Beier and Beier-

Hellwig (1973) draw attention to the difficulty of differen-

tiating between a specific postalbumin fraction and hemo-

globin artifacts using electrophoretic techniques.

Recently F. W. Bazer and W. W. Thatcher (unpublished

data) collected bovine uterine fluid on days 5 and 15 of the

estrous cycle using the same procedure described in this chap-

ter, but the samples were concentrated by freeze-drying

instead of vacuum dialysis. Sephadex G-200 gel filtration








of these freeze-dried samples and day 35 bovine allantoic

fluid yielded protein profiles which were quite similar to

each other, but these profiles had considerably larger Frac-

tion V peaks than did the Sephadex G-200 protein profiles of

uterine flushing obtained in this study. This fraction is

comparable to the "low-profile" Fraction V (125 to 150 ml)

of this study and they both were eluted near the total volume

of the column. Mien this fraction was reconcentrated and

subjected to polyacrylamide gel electrophoresis the resulting

gels apparently consisted predominantly of a prealbumin band.

This fraction was yellow in color and the major portion of

the protein was lost upon dialysis, thus indicating a very

small protein molecule (<10,000 MW). The question is immedi-

ately raised as to the accuracy of the evaluation of quanti-

tative and qualitative changes in the actual protein milieu

of the uterine lumen in the present study because of the

possibility that low molecular weight proteins may have been

lost due to concentration by vacuum dialysis and subsequent

dialysis using dialysis tubing that allowed loss of <10,000

MW protein molecules.

Total uterine protein from samples concentrated by freeze-

drying (F. W. Bazer and W. W. Thatcher, unpublished data;

Roberts and Parker, 1974a) was not different from total uter-

ine protein from samples in this study concentrated by vacuum

dialysis. Roberts and Parker (1974a) found that differences

between fresh, untreated serum and uterine washings largely

disappeared when the uterine washings were compared with









dialyzed and freeze-dried serum. Thus, it is possible that

the Fraction V protein(s) discussed above may be the result

of some change in a uterine proteins) that is brought about

by concentration, with freeze-drying causing a greater change

than vacuum dialysis. The above observations also raise the

question of whether or not differences between untreated

plasma and uterine flushings found in this study would have

been the same if the plasma samples had been diluted in 0.33 M

saline and processed the same as the uterine flushing samples.

Answers to these questions will come after additional research

determines the actual make-up of Fraction V; i.e., whether it

is composed of an original uterine protein(s) or whether it

is made up of altered uterine protein due to changes in the

original proteins) brought about by processing of uterine

flushJ ngs.

This study indicated that both qualitative and quantita-

tive aspects of bovine uterine protein secretion changed dur-

ing the luteal phase of the estrous cycle. Correlation of

total uterine protein and presence of the Rf 0.35 protein in

uterine flushings with plasma progesterone concentration

suggested that progesterone is inducing these quantitative

and qualitative changes during the luteal phase of the estrous

cycle.

The Rf 0.35 protein was present in the bovine uterus in

this study at a time which corresponds with rapid expansion

of the blastocyst and specifically elongation of the tropho-

blast (Winters, Green and Comstock, 1942; Chang, 1952). In








the pig, trophoblast growth and development has been associ-

ated with a purple uterine protein which increases quanti-

tatively during the latter half of the luteal phase of the

porcine estrous cycle (between days 12 and 16) and which

appears in allantoic fluid by day 30 of pregnancy (Murray,

1971; Murray et al., 1972a; Chen et a]., 1973a; Schlosnagle

et al., 1974; Chen et al., 1975). Thus, it is possible that

the Rf 0.35 protein could be involved in rapid expansion of

the blastocyst that begins to occur about day 12 of pregnancy

in the bovine, and the subsequent rapid elongation of the

trophoblast, which occurs at a time corresponding to the

latter one-third of the normal estrous cycle.

It is also possible that the Rf 0.35 protein may in some

way be associated with the luteolytic process as it was pres-

ent prior to corpus luteum regression when progesterone con-

centration was high and was absent after corpus luteum regres-

sion when progesterone concentration was low. A possible

function, if the protein is associated with CL regression,

couldbe binding of PGF2a~ which is luteolytic in the bovine

(Lauderdale, 1972; Rowson et al., 1972; Hansel et al., 1973;

Inskeep, 1973; Chenault, 1973; Lauderdale et al., 1974), and

transfer of PGF2 into the chorioallantoic membranes and/or

allantoic fluid. The Rf 0.35 protein is apparently present

in day 35 bovine allantoic fluid (F. W. Bazer and W. W.

Thatcher, unpublished data). Thus, if an embryo is present

in the uterus, sufficient PGF2U could be transferred into the

chorioallantoic membranes and/or allantoic fluid to prevent








luteolysis. It is interesting to note that Knight (1972)

suggests the possibility of luetolytic prostaglandin being

attached to an electrophoretic component of porcine uterine

flushing Fraction II (Murray, 1971), changing its mobility.

He noted the presence of an electrophoretic protein band

(Rf=0.53) on day 13 in uterine flushings of progesterone

treated ovariectomized gilts that was not present on days 7,

9 or 11. This component of Fraction II was shown to be main-

tained as long as progesterone treatment was continued. The

appearance of this band occurred when luteolytic activity

would be expected to occur (Schomberg, 1967), and it was

present in a fraction reported to have luteolytic activity

(Barber, 1972) with a similar estimated molecular weight to

the luteolytic component of porcine uterine fluid reported

by Schomberg (1969).

The Rf 0.35 protein could also be associated with a pos-

sible adherence of the expanding blastocyst to the uterine

wall during trophoblast expansion. It is logical to assume

that a certain amount of adhesion must occur between the

trophoblast and the uterine wall to permit the extensive

elongation of the trophoblast through the uterine horn con-

taining the blastocyst by about day 18 and into the adjacent

horn by about day 20 (Winters et al., 1942; Chang, 1952).

This speculation as to function of the Rf 0.35 protein found

in this study is supported by the work of Mintz (1970) and

Pinsker and Mintz (1973). They found a factor (estrogen

dependent) in mice which is believed to be a proteolytic









enzyme that induces blastocysts to adhere to the uterine wall

prior to implantation. In the rabbit, protease activity has

been found in the B-glyceprotein fraction of uterine secre-

tion 24 hours before implantation, and a relation has been

assumed between uterine protease (B-glycoprotein) and the

implantation of the blastocyst (Kirchner ct al., 1971). The

rabbit uterine specific B-glycoprotein described by Beier

(1974b) is a protease (MW=100=,000; similar electrophoretic

mobility to Rf 0.35 protein in this study), is present in

blastocyst fluid around implantation and is associated with

the outer surface of the trophoblast. Thus, with similar

electrophoretic mobility and molecular weight linking the

rabbit uterine specific h-glycoprotein to the Rf 0.35 uter-

ine protein of this study, this speculation as to function

of the Rf 0.35 protein is offered. Rapid expansion of the

rabbit trophoblast prior to and during implantation can also

be compared to rapid expansion of the bovine trophoblast

coincident with the presence of the Rf 0.35 uterine protein

during the latter half of the luteal phase of the estrous

cycle.













GENERAL DISCUSSION


Considerable research has established the importance of

the uterus and its secretions in embryonic development (Beier,

1974a,b). Sufficient data have accumulated on restriction of

embryos to the oviduct, in vitro culture, synchrony of

embryonic and uterine development and delayed implantation

to support the concept that some uterine components) is

extremely important for continued normal development of the

embryo past the early blastocyst stage.

Rabbit uterine fluid contains considerable protein. The

volume of this fluid, total concentration of its macromolecu-

lar components and its protein patterns change continuously

from ovulation to implantation (Beier, 1974a,b). Porcine

uterine protein secretions change qualitatively and quanti-

tatively during the estrous cycle (Murray, 1971). These

rabbit and porcine uterine proteins (predominately glycopro-

teins of less than 50,000 MW) are a product of selective

filtration from plasma proteins and of biosynthesis by epi-

thelial cells of uterine endometrial glands and endometrial

surface epithelium (Beier, 1974a,b; Chen et al., 1975).

Rabbit blastocyst fluid contains the same proteins as those

in uterine fluid (Beier, 1974b). By day 30 of pregnancy,

porcine allantoic fluid contains a purple protein (Chen et al.,








1973a,b) specific to the uterine protein milieu and poly-

acrylamide gels of porcine day 35 allantoic fluid are very

similar to those of day 15 uterine fluid (F. IW. Bazer, unpub-

lished data). Thus, it would appear that uterine proteins of

the rabbit and pig are transferred from the uterine lumen

into the fluids and membranes of the early embryo. Secretion

of these uterine proteins has been shown to be regulated by

steroids, specifically progesterone in the rabbit and pig

(Urzua et al., 1970; Knight et al., 1973b; Goswa:ii and Foigel

son, 1974).

Early work with cattle has supported the concept that

bovine uterine fluid is the result of active secretion and

not merely a product of simple diffusion frre blood (Schultz

et al., 1971). Variation in its chemical composition with

stage of estrous cycle (highest levels reported during the

luteal phase) suggested hormonal control of bovine uterine

fluid composition.

In this study the first approach was to characterize

uterine protein secretions of the bovine using a nonsurgical

technique; however, blood contamination of the uterine fluid

samples was indicated. Consequently, any uterine protein

secretions that might have been present in the samples were

probably masked by contaminating plasma proteins. Average

total uterine protein collected in this first study (46.93

41.31 mg) was similar to average total uterine protein

recovered by earlier workers (Gupta, 1962; Heap, 1962;

Schultz et al., 1971). Furthermore, the qualitative








indications of blood contamination indicated that uterine

protein samples collected by this nonsurgical method did not

represent the actual intraluminal uterine protein milieu.

This was further indicated by the lower total uterine protein

recovered by the "surgical-like" postslaughter flushing tech-

nique used in the second half of this study (4.64 4.35 mg).

This compares favorably with the 7.5 mg average total uterine

protein collected during the first 2 to 3 weeks of pregnancy

by Roberts and Parker (1974a) who used a similar postslaughter

flushing technique.

In the second half of this study (Chapter 2) total uter-

ine protein recovered was highest on days 0 and 20 of the

estrous cycle and daily variation in recoverable protein was

significant (P<.025). A correlation coefficient of .73

between estradiol and total uterine protein during estrus and

metestrus (day 0 to 3), the similarity between electrophore-

tic patterns of polyacrylamide gels and Sephadex G-200 protein

profiles of uterine protein and plasma collected on day 0 and

the fact that highest total uterine protein collected was on

day 0 indicated the possible movement of blood proteins into

the uterus by diapedesis at this time when blood flow to the

uterus is elevated as a result of elevated estrogen levels.

During the luteal phase of the estrous cycle, total uterine

protein tended to be higher during the middle and late luteal

phase than during the early luteal phase. Significant (P<.05)

correlation coefficients between plasma progesterone concen-

tration and total uterine protein of .23, for days 0 to 20,








and .44,for days 4 to 18 of the estrous cycle, suggest that

progesterone may be influencing the quantitative change in

bovine uterine protein secretion during the luteal phase.

This would agree with data reported for pigs (Knight et al.,

1973b).

Electrophoretic patterns on polyacrylamide gels and

Sephadex G-200 protein profiles of uterine protein and plasma

collected on days 0 to 20 of the bovine estrous cycle indi-

cated qualitative changes in the uterine protein milieu.

There was a highly significant (P<.01) difference in the pres-

ence of one uterine protein having an Rf of 0.35 from days 0

to 20 of the estrous cycle. This protein was present only

between days 13 and 20 (present in all samples on days 15 and

16), but was not present in corresponding plasma. Its pres-

ence was correlated (r=.46) with plasma progesterone concen-

tration (P<.01). This correlation suggested that progester-

one may be influencing the qualitative changes in bovine uter-

ine protein secretions during the luteal phase of the estrous

cycle. Again, similar observations have been made for pigs

(Knight et al., 1973b). Another protein present in bovine

uterine samples, but not in corresponding plasma, had an Rf

of 0.72 and was present throughout the estrous cycle. Elec-

trophoretic mobility of this protein was similar to that

reported for the rabbit uterine specific protein "blastokinin"

(Bullock and Connell, 1973). Laster (1974) reported the

absence of a protein having an Rf of 0.76 in uterine endo-

metrium of most nonpregnant heifers on days 3 and 12 of the









estrous cycle as opposed to its presence in pregnant heifers.

There were also one or two prealbumin protein bands present

on many of the electrophoretic gels of uterine samples, but

not plasma, between days 0 and 20 of the estrous cycle that

compare with prealbumin bands reported in the rabbit (Beier,

1974a,b). However, data were presented and discussed in

Chapter 2 which suggested that these prealbumin electropho-

retic bands might be the result of processing procedures,

i.e., concentration by lyophilization or vacuum dialysis.

This study has indicated that both qualitative and quan-

titative aspects of bovine uterine protein secretion change

during the luteal phase of the estrous cycle. Correlation of

total uterine protein and presence of the Rf 0.35 protein in

uterine flushing with plas~ii progesteloe concentration

suggests that progesterone is inducing these qualitative and

quantitative changes during the luteal phase of the estrous

cycle.

Porcine uterine protein secretions were first suggested

to be related to growth and expansion of the trophoblast by

Murray (1971). Subsequent work by Chen (1973) indicated that

porcine uterine proteins were involved in some aspect of pla-

cental development. More recent work has associated tropho-

blast growth and development with a purple, uterine specific,

protein which increases quantitatively during the latter half

of the-luteal phase of the porcine estrous cycle (between

days 12 and 16) and appears in allantoic fluid by day 30 of








pregnancy (Chen et al., 1973a; Schlosnagle et al., 1974;

Chen et al., 1975).

Knight et al. (1974)) reported that increased uterine

protein secretions, resulting from high progesterone levels,

enhanced placental development and allantoic fluid volume.

In more recent work, Knight (1975) concluded that development

of adequate placental mass was apparently the key factor

necessary for adequate and sustained fetal growth and develop-

ment. In cattle, Bellows et al. (1974) reported that twin

fetuses were associated with lower cotyledon weights than

singles and suggested that the ability to exchange metabolites

between dam and fetus could affect birth weight or become a

limiting factor in maintenance of multiple pregnancies.

The Rf 0.35 protein of this study was present in the

bovine uterus at a time which corresponds with rapid expansion

of the blastocyst and, specifically, elongation of the tropho-

blast (W'inters et al., 1942; Chang, 1952). Therefore, it is

possible that the Rf 0.35 protein could be involved with

trophoblast growth and development as is the purple uterine

specific protein in pigs.

Data were presented and discussed at the end of Chapter 2

which provide a basis for speculation relative to two other

possible functions or roles for the Rf 0.35 uterine protein.

These were (1) a possible association with the luteolytic

process and (2) a possible adherence of the expanding blasto-

cyst to the uterine wall during trophoblast expansion.








Embryo transfer in the cow (Rowson et al., 1969) has

shown that a uterus which has not contained an embryo for

approximately the first half of the estrous cycle will sup-

port growth and development of a transferred blastocyst.

Thus, the environment supplied by the normal nonpregnlant

uterus must be similar to that supplied by the pregnant

uterus. In light of this and the relation between total

uterine protein and the Rf 0.35 protein and progesterone con-

centration, it is not surprising that Boyd et al. (1969)

found higher plasma progesterone levels in pregnant than in

nonpregnant cows on day 16. They also reported a direct rela-

tionship between blastocyst length of day 16 embryos and

plasma progesterone concentration. This supports the possi-

bility of an association between bovine uterine protein secre-

tions, regulated by progesterone, and trophoblast growth and

development.

The importance of synchrony between uterine protein

secretions and embryonic growth and development may help to

explain the poor viability of embryos resulting from super-

ovulation of the cow. Use of PMSG in the cow and pig results

in very high levels of estrogen before estrus and a delayed

decline to normal levels after estrus (Lamond and Gaddy, 1972;

Spilman et al., 1973; Henricks et al., 1973; Guthrie, Ilenricks

and Handlin, 1974). Estrogen treatment performed shortly

after mating in the rabbit caused retardation of endometrial

proliferation and secretion to such an extent that the normal

pattern of protein secretion was delayed 3 to 4 days and








fertility was nil (Bcier, 1974b). High estrogen levels in

superovulated cows could be causing a delay in normal uterine

protein secretion patterns which would result in asynchrony

of the uterus. (behind) and embryo.

Additional research on uterine protein secretions of the

cow may shed light on reasons for early embryonic death loss.

Also, the possibility of a delay in uterine protein secretion

of superovulated cows may be determined. Such information

could lead to more successful embryo transfer techniques. The

identification of uterine specific proteins in the bovine may

lead to development of better embryo culture media for short-

term storage of embryos. The overall effect of future research

on bovine uterine secretions will, hopefully, result in a

better understanding of those factors which affect develop-

ment of the bovine concepts. This information is essential

if improvements in animal fertility are to be realized.














SUMMARY


During the course of this study bovine uterine flushings

were collected throughout the estrous cycle and examined for

changes in total recoverable protein, Sephadox G-200 gel fil-

tration protein profiles and electrophoretic protein patterns.

An attempt was made to correlate plasma progesterone and

estradiol concentrations with qualitative and quantitative

changes in uterine protein secretions.

In a preliminary study approximately seven uterine flush-

ing samples per day of the estrous cycle were collected non-

surgically (total = 144) using a urethral catheter inserted

through the cervix. An average of 84 20% of saline put

into the uterus was recovered and the samples appeared to be

relatively free of blood contamination during most of the

estrous cycle. However, Sephadex G-200 gel filtration protein

profiles and clectrophoretic polyacrylanide gels of the uter-

ine flushings were very similar to those of plasma collected

at the same time. Blood contamination of the uterine fluid

samples was indicated and it was thought that any uterine

protein secretions that might have been present in the samples

had been masked by contaminating plasma proteins. Therefore,

in a second study uterine flushing were obtained from heifers

and cows immediately following slaughter.








In the second study, preslaughter plasma and immediate

postslaughter 0.33 M saline uterine flushings were collected

from 67 cattle of mixed breeding (Angus, Hereford, Holstein

and Brahman X British crosses). Three samples were collected

for each day of the estrous cycle and 4 samples each were

obtained on days 3, 5, 13 and 15. Plasma samples were assayed

for progesterone (competitive protein binding procedure) and

estradiol (radioimmunoassay). Total uterine protein was

determined by Lowry's method. Qualitative polyacrylamide gel

electrophoresis and Sephadex G-200 gel filtration protein pro-

files were obtained for each sample.

During days 0 to 20 of the estrous cycle a significant

(P<.025) difference was found in total uterine protein pres-

ent in uterine flushing. Flushings from cows yielded sig-

nificantly (P<.05) more total uterine protein than those from

heifers on days 4 to 18 of the estrous cycle. After correc-

tion for breed, parity and day effects by least-squares analy-

sis, on days 0 to 20 of the estrous cycle a significant

(P<.05) correlation (r=.23) was found between peripheral

plasma progesterone concentration and total uterine protein.

On days 4 to 18 of the estrous cycle the correlation was .44

(P<.05), indicating a greater effect of progesterone on total

uterine protein recovered during the luteal phase of the

estrous cycle. No significant correlation was found between

estradiol concentration and total uterine protein after cor-

rection for breed, parity and day effects by least-squares

analysis. Before correction, the correlation between








estradiol concentration and total uterine protein was .26

(P<.05) for days 0 to 20 and .73 (P<.01) for days 0 to 3 of

the estrous cycle. The .73 correlation between estradiol and

total uterine protein during estrus and metestrus, the simi-

larity between electrophoretic patterns of polyacrylanide gels

and Sephadex G-200 protein profiles of uterine protein and

plasma collected on day 0 and the fact that highest total

uterine protein collected was on day 0 indicated a possible

movement of blood proteins into the uterus by diapedesis at

this time when blood flow to the uterus is elevated as a

result of elevated estrogen levels.

Electrophoretic data indicated qualitative changes in

proteins of the uterine flushing during the estrous cycle.

Sixteen (62%) polyacrylarmide gels of uterine protein samples

obtained between days 13 and 20 of the estrous cycle had an

electrophoretic protein band (Rf=0.35) that was not present in

corresponding plasma gels. This band was present in all gels

of samples on days 15 and 16. There was a highly significant

(P<.01) difference in the presence of this protein band from

days 0 to 20 of the estrous cycle and a correlation of .46

(P<.01) was found between the plasma progesterone concentra-

tion and its presence. Based on its Sephadex G-200 gel fil-

tration elution volume elutedd just prior to albumin), the

molecular weight of the protein represented by this Rf 0.35

electrophoretic protein band was estimated to be approximately

100,000. A second uterine protein electrophoretic band

(Rf=0.72) not appearing in corresponding plasma gels was








found in all uterine protein gels from days 0 to 20. It

migrated just ahead of the transferring. The electrophoretic

mobility of this protein was similar to that of the rabbit

uterine specific protein "blastokinin." One or two prealbunin

protein bands (Rf=1.09 and 1.12) were present in 59 (58%) of

the uterine protein gels between days 0 and 20 but not in

plasma gels. There was also a prealbumin band (Rl=1.27)

present in all uterine protein gels and some plasma gels

(very faint) that migrated with the marker dye used in the

electrophoretic procedure.

No significant differences in size classes of proteins

present between days 0 and 20 of the estrous cycle were found

in Sephadex G-200 gel filtration uterine protein profiles.

There was, however, some slight trailing of two low molecular

weight protein fractions which were never present in plasma

samples. These two 'low-profile"fractions were in the 10,000

to 15,000 and <10,000 molecular weight range.

In this study the data indicate that there are both quan-

titative and qualitative changes in the bovine uterine pro-

tein secretions during the estrous cycle. During the luteal

phase these changes appear to be under the influence of pro-

gesterone.













LIST OF REFERENCES


Abrams, R.M., D. Caton and F.W. Bazer. 1972. Effect of
estrogen on vaginal blood flow in ewes. Am. J. Obstet.
Gynecol. 103:629.

Abrams, R.M., W.W. Thatcher, F.W. Bazer and C.J. Wilcox.
1973. Effect of estradiol-17g on vaginal thermal con-
ductance in cattle. J. Dairy Sci. 56:1058.

Adams, C.E. 1958. Egg development in the rabbit: the influ-
ence of post-coital ligation of the uterine tube and of
ovariectomy. J. Endocrinol. 16:283.

Adams, C.E. 1965. The influence of maternal environment on
preimplantation stages of pregnancy in the rabbit, p. 345.
In: G.E.W. Wolstenholme and M. O'Connor (Eds.) Preimplan-
tation Stages of Pregnancy. Little, Brown and Co.,
Boston.

Adams, C.E. 1969. Egg-uterus interrelationships. Adv.
Biosci. 4:149.

Adams, C.E. 1971. The fate of fertilized eggs transferred
to the uterus or oviduct during advancing pseudopregnancy
in the rabbit. J. Reprod. Fert. 26:99.

Adams, C.E. 1973. Asynchronous egg transfer in the rabbit.
J. Reprod. Fert. 35:613.

Albers, H.J. and M.N. Castro. 1961. The protein components
of rat uterine fluid. An analysis of its antigens by
immunoelectrophoresis and Ouchterlony gel diffusion tech-
nique. Fert. Steril. 12:142.

Alden, R.H. 1942. Aspects of the egg-ovary-oviduct relation-
ship in the albino rat. II. Egg development within the
oviduct. J. Exp. Zool. 90:171.

Anderson, L.L. 1966. Pituitary-ovarian-uterine relation-
ships. J. Reprod. Fert., Suppl. 1:21.

Anderson, L.L., K.P. Bland and R.M. Melampy. 1969. Compara-
tive aspects of uterine-luteal relationships. Rec. Prog.
Hor. Res. 25:57.




Full Text

PAGE 1

cyclic nature of eovixe uterine luminal proteins Ajnd their relationship to peripheri\l plasma progesierone .^nd estrogen levels By ALBERT CARTER MILLS, III A DISSERTATION PRESENTED TO THE GR.ADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE FlEQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 19 75

PAGE 2

ACKNOWLEDGMENTS The author wishes to express sincere gratitude to the members of his supervisory committee: Dr. A. C. Warnick , chairman; Dr. Fuller V/. Bazer; Dr. D. E. Franke ; Dr. D. H. Barron and Dr. P. T. Cardeilhac. Sincere appreciation is extended to Dr. A. C. Warnick for his constant support, guidance and assi.stancc during th.e author's graduate program. Special thanks are expressed to Dr. Fuller W. Bazer for his continual willingness to be of assistance in the planning and conduct of this research. Deep appreciation is expressed to Dr. W. W. Thatcher for his constant help and encoaragem.ent with the hormonal and statistical ar.alyses conducted during the course of this research. Thanks are also due Dr. C. J. Wilcox, Dr. P. S. Kalra, Dr. F. C. Gv;azdauskas and Susan Acree for their help with these analyses. The author wishes to thank his fellow graduate students, Dr. James W. Knight, Thomas T. Chen and Eddy Muljono for their willing assistance in the collection and processing of samples for this research. The valuable technical assistance of Mary Bates Smith during the many hours of electrophoresis and protein determination is also appreciated. 11

PAGE 3

The handling of slaughter of experimental animals by Dr. A, Z. Palmer and Jerry S. Scott is greatly appreciated. Thanks are due Mr. Dean E. Pogue for care and feeding of the animals . The author wishes to thank his wife, "Dotty," for her constant encouragement and help during the writing of this dissertation .

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGMENTS ii LIST OF TABLES v LIST OF FIGURES vi ABSTRACT vii INTRODUCTION 1 REVIEW OF LITERATURE 4 CHAPTER 1 NONSURGICAL COLLECTION AND STUDY OF BOVINE UTERINE LUMINAL PROTEINS 33 Materials and Methods 34 Results and Discussion 39 2 CHAN'GES IN BOVIN}', UTERINE LUMINAL PROTEINS DURING THE ESTROUS CYCLE AND THEIR RELATIONSHIP TO PLASMA PROGESTERONE AND ESTRADIOL LEVELS 44 Materials and Methods 45 Results and Discussion 50 GENER.AL DISCUSSION 79 SUN&IARY 8 7 LIST OF PJiFERENCES 91 BIOGR^APHICAL SKETCH 10 7 IV

PAGE 5

LIST OF TABLES Table Page 1 Average total protein recovered from the bovine uterus during the estrous cycle 40 2 Average total uterine protein recov^ered and pla5iaa progesterone and estradiol concentration from cattle on each day of the estrous cycle 46 3 Least-squares analysis of variance: data obtained betv:een days and 20 of the estrous cycle 52 4 Least-squares analysis of variance: data ^ . obtained between days 4 and 18 of the estrous cycle 5 3 5 Least-squares analysis of variance: data obtained between days and 3 of the estrous cycle 5 4 6 Correlation of plasm.a progesterone and estradiol concentrations with total uterine protein and presence of Rf 0.35 uterine protein band following electrophoresis 58 7 Number and percent of poly acryl amide gels of uterine protein containing elcctrophoretic bands not present in corresponding plasma gels 65

PAGE 6

LIST OF FIGURES Figure Page 1 Averages for total uterine protein recovered on days to 2 of the estrous cycle 56 2 Averages for plasma progesterone concentration on days to 20 of the estrous cycle 6 3 Averages for plasma estradiol concentration on days to 2 of the estrous cycle 63 4 Polyacrylamide gel electrophoresis to'.vard the anode of h ovine uterine protein and plasma collected on days to 20 of the estrous cycle 67 5 Typical Sephadex G-200 gel filtration protein profiles of bovir^Luterine protein collected on days 0, 5 and 15 of the estrous cycle and bovine plasma 72

PAGE 7

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy CYCLIC NATURE OF BOVINE UTERINE LUMINAL PROTEINS AND THEIR RELATIONSHIP TO PERIPHER^^L PLASMA PROGESTERONE .AND ESTROGEN LEVELS by Albert Carter Mills, III June , 19 75 Chairman: A. C. h'arnick Major Department: Animal Science Bovine uterine protein secretions were examined during the estrous cycle for changes in total recoverable protein, Sephadex G-200 gel filtration protein profiles and electrophoretic protein patterns. QLalitative and quantitative changes were cone 1 ate d with plasma progesterone and estradiol concentrations. Preslaughter plasma and immediate posts laughter 0.35 M saline uterine flushings were collected from 67 cattle of mixed breeding (Angus, Hereford, Holstein and Brahman X British crosses). Three samples were collected for each day of the estrous cycle and 4 samples each were obtained on days 3, 5, 15 and IS. Plasma samples were assayed for progesterone (competitive protein binding procedure) and estradiol (radioimmunoassay). Total uterine protein was determined by Lowry's method. Qualitative poly acrylamide gel electrophoresis and Sephadex G-200 gel filtration protein profiles were obtained for each sample. vii

PAGE 8

There was a significant (P<.025) difference in total uterine protein recovered on days to 20 of the estrous cycle. From days 4 to 18, cows yielded significantly (P<.05) greater amounts of total uterine protein than heifers. Plasma progesterone concentration was significantly CP<-05) correlated with total uterine protein (r=.23} from days to 20 after correction for breed, parity and day effects by leastsquares analysis. This correlation coefficient was .44 (P<.05) when data collected during the luteal phase (days 4 to 18) of the estrous cycle were considered separately. The correlation coefficient between estradiol and total uterine protein was .26 (P<.05) when all data were considered (days to 20) and .75 (P<.01) when data collected during estrus and metcstrus (days to 5) were considered separately. However, after correction for breed, parity and day effects by leastsquares analysis, the correlations between estradiol and total uterine protein were not significant (P>.10). A protein band, not present in plasma samples, was found in 16 '(62?0 of the polyacrylamide gels following electrophoresis of uterine samples obtained between days 13 and 20 of the estrous cycle. There was a highly significant (P<.01) difference in the frequency with which this protein band was present from days to 20. This band (Rf=0.35) was present in all samples on days 15 and 16. Based on its Sephadex G-200 elution volume (eluted just prior to albumin), molecular weight of the protein represented by this band was estimated to be approximately 100,000. A highly significant (P<.01) VI 11

PAGE 9

correlation (r=.46) was found between plasma concentration of progesterone and the presence of this protein. A second protein band (Rf=0.72) which migrated just ahead of the transferrins appeared in all uterine protein samples analyzed from the entire estrous cycle, but it was absent in plasma. The electrophoretic mobility of this protein was similar to that of the rabbit uterine specific protein "b lastokinin. " One or tivo prealbumin protein bands (Rf=1.09 and 1.12) were present in 39 (5 8°^) of the uterine protein gels between days and 20 but not in plasma gels. Sephadex G-200 gel filtration uterine protein profiles did not reveal any significant differences in the size classes of proteins present between days and 20 of the estrous cycle However, there was some slight trailing in uterine samples of two lovj molecular weight protein fractions which were never present in plasma samples. These two "lov7-profile" fractions were in the 10,000 to 15,000 and <10,000 molecular weight range . This study indicated that both the qualitative and quantitative aspects of bovine uterine protein secretion changed during the luteal phase of the estrous cycle. Correlation of these changes with plasma progesterone concentration suggested that these changes were associated with plasma progesterone concentration.

PAGE 10

INTRODUCTION In recent years there have been several studies on the protein composition of uterine fluids in mice (Homburger et_ al_. , 1963; Mintz , 1970), hamsters (Noske and Daniel, 1974), rats (Junge aiid Blandau, 1958; Howard and DeFeo, 1959; Albers and Castro, 1961; Ringler, 1961; Heap and Lamming, 1962; Kunitake et_ al . , 1965), rabbits (Stevens, Hafs and Hunter, 1964; Krishnan and Daniel, 1967; Beier, 1968; Daniel, 1971a; Murray and Daniel, 1973; Beier, 1974a), swine (Murray, 1971; Murray ej^ al . , 1972a; Squire, Bazer and Murray, 1972), sheep (Heap and Lamming, 1960,1963; Heap, 1962; Perkins et al . , 1965; Iritani, Gomes and VanDemark, 1969; F. W, Bazer, unpublished data) and cattle (Fahning, Schultz and Graham, 1967; Schultz , Fahning and Graham, 19 71; Roberts and Parker, 1974a, b) These reports support the concept that uterine fluid proteins are a product of active secretion and not simple diffusion from blood. Synchronization of uterine secretions and stage of em.bryonic development has been shown to be critical to normal growth and implantation of the embryo (Dickmann and Noyes , 1960; Rowson and Moor, 1966; Rowson, Moor and Lawson, 1969; Adams, 1969,19 71,19 73; Rowson et_ aj^. , 1972a; Beier, Mootz and Kiihnel , 1972). It has been established that the embryo must reach the uterine environment

PAGE 11

at the proper time for continued and normal development to occur beyond the early blastocyst stage. This suggests that uterine protein secretions play an actii^e role in early embryonic growth and development. Work with mice (Mintz, 19 70; Pinsker and Mintz, 19 73) has indicated that estrogen controls the secretion of a uterine factor which affects implantation. In the rabbit (Krishnan and Daniel, 1967; Beier, 1968; Urzua e^ al . , 1970; Arthur and Daniel, 19 72; Johnson, 19 72; UTiitson and Murray, 19 74; Goswami and Feigelson, 19 74) and pig (Murray, 19 71; Knight, Bazer and Wallace, 1973a, b, 1974b; Knight et_ al . , 1974c; Chen, 1975; Chen e^ al . , 1973a; Schlosnagle et al . , 1974) recent work has established that uterine protein secretions are under the regulation of pri^ges terone and that these proteins are associated with early development of the embryo and trophoblast. Several workers have also suggested that the composition of box'ine uterine fluid is under hormonal control (Olds and VanDemark , 1957; Heap and Lamming, 1960; Heap, 1962; Fahnin'g e^ al_. , 196 7; Schultz et_ a]^. , 19 71). The working hypothesis which served as the basis for this study was that the uterus of the cow secretes specific proteins in response to changing plasma progesterone and/or estradiol concentrations, which are required by the embryo for complete development. Therefore, this study was designed to examine bovine uterine protein secretions during the estrous cycle for changes in total recoverable protein, Sephadex

PAGE 12

G-200 gel filtration protein profiles and electrophoretic protein patterns, and to correlate these changes with plasma progesterone and estradiol levels.

PAGE 13

REVIEW OF LITER.\TURE Recent research lias made it clear that uterine factors play a very important part in mammalian reproduction. The evidence indicates that these factors influence embryonic development in very specific ways. Some component (s) of the uterus appears to be important for embryonic development to the early blastocyst stage and absolutely essential for continued normal development past the early blastocyst stage. Evidence supporting the presence and importance of uterine protein secretions and their hormonal regulation will be considered herein. Embryos restricted to the oviductal environment develop to the early blastocyst stage in mice (Kirby, 1962; Orsini and McLaren, 1967; Kliittingham, 1968), rats (Alden, 1942), rabbits (Pincus and Kirsch, 19 36; Adams, 19 5 8) and swine (Murray et al^. , 1971; Pope and Day, 1972). However, continued development beyond the early blastocyst stage in the oviduct is retarded (Pope and Day, 19 72) or the embryos degenerate (Adams, 1958; Kirby, 1962; Orsini and McLaren, 1967; Murray et_ al_. , 1971). Kirby (1962) suggested the requirement of a "uterine factor" for development of mouse embryos beyond the blastocyst stage. He observed that mouse embryos recovered from the uterus continued to

PAGE 14

develop when transplanted beneath the kidney capsule as opposed to only trophoblast development when tube-locked embryos were treated similarly. Thus, the conclusion was reached that the uterus must provide some £actor(s) which is absolutely essential for continued normal development to occur past the early blastocyst stage of embryonic development. Early research with iri vitro culture of embryos demonstrated the importance of the uterine environment and its secretions. Lewis and Gregory (1929) found that rabbit ova developed no further than the blastocyst stage when cultured in rabbit blood plasma. It did not appear from earlier studies that simple nutrients constituted the uterine factors which enabled uterine embryos to develop continuously. Failure of development of the embryo past the early blastocyst stage ill vitro was the general rule regardless of the culture media or the species studied (Chang, 1949; Hammond, 1949; Whitten, 1956,1957; McLaren and Biggers, 1958; Tarkowski, 1961; Brinster, 1963; Adams, 1965; Cole and Paul, 1965; Maurer, VVfhitener and Foote , 1969; Rundell, 1969). According to Krishnan and Daniel (1967) and Beier (1968), a specific uterine protein secretion (termed "b las tokinin" and "uteroglobin" by the respective authors) is involved in blastocyst formation by rabbit morulae during culture in_ vitro in a chemically defined medium. More recent data have provided evidence that successful in vitro embryo culture can be achieved in chemically defined culture media up to the expanding blastocyst stage in mice

PAGE 15

(IVhitten and Biggers, 1968; IVhitten, 19 70), rats (Folstad, Bennett and Dorfman, 1969), rabbits (Onuma, Maurer and Foote, 196 8; Maurer, Onuma and Foote, 19 70; Kane and Foote, 19 70a,b,c) and sheep and cattle (Tervit, IVhittingham and Rows on, 19 72). When transferred to recipients many of these cultured blastocysts developed to form normal young (WTiitten, 19 70; Maurer et al , , 1970; Tervit et_ al . , 1972). Tubal mouse embryos cultured In vitro to the blastocyst stage and transferred to an ectopic site were also reported to be capable of forming welldifferentiated embryos (Billington, Graham and McLaren, 1968). Therefore, it would seem that blastokinin is not essential for blastulation to occur as originally proposed by Krishnan and Daniel (1967) and Beier (1968), It was later indicated that blastokinin m.ay have a more general function in embryogenesis after the free -bias tocyst stage (El-Banna and Daniel, 1972a, b; Daniel, 1972c). Development of embryos past the blastocyst stage iii vitro has not been reported. This supports the conclusion that uterine secretions, the protein milieu in particular, are necessary for continued embryonic development. A possible explanation of differences found in the extent of blastocyst development in. vitro in the absence of the uterine environment is found in the work of Nicholas (1942) , Fawcett, h'islocki and Waldo (1947), Runner (1947), Fawcett (1950), and Kirby (1962). Their data indicated that trophoblast of transplanted oviducal embryos expand in a morphologically normal manner, but the embryonic disc cells fail

PAGE 16

to develop. Also, Daniel (19 71rO acknowledged that blastocyst growth, presumably promoted by blastokinin, might be only volume change due to fluid accumulation in the blastocele rather than an increase in the number of cells or rate of mitosis by trophoblast cells. Fluid accumulation in the blastocyst is an energy-coupled process and not a simple process of osmosis (Tuft and Boving, 1970; Gamow and Daniel, 19 70). Fluid accumulation could, therefore, be an important function associated with the presence of blastokinin. Synchrony between stage of embryonic and uterine development has been demonstrated to be essential for normal embryonic development in mice (Kirby, 1962), rats (Dickmann and Noyes , 1960), rabbits (Adams, 1969,19 71,19 73; Beier et al . , 1972), sheep (Rowson aiid Moor, 1966), swine (Hunter, Polge d Rowson, 1967; Bazer et_ al . , 1969) and cattle (Rowson ^ al. , 1969,1972a). The work of Bazer ejt al. (1969) indicated that embryos in synchrony with the uterus were at an advantage in competing for a possible uterine factor(s) as compared to an asynchronous embryo. This, along with Kirby 's (1962) work with mice, suggests that embryos must reach a precise stage of development before they can utilize the uterine factor(s) apparently necessary for their continued development and that the uterine factor (s) is secreted only at a specific time. Therefore, the embryo must be present in the uterus at the proper time in order to utilize the secretion (s) . Additional information on the relationship between the uterus and embryonic development is found in experiments with an e

PAGE 17

animals having delayed implantation (Dickmann and DeFeo, 1967; Qiang , 196 8; Daniel, 19 70), The work o£ Dickmann and DeFeo (1967) and Chang (1968) demonstrated that uterine rather than embryonic factors determine whether or not implantation will be delayed. Therefore, it is obvious that the state of the uterus determines the state of the blastocysts it contains. This is also additional support for the view that uterine secretions are necessary for continuous em.bryonic development and that secretion of the required factor(s) is not constant but occurs at a specific time. In recent years , uterine luminal fluids from several species have been examined and found to consist primarily of proteins. In the rat, in contrast to early work by Warren (1938) which describes uterine fluids only as the necessary medium for movement of spermatozoa and studies by others on the chemical composition of uterine secretions (Shih, Kennedy and Huggins , 1940; Howard and DeFeo, 1959; Heap and Lamming, 1960,1962,1963; Heap, 1962), the first studies of protein components of uterine luminal fluid were those of Bredeck and Mayer (1955), Junge and Blandau (1958), Ringler (1961), Albers and Castro (1961) and Kunitake et_ al. (1965). Marked differences between potassium content of ucerine fluid and plasma indicated to Howard and DeFeo (1959) that the fluid in the uterine lumen was a secretion rather than a transudate from plasma. Other chemical composition studies (Shih et_ al . , 1940; Heap and Lamming, 1960,1962,1963; Heap, 1962) revealed that certain chemical constituents of the rat uterine li Lumen

PAGE 18

were greater at estrus than at diestrus and that they were under the influence of estrogen. Junge and Blandau (1958) found low levels of four niajor electrophoretic protein components in rat uterine fluid. Ringler (1961) reported that rat uterine fluid contained proteins with five different electrophoretic mobilities and thatonly one, a prealbumin fraction, was specific to uterine fluid. Based on this finding he suggested that uteiine luminal fluid was essentially an ultrafiltrate of plasma supplemented by specific uterine secretions. Albers and Castro (1961) also found that rat uterine fluid contained at least five protein comiponents when they used the techniques of Ouchterlony gel diffusion and Immunoelectrophoresis. Only one of their five proteins was specific to uterine fluid, and it migrated simiilar to serum B-globulins. Kunitake et_ ajl. (1965) demonstrated nine protein components of rat uterine fluid by disc electrophoresis. Using Ouchterlony gel -diffusion analysis with polyacrylamide discs and rabbit anti-rat sera, they showed that at least four of these proteins were common to serum. Five of the nine protein components were thought to be specific to uterine fluid. These results added further evidence to the concept that uterine fluid is, in part, a secretory product of the uterine endometrium. Using immunofluorescence analysis, Joshi and Murray (19 74) recently reported data which indicated that rat uterine fluid contained a peptidase unique to uterine secretion, not found in blood.

PAGE 19

10 Studies on the chemical composition of rabbit uterine secretions revealed greater levels of certain constituents during the luteal phase (pseudopregnancy) than at estrus which was contrary to data from the rat (Shih et_ al_. , 1940; Heap and Lamming, 1960,1962,196 3; Heap, 1962). Also, these chemical constituents appeared to be under control of progesterone rather than estrogen. The first study on proteins of rabbit uterine fluid v/as conducted by Stevens et al. (1964} using diffusion in agar gel, moving boundary electrophoresis and imp.iunoelect rophores is to characterize the proteins in fluid from, the ligated uteri of estrous rabbits. Eight electrophoretic components were separated using moving boundary electrophoresis. Two of these were specific to uterine fluid and v/ere not found in serum, one migrating as a prealbumin and the other as an ct-globulin. Diffusion of uterine fluid in agar gel revealed 13 antigenic components, three of which appeared to be specific to uterine fluid. At least five antigens, not found in serum, were identified by means of Immunoelectrophoresis. Two of these had electrophoretic m.obilities simiilar to prealbum.ins and three similar to 3-globulins. Kirchner, Hirschhauser and Kionke (19 71) demonstrated protease activity in the p-glycoprotein fraction of rabbit uterine secretion. They assumed a relation between uterine protease and im.plantation of the blastocyst. Using Sephadex G-200 gel filtration, Krishnan and Daniel (1967) studied uterine fluid protein constituents of rabbits in early pregnancy. They were able to demonstrate the

PAGE 20

11 presence of five major protein fractions. One of these fractions, Fraction IV, was not present until day 3 post coitum ; it reached a inaximum concentration on day 5 and then declined until day 9-. It was also found in uterine fluid of pseudopregnant rabbits on day 7 post coitum and in blastocelic fluid, but was not observed in maternal serum, fetal serum, fetal amniotic fluid or in uterine fluid accumulated by ligation of the rabbit uterus during days 3 to 10 of gestation. KTien used to supplement Ham's FIO culture medium in rabbit embryo culture, it promoted embryonic development to the expanding blastocyst stage and was thus termed "blastokinin. " Beier (1968), at approximately the same time as Krishnan and Daniel (1967), demonstrated that flushed rabbit uterine fluid and blastocelic fluid obtained on day 6 of pregnancy contained both plasma and uterine specific proteins. He described one of the uterine specific proteins present in blastocelic fluid as "uteroglobin" and suggested that it was involved in blastocyst development. There is no doubt that "uteroglobin" and "blastokinin" are the same protein (Kirchner, 1972; Beier, 19 70; Hamana and Hafez, 19 70; Beier, Kiihnel and Petry, 19 71; Daniel, 1971a). Hamana and Hafez (1970) demonstrated the presence of this protein in rabbit blastocelic fluid between days 5 and 8 of pregnancy ivith a maximum intensity at 6.5 to 7 days post coitum . Petzoldt (1974) recently reported that micro-disc electrophoretic protein patterns of day 4, 5, 6 and 7 blastocyst fluids were similar to corresponding

PAGE 21

12 patterns of uterine secretion samples. Using immunof lucres cent techniques, Kirchner (19 72) was able to show that uteroglobin was a product of the uterine endometrium and that it diffused across the blastocyst coverings into the blastocele. Labeled amino acid work by Murray and Daniel (19 73) showed that blastokinin, as well as other macroglobulin fractions, was produced by the uterine endometrium. Bullock and Connell (1973) recently demonstrated the presence of a protein, similar to blastokinin, in uterine flushings from nonpregnant rabbits. They also observed an electrophoretic band with mobility similar to blastokinin in flushings from nonpregnant rabbits and pregnant rabbits on days 1 and 2. Thus, it appears that blastokinin is present in minute quantities in the nonpregnant state and that its secretion is greatly accelerated during early pregnancy. Blastokinin is a glycoprotein with amino acids constituting approximately 741 of its weight and carbohydrates 6% (Krishnan and Daniel, 1968). Beier (1968) also found it to be a glycoprotein. Beier (1968) showed that estradiol and progesterone had a profound effect on the level of this uterine glycoprotein in day 6 post coitum uterine fluid. He suggested that uteroglobin secretion was under the control of progesterone and that excessive amounts of estradiol caused a reduction in the amount of uteroglobin produced. The molecular weight of blastokinin was originally estimated by Sephadex G-200 gel filtration to be 27,000 (Krishnan

PAGE 22

13 and Daniel, 1967,1968) and by ultracentrifugation (Beier, 1968,1970) to be about 30,000. However, Murray, McGaughey and Yarus (19 72b) found the above figures to be overestimated. Using gel filtration, SDS polyacrylamide gel electrophoresis and equilibrium sedimentation centrifugation , they reported the molecular weight to be approximately 15,000. Amino acid and spectrophotometric analyses of blastokinin suggested that the protein had a minim.al molecular weight of approximately 14,200 (McGaughey and Murray, 19 72). Bullock and Connell (1973) confirmed the latter reports and estimated the molecular weight of blastokinin to be 14,525 by gel filtration and 10,045 by polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. The possibility of subunits for blastokinin was suggested by McGaughey and Murray (19 72). The existence of subunits was given as a possible explanation for the difference between the two methods of molecular weight determination used by Bullock and Connell (19 73). An immunoassay for blastokinin using radial immunodiffusion with goat anti-blastokinin antiserum was developed by Johnson, Cowan and Daniel (19 72). They presented evidence that blastokinin possesses at least two antigenic determinants which may occur on separate subunits. This was added support for the possible subunit structure for blastokinin. Recent data have verified most of the earlier observations that blastokinin is the predominant uterine specific glycoprotein in rabbits (Hamner, 1970; Daniel, 1971a; Beier et al. , 1971; Beier and Beier-Hellwig , 1973; Beier, 1974a,b).

PAGE 23

14 ume Blastokinin was presumed to be the most important because of its progesterone dependent accumulation and its possible influence on development of the embryo. Fluid in the uterine lumen of rabbits contains considerable protein, and the vol of this fluid, total concentration of its macromolecular components and its protein patterns change continuously from ovulation to implantation. These proteins (predominantly glycoproteins of 25,000 to 50,000 MW) are a product of selective filtration from plasm.a proteins and of biosynthesis by epithelial cells of uterine endometrial glands and endometrial surface epithelium. Data also indicate that ovarian hormones regulate secretion and synthesis of the proteins. The most recent paper by Beier (19 74b) included a summary of recent knowledge oa iiumunologically identical proteins of uterine secretion and blastocyst fluid in the rabbit during the late pieimplantation period (days 6 and 7 post coitum) . He listed serum identical proteins as albumin, transferrin, immiunoglobulin and a-macroglobulin. Uterine "secreto-proteins" were listed as uterine prealbumin, uterine postalbumin, uteroglobin, uterine 3 glycoprotein and £ -uterus -macroglobulin. The serum identical transferrin and a-macroglobulin were not present in blastocyst fluid until day 8 post coitum . Added support for the importance of uterine specific proteins in embryonic development, specifically blastokinin, is found in studies on the relationship between uterine fluid proteins and the diapausing state of blastocysts from mammals having delayed implantation (Daniel, 1968 ,19 70 ,19 71b ,19 72a ;

PAGE 24

15 Daniel and Krishnan, 1969). Comparison of uterine fluid during the free-blastocys t period revealed a much lower concentration of protein from mammals with delayed implantation. The rabbit had much higher levels of total uterine protein than did the lactating rat during facultative delayed implantation, or the m.ink , fur seal and armadillo during obligate delayed implantation. Polyacrylamide gel disc electrophoresis revealed no protein with similar mobility to rabbit blastokinin in the lactating rat and armadillo; however, a similar band was found in low concentration in uterine fluid of mink (Daniel, 1968) and fur seal (Daniel, 19 71b , 19 72a) . In fur seal this component of uterine fluid differed in immunologic properties from rabbit blastokinin. Similarity between an electrophoretic band of any of these species and blastokinin must be taken in light of the discussion by Beier and Beier-Hellwig (1973) which draws attention to the electrophoretic difficulty of differentiating between a specific postalburain band and hemoglobin artifacts. It was obvious to Daniel" (1971b) that the uterine fluid proteins of the fur seal differed both qualitatively and quantitatively between the time when the embryo was dormant and the time it became reactivated to implantation. Daniel and Krishnan (1969) also observed gross qualitative and quantitative differences in the protein content in uterine secretions of certain mamjnals depending on whether or not they exhibited delayed implantation. Diapause embryos of animals with .obligate delayed implantation (mink, fur seal and armadillo) were stimulated

PAGE 25

16 to grow (as measured by total blastocyst expansion, and increased mitotic index) in^ v itro in the presence of rabbit blastokinin, but not in media containing serum. Blastocysts from heavy lactating rats, where a facultative delayed implantation had been produced, expanded in FIO culture medium alone or media containing serum or macromolecular components of uterine secretions from day 5 pregnant rats. However, supplementation of the medium with rabbit blastokinin had no effect. The mitotic index was higher than in uncultured controls in all conditions. Growth of active day 5 rat blastocysts, as evidenced by increased m.itotic activity, was stimulated by culture medium supplemented with uterine macromolecular components. Thus, there was evidence that in the case of facultative delayed iiiiplaiitation there apparently was some other uterine condition inhibiting grov/th of blastocysts, in addition to a protein deficiency. It was concluded that delayed implantation resulted from failure of the mother to provide sufficient protein and/or certain specific proteins as needed for active growth of the blastocysts. Quantitative and qualitative variation in protein constituents of sv:ine uterine fluid v/ere studied on days 2 to 18 and day 20 of the estrous cycle (Murray, 1971; Murray et_ al . , 1972a; Squire e_t aJ . , 1972). The maximum average total uterine protein level was found on day 15 v/ith levels increasing from day 10 to day 15. The uterine protein level decreased sharply after day 15 to levels on days 17, 18 and 20 which were similar to those observed before day 10 of the estrous

PAGE 26

17 cycle. Sephadex G-200 gel filtration revealed two protein fractions (IV and V) which were present only during the luteal phase of the estrous cycle. Three additional fractions (I to III) were present throughout the estrous cycle. Estimated molecular weights of Fractions I to V were 400,000 and greater; 200,000; 90,000; 45,000; and 20,000, respectively. Fraction IV contained a lavender-colored protein and was present only between days 12 and 16 of the estrous cycle. Polyacrylam.ide gel disc electrophoresis demonstrated that the lavender protein was basic, as it migrated toward the cathode at pH 8.0. Fraction V appeared between days 9 and 16 of the estrous cycle ^^/ith peak concentration of total uterine protein occurring on day 15. This fraction constituted greater than 20"o of the total recoverable protein on days 9 to 16. Polyacrylamide gel disc electrophoresis showed that this fraction was made up of six acidic proteins migrating toward the anode at pH 8.0 with Rf values of 0.91, 0.84, 0.80, 0.74, 0.19 and 0.18 (albumin Rf=1.00). Murray (1971) suggested that these two fractions appearing during the luteal phase of the estrous cycle might be related to rapid growth and expansion of the trophoblast of the more advanced embryo. These data on the cyclic nature of porcine uterine protein secretion indicated quantitative and qualitative changes during the estrous cycle. In subsequent work with the purple Fraction IV discussed above, it was demonstrated that the Fraction IV basic proteins (one to three) were present in allantoic fluid after day 30 of

PAGE 27

18 pregnancy, and evidence was presented vhich suggested that these proteins were involved in some aspect of placental development which might, in some way, have affected fetal development (Chen, 19 73; Chen and Bazer, 19 73; Chen, Bazer and Roberts, 1973b'J. Intravenous administration of sheep anti-Fraction IV antisera on days 7, 9, 11, 13 and 15 following mating decreased placental length and allantoic fluid protein concentration (P<.01). Administration of the antisera on days 34, 36, 38, 40 and 42 of pregnancy significantly (P<.01J reduced placental weight and length, and fetal wet weight and crownrump length, Krishnan (19 71) administered chicken anti -bias tokinin antisera to rabbits during early pregnancy (days 2, 4 and 6 post^ coitum) and reported either abnormal embryoiiic dc;velcj..:nent or complete cessation of pregnancy. Daniel (1972b), in some preliminary work with rabbit antiserum to swine luteal phase uterine protein, substantiated, in part, Krishnan' s observations. Of three sows treated with antiserum, two did not give birth to any offspring and did not return to estrus until 173 and more than 2 37 days after breeding. More recent work demonstrated that Fraction IV contained only one protein component, represented by a single electrophoretic band which moved toward the cathode at pH 7.0, with a molecular weight of 32,000 and an isoelectric point at approximately pH 9.7 (Chen e^ a]_. , 19 73a). The protein was specific to the uterus as antiserum prepared against it did not cross-react with extracts from other tissues. It was

PAGE 28

19 made up of 12.51 carbohydrate by weight and large amounts of basic amino acids. It was pointed out that the purple protein evidently self associated to form dimers, under certain conditions, which easily dissociated in the presence of high salt concentration. Schlosnagle et al. (1974) reported that the purple protein contained one atom of iron per 32,000 molecular weight polypeptide and showed acid phosphatase activity. Using fluorescent antibody techniques, Chen ejt a_l_. (1975) showed that the uterine purple acid phosphatase was formed in the epithelial cells of the endometrial surface and glands, and was detectable within the areolae by about day 30 of gestation. It was suggested that the purple protein was synthesized and secreted by the uterine surface and glandular epithelial cells and that the placental areolae served as sites of absorption and transport into the chorioal lantoic membranes and allantoic fluid during pregnancy. As in the rabbit. Heap (1962) and Heap and Lamming (1960, 1963) reported a significant increase in certain chemical constituents of the sheep uterus during the luteal phase of the estrous cycle. The work with cannulated sheep uterine fluid by Perkins e^ al. (1965) and Iritani et al_. (1969) showed an increase in volume around estrus, but no luteal phase increase was noted. More recent work with sheep uterine flushings (F. W. Bazer, unpublished data) indicated a quantitative and qualitative change in uterine protein secretion during the estrous cycle. There was an increase in total recoverable protein (also, some low molecular weight

PAGE 29

20 proteins were present) during the luteal phase of the estrous cycle . Early work on uterine fluid of cattle (Olds and VanDemark , 195 7; Heap,. 1962; Heap and Lamming, 1962; Fahning et al. , 1967} indicated that its chemical composition varied with stage of estrous cycle, with highest levels being reported during the luteal phase. This indicated hormonal control of bovine uterine fluid composition. Schultz et_ al . (1971) further illustrated this when they found that concentrations of reducing substances, total protein, potassium, chloride, inorganic phosphate, and alkaline and acid phosphatase activities all varied significantly with stage of the estrous cycle. These above reports all support the concept that bovine uterine fluid is the result of active secretion and not merely a product of diffusion from the blood. Until the present study was conducted no attempt had been made to characterize the proteins present in bovine uterine fluid nor to relate them to actual plasma steroid concentrations. After the present study had been completed, Roberts and Parker (19 74a] reported that the protein components of bovine luminal fluid were mainly serum proteins, but that small amounts of uterinespecific proteins had been detected by disc electrophoresis at pH 4.5. They had examined the luminal fluid from uterine horns of cows at different stages of the estrous cycle or early pregnancy by polyacrylamide gel electrophoresis at pH 4,5 and 8.9, isoelectric focusing, Immunoelectrophoresis, and gel filtration.

PAGE 30

21 In preliminary studies of human uterine secretions, Daniel Cl971a) was not able to find a protein comparable to blastokinin. Beier et^ al_. (1971) state that the intra-tubal and intra-uterine protein pattern apparently develops not only in the rabbit but in the human through selection of individual plasma proteins on the one hand and through biosynthesis of uterus-specific proteins on the other. Shirai, lizuka and Notake (1972) presented disc-ele ctrophoretic evidence for a postalbumin "b lastokininlike" fraction which exhibited a prominent peak during the mid-secretory phase of the human menstrual cycle. Noske and Daniel (19 74) also reported the appearance of a postalbumin protein band in the ham.ster on day 5 p os t coi turn , which had an electrophoretic mobility similar to that of blastokinin. Both of these reports must be taken in light of the electrophoretic difficulty of differentiating between a specific postalbumin band and hemoglobin artifacts (Beier and Beier-Hellwig , 19 73), Composition of bovine uterine fluid was suggested to be under hormonal control (Olds and VanDemark , 195 7; Heap and Lamming, 1960; Heap, 1962; Fahning e_t aJ . , 1967; Schultz ejt al. , 1971) because of observed cyclical changes in certain chemical constituents of bovine uterine fluid during the estrous cycle. Schultz et_ aj^. (1969) reported that progesterone administered to ovariectomized cows resulted in a greater increase than did estradiol in the size of nuclei of epithelial cells lining the mucous glands of the uterine endometrium. They suggested that the increase in nuclear

PAGE 31

22 size might have been associated vith an increased rate of cell secretion. In the ewe (Murdoch, 1972), progesterone stimulated activity of acid and alkaline phosphatases in the intercotyle Jonary endcmetrium. Roblero (19 73) found that progesterone administered to ovariectomized pregnant mice resulted in embryos which had significantly more cells than embryos from ovariectomized females not receiving progesterone. Other work with the mouse (Mintz, 19 70; Pinsker and Mintz , 19 73) has indicated that estrogen controls the secretion of a uterine factor which affects implantation. Yasukawa and Meyer (1966) , in a study on the effect of progesterone and estrone on preimplantation and im.plantation stages of embryo development in the rat, reported tbat changes necessary for and indicative of impending implantation had been induced by the synergistic action of estrone and progesterone on the blastocyst. Several cheniical composition studies with rat uterine fluid (Shih e_t al . , 1940; Heap and Lamming, 1960,196 2,196 3; Heap, 1962) revealed tliat certain chemical constituents were present in greater amounts at estrus than diestrus and were therefore considered to be influenced by estrogen. In the rabbit, concentration of chemical constituents was greater during the luteal phase (pseudopregnancy) than at estrus and was considered to be under the control of progesterone rather than estrogen. Wu and Allen (1959) reported graded effects of progesterone on pregnancy maintenance in castrated rabbits. Beier (1968,1970) reported that estradiol and progesterone

PAGE 32

23 had a profound effect on uteroglobin on day 6 of pregnancy in rabbits. He reported that excess estradiol caused a reduction in the amount of uteroglobin produced, which suggested to him that uteroglobin might be under the control of progesterone and that estrogen antagonized the effect of progesterone. Pincus and Kirsch (1956) reported a detrimental effect of estrogen on blastocyst growth and development. Administration of estrogen postcoitally in the rabbit has been reported to lower uterine carbonic anhydrase after 2 4 hours (Makler and Morris, 1971). El-Banna and Daniel (1972a) found that progesterone stimulated rabbit blastocysts in vitro when in combination with uterine proteins. Urzua e_t a]_. (1970) found that blastokinin v/as present in uterine fluid of ovariectomized rabbits that had received progesterone, or progesterone and estradiol in com.binat ion , but it was absent in rabbits treated with estradiol alone. Arthur and Daniel (1972) found blastokinin in the uterine fluids of ovariectomized rabbits following administration of progesterone but not estrogen, and the kinetics of the doseand timeresponses obtained indicated that this was the normal relationship during pregnancy. Blastocysts transferred to uteri of castrate rabbits given progesterone grew and differentiated up to the time of implantation. IVhitson and Murray (1974) found that endometrial cells from mature female estrous rabbits were capable of synthesizing blastokinin iii vitro following treatment with progesterone for 48 hours. Urzua et_ a]_. (1970) and Arthur, Cowan and Daniel (19 72) reported that blastokinin

PAGE 33

24 bovmd progesterone and estradiol, but binding of progesterone to blastokinin was inhibited by estradiol and vice versa. Recently, Gosv;ami and Feigelson (19 74) reported on the differential regulation of a low molecular weight protein in rabbit oviductal and uterine fluids by progesterone and estradiol-17B. Based on similar molecular weight and electrophoretic njobility, this protein was thought to be the same as blastokinin. It was present in uterine and oviductal fluids of intact estrous rabbits and assumed a unique cone-shaped profile, upon polyacrylamide gel electrophoresis followed by a modified Amido Black staining and destaining procedure. The p7'otein was under hormonal control and was absent from both uterine and oviductal fluid following ovariectomy. Exogenously admdnis tered progesterone strongly induced it in uterine fluid (very slightly in oviductal fluid) of ovariectomized rabbits. Estradiol had a much greater effect than progesterone on the induction of this coneforming protein in oviductal fluid, although of lesser absolute magnitude than the effect of progesterone on its induction in uterine fluid. Contrary to the work of Urzua e_^ al_. (19 70) and Arthur e^ al_. (19 72), neither the uterine nor the oviductal fluid coneforming protein was demonstrated to bind progesterone in vitro to any discernable extent. More extensive Sephadex G-50 gel column chromatography was used by Goswami and Feigelson (19 74), which clearly indicated that the coneforming protein was eluted in a fraction widely separated from the single peak of free progesterone. Thus, it was

PAGE 34

25 concluded that rabbit genital tract fluid "cone protein" does not detectably bind progesterone. Additional immunological work will have to establish whether blastokinin and this coneforming protein are the same. Data in the pig (Murray, 19 71) indicated that maxim.um uterine secretory activity occurred in gilts during the luteal phase of the estrous cycle when the uterus was under the influence of progesterone. He reported an abrupt decrease in protein content of uterine flushings which coincided with regression of the corpora lutea. Knight et_ a_l_. (1975a) reported a positive relationship between quantity of recoverable uterine protein and quantity of luteal tissue in superovulated and unilaterally ovariectomized-hysterectomized gilts. It was suggested that progesterone, possibly in synergism with estrogen, was primarily responsible for inducing quantitative and qualitative changes in the protein milieu of uterine flushings. In a subsequent study. Knight et^ al . (1973b) indicated that progesterone was the essential hormone which ' regulated the quantitative and qualitative aspects of porcine uterine protein secretion during the estrous cycle and presumably during pregnancy. Work with the porcine purple uterine protein (Chen, 1973; Chen et aJ . , 1973a; Schlosnagle et al . , 1974) also established that its secretion was regulated by progesterone. Uterine protein secretions of ovariectomized gilts treated with progesterone until either day 7, 9, 11, 13, 15, 17 or 19 after onset of estrus were quantitatively and qualitatively similar to uterine protein

PAGE 35

26 secretions o£ intact nonmated gilts up to day 15 of the estrous cycle (Knight et_ al . , 1974c), However, on days 17 and 19 total protein was greater in the treated gilts and the qualitati\'e aspects were maintained. Knight e_t_ al^. (19 74b) reported that increased uterine protein secretions, resulting from high progesterone levels, enhanced placental development and allantoic fluid volume. This led to an increase in the placental surface area which was in contact with the maternal endometrium. It was suggested that the establishment of maxi mum placental surface area early in gestation might be of critical im.portance with respect to fetal growth and survival as pregnancy progressed towards term. In more recent work, Knight (19 75) presented data which indicated that the develop ment of adequate placental mass was apparently the key factor necessary for adequate and sustained fetal grov/th and develop ment. The role of the uterus is not limited solely to control of embryonic development. It is also involved in control of luteal function. Wiltbank and Casida (1956) reported that hysterectomy during the luteal phase of the estrous cycle prolonged the lifespan of corpora lutea in cows and ewes. Since this report, it has been clearly demionstrated that the uterus is responsible for regression of corpora lutea of estrous cycles in cattle, sheep, swine, horses and other species (reviews by Ariderson, Bowerman and Melampy, 1963; Bland and Donovan, 1966; Anderson, Bland and Melampy, 1969; Schomberg, 1969; Rowson, 19 70). These reviews point out that

PAGE 36

27 hysterectomy results in maintenance of the corpora lutea for approximately the length of normal gestation. Cyclical luteal regression appears to be under local control of the uterine horn adjacent to the corpora lutea of cattle, sheep and swine (Hansel and Echternkamp, 1972). l\Tien one uterine horn was removed in sheep, swine (.^derson, 1966) or cows (Ginther e_^ al_. , 1967) corpora lutea on the ovary adjacent to the retained horn regressed at the expected time, but corpora lutea on the opposite ovary were maintained beyond the normal estrous cycle length. Luteolytic effects of endometrial extracts have been demonstrated. Williams e_^ aJ. (1967) reported that acetonedried extracts of bovine uteri caused luteal regression when injected into pseudopregnant rabbits. Schomberg (196 7) found that sv.'ine uterine flushings obtained on days 1 to 10 and day 20 of the estrous cycle had no detrimental effect on in vitro growth of granulosa cells or their progesterone production, but day 12 uterine flushings sometimes showed a luteolytic effect. Flusliings obtained between days 14 and 18 of the estrous cycle were markedly luteolytic and destroyed the granulosa cells within 6 to 8 hours. The uterine fluid component with luteolytic activity was later reported to be thermolabile and nondialyzable with an estimated molecular weight of about 200,000 (Schomberg, 1969). Barber (1972) demonstrated that the component of swine uterine flushings showing the greatest in vitro luteolytic effect on granulosa cells was the Sephadex G-200 gel filtration Fraction II

PAGE 37

28 (Murray, 1971), which also had a molecular weight of approximately 200,000. Mazer and Vv'right (1968) reported that a nondialyzable luteolytic factor was present in the uterus of the hamster on days 6 and 7 of pseudopregnancy . Lukaszewska and Hansel (1970) found that aqueous extracts (precipitable in 551 ammonium sulfate) of day 10 to 13 bovine endometrium were luteolytic when injected intraperitoneally into pseudopregnant hysterectomized hamsters. The active factor v;as thought to be either a large molecular weight protein or a smaller molecule bound to protein. Further studies showed that lipid extraction of the active protein fraction removed the luteolytic activity. Wlien the lipid extract was subjected to thin layer chromatography, the fraction containing prostaglandins (discussed below) did not have luteolytic activity. Hansel, Concannon and Lukaszewska (19 73) showed clearly that luteolytic activity was separated from prostaglandin activity and that the luteolytic factor studied had an electrophoretic Rf value similar to that of the arachidonic acid standard. It was suggested that the bovine endometrium exerts its local luteolytic effect by providing the corpus luteum with one or more precursors (arachidonic acid) which could be converted into prostaglandin or other luteolysin in situ by the corpus luteum. Much evidence has accumulated on the luteolytic properties of PGF^ , a prominent uterine prostaglandin. It has been 2a ' ^ -^ identified by Coding e^ al. (1972) as "the" luteolytic hormone and by McCracken et al. (1972) as a luteolytic hormone

PAGE 38

29 in sheep. Prostaglandin F^ ^ has also been shown to cause luteal regression in cattle [Lauderdale, 19 72; Rowson, Teruit and Brand, 19 72b, Hansel e_t al_. , 1975; Inskeep , 1973; Cher.ault, 19 73; Lauderdale et a_l . , 19 74), horses (Noden, Hafs and Oxender, 1973), swine (Muljono et_ aJ . , 1974), rats (Pharriss and Wyngarden, 1969), hamsters (Gutknecht, Wyngarden and Pharriss, 1971) and guinea pigs (Blatchley and Donovan, 1969). Several facts 'nave prevented acceptance of PGF^ as "the" luteolytic hormone by all investigators. Hansel ejt al . (19 73) argued that measurements of rGF„ in uterine venous plasma on any given day of the estrous cycle were highly variable, and that agreement between different laboratories v/as very poor. They stated that it v;as not clear whetlier the rise in plasma PCF^ preceded the fall in progesterone in every case studied. Some iii vitr o studies indicated that PGFo vs\-is luteotrophic rather tlian luteolytic (Hansel et al., 1973; Speroff and Ramwell, 1970; Sellner and IVicke rsham, 1970). Hansel et^ a]_. (1973) also found prostaglandins E^ , E, , A. and A„ to be luteotrophic when incubated with bovine luteal tissue. Wilson, Butcher and Inskeep (1972) reported levels of PGF^ in the endometrium and uterine venous blood of preg2a nant ewes to be higher than in nonpregnant ewes on days 13, 14, 15, 16 and 18 after onset of estrus, .. In maintenance of luteal tissue during pregnancy, it is possible that some luteotrophin may be produced by the embryo which overrides the uterine luteolytic factor (probably PCF^^^ or a precursor such as arachidonic acid) , or the embryo may

PAGE 39

30 exert a direct antiluteolytic effect (i.e., absorption or inactivation of the luteolytic factor) . Evidence supporting the latter mechanism was presented by Warren, Hawk and Williams (1971). They found that infusion of homogenized embryos into the uterus of the ewe prevented lUD-induced luteal regression. That the embryo or trophoblast could be producing a luteotrophin or overriding the effect of the uterine luteolytic factor is suggested by much recent work with the blastocyst. The steroids, estrogen and progesterone, are present in rabbit blastocysts and uterine fluid (Seamark and LutwakMann , 1972; Beier, 1974b); and estradiol bound to rabbit blastocysts has been implicated in the development or implantation of the blastocysts, consistent with the hypothesis that estradiol may act as a local signal from the blastocyst to the uterus (Bhatt and Bullock, 19 74). In the rat (Dickman and Dey , 19 73 ,19 74a, b ; Dey and Dickman, 19 74a) and mouse (Dey and Dickman, 19 74b) data were presented which suggested that blastocysts can synthesize steroid hormones which are critical for morula to blastocyst transformation and implantation. The results suggested that one of the hormones synthesized was estrogen. Perry, Heap and Amoroso (1973) reported production of both progesterone and estrogen by pig blastocysts (days 14 to 16). Extremely high estrone and estradiol concentrations were reported in allantoic fluid on various days of pregnancy in the gilt (Knight et^ al . , 1974a; Knight, 1975), suggesting a placental source for the estrogens. Research on the

PAGE 40

31 influence of the marsupial embryo on the uterus (Renf ree , 1972) provided evidence that the embryo or placenta exerts a twofold effect on the uterus due to a hormone produced by the embryo "or its membranes or to an immunological response. The suggested effects of the embryo were: to exceed the influence of the corpus luteum by increasing the weight of the endometrium in the uterus which carries it, and to stimulate production of uterine -specif ic proteins. Of considerable interest are the recent reports that a substance similar to HCG or LH was found in human plasma within 6 days of fertilization (Saxena et a]^. , 1974), and in rabbit blastocysts on days 5 and 6 post coitum (prior to implantation) at concentrations ten times higher than in pregnant rabbit plasma (Haour and Saxena, 1974). Experiments involving the surgical transfer of blastocysts in sheep (Moor and Rowson, 1966a, b) and pigs (Dhindsa and Dziuk, 1968) have shown that embryos must be present in the uterus before day 13 for pregnancy to be established and for the corpora lutea of the estrous cycle to be converted into corpora lutea of pregnancy. This apparent ability of the blastocyst to signal its presence and to modify corpus luteum function prior to attachment to the uterine endometrium implies production of some hormonal factor. Such an antiluteolytic factor must be suppressing endometrial synthesis of the uterine luteolysin, blocking its direct transfer from the uterus to the site of action in the corpus luteum, or having a luteotrophic effect on the corpus luteum in the presence of secretion of the uterine luteolysin.

PAGE 41

32 Heap and Perry (1974) speak o£ estrogens being luteotrophic in the pig, and postulate that the uterine endometrium is able to conjugate estrogens produced by the blastocyst which might provide a source of circulating estrogens of low biological potency, yet readily metabolizable to an active form, in other tissues (such as the corpus luteuin) where their stimulatory (luteotrophic) effect might be expressed. Sufficient data have been reviewed on restriction of embryos to the oviduct, rn vitro culture, synchrony of embryonic and uterine development and delayed im.plantation to support the concept that some uterine component (s) is extremely important for continued noriaal development of the embryo past the early blastocyst stage. Uterine secretions in many species have been reviewed and it has been established that these secretions (predominantly proteins) are the product of active secretion by the uterine endometrium supplemented by an ultrafiltrate of plasma. It has been established that these uterine protein secretions change quantitatively and qualitatively during the estrous cycle in most animals studied and that their secretion is regulated by steroids. They have also been associated with early development of the embryo and trophoblast. The relationship between the uterus and maintenance of corpus luteum function has been established, and much data presented and discussed which implicates the embryo in a definite role, in some cases, in prevention of luteal regression.

PAGE 42

CHAPTER 1 NONSURGICAL COLLECTION. AND STUDY OF BOVINE UTERINE LUMINAL PROTEINS Bovine uterine fluids obtained following slaughter and from the live cow liavc been examined by several researchers. Olds and VanDemark (195 7) obtained fluid by individually passing each uterine horn through a hand-operated clothes wringer under moderate pressure. Roberts and Parker (19 74a) obtained uterine saline washings within a few minutes of stunning at slaughter. Gupta (1962) used a return flow uterine catheter attached to a 250 r.il Erleume/er flask and a vacuum pump to collect uterine fluid samples during estrus. A technique described by Fahning, Schultz and Graham (1966) was used to aspirate uterine fluid from cattle during the estrous cycle by Fahning e_t al_. (1967) and Schultz et al. (1971). Loe (1970) modified the procedure of Fahning et_ al. (1956) and aspirated fluids on days and 2. Heap and Lamming (1960) and Heap (1962) obtained uterine washings from cows by using an endotraclieal tube with an inflatable cuff placed into the uterus via the cervix. They injected an isotonic solution into the uterus and recovered it by massaging the uterus per rectum . The procedure described in this chapter for nonsurgically flushing the uteri of cows and heifers was developed using 33

PAGE 43

34 information gained by C. K. Vincent, J. W, Rundell and A. C. Mills (unpublished data) and A. C. Mills (unpublished data) in attempts to nonsurgically recover embryos from cattle, and from information in the procedures described by Rowson and Dowling (1949), Fahning et_ aJ. (1966) and Heap (1962). Initially, it was determined from nonsurgical embryo recovery that most of the fluid injected into the uterus could be recovered (A. C. Mills, unpublished data). The study described below v/as undertaken to determine if there were proteins specific to the bovine uterus and if any variation in these proteins might be related to plasma steroid concentration and to the estrous cycle. Materials and Methods In this preliminary study of bovine uterine protein secretions approximately seven uterine flushing samples per day of the estrous cycle (total of 144) were collected from live cows and heifers. Approximately three samiples per day were collected from a group of crossbred heifers, three per day from dairy cows (greater than 35 days postpartum) and one per day from a set of triplet Angus X Holstein heifers. More than one sample was collected from the triplets and from several of the crossbred heifers. In most cases the animals had been in estrus at least twice with a normal estrous cycle preceding each collection.

PAGE 44

35 The cattle were checked visually twice daily for estrus with the aid of Kamar heat detector patches and/or two penectomized Angus bulls fitted with marking collars. Day of standing estrus was designated as day 0. The cattle were penned on the morning of the randomly selected day of the estrous cycle on which a sample was to be collected and then restrained in a standing position immediately before collection. Uterine fluid samples were collected by inserting an 18 Fr urethral catheter through the cervix (rectal manipulation) and inflating the 5 cc cuff just inside the body of the uterus. Tension was kept on the catheter so that the cuff was kept firmly against the cervix. A 50 ml glass syringe filled with 0.33 M saline was connected to the catheter and the saline was then carefully injected into the uterus. Care was taken to make sure that the fluid filled the entire uter and not just one horn, and that excess pressure was not put on the uterus by tlie injection of too much fluid. In some cases the entire 50 ml was not injected into the uterus. After momentarily massaging the uterus (less than 1 minute) the fluid was allowed to flow back into the syringe under slight negative pressure with rectal manipulation of the uterus as needed. Ivlien no more fluid could be recovered a pinch c].amp was placed on the catheter, the syringe containing the uterine flush was disconnected and a second 50 ml glass syringe filled with 0.33 M saline v;as connected and the us

PAGE 45

36 procedure was repeated. The uterine flushings were placed into stoppered bottles and put into ice water until processing Immediately after a uterine flush was completed in the group of crossbred heifers a 10 ml jugular blood sample was taken (heparinized vacuum tubes) and placed in ice water. In the group of dairy cows a heparinized Erlenmeyer flask was used. No blood was collected from the triplets. The blood samples were centrifuged for approximately 10 minutes after which the plasma was removed and stored at -20°C for possible steroid analysis and for comparison of the plasma and uterine proteins . Uterine flushings were taken from the ice water and centrifuged at 10,000 rpm for 10 minutes at 3°C. They were sterilized by filtration of the supernatant through a 0.45 y Millipore filter and concentrated to 1 to 2 ml by vacuum dialysis. Next the samples were dialyzed against 0.05 M phosphate -citrate buffer, pH 7.4, for 36 to 48 hours at 3°C and then stored at -20°C until further analysis. Protein concentration of each stored uterine sample was determined by the method of Lowry et_ ad^. (1951) and the total quantity of protein in each sample calculated by multiplying that value by sample volume. Total recovered protein was adjusted (divided by percent recovery) if the saline flush recovery was less than 100%. Sephadex G-200 gel filtration was used to fractionate a portion of each uterine protein sample to obtain a protein profile. The columns used were 1.5 x 90 cm with a bed height

PAGE 46

37 of approximately 85 cm. They were calibrated using Blue Dextran (MW=2 ,000 ,000) , to determine the void volume, and aldolase (MW=158,000) , ovalbumin (MVv = 45,000) and ribonuclease A (MW=13,700) as standard globular proteins of known molecular weight. Gel filtration was carried out at 3°C. The eluent was 0.05 M phosphate-citrate buffer, pH 7.4, with flow rates of approximately 6 ml per hour. .An aliquot of each fraction (approximately 2.5 ml) was used for protein determination and the resulting optical density of each fraction was plotted against the elution volume, at which the fraction occurred, to construct a protein profile. Sephadex G-75 gel filtration was used to construct a uterine protein profile for each day of the estrous cycle from a pooled aliquot of the crossbred heifer samples (approximately three per day). The procedures were the same as for the Sephadex G-200 protein profiles above. Bovine plasma Sephadex G-200 and G-75 protein profiles were also obtained for comparison with the uterine protein Sephadex G-200 and G-75 profiles. Polyacrylamide gel disc electrophoresis, using the basic procedure of Clarke (1964) , was used to further study the protein components of the uterine flushings. A 1% polyacrylamide gel, 5 mm in diameter and 70 mm long, was used ;\ithout spacer or sample gels. Samples containing about . 3 mg protein were applied in 0.1 ml of 1 M sucrose and two drops of bromophenol blue directly on the separating gel. The samples were run in 0.05 M Tris 0.38 M glycine buffer, pH 8.0, at a constant

PAGE 47

38 current of 2.5 mA per tube and a nominal v^oltage of 90 volts for about 1.5 hours or until the marker dye had migrated about 65 mm. The gels were stained with amido black solution (1 gm per 100 ml of 1% acetic acid) for 1 hour and destained electrophoretically in 31 acetic acid. The e lectrophoretic mobility of uterine proteins was expressed in terms of their mobility relative to that of albumin (Rf) when albumin was assigned an Rf value of 1.0. Electrophoresis of bovine plasma was also conducted, using the same procedure, for comparison with uterine protein polyacrylamide gels. Prior to making the conclusions discussed below concerning the quality of the uterine protein flushings obtained nonsurgically , plasma progestin concentrations were determined for the group of crossbred lieifcrs (approximately three per day). For analysis, approximately 2,000 CPM of progesterone1,2-"^H (Amersham/Searle , 31.7 Ci/mM) were added to 2 ml of plasma. This isotopic steroid served as an internal standard for correction for procedural losses. The mixture was extracted vigorously twice with 10 ml of isooctane and aliquots of the extracted progestins were quantified by competitive protein binding assay (Murphy, 1967). Details of the procedures used were described by Gwazdauskas (19 72) and Chow (1972).

PAGE 48

39 Results and Discussion An average o£ 84.3 ± 19.7-0 of the 0.33 M saline put into the uterus v/as recovered and the samples appeared to be relatively free of blood contamination during most of the estrous cycle. Total uterine protein recovered averaged 46.93 ± 41.31 mg , but there v;ere no apparent differences associated with day or stage of the estrous cycle. Data presented in Table 1 compare average total uterine protein recovered during the estrous cycle in this study with that recovered by various methods in other studies. Average total uterine protein collected with the nonsurgical flushing procedure of this study is comparable to that collected by Heap (19G2) who used a similar type procedure. It is also similar to that found by Olds and VanDemark (1957), using a pos tslaughter stripping technique, and by Gupta (1962) and Schultz et_ al_. (1971) who nonsurgically aspirated fluid from the uterus. Schultz et_ al . (1971) reported protein concentration of plasma and their aspirated uterine fluid to be practically the same. The "surgical-like" posts laughter flushing that will be discussed in Chapter 2 of this study and the pos ts laughter flushing reported by Roberts and Parker (1974a), published after the present study was completed, yielded much lower quantities of average total uterine protein. The high levels of uterine protein recovered in nonsurgical flushings in this study and in those of Olds and VanDemark (1957), Heap (1962), Gupta

PAGE 49

40 Table 1. Average total protein recovered from the bovine uterus during the estrous cycle. Study

PAGE 50

41 (1962) and Schultz e;^ al. (1971) indicate varying degrees of blood contamination due to techniques employed. Sephadex G-200 and G-75 gel filtration protein profiles and electrophoretic polyacrylamide gels of uterine flushings were similar to those of bovine plasma collected at the same stage of the estrous cycle. However, uterine flushings contained varying amounts of hemoglobin throughout the estrous cycle as evidenced by a hemoglobin band(s) on the uterine polyacrylamide gels and a hem.oglobin fraction on many of the uterine Sephadex gel filtration protein profiles. Practically all of the uterine electrophoretic gels had a distinct hemoglobin band not present on plasma gels. Sephadex G-200 protein profiles of uterine flushings revealed varying proportions of hemoglobin which resulted in either a fourth fraction beyond the albumin peak (third fraction) or a distinct extended shoulder on the albumin peak. This variation may be explained by relative differences in the proportion of album.in and hemoglobin present in each sample. Plasma Sephadex G-200 protein profiles contained only three distinct fractions (albumin being the third). There was an obvious increase in the proportion of hemoglobin relative to the albumin peak in Sephadex G-200 protein profiles of uterine flushings recovered on days -1, 0, 1 and 2 as compared to the rest of the estrous cycle. This could possibly be explained by a combination of metestrus bleeding (Schultz e_t al , , 1971, reported samples aspirated from the uterus during metestrus to be consistently red and brownish-

PAGE 51

42 red with blood elements present) and increased blood flow to the reproductive tract under the influence of estrogens (Abrams, Caton and Bazer, 1972; Abrams et_ al_. , 1973; Gwazdauskas et_ al_. , 1974) around estrus. With increased blood flow to the reproductive tract it would be expected that the quantity of hemoglobin present due to collection procedure might increase along with blood flow. Progestin analyses were discontinued on plasma samples collected inuaediately after successful uterine flushing because of the possibility that stress of the uterine fluid collection procedure might have altered plasma progestin concentration (Gwazdauskas , 1972), Also, after it was decided that the uterine flushing technique had resulted in varying degrees of blood contamination, it was evident that any attempt to correlate progestin concentration vvith total uterine protein recovered would be meaningless. The levels of uterine protein in nonsurgical flushings, the presence of hemoglobin in uterine flushings, and the similarity between Sephadex G-200 gel filtration and polyacrylamide gel electrophoretic protein profiles of plasma and uterine flushings collected in this study led to the conclusion that the nonsurgical collection technique resulted in varying degrees of blood contamination. Thus, tlie attempt to characterize uterine protein secretions of the bovine using this technique precluded any attempt to study bovine uterine protein secretions. Any uterine protein secretions present in the uterine flushings were obviously masked by the presence

PAGE 52

43 of much greater quantities of plasma proteins as a result of blood contamination.

PAGE 53

. CHAPTER 2 CHANGES IN BOVINE UTERINE LUMINAL PROTEINS DURING THE ESTROUS CYCLE AND THEIR RELATIONSHIP TO PLASMA PROGESTERONE AND ESTRADIOL LEVELS In Chapter 1, data were presented which indicate blood contamination of uterine flushings collected nonsurgically . These data and similar average total uterine protein recovered by previous workers (Gupta, 1962; Heap, 1962; Schultz et al. , 1971) indicated that uterine protein samples collected by nonsurgical methods through cervical penetration do not represent the actual intraluminal uterine protein milieu. That this is the case was further evidenced by much less total uterine protein being recovered by the "surgical like" pos tslaughter flushing technique, described in this section, from several heifers slaughtered on various days of the estrous cycle during a preliminary investigation. The following study was initiated in an attempt to accurately determdne quantitative and qualitative changes in the intraluminal protein milieu of the bovine and to correlate these changes v.'ith plasma progesterone and estradiol concentrations . 44

PAGE 54

45 Materials and Methods Sixty-seven cycling heifers and cows of mixed breeding (Angus, Hereford, Holstein and Brahman X British crosses) served as the experimental animals. Heifers were approximately three years of age and cows were greater than 90 days postpartum. Animals were checked visually twice daily for estrus with the aid of Kamar heat detector patches and two penectomized Angus bulls fitted with marking collars. Day of standing estrus was designated as day 0. Cattle v.-ere randomly slaughtered within each breed and parity (heifers or cows) group on days to 20 of the estrous cycle. Each animal had a normal estrous cycle prior to slaughter. Feed was withheld for 24 hours prior to slaughter. The number of animals slaughtered on each day of the estrous cycle is presented in Table 2. Prior to slaughter, blood samples (100 ml) were collected via jugular venipuncture into heparinized flasks. The flasks were immediately placed into an icebath, centrifuged at 10,000 rpm for 10 minutes at 3°C and plasma was removed and stored at -20°C until analyzed for progesterone and estradiol. Immediately postslaughter , after removal of the reproductive tract and recording of ovarian data, each uterine horn was flushed with 25 ml of 0.33 M NaCl. This was completed in less than 30 minutes post-stunning. Each uterine horn was clamped as near as possible to the bifurcation. A small incision was made in the oviduct approximately 1 cm above the tubo-uterine junction and a poly-vinyl catheter (I.D.=

PAGE 55

46 Table 2. Average total uterine protein recovered and plasma progesterone and estradiol concentration from cattle on each day of the estrous cycle.

PAGE 56

47 1.25 mm) was inserted through this incision into the oviduct, through the tubo-uterine junction and into the uterine lumen. A glass syringe was connected to the catheter and 25 ml of 0.33 M NaCl was carefully injected into the lumen of the uterine hona. The uterine horn containing the hypertonic saline was massaged prior to the saline being withdrawn into the syringe under slight negative pressure. Flushings from both uterine horns were pooled and stored in an ice bath until centrifugation at 10,000 rpm for 10 minutes at 3°C. The flushings were sterilized by filtration through a 0.20 u Millipore filter, concentrated to 1 to 2 ml by vacuum dialysis and dialyzed against 0.05 M phosphate citrate buffer, pH 7.4, for 36 to 48 hours at 3°C. The samples were then stored at -20°C until further analysis. Protein concentration of each sample was determined by the method of Lowry et al. (1951), and the total quantity of protein in each sample calculated. Total recovered protein for each sam.ple was adjusted if percent recovery of the saline flush was less than lOO'o (total recovered protein divided by percent recovery). Sephadex G-200 gel filtration was used to fractionate a portion of each uterine protein sample to obtain a protein profile. Sephadex G-200 protein profiles were also obtained from plasma samples for comparison with the Sephadex G-200 uterine protein profiles. The columns used (1.5 x 90 cm with a bed height of approximately 85 cm) were calibrated using Blue Dextran (MW=2 ,000 ,000) , to determine the void volume,

PAGE 57

48 and aldolase (MW=15 8 ,000) , ovalbumin (MW=45,000) and ribonuclease A C^^V=13,700) as standard globular proteins of known molecular weight. Gel filtration was carried out at 3°C. The eluent was 0.05 M phosphate ci trate buffer, pH 7.4, with flow rates of approximately 6 ml per hour. An aliquot of each fraction (approximately 2,5 ml) was used for protein determination, and the resulting optical density of each fraction was plotted against the elution volume, at which the fraction occurred, to construct a protein profile. Polyacryl amide gel disc electrophoresis, using the basic procedure of Clarke (1964) , was used to further study the protein components of the uterine flushings. A 1% polyacrylamide gel, 5 m^m in diameter and 70 mm lor.g, was used without spacer or sample gels. Samples containing about . 5 mg protein were applied in 0.1 ml of 1 M sucrose and two drops of bromophenol blue directly on the separating gel. The samples were run in 0.05 M Tris 0.38 M glycine buffer, pH 8.0, at a constant current of 2.5 mA per tube and a nominal voltage of 90 volts for about 1,5 hours or until the marker dye had migrated about 65 mm. The gels were stained with amido black solution (1 gm per 100 ml of 1% acetic acid) for 1 hour and destained electrophoretically in Vo acetic acid. The electrophoretic mobility of uterine proteins was expressed in terms of their mobility relative to that of albumin (Rf) when albumin was assigned an Rf value of 1.0. Electrophoresis of corresponding bovine plasma sam.ples from each female was also conducted, using the same procedure, for direct comparison with each uterine protein polyacrylamide gel.

PAGE 58

49 For analysis of steroids, approximately 2,000 cpm each 3 of progesterone1, 211 (Amersham/Searle , 51.7 Ci/imM) and estradiol-173-6 , 7 "^H (xNew England Nuclear, 40 Ci/mM) were added to 10 ml of plasma. These isotopic steroids served as internal standards for correction for procedural losses. The mixture was extracted vigorously three times with two volum.es of freshly distilled diethyl etlier. Progesterone and estradiol in the ether extract were isolated by chromatography on 1 X 40 cm Sephadex LH 20 columjis , as described by Chenault (1973) , and stored at 4°C until assayed. The solvent elution system was chloroform: ethanol (96:4). Aliquots of isolated progesterone w'ere quantified by competitive protein binding assay (Murphy, 1967). Details of the procedure were described by Gwazdauskas (19/2) and Chow (19 72). The isolated estradiol was quantified by radioimmunoassay (Hotchkiss, Atkinson and Knobil, 19 71) using an antiserum produced by immunization of sheep with tlie conjugate 1, 5, 5 (10) es tratriene3 , 176-diol, 17B -succinyl-bovine serum albumin. Details of the procedure were described by Chenault (19 73). Least-squares procedures (Harvey, 1960) were used to evaluate the effects of da.y of estrous cycle, breed and parity on the variables measured (total uterine protein and plasma progesterone and estradiol concentration) for days to 20, to 3 and 4 to IS of the estrous cycle. Regression curves were calculated and plotted using the individual The estradiol antiserum was generously supplied by Drs T. Nett and L. Estergreen of Washington State University, Pullman .

PAGE 59

50 observations for total uterine protein recovered and for plasma progesterone concentration on days to 20 of the estrous cycle. Correlations were determined for progesterone and estradiol concentration with total uterine protein and with the presence of a Rf 0.35 uterine protein band following electrophoresis. Results and Discussion Volumes of uterine flushings recovered did not vary significantly due to day of the estrous cycle. An average of 97.2 ± 9.7% of the volume of saline introduced into the uterine lumen was recovered (Table 1). This was comparable to the surgical recovery of uterine flushings in gilts by Murray et al. (19 72a) who used a similar flushing technique in anesthetized pigs . Concentrated uterine flushings obtained during proestrus , estrus and early metestrus had a faint yellow color, but samples collected during diestrus were clear. The yellow color was thought to be due to albumin and/or a low concentration of hemoglobin in the sample as would be expected during metestrous bleeding. Schultz et al^. (1971) reported uterine samples during metestrus as being consistently red and brownish-red with blood elements present. Average total uterine protein recovered with the "surgical-like" postslaughter flushing technique used in this study (Table 1) was 4.64 ± 4.35 mg. This compares favorably with the 7.5 mg average total uterine protein recovered by Roberts

PAGE 60

51 and Parker (19 74a) with a sirailar postslaughte r flushing technique during the first 2 to 3 weeks of pregnancy. Average total uterine protein recovered and average plasma progesterone and estradiol concentrations from the cattle on each day of the estrous cycle are presented in Table 2. During days to 20 of the estrous cycle a significant (P<.025) difference was found in total uterine protein recovered (Table 5). The differences in plasma progesterone concentration during days to 20 approached significance (P<.10), but there were no significant differences in plasma estradiol concentration during any period of the estrous cycle (Tables 3, 4 and 5). Flushings from cows yielded significantly (P<.053 more total uterine protein than those from heifers on days 4 to 18 of the estrous cycle as shown by the significant parity effect in Tabic 4. During days to 3 of the estrous cycle there was a si gnif i caiit (P<.02 53 breed effect on plasm.a progesterone concentration. Figure 1 shows the total uterine protein averages on days to 20 of the estrous cycle. Also, the fifth-order regression curve best describing the individual observations is shovv'n. Total uterine protein tended to be higher during the middle and late luteal phase than during the early luteal phase of the estrous cycle. Least-squares analysis of data obtained between days 4 and 18 of the estrous cycle (Table 4) revealed that differences in total uterine protein recovered during this period approached significance (P<.10}.

PAGE 61

52 rt T3 O M-l C O rt •H O > t3 O o in •H V) CO !fl S O 0) +-) I— I ^ o u D O CO O !/) P -H O t/1 nj s-H OJ +J 4-1 .-] o a> •H o 13 o o cd ci! ::o a, o -p o

PAGE 62

53 tS

PAGE 63

54 Table 5. Least-squares analysis of variance: data obtained between days and 3 of the estrous cycle.

PAGE 65

56 T(0 _ o w CD s^ (9IAI) NGlOdd IVIOI

PAGE 66

57 Total uterine protein recovered was highest on days and 20 of the estrous cycle. The Sephadex G-200 protein profiles and polyacry lamide electrophoret ic gels of uterine samples with lovv corresponding plasma progesterone concentrations on these two days closely resembled those of corresponding plasma. Two cows had apparent functional luteal tissue on day 20 as evidenced by gross appearance of corpus luteum and plasma progesterone concentration. The uterine fluid vSephadcx G-200 protein profiles and polyacrylamide elect rophoretic gels of these two cows did not resemble those of corresponding plasma. These day and 20 quantitative and qualitative observations could possibly be explained by increased blood flow to the reproductive tract under the influence of elevated estradiol concentration (Abrams et_ aJ . , 1972; Abrams et_ a]_. , 1975; Gwazdauskas et_ ru . , 1974} if there was any movement of plasma proteins by diapedesis into the uterus. The .73 (P<.01) correlation found between plasma estradiol concentration and total uterine protein recovered on days to 3 (Table 6) also supports this explanation of the similarity between uterine fluid and plasma proteins on day of the estrous cycle. Figure 2 shows the plasma progesterone concentration averages on days to 20 of the estrous cycle. In spite of variable progesterone concentrations among females within days, the fourth-order regression curve shown in this figure which best describes the individual observations is similar to those reported by other workers

PAGE 67

58 Table 6. Correlation of plasma progesterone and estradiol concentrations with total uterine protein and presence of Rf 0.35 uterine protein band following electrophoresis . Hormone

PAGE 68

o

PAGE 69

60 8 00 to ^ UJ CJ CX) CO CvJ > o CO o q: (CO LxJ liO niAI/9N) 3N0d31S390dd

PAGE 70

61 (Stabenfeldt , Ewing and McDonald, 1969; Shemesh, Lindner and Ayalon, 1971; Wettemann et_ aJ . , 1972). The high mean plasma progesterone concentration on day 20 was due to inclusion of samples from two cows having apparently functional luteal tissue (based on ovarian examination at slaughter and corresponding lew plasma estradiol and high plasma progesterone concentrations). The plasm.a estradiol curve of average concentration on days to 20 of the estrous cycle is shov/n in Figure 3. As previously mentioned, there were no significant day effects on estradiol concentration during any period of the estrous cycle. Least-squares analysis of variance did not indicate significant breed or parity effects on estradiol concentration Variation in estradiol concentration between females (as evidenced by large standard deviations) could account for the dissimilarity between the estradiol curve in Figure 3 and that reported by Wettemann et_ al. (1972). Their cattle were sampled each day over the entire estrous cycle, whereas, because of slaughter, in the present study each observation was from a different animal. After correction for breed, parity and day effects by least-squares analysis on days to 20 of the estrous cycle, a significant (P<.05) correlation (r=.23) was found between plasma progesterone concentration and total uterine protein (Table 6). On days 4 to 18 the correlation was .44 (P<.05), indicating a greater effect of progesterone on total uterine protein recovered during the luteal phase of the estrous

PAGE 71

0)

PAGE 72

63 ("ll^/9d) 10iaVdlS3

PAGE 73

64 cycle. These correlations between progesterone concentration and total uterine protein suggest that progesterone may be causing the quantitative changes in bovine uterine protein secretion during the luteal phase of the estrous cycle as reported for pigs (Knight, 1972; Knight et_ a]^. , 1973b). A significant correlation was not found between estradiol concentration and total uterine protein after correction for breed, parity and day effects by leas tsquares analysis. Before correction, however, the correlation betv.'een estradiol concentration and total uterine protein was ,26 (P<.05) for days to 20 and .73 (P<.01) for days to 3 of the estrous cycle. As previously mentioned, uterine protein samples qualitatively resembled those of corresponding plasma on day 0. Therefore, tliis .73 correlation between plasma estradiol concentration and total uterine protein on days to 3 would be expected if, in fact, the uterine samples collected around estrus consisted primarily of plasma proteins moving into the uterine lum.en by diapedesis under the influence of an estrogeninduced increased blood flow at estrus. Electrophoretic data in this study indicated qualitative changes in the protein? of the uterus during the estrous cycle. Table 7 gives the number and percent of polyacrylamide gels of uterine protein containing electrophoretic bands not present in corresponding plasma gels. Figure 4 gives a representative sample of polyacrylamide gel electrophoresis toward the anode of uterine protein collected on days to 20 of the estrous cycle and of plasma from each third of the estrous

PAGE 74

65 Table 7. Number and percent of polyacryl amide gels of uterine protein containing electrophoretic liands not present in corresponding plasma gels. Day of Estrous

PAGE 75

Figure 4. Poly acrylamide gel electrophoresis toward the anode (+} of boidne uterine protein and plasma (P) collected on davs t( the estrous cycle. I'wo uterine proteins not appearing in plasna are indicated by the arrows U-^i • Albumin (A) and transferrins (T) are also indicated.

PAGE 76

© © 67 14 15 16 17 18 19 20 P © S

PAGE 77

68 cycle. Sixteen, (62^o) polyacryl amide gels of uterine protein samples obtained between days 13 and 20 of the estrous cycle had an electrophorctic protein band (Rf=0.55) which migrated just behind the transferrins (Figure 4). A similar protein was not present in corresponding plasma gels. This band was present in all gels of samples collected on days 15 and 16. Two uterine samples (both from cows having functional luteal tissue) on day 20 yielded this Rf 0.35 electrophoreti c protein band. The corresponding plasma progesterone concentrations from both cows were still elevated. Neither the day 19 samples nor the remaining day 20 uterine sample yielded this band; however, they had low corresponding progesterone concentrations. There was a highly significant fP<.01) difference in the presence of this protein band from days to 20 of the estrous cycle (Table 3), and a correlation of .46 (P<.01) was found between plasina progesterone concentration and its presence (Table 6). Based on its Sephadex G-200 gel filtration elution volume (eluted just prior to albumin), the miolecular weight of the uterine protein having an Rf of 0.35 was estimated to be approximately 100,000, The above correlation between progesterone concentration and the presence of the Rf 0.35 protein suggests that progesterone may also be causing the qualitative cb.anges in bovine uterine protein secretion during the luteal phase of the estrous cycle as reported for pigs (Knight, 19 72; Knight et_ a]_. , 1973b). A second uterine protein electrophoretic band (Rf=0.72) migrated between the transferrins and albumin (Figure 4) and

PAGE 78

69 did not appear in corresponding plasma gels. This protein band was found in all uterine protein samples collected from days to 20 of the estrous cycle. The electrophoretic mobility (Rf) of the protein represented by this band was similar to that reported for the rabbit uterine specific protein "blastokinin" (Bullock and Connell, 1973). Also, the electrophoretic band(s) representing hemoglobin found in the first study (Chapter 1) migrated to approximately the same position, i.e., between transferrin and albumin, as did this Rf 0. 72 band. One or two prealbumin protein bands (Rf=1.09 and 1.12) were present in 39 (58°6) of the uterine protein gels between days and 20 but not in plasma gels. An Rf 1.12 band is clearly visible on the day 11 gel just in front of the albumin baud at tiie bottom edge of tiie photograph in Figure 4. Most of the prealbumin electrophoretic bands are not visible in the pliotographs or else appeared below the photographic field. There was also a prealbumin band (Rf=^1.27) present in all uterine protein gels and some plasma gels (very faint) that migrated with the marker dye used in the electrophoretic procedure. It was just as prominent in polyacry lamide gels of nonsurgical uterine flushings in the first study. This Rf 1.27 band appeared at the bottom of the gels and, therefore, is not visible in Figure 4. For each major protein band on polyacrylamide gels of day 70 bovine allantoic fluid (F. W. Bazer and W. W, Thatcher, unpublished data) there was a similar protein band present in

PAGE 79

70 gels of uterine protein samples obtained in this study. The bands included were the Rf 0,35 protein, transferrin, Rf 0.72, albumin (Rf=1.0) and a band tliat migrated between the Rf 0.72 band and albumin. In some uterine protein and plasma gels there were two bands instead of one visible in the area between the Rf 0.72 band and the albumin band. The polyacrylamide gels of bovine day 70 allantoic fluid were also similar to day 35 porcine allantoic fluid (F. V/. Bazer and W. W. Thatcher, unpublished data). There were no prealbumin bands present on poly aery lamide gels of bovine or porcine allantoic fluid which had not been processed, i.e., lyophilized or vacuum dialyzed. No significant differences in the size classes of proteins present between days and 20 of the estrous cycle were found in Scphadex G-200 gel filtration uterine protein profiles. There was, however, some slight trailing of two low molecular weight protein fractions which were never present in plasma samples. These two "low-profile" fractions were estimated to be in the 10,000 to 15,000 and <10,000 molecular weight range. Figure 5 shows typical protein profiles following Sephadex G-200 gel filtration of uterine samples collected on days 0, 5 and 15 of the estrous cycle and of bovine plasma. The Rf 0.35 protein was eluted by Sephadex G-200 gel filtration just ahead of Fraction III (predominantly albumin, 75 to 95 ml) and probably contributed to the indistinct separation between Fractions II and III (75 to 85 ml) on the day 15 protein profile in Figure 5. Based on polyacrylamide gels of

PAGE 80

Figure 5. Typical Sephadcx G-200 gel filtration protein profiles of bovine uterine protein collected on days 0, 5 and 15 of the estrous cycle and boviiie plasma.

PAGE 81

72 UJ o a. -J o o UJ o UJ Q_ lOl 510DAY-15 50 60 70 80 90 100 110 120 130 140 150 T 1 1 1 1 \ 1 1 1 T 50 60 70 80 90 100 110 120 130 140 150 •i 1 1 1 — I 1 1 1 1 — I r50 60 70 80 90 100 110 120 130 140 150 lOH PLASMA 50 60 70 80 90 100 110 120 130 140 150 ELUTION VOLUME (ML)

PAGE 82

73 concentrated (vacuum dialysis) pooled Sephadex G-200 fractions of the uterine protein samples, the Rf 0.72 protein apparently contributed predominantly to Fraction IV (95 to 125 ml portion of the uterine protein profiles in Figure 5), and the prealbumin electrophoretic proteins (Rf=1.09, 1.12 and 1.27) apparently contributed predominantly to Fraction V (125 to 150 ml portion of the uterine protein profiles in Figure 5). On some poly acrylamide gels of these pooled Sephadex G-200 fractions there were two electroplioreti c bands in the Rf 0.72 area. In the first study (Chapter 1) hemoglobin occurred on the polyacrylamide gels as either one or two bands in this same area (some cattle in this study apparently had two molecular species of hemoglobin). This observation could be an indication that the Rf 0.72 protein is hemoglobin or, at least, that in some samples in this study there was some hemoglobin contamination. It was noted that if the Rf 0.72 protein was not hemoglobin it would be very difficult to separate the two by either Sephadex G-200 gel filtration or by polyacrylamide gel electrophoresis. Beier and BeierHellwig (19 73) draw attention to the difficulty of differentiating bet^v'een a specific postalbum.in fraction and hemoglobin artifacts using electrophoretic techniques. Recently F. W, Bazer and W. W. Thatcher (unpublished data) collected bovine uterine fluid on days 5 and 15 of the estrous cycle using the same procedure described in this chapter, but the samples were concentrated by f reeze-drying instead of vacuum dialysis. Sephadex G-200 gel filtration

PAGE 83

74 of these freeze-dried samples and day 35 bovine allantoic fluid yielded protein profiles which were quite sir.ilar to each other, but these profiles had considerably larger Fraction V peaks than did the Sephadex G-200 protein profiles of uterine flushings obtained in this study. This fraction is comparable to the "low-profile" Fraction V (125 to 150 ml) of this study and they both were eluted near the total volume of the column. \\l\en this fraction was reconcentrated and subjected to polyacrylamide gel electrophoresis the resulting gels apparently consisted predominantly of a prealbumin band. This fraction was yellow in color and the iiiajor portion of the protein was lost upon dialysis, thus indicating a very small protein molecule (<10,000 MV;j , The question is immediately raised as to the accuracy of the evaluation of quantitative and qualitative changes in the actual protein milieu of the uterine lumen in the present study because of the possibility that low molecular weight proteins may have been lost due to concentration by vacuum dialysis and subsequent dialysis using dialysis tubing that allowed loss of <10 ,000 MW protein m.olecules. Total uterine protein from samples concentrated by freezedrying (F. V/. Bazer and VL W. Thatcher, unpublished data; Roberts and Parker, 19 74a) was not different from total uterine protein from samples in this study concentrated by vacuum dialysis. Roberts and Parker (1974a) found that differences between fresh, untreated serum and uterine washings largely disappeared when the uterine washings were compared with

PAGE 84

75 dialyzed and freeze-dried serum. Thus, it is possible that the Fraction V protein(s) discussed above may be the result of some change in a uterine protein (s) that is brought about by concentration, with freezedrying causing a greater change than vacuum dialysis. The above observations also raise the question of whether or not differences between untreated plasma and uterine flushings found in this study would have been the same if the plasma samples had been diluted in 0.33 M saline and processed tlie same as the uterine flushing samples. Answers to these questions will come after additional research determines the actual make-up of Fraction V; i.e., whether it is composed of an original uterine protein (s) or whether it is made up of altered uterine protein due to changes in the original protein (s) brought about by processing of uterine flushings . This study indicated that both qualitative and quantitative aspects of bovine uterine protein secretion changed during the luteal phase of the estrous cycle. Correlation of total ' uterine protein and presence of the Rf 0.35 protein in uterine flushings with plasma progesterone concentration suggested that progesterone is inducing these quantitative and qualitative changes during the luteal phase of the estrous cy c 1 e . The Rf 0.35 protein was present in the bovine uterus in this study at a time which corresponds with rapid expansion of the blastocyst and specifically elongation of the trophoblast (Winters, Green and Comstock, 1942; Chang, 1952). In

PAGE 85

76 the pig, trophoblast grov/th and developnient has been associated with a purple uterine protein which increases quantitatively during the latter half of the luteal phase of the porcine estrous cycle (between days 12 and 16) and which appears in allantoi.c fluid by day 30 of pregnancy (Murray, 19 71; Murray e_t_ £!_. , 1972a; Chen e_t a]_. , 1973a; Schlosnagle et_ al_, , 19 74; Chen et_ a]^. , 19 75). Thus, it is possible that the Rf 0.35 protein could be involved in rapid expansion of the blastocyst that begins to occur about day 12 of pregnancy in the bovine, and the subsequerit rapid elongation of the trophoblast, which occurs at a time corresponding to the latter one-third of the normal estrous cycle. It is also possible that the Rf 0.35 protein may in some way be associated w-ith the ]uteolytic process as it v/as present prior to corpus luteum regression when progesterone concentration was high and was absent after corpus luteum regression when progesterone concentration v.'as low. A possible function, if the protein is associated v;ith CL regression, couldbe binding of PGF^ , which is luteolytic in the bovine ^ 2a ' (Lauderdale, 1972; Rowson et al_ . , 1972; Hansel et al_. , 1973; Inskeep, 19 73; Chenault , 19 73; Lauderdale et_ al_. , 19 74), and transfer of PCF-, into the chorioallantoic membranes and/or 2a allantoic fluid. The Rf 0.35 protein is apparently present in day 35 bovine allantoic fluid (F, W, Bazer and W. W. Thatcher, unpublished data). Thus, if an embryo is present in the uterus, sufficient PGF^_ could be transferred into the chorioallantoic membranes and/or allantoic fluid to prevent

PAGE 86

77 luteolysis. It is interesting to note that Knight (1972) suggests the possibility o£ luetolytic prostaglandin being attached to an electrophoretic coj?.ponent of porcine uterine flushing Fraction II (Murray, 1971), clianging its mobility. He noted the presence of an electrophoretic protein band (Rf=0,53) on day 13 in uterine flushings of progesterone treated ovariectomized gilts that was not present on days 7, 9 or 11. This coraponent of Fraction II was shown to be maintained as long as progesterone treatment was continued. The appearance of this band occurred when luteolytic activity would be expected to occur (Schomberg, 1967), and it was present in a fraction reported to have luteolytic activity (Barber, 19 72) with a similar estimated molecular v/eight to t}ie luteolytic component of porcine uterine fluid reported by Schomberg (1969), The Rf 0.35 protein could also be associated with a possible adherence of the expanding blastocyst to the uterine wall during trophoblast expansion. It is logical to assume that a certain amount of adhesion must occur between the trophoblast and the uterine wall to permit the extensive elongation of the trophoblast through the uterine horn containing the blastocyst by about day 18 and into the adjacent horn by about day 20 (Winters et_ al_, , 1942; Chang, 1952). This speculation as to function of the Rf 0.35 protein found n this study is supported by the work of Mintz (19 70) and Pinsker and Mintz (19 73). They found a factor (estrogen dependent) in mice which is believed to be a proteolytic 1

PAGE 87

78 enzyme that induces blastocysts to adhere to the uterine wall prior to implantation. In the rabbit, protease activity has been found in the 6 -glycoprotein fraction of uterine secretion 24 hours before implantation, and a relation has been assumed betv/cen uterine protease (3-g]ycoprotein3 and the implantation of the blastocyst (Kirchner et^ a_l, 1971). The rabbit uterine specific B -glycoprotein described by Beier (19 74b} is a protease (MK= 10 0,000; similar electrophorctic mobility to Rf 0.35 protein in this study), is present in blastocyst fluid around implantation and is associated with the outer surface of the trophoblast. Thus, with similar electrophorctic mobility and molecular weight linking the rabbit uterine specific 3 -glycoprotein to the Rf 0.35 uterine protein of this study, tills speculation as to function of the Rf 0.35 protein is offered. Rapid expansion of tlie rabbit trophoblast prior to and during im.plantation can also be compared to rapid expansion of the bovine trophoblast coincident with the presence of the Rf 0.35 uterine protein during the latter half of the luteal phase of the estrous cy c 1 e .

PAGE 88

GENER.\L DISCUSSION Considerable research has established the importance of the uterus and its secretions in embryonic development (Beiei', 1974a, b). Sufficient data have accumulated on restriction of embryos to the oviduct, in^ vitro culture, synchrony of embryonic and uterine devclopm.ent and delayed implantation to support the concept that some uterine com.ponent (s ) is extremely important for continued normal de velopraent of the emibryo past the early blastocyst stage. Rabbit uterine fluid contains considerable protein. The volume of this fluid, total concentration of its macromolecular components and its protein patterns change continuously from ovulation to implantation (Beier, 1974a, b). Porcine uterine protein secretions change qualitatively and quantitatively during the estrous cycle (Murray, 1971). These rabbit and porcine uterine proteins (predomanately glycoproteins of less than 50,000 MW) are a product of selective filtration from plasma proteins and of biosynthesis by epithelial cells of uterine endometrial glands and endometrial surface epithelium (Beier, 1974a, b; Chen e_^ aj^, 1975). Rabbit blastocyst fluid contains the same proteins as those in uterine fluid (Beier, 1974b). By day 30 of pregnancy, porcine allantoic fluid contains a purple protein (Chen e_t al^. , 79

PAGE 89

80 1973a, b) specific to the uterine protein nilieu and polyacrylamide gels of porcine day 35 allcXntoic fluid are very similar to those of day 15 uterine fluid (F. W. Bazer, unpublished data). Thus, it would appear that uterine proteins of the rabbit and pig are transferred from the uterine lumen into the fluids and membranes of the early embryo. Secretion of these uterine proteins has been shov/n to be regulated by steroids, specifically progesterone in the rabbit and pig (Urzua e_t al . , 1970; Knight ej^ a]_. , 1973b; Gosivami and Feigelson , 19 74) . Early work with cattle has supported the concept that bovine uterine fluid is the result of active secretion and not merely a product of simple diffusion from blood (Schultz ^. ^L' 1971). Variation in its chemical composition with stage of estrous cycle (highest levels reported during the luteal phase) suggested hormonal control of bovine uterine fluid composition. In this study the first approach was to characterize uterine protein secretions of the bovine using a nonsurgical technique; however, blood contamination of the uterine fluid samples was indicated. Consequently, any uterine protein secretions that miglit have been present in tlie sam.ples -were probably masked by contaminating plasma proteins. Average total uterine protein collected in this first study (46,93 ± 41.31 mg) was similar to average total uterine protein recovered by earlier workers (Gupta, 1962; Heap, 1962; Schultz ej^ aj^. , 19 71). Furthermore, the qualitative

PAGE 90

81 indications of blood contamination indicated that uterine protein samples collected by this nonsurgical method did not represent the actual intraluminal uterine protein milieu. This was further indicated by the lower total uterine protein recovered by the "surgicallike" posts laughter flushing technique used in the second half of this study (4.64 ±4.35 mg) . This compares favorably with the 7.5 mg average total uterine protein collected during the first 2 to 3 weeks of pregnancy by Roberts and Parker (19 74a) wlio used a similar posts laughter flushing technique. In the second half of this study (Chapter 2) total uterine protein recovered was higliest on days and 20 of the estrous cycle and daily variation in recoverable protein was significant (P<.025}. A correlation coefficient of .73 between estradiol and total uterine protein during estrus and metestrus (day to 3) , the similarity between electrophore tic patterns of polyacrylamide gels and Sephadex G-200 protein profiles of uterine protein and plasma collected on day and the fact that highest total uterine protein collected was on day indicated the possible movement of blood proteins into the uterus by diapedesis at this tim>e when blood flow to the uterus is elevated as a result of elevated estrogen levels. During the luteal phase of the estrous cycle, total uterine protein tended to be higher during the middle and late luteal phase than during the early luteal phase. Significant (P<.05) correlation coefficients between plasma progesterone concentration and total uterine protein of .23, for days to 20,

PAGE 91

a ma 82 and .44, for days 4 to 18 o£ the estrcus cycle, suggest that progesterone may be influencing the quantitative change in bovine uterine protein secretion during the luteal phase. This would agree with data repoi'ted for pigs (Knight e_t al . , 1973b). Electrop?ioretic patterns on polyacrylamide gels and Sephadex G-200 protein profiles of uterine protein and plas collected on days to 20 of the bovine estrous cycle indicated qualitative changes in the uterine protein milieu. There was a highly significant (P<,01) difference in the pres ence of one uterine protein having an R£ of 0.55 from days to 20 of the estrous cycle. This protein was present only between days 15 and 20 (present in all samples on da)'s 15 and 16), hut was not present in corresponding plasma. Its presence was correlated (r=.46) with plasma progesterone concentration (P<.01). This correlation suggested that progesterone may be influencing the qualitative changes in bovine uter ine protein secretions during the luteal phase of the estrous cycle. Again, similar observations have been made for pigs (Knight e^ al^. , 1975b). Another protein present in bovine uterine samples, but not in corresponding plasm.a, had an Rf of 0.72 and was present throughout the estrous cycle. Electrophoretic miobility of this protein was similar to that reported for the rabbit uterine specific protein "blastokinin (Bullock and Connell, 1973). Taster (1974) reported the absence of a protein having an Rf of 0.76 in uterine endometrium of most nonpregnant heifers on days 3 and 12 of the

PAGE 92

83 estrous cycle as opposed to its presence in pregnant heifers. There were also one or two prealbumin protein bands present on many of the elect ropiioretic gels of uterine samples, but not plasma,between days and 20 of the estrous cycle that compare with prealbumin bands reported in the rabbit (Beier, 1974a, b). However, data were presejited and discussed in Chapter 2 which suggested that these prealbumin electrophoretic bands might be the result of processing procedures, i.e., concentration by ]yophi lii:ation or vacuum dialysis. This study has indicated that both qualitative and quantitative aspects of bovine uterine protein secretion change during the luteal phase of the estrous cycle. Correlation of total uterine protein and presence of the Rf 0.55 protein in uterine flushings with plaSiiia progesterone concentratioii suggests that progesterone is inducing these qualitative and quantitative changes during the luteal pliase of the estrous cycle. Porcine utei-ine protein secretions were first suggested to be related to growth and expansion of the trophoblast by Murray (19 71). Subsequent work by Chen (19 73} indicated that porcine uterine proteins were involved in some aspect of placental development. More recent work has associated trophoblast growth and development with a purple, uterine specific, protein which increases quantitatively during the latter half of the luteal phase of the porcine estrous cycle (between days 12 and 16) and appears in allantoic fluid by day 30 of

PAGE 93

84 pregnancy (Chen e;^ al_. , 1973a; Schlosnagle et_ al . , 1974; Chen ejt al . , 19 75) . Knight e_t a_l. (19 74b) reported that increased uterine protein secretions, resulting from higii progesterone levels, enhanced placental development and allantoic fluid volume. In more recent work. Knight (19 75) concluded that development of adequate placental mass was apparently the key factor necessary for adequate and sustained fetal growth and development. In cattle, Bellows et_ aJ . (1974) reported that twin fetuses were associated with lower cotyledon weights than singles and suggested that the ability to exchange metabolites between dam and fetus could affect birth weight or become a limiting factor in maintenance of multiple pregnancies. The Rf 0,3? protein cf this study was present in the bovine uterus at a tim.e which corresponds with rapid expansion of the blastocyst and, specifically, elongation of the trophoblast (Winters ej^ al_. , 1942; Chang, 1952). Therefore, it is possible that tlie Rf 0.35 protein could be involved with trophoblast grov/th and development as is the purple uterine specific protein in pigs. Data were presented and discussed at the end of Chapter 2 which provide a basis for speculation relative to two other possible functions or roles for the Rf 0.35 uterine protein. These were (1) a possible association with the luteolytic process and (2) a possible adlierence of the expanding blastocyst to the uterine wall during trophoblast expansion.

PAGE 94

85 Embryo transfer in the cow (Rowson et_ al,, 1969) has shown that a uterus which has not contained an embryo for approximately the first half of the estrous cycle will support growth and development of a transferred blastocyst. Thus, the environment supplied by the normal nonpregnant uterus must be similar to that supplied by the pregnant uterus. In light of this and the relation between total uterine protein and the Rf 0.35 protein and progesterone concentration, it is not surprising that Boyd et al. (1969) found higher plasma progesterone levels in pregnant than in nonpregnant cows on day 16. They also reported a direct relationship between blastocyst length of day 16 em.bryos and plasma progesterone concentration. This supports the possibility of an association between bovine uterine protein secretions, regulated by progesterone, and trophoblast growth and development . The importance of synchrony between uterine protein secretions and embryonic growth and development may help to explain the poor viability of embryos resulting from superovulation of the cow.. Use of PMSG in the cow and pig results m very high levels of estrogen before estrus and a delayed decline to normal levels after estrus (Lamond and Gaddy, 19 72; Spilman ej^ al . , 1973; Henricks et al . , 1973; Guthrie, Henricks and Handlin, 19 74), Estrogen treatment performed shortly after mating in the rabbit caused retardation of endometrial proliferation and secretion to such an extent that the normal pattern of protein secretion was delayed 3 to 4 days and

PAGE 95

86 fertility was nil (Beier, 19 74b3. High estrogen levels in superovulated cows could be causing a delay in normal uterine protein secretion patterns which would result in asynchrony o£ the uterus, (behind) and embryo. Additional research on uterine protein secretions of the cov/ may shed light on reasons for early embryonic death loss. Also, the possibility of a delay in uterine protein secretion of superovulated cows JTiay be determined. Such information could lead to more successful embryo transfer techniques. The identification of uterine specific proteins in the bovine m.ay lead to development of better embryo culture media for shortterm, storage of embryos. The overall effect of future research on bovine uterine secretions will, hopefully, result in a better understanding of those factors v'liich affect development of the bovine conceptus . This information is essential if improvements in animal fertility are to be realized.

PAGE 96

SUMMARY During the course of this study bovine uterine flushings were collected throughout the estrous cycle and exairdned for changes in total recoverable protein, Sephadex G-200 gel filtration protein profiles and elect ronhoreti c protein patterns. An at'cerfipt v.'as Fiade to correlate plasma progesterone and estradiol concentrations wiili qualitative and . quantitative changes in uterine protein secretions. In a preliir.inary study approxirr.ately seven uterine flushing samples per day of the estrous cycle were collected nonsurgjcally (total = 144) using a urethral catheter inserted through the cervix. An average of 84 ± 20-o of saline put into the uterus was recoi^ered and the samples appeared to be relatively free of blood contamination during most of the estrous cycle. However, Sephadex G-200 gel filtration protein profiles and electrophoretic polyacrylamide gels of the uterine flushings were very similar to those of plasma collected at the same time. Blood cont am.ination of the uterine fluid samples v/as indicated and it was thought that any uterine protein secretions that might have been present in the samples had been masked by contaminating plasma proteins. Therefore, in a second study uterine flushings were obtained from heifers and cows immediately following slaughter. 87

PAGE 97

88 In the second study, preslaughter plasma and iminediate postslaughter 0.33 M saline uterine flushings were collected from 67 cattle of mixed breeding (.Angu;" , Hereford, Holstein and BrahmanX British crosses). Three samples were collected for each day of the estrous cycle and 4 samples each were obtained on days 3, 5, 13 and 15. Plasmia samples were assayed for progesterone (competitive protein binding procedure) and estradiol (radioimrriunoassay) . Total uterine protein v/as determined by Lowry's method. Qualitative polyacryla'uide gel electrophoresis and Sephadex G-200 gel filtration protein profiles were obtained for each sample. During days to 20 of tlie estrous cycle a significant (P<.025) difference was found in total uterine protein present in uterine flushings. Flushings froii. cows yielded significantly (P<.05) more total uterine protein than those from heifers on days 4 to IS of the estrous cycle. After correction for breed, parity and day effects by leas tsquares analysis, on days to 20 of tlie estrous cycle a significant (P<.05) correlation (r=.23) was found between peripheral plasma progesterone concentration and total uterine protein. On days 4 to 18 of the estrous cycle the correlation was .44 (P<.05), indicating a greater effect of progesterone on total uterine protein recovered during the luteal phase of the estrous cycle. No significant correlation was found between estradiol concentration and total uterine protein after correction for breed, parity and day effects by least-squares analysis. Before correction, the correlation between

PAGE 98

89 estradiol concentration and total uterine protein was .26 (P<.05) for days to 20 and .73 (P<.01) for days to 3 of the estrous cycle. The .73 correlation between estradiol and total uterine protein during estrus and Rietestrus, the similarity between electrophoretic patterns of polyacrylamide gels and Sephadex G-200 protein profiles of uterine protein and plasma collected on day and the fact that highest total uterine protein collected vas on day indicated a possible movement of blood proteins into the uterus by diapedesis at this time when blood flow to the uterus is elevated as a result of elevated estrogen levels. Electroplioretic data indicated qualitative changes in proteins of the uterine flushings during the estrous cycle. Sixteen (e2''o) polyacrylaiaide gels of uterine protein samples obtained between days 13 and 20 of the estrous cycle had an electrophoretic protein band (Rf=0.35) that v/as not present in corresponding plasma gels. This band was present in all gels of samples on days 15 and 16. There was a highly significant (P<.01) difference in the presence of this protein band from days to 20 of the estrous cycle and a correlation of .46 (P<.01) was found betv/een the plasma progesterone concentration and its presence. Based on its Sephadex G-2G0 gel filtration elution volume (eluted just prior to albumin) , the molecular weight of the protein represented by this Rf 0.35 electrophoretic protein band was estimated to be approximately 100,000. A second uterine protein electrophoretic band (Rf=0.72) not appearing in corresponding plasma gels was

PAGE 99

90 found in all uterine protein gels from .days to 20. It migrated just ahead of the transferrins. The electrophoretic mobility of this protein was similar to that of the rabbit uterine specific protein "b las tokinin . " One or two prealbumin protein bands (Rf=1.09 and 1.12} were present in 59 (58-^) of the uterine protein gels between days and 20 but not in plasma gels. There was also a prealbumin band (Rf=^1.27) present in all uterine protein gels and some plasma gels (very faint) that migrated with the marker dye used in the electrophoretic procedure . No significant differences in size classes of proteins present betvveen days and 20 of the estrous cycle were found in Sephadex G-200 gel filtration uterine protein profiles. There was, however, some sliglit trailing of two low molecular weight protein fractions which were never present in plasm.a samples. These two 'low-profi le" fractions were in the 10,000 to 15,000 and <10,000 m.olecular weight range. In this study the data indicate that there are both quantitative and qualitative changes in the bovine uterine protein secretions during the estrous cycle. During the luteal phase these changes appear to be under the influence of proges terone .

PAGE 100

LIST OF REFERENCES Abranis, R.M. , D. Caton and F.V/. Bazer. 19 72. Effect of estrogen on vaginal blood flow in ewes. M\. J. Obstet. Gynecol. 103:629. Abraras, R.M. , "^ A^i . Thatclier, F.W. Bazer and C.J. Wilcox. 19 73. Effect of estradiol173 on vaginal thermal conductance in cattle. J. Daii'V Sci. 56:105 8. Adams, C.E. 1958. Egg development in the rabbit: the influence of postcoital ligation of the uterine tube and of ovariectoi.'iy . J. Endocrinol. 16:283. Adams, C.E. 1965. The influence of maternal environment on preimplan tation stages of pregnancy in the rabbit, p. 345 In: G.E.W. Wolstenholme and M. O'Connor (Eds.) Preimplant at ion Stages of Pregnancy. Little, Brown and Co., Boston. Adams, C.E. 1969. Egg-uterus interrelationships. Adv. Biosci. 4:149, Adams, C.E. 1971. The fate of fertilized eggs transferred to the uterus or oviduct during advancing pseudopregnancy in the rabbit. J. Rcprod. Fort. 26:99. Adams, C.E. 1973. Asynchronous egg transfer in the rabbit. J. Re prod. Fert. 35:613. Albers , H.J. and M.N. Castro, 1961. The protein components of rat uterine fluid. An analysis of its antigens by Immunoelectrophoresis and Ouchterlony gel diffusion technique. Fert. Steril. 12:142. Alden, R.H. 1942. Aspects of the egg-ovaryoviduct relationship in the albino rat. II. Egg development within the oviduct. J. Exp. Zool. 90:171. Anderson;, L.L. 1966. Pituitary -ovarian-uterine relationships. J. Reprod. Fert., Suppl. 1:21. Anderson, L.L., K.P. Bland and R.M. Melampy. 1969. Comparative aspects of uterine -luteal relationships. Rec. Prog. Hor. Res. 25:57. 91

PAGE 101

92 Anderson, L.L., A.M. Eowernian and R.M. Melampy. 196 3. Neuroutero-ovarian relationships, p. 345. In: A. Nalbandov (Ed.) Advances in Neuroendgcrinology . University of Illinois Press, Urbana. Arthur, A.T., B.D. Cowan aiid J.C. Daniel, Jr. 1972. Steroid binding to blastokinin. Pert, Steril. 25:85. Arthur, A.T. and J.C. Daniel, Jr. 1972. Progesterone regulation of blastokinin production and maintenance of rabbit blastocysts transferred into uteri of castrate recipients. Pert. Steril. 25:115, Barber, Y.S. 1972. The effects of porcine uterine protein secretions en ovarian luteal cell function. M.S. Thesis, University of Florida, Gainesville, PI a. Bazer, F.IV., A.J. Cla^vson, O.IV. Robison and L.C. Ulberg. 1969 Uterine capacity in gilts. J. Aniia. Sci. 18:121.'' Beier, H.M. 1968. Uteroglobin: a hormone sensitive endometrial protein involved in blastocyst development. Biochim, Biophys. Acta. 160:2 89. Beier, H.M. 19 70. Protein patterns of endometrial secretion in the rabbit, p. 15 7. In: P.O. Hubinot, F. Lerov, C. Robyn and P. Peleux (Eds.) Ovo Implantation . Human Gonadotropins and Prolactin. Karger, New York. Beier, h'.M. 19 74a. Oviducal and uterine fluids. J. Reprod Pert. 37:221. Beier, H.M. 1974b. Ovarian steroids in em.bryonic development before nidation. Adv. Biosci. 15:199. Beier, H.M. and K. Beier-H'ol Iv/ig. 19 75. Specific secretory protein of the female genital tract, p. 404. In: E. Diczfalusy (Ed.) Karolinska Symposia on Research Methods in Reproductive Endocrinology. Karolinska Institutet Stockholm. ' Beier, H.M., W. Kuhnel and G. Fetry. 1971. Uterine secretion proteins as extrinsic factors in preimplantation development. Adv. Biosci, 6:165. Beier, H.M., U. Mootz and W. Kiihnel. 1972. Asynchronous egg transfer during delayed uterine secretion in the rabbit. The 7th Int. Congr. Anim. Reprod. Artif. Insem. . Miinchen 3:1891. Bellows, R.A., G.P. Kitto, R.D. Randel , R.E. Short and L.W. Vamer. 19 74. Conceptus development in superovulated beef heifers, J. Mim. Sci. 39:198. (Abstr.).

PAGE 102

93 Bhatt, B.M. and D.V/. Bullock. 1974. Binding of oestradiol to rabbit blastocysts and its possible role in implantation. J. Reprod. Pert. 39:65, Billington, iv'.D., C.F. Graham and A. McLaren. 1968. Extrauterine development of mouse blastocysts cultured in yJ-L^SL fi'om early cleavage stages. j'. Embrvol. Exp'.~~ Morph. 20:591. Bland, K.P. and B.T. Donovan. 1966. Uterus and control of ovarian function, p. 179. In: A, McLaren (Hd.J Advances in Reproductive i'hysiology. Vol. 1. Academic Press, New York. Blatchley, F.R. and B.T. Donovan. 1969. Luteolytic effect of prostaglandin in tiie guinea-pig. Nature 221:1065. Boyd, H. , P. Bacsich, A. Ycung and J. A. McCracken. 1969. Fertilization and embryonic survival in dairy cattle. Br. Vet. J. 12 5:87. Bredeck, }I.f;. and D.T. Mayer. 1955. Uterine phospliatase concentrations and tlieir relationsli j ;i to nu:uber and weight of embrvos in the rat. Mo. Aer. Exp. Sta. Res. Bui. 591. Brinster, R.L. 1965. A method f^r in vitro rvltivation of mouse ova from two-cell to b 1 as to c7Jt7~ E xp . Cell Res. 32:205, Bullock, D.W. and K.M, Connell. 1973. Occurrence and molecular v.eight of rabbit uteri.ne "b lastokinin. " Biol. Reprod. 9:125. Chang, M.C. 1949. Effects of heterologous sera on fertilized ra])bit ova. J. Gen. Physiol. 32:291. Chang, M.C. 1952. Develop)nent of bovine blastocyst with a note on implantation. Anat. Rec. 113:143. Chang, M.C. 1968. Reciprocal insemination and egg transfer between ferrets and mink. J. Exp. Zool. 168:49. Chen, T.T. 19 73. An iimiiunological approach for studying the effect of jjorcine uterine protein secretions on fetal and placental development. M.S. Thesis. University of Florida, Gainesville, Fla. Chen, T.T. and F.W. Bazer. 1973. Effect of antiserum to porcine fraction IV uterine protein on the conceptus. J. Anim. Sci. 37:304. (Abstr'.}.

PAGE 103

94 Chen, T.T., F.W. Bazer, J.J. Cetorelli, W.E. Pollard and R. Michael Roberts. 19 73a. Purification and properties of a progesteroneinduced basic glycoprotein from the uterine fluids of pigs. J. Biol. Chem. 248:8560. Chen, T.T., F.W. Bazer, B. Gebhardt and R.M. Roberts. 1975. Synthesis and movement of porcine purple uterine protein. J. Anim. Sci . 40:170. (Abstr.). Chen, T.T., F.iV. Bazer and R.M. Roberts. 1973b. A study of porcine lavender protein fraction IV. J. Anim. Sci. 37:30 4. (Abstr.). Chenault , J.R. 19 73. Transitory changes in plasm^a progestins, estradiol and L!I approaching ovulation and after pi'ostag] andin F2a injection in the bovine. M.S. Thesis. University of Florida, Gainesville, Fla. Chow, L.A. 19 72. Studies on reproductive and endocrine functions follovv'ing synclironis ation of estrus v/ith melengestrol acetate (MGA) in dairy heifers. M.S. Thesis. University of Florida, Gainesville, Fla. Clarke, J. T. 1964. Simplified "disc" (polyacryl amide gel) e lectropliores is . Mm. N.Y. Acad. Sci. 121:428. Cole, R.J. and J. Paul. 196S. Properties of cultured preimplantation mouse and rabbit emliryos, and cell strains derived from them, p. 82. In: G.E.W. Wolstenholme and M. O'Connor (Eds.) Pre imp Ian tat ion Stages of Pregnancy. Little, Brown and Co., Boston. Daniel, J.C. 1968. Comparison of e lectrophoreti c patterns of uterine fluid from rabbits and mamjnals having delayed implantation. Comp. Biochem. Physiol. 24:297. Daniel, J.C. 19 70. Dormant emlTrvos of mammals. BioScience 20:411. Daniel, J.C. 19 71a. Uterine proteins and embryonic development. Adv. Biosci. 6:191. Daniel, J.C. 19 71b. Growth of the preimplantation emhryo of the northern fur seal and its correlation with changes in uterine protein. De\'. Bio. 26:316. Daniel, J.C, 1972a. Blastokinin in the northern fur seal. Pert. Steril. 23:78. Daniel, J.C. 19 72b. Preliminary attem.pts to term.inate pregnancy by immunological attack on uterine protein. Experientia 28:700.

PAGE 104

95 Daniel, J.C. 19 72c. Local production of protein durin*? implantation in the rabbit. J. Reprod. Pert. 31:305, Daniel, J.C. and R.S. Krishnan. 1969. Studies on the relationship between uterine fluid components and the diapausing state of blastocysts from ir.ammals having delayed implantation. J. Exp. Zool. 172:267. Dey, S.K. and Z. Dickmann. 1974a. Estradiol -1 7g -hydroxysteroid dcliydrogenase activity in preimpl antation rat embryos. Steroids 24:57. Dey, S.K. and Z. Dickmann. 1974b. A^. 33 ^j.^^^.^^.^^^,^^^^^^ . ^ dehydrogenase activity in mouse morulae and' bias tocysts . Proc, 7th An. Meet. Soc, Study Reprod., p. 149. (Abstr,). Dhindsa, D.S. and P.J. Dziuk. 196S. Effect on pregntmcy in the pig alter killing embryos or fetuses in one uterine hor-n in early gestation. J. Anim, Sci. 27:122. Dickmann, Z. and V.J. DeFeo. 1967. The rat blastocyst during normal pregnancy and during delayed implantation, mcluamg an observation on the shedding of the zona pellucida. J. Reprod. Pert. 13:3. Dickmann, Z. and S.K. Dey. 1973. Two theories: the preimplantntion embryo is a source of steroid hormones controlling (1) morula-blastocyst transformation, and (2) implantation. J. Reprod. Pert. 35:615. Dickmann, Z. and S.K. Dey. 1974a. Evidence that t^-3Qhydroxysterci d dehydrogenase activity in rat blastocysts is autonomous, J. Endocrinol. 61:513. Dickmann, Z, and S,K. Dey. 19 74b, Steroidogenesis in the preiiiiplantation rat embryo and its possible influence on morula-blastocyst transformation and im.plantation, J. Reprod. Pert, 37:91. Dickmann, Z. and R.IV. Noyes. 1960. The fate of ova transferred into the uterus of the rat, J. Reprod. Pert, 1:19 7, El-Banna, A. A. and J.C. Daniel, Jr, 1972a. Stimulation of rabbit blastocysts in vitro by progesterone and uterine proteins in combination. Pert. Steril. 23:101. El-Banna, A. A. and J.C. Daniel, Jr. 1972b. The effect of protein fractions from rabbit uterine fluids on embryo growth and uptake of nucleic acid and protein precursors Pert. Steril. 23:105.

PAGE 105

96 Fahning, M.L., R.H. Schultz and E.F. Graham. 1966. A technique for the collection of uterine fluids from the live cow. Vet. Rec. 79:230. Fahning, M.L., R.H. Schult?. and F.F. Graham. 1967. The free amino acid content of uterine fluids and blood serum in the cow. J. Reprod. Fert. 15:229. Fawcett, D.W. 1950. The developivient of m.ouse ova under the capsule of the kidney. Anat. Rec. 108:71. Fawcett, D.IV., G.B. Ivislocki and C.M. Waldo. 1947. The development of mouse ova in the anterior cliamber of the eye and in tlie abdominal cavity. Amer. J. Anat. 81:413, Folstad, L. , J. P. Beiinett and R.I. Dorfman. 1969. The in vitro culture of rat ova. J. Reprod. Fert. 18:145. Gamow, F. and J.C. Daniel. 19 70. Fluid transport in the rabbit blastocyst. V/ilhelm Roux' Archiv 164:261. Ginther, O.J.,, CO. U'oody, S, Makaion, S. Janakiraman and L.E. Casida. J967, Effect of oxytocin administration on the oestrous cycle of unilaterally hys terectomi ::ed he i f e rs . J . Ren rod . Fert . 14:225. Coding, J.R., M.D. Cain, J. Cerini, M. Cerini , W.A. Chamley and I, A. Gumming. 1972. Prostaglandin F2a "the" luteolytic hormone in the ewe. J. Reprod. Fert. 28:146. Goswarai , A. and Feigelson. 19 74. Differential regulation of a low-molecular-weight protein in oviductal and uterine fluids by ovarian hormones, Endocrinol. 95:569. Gupta, H.C. 1962. Biochemical and physiological properties of the cervical and uterine fluids of the cov: duri.ng estrus. Diss. Abstr, 25:803. Guthrie, H.D., D.M. Henricks and D.L. Hardlin. 1974. Plasma hormone levels and fertility in pigs induced to superovulate with PMSG. J, Reprod. Fert. 41:361. Gutknecht, G.D., L.J. U'yngarden and B.B. Pharriss. 1971. The effect of prostaglandin F2.::i on ovarian and plasma progesterone levels in the pregnant hamster. Proc. Soc. Exp. Biol. Med. 136:1151. Gwazdauskas, F.C. 1972. Characterization of bovine adrenal responsiveness to adrenocorticotrophin and various hormonal, physiological and environmental interrelationships at insemination affecting conception. M.S. Thesis. University of Florida, Gainesville, Fla.

PAGE 106

97 Gwazdauskas , F.C., R.M. Ab rams , W.W. Thatcher, F.W. Bazer and D, Caton. 19 74. Thermal changes of the bovine uterus following administration of estradiol-173-'»2 . J. Anim. Sci. 39:87. Hamana, K. and E.S.E. Hafez. 1970. Disc electrophoretic pattern? of uteroglobin and serum proteins in rabbit blastocoelic fluid. J. R.eprod. Pert. 21:555. Hammond, J., Jr. 1949. Recovery and culture of tubal mouse ova. Nature, bond. 163:28. Hamner, C.E. 1970. Composition of oviductal and uterine fluids. Adv. Biosci. 6:143. Hansel, V/. , P.W. Concannon and J.H. Lukaszev/ska. 1973. Corpora lutea of the large domestic animals. Biol. Reprod. 8:222. Hansel, W. and S.E. Echternkamp. 1972. Control of ovarian function in domestic animals. Am. Zoologist 12:225. Haour, F. and B.B. Saxena. 1974. Detection of a gonadotropin in rabbit blastocyst before implantation. Science 185:444. Harvey, W.R. 1960. Least-squares analysis of data with unequal subclass frequencies. USDA, ARS 20. Heap, R.B. 1962. Some chemical constituents of uterine washings: a method of analysis with results from various species. J. Endocrinol. 24:367. Heap, R.B. and G.E. Lamming. 1960. Studies of the uterine environment of different species. I. Influence of ovarian hormones on the chemical composition of uterine secretions. J. Endocrinol. 20:23. Heap, R.B. and G.E. Lamming. 1962. The influence of ovarian hormones on some chemical constituents of the uterine washing of rat and rabbit. J. Endocrinol. 25:57. Heap, R.B. and G.E. Lamming. 1963. An acid-soluble component of uterine washings. J. Endocrinol. 27:265. Heap, R.B. and J.S. Perry. 19 74. The maternal recognition of pregnancy. Brit. J. of Hosp. Med. 12:8. Henricks , D.M. , J.R. Hill, Jr., J.F. Dickey and D.R. Lamond. 1973. Plasma hormone levels in beef cows with induced multiple ovulations. J. Reprod. Fert. 35:225.

PAGE 107

98 Homburger, F. , P, Berfeld, A. Tregier, M.S. Grossman and P. Harpel. 1963. Endometrial secretions. Ann. N.Y. Acad. Sci. 106:685. Hotchkiss, J., L.E. Atkinson and E. Knobil. 1971. Time course of serum estrogen and luteinizing hormone (LH) concentrations during the menstrual cycle of the rhesus monkey. Endocrinol. 89:177. Howard, E. and V.J. DeFeo. 1959. Potassium and sodium content of uterine and seminal vesicle secretions. Am. J. Physio]. 196:65. Hunter, R.H.F., C. Polge and L.E. A. Rowson. 1967. The recovery, transfer and survival of blastocysts in pigs. J. Reprod. Fert. 14:501. Inskecp, E.K. 1973. Potential uses of prostaglandins in control of reproductive cycles of domestic animals. J. Anim. Sci. 36:1149. Iritani, A., W.R. Gomes and N.L. VanDemark. 1969. Secretion rates and chemical composition of oviduct and uterine fluids in ewes. Biol. Reprod. 1:72. Johnson, M.H. 1972. The protein composition of secretions from pregnant and pseudopregnant rabbit uteri with and without a copper intrauterine device. Fert. Steril. 23:12 3. Johnson, M.H. , B.D. Cowan and J.C. Daniel, Jr. 1972. An immunologic assay for blastokinin. Fert. Steril. 23:93. Joshi, M.S. and I.M. Murray. 1974. Immunological studies of the rat uterine fluid peptidase. J. Reprod. Fert. 37:361. Junge, J.M. and R.J. Blandau. 1958. Studies on the electrophoretic properties of the cornual fluids of rats in heat. Fert. Steril. 9:353. Kane, M.T. and R.K. Foote. 1970a. Culture of twoand fourcell rabbit embryos to the blastocyst stage in serum and serum extracts. Biol. Reprod. 2:245. Kane, M.T. and R.H. Foote. 1970b. Culture of twoand fourcell rabbit embryos to the expanding blastocyst stage in synthetic media. Proc. Soc. for Exp. Biol. Med. 133:921. Kane, M.T. and R.H. Foote. 1970c. Fractionated serum dialysate and synthetic media for culturing 2and 4-cell rabbit embryos. Biol. Reprod. 2:356.

PAGE 108

99 Kirby, D.R.S. 1962. The influence of the uterine environment on the development of mouse eggs, J. Embryol Exp Morph. 10:496. Kirchner, C. 1972. Immune histologic studies on the synthesis of a uterine-specific protein in the rabbit and its passage through the blastocyst coverings. Pert. Steril 23:131. Kirchner, C. , C. Hirschhauser and M. Kionke. 1971. Protease activity in rabbit uterine secretion 24 hours before implantation. J. Reprod. Pert. 27:259. Knight, J.W. 1972. Effect of superovulation , unilateral ovariectomy-hysterectomy and progesterone-estrogen therapy on qualitative and quantitative aspects of porcine uterine protein secretions. M.S. Thesis. University of Florida, Gainesville, Fla. Knight, J.l\'. 19 75. Conceptus development in intact and unilaterally hysterectomized-ovariectomized gilts: interrelationships between hormonal status, placental development, fetal fluids and fetal growth. Ph.D. Dissertation. University of Florida, Gainesville, Fla. Knight, J.lv., P. IV. Bazer, V^ .Vi . Thatcher and C.J. Wilcox. 1974a. Steroid levels in gilts during gestation. J. Anim. Sci. 39:215. (Abstr.). Knight, J.W., P.W. Bazer andH.D. Wallace. 1973a. Effects of superovulation and unilateral ovariectomy-hysterectomy on porcine uterine protein secretions. J. Anim. Sci. 36:61. Knight, J.W., F.W. Bazer and H.D. Wallace. 1973b. Hormonal regulation of porcine uterine protein secretion. J. Auim. Sci. 36:546, Knight, J.W. , F.W. Bazer and H.D. Wallace. 1974b. Effect of progesterone induced increase in uterine secretory activity on development of the porcine conceptus. J. Anim. Sci. 39:743, Knight, J.W. , F.W. Bazer, H.D. Wallace and C.J. Wilcox. 1974c. Dose-response relationships between exogenous progesterone and estradiol and procine uterine protein secretions. J. Anim. Sci. 39:747.

PAGE 109

100 Krishnan, R.S. 1971. Effect of passive administration of antiblastokinin on blastocyst development and maintenance of pregnancy in rabbits. Experientia 27:955. Krishnan, R.S, and J.C. Daniel. 1957. "Bias tokinin" : inducer and regulator of blastocyst development in the rabbit uterus. Science 158:490. Krishnan, R.S. and J.C. Daniel, Jr. 1968. Composition of "blastokinin" from rabbit uterus. Biochim. Biophys. Acta. 168:579. Kunitake, G.M., R.M. Nakamura, B.C. Wells and D.L. Moyer. 1965. Studies on uterine fluid: I. Disc electrophoretic and disc-gel Oucl^terlony analysis of rat uterine fluid. Pert. Steril. 16:120, Lamond, D.R. and R.G. Gaddy, 1972, Plasma progesterone in cows vi/ith multiple ovulations. J. Reprod, Pert. 29:307. Laster, D.B. 1974. Uterine proteins and pregnancy in cattle, J, Anim, Sci, 39:216, (Abstr,). Lauderdale, J.W. 19 72. Effects of PGF2a on pregnancy and estrous cycle of cattle. J. Anim. Sci. 35:246, (Abstr,). Lauderdale, J,W., B,E, Seguin, J.N. Stellflug, J.R. Chenault , W.W. Thatcher, C.K. Vincent and A.F. Loyancano. 1974, Fertility of cattle follovdng FGF-?^ injection, J. Anim, Sci. 38:964. Lewis, W.H, and P,W. Gregory, 1929. Cinematographs of living developing rabbit-eggs. Science 69:226. Loe , W.C. 1970. Protein and amino acid content of uterine and oviductal fluid of dairy heifers. Ph.D. Dissertation. Louisiana State University, Baton Rouge, La. Lowry, O.H. , N.J. Rosebrough, A.L. Farr and R.J. Randall. 1951. Protein measurement with the Folin phenol reagent, J. Biol. Chem. 193:265. Lukaszewska, J.H. and W. Hansel. 1970. Extraction and partial purification of luteolytic activity from bovine endometrial tissue. Endocrinol, 86:261. Makler, A. and J.M. Morris, 1971, Effect of postcoital estrogen on uterine carbonic anhydrase. Pert, Steril, 22:204, Maurer, R,R., H, Onuma and R,H. Foote. 1970. Viability of cultured and transferred rabbit embryos. J. Reprod. Pert. 21:417,

PAGE 110

101 Maurer, R.R., R.H. IvTiitener and R.H. Foote. 1969. Relationships of in_ vivo gamete aging and exogenous hormones to early embryo development in rabbits. Proc. Soc. Exp. Biol. Med. 131:882. Mazer, R.S. and P. A. Wright. 1968. A hamster uterine luteolytic extract. Endocrinol. 83:1065. McCracken, J.A. , J.C. Carlson, M.E. Glevv^, J.R. Coding, D.T. Baird, K. Green and B. Samuelsson. 1972. Prostaglandin F2a identified as a luteolytic hormone in sheep. Nature 238:129. McGaughey , R.W. and F.A. Murray. 19 72. Properties of blastokinin: amino acid composition, evidence, for subunits, and estimation of isoelectric point. Pert. Steril. 23:39 9. McLaren, A. and J.D, Biggers. 1958. Successful development and birth of mice cultivated in_ vitro as early embryos. Nature, Lond, 182:877. Mintz, B. 1970. Control of embryo implantation and survival. Adv. Biosci. 6:317. Moor, R.M. and L.E.A. Rovvrson. 1966a. The corpus luteum of the sheep: functional relationship between the embryo and the corpus luteum., J. Endocrinol. 34:233. Moor, R.M. and L.E.A. Rowson. 1966b. The corpus luteum of the sheep: effect of the removal of embryos on luteal function. J. Endocrinol. 34:497. Muljono,M.P.E. , \{ .VI . Thatcher, F.W. Bazer and A.C. Warnick. 1974. Effect of PGF7a in hysterectomized gilts, J. Anim. Sci. 39:219. (Abstr.j. Murdoch, R.N. 19 72. Phosphomonoes terases and histamine in the uterus of the ewe during early pregnancy. Aust. J, Biol. Sci. 25:1289. Murphy, B.E.P. 1967. Some studies of the protein binding of steroids and their application to the routine micro and ultra-micro measurement of various steroids in body fluids by competitive protein-binding radioassay. J. Clin. Endocrinol. 27:973. Murray, F.A., Jr. 1971.' Characterization of protein secretions by the porcine uterus and their relationship to reproductive physiology. Ph.D. Dissertation. University of Florida, Gainesville, Fla.

PAGE 111

102 Murray, F.A. , F.W. Bazer, J.W. Rundell, C.K. Vincent, H,D. Wallace and A.C. Warnick. 1971. Developmental failure of swine embryos restricted to the oviducal environment. J. Reprod. Pert. 24:445. Murray, F.A., F.lv. Bazer, H.D. Wallace and A.C. Warnick. 19 72a. Quantitative and qualitative variation in the secretion of protein by the porcine uterus during the estrous cycle. Biol. Reprod. 7:314. Murray, F.A. and J.C. Daniel. 1973. Synthetic pattern of proteins in rabbit uterine flushings. Fert. Steril. 24:692. Murray, F.A., R.W. McGaughey and M.J. Yarus . 1972b. Blastokinin: its size and shape, and an indication of the existence of subunits. Fert. Steril. 23:69. Nicholas, J.S. 1942. Experiments on developing rats. IV, The growth and differentiation of eggs and eggcylinders when transplanted under the kidney capsule. J. Exp. Zool. 90:41. Noden, P. A., H.D. Hafs and W.D. Oxender. 1973. Progesterone estrus and ovulation after prostaglandin F2a in horses. Fed. Proc. 32:229. (Abstr.). Noske , I.G. and J.C. Daniel. 1974. Changes in uterine and oviducal fluid proteins during early pregnancy in the golden hamster. J. Reprod. Fert. 3 8:173. Olds, D.M. and N.L. VanDemark. 1957. Composition of luminal fluids in bovine female genitalia. Fert. Steril. 8:345. Onuma, H. , R.R. Maurer and R.H. Foote. 1968. In vitro culture of rabbit ova from early cleavage stages to tTie blastocyst stage. J. Reprod. Fert. 16:491. Orsini, M.W. and A. McLaren. 1967. Loss of the zona pellucida in mice, and the effect of tubal ligation and ovariectomy. J. Reprod. Fert. 13:485. Perkins, J.L., L. Goode, W.A. Wilder, Jr. and D.B. Henson. 1965. Collection of secretions from the oviduct and uterus of the eive. J. Anim. Sci. 24:383. Ferry, J.S., R.B. Heap and E.G. Amoroso. 1973. Steroid hormone production by pig blastocysts. Nature 245:45. Petzoldt, V. 1974. Micro-disc electrophoresis of soluble proteins in rabbit blastocysts. J. Embryol. Exp. Morph. 31:479.

PAGE 112

103 Pharriss, E.B. and L.J. V/yngarden . 1969. The effect of prostaglandin F2rx on the progestogen content of ovaries from pseiidopregnant I'ats. Proc, Soc. Lxp. Biol. Med. 130:92. Pincus , G. and R.E. Kirsch. 1936. The sterility in rabbits produced by injections of oestrone and related coiiipounds , Am. J. Physiol. 115:219. Pinsker, M.C. and B. Mintz. 1975. Change in cell-surface glycoproteins of mouse embryos before implantation. Proc. Nat. Acad. Sci. 70:1645. Pope, C.E. and B.N. Day. 19 72. Developm.ent of pig emibryos follovring restriction to the ampullar portion of the oviduct. J. Fveprod. Pert. 31:135. Renf ree , M.B. 19 72. Influence of the embryo on the marsupial uterus. Nature 240:4 75. Ringler, I, 1961. The composition of rat uterine luminal fluid. Endocrinol. 68:281. Roberts, G.P. and J.M. Pai"ker, 1974a. Macrom.olecular components of the luminal fluid from, the bovine uterus. J. Reprod. Pert. 40:291. Roberts, G.P. and J.M. Parker. 1974b. An investigation of enzymes and hormione -b inding proteins in the lumiinal fluid of the bovine uterus. J. Reprod. Pert. 40:305. Roblero, L. 19 73. Effect of progesterone i^n viyo_ upon the rate of cleavage of mouse embryos. J. Reprod. J-ert. 35 :153. Rowson, I.E. A. 19 70. Evidence for luteolvsin. Br. Med. Bull. 26:14. Rowson, I.E. A. and D.F. Bowling. 1949. An apparatus for the extraction of fertilized eggs from the living cow. Vet. Record 61 : 191. Rowson, L.E.A., R.A.S. Lawson, R.M. Moor and A. A. Baker. 19 72a. Egg transfer in the cow: synchronization requirements. J. Reprod. Pert. 28:427. Rowson, L.E.A. and R.M. Moor. 1966. Emibryo transfer in the sheep: the significance of synchronizing oestrus in the donor and recipient animal. J. Reprod. Pert. 11:207. Rowson, L.E.A. , R.M. Moor and R.A.S. Lawson. 1969, Fertility following egg transfer in the cow: effect of method, medium and synchronization of estrus. J. Reprod. Pert. 18:517.

PAGE 113

104 Rowson, L.E.A., R. Teruit and A. Brand. 19 72b. The use of prostaglandins for synchronization of oestrus in cattle J, Reprod. Pert. 29:145. Rundell, J . iv . 1969. I_n vitro culture of swine ova. M.S. Thesis. Louisiana State University, Baton Rouge, La. e Runner, M.N. 1947. Development of the mouse eggs in th anterior chamber of the eye. Anat. Rec. 98:1. Saxena, f i . B . , S.H. Flasan, F. llaour and M. Schmidt-Gollwitzer . 19 74. Radioreceptor assay of human chorionic gonadotropin: detection of pregnancy. Science 184:793. Schlosnagle, D.C., F.W. Bazer, J.C.Ni. Tsibris and R.M. Roberts 19 74. An iron-containing phosphatase induced by progesterone in the uterine fluids of pigs. J. Biol. Chem. 249: 7S74. Scliombcrg, D . 1'.' . 1967. A dem.onstr:^ tion i_2i vi tro of luteolytic activity in pig uterine flushings. J. Endocrinol. 38:359. Schomiberg, D.IV. 1969. The concept of a uterine luteolytic hormone, p. 38 3. In: K.W. McKern (Ld.) The Gonads. Appleton CenturyCrofts , New York. Schultz , R.H., H.B. Burcalow, M.L. Fahning, E.F. Graham and A.F. V,'ebei-. 1969. A karyometric study of epithelial cells lining the glands of the bovine eridomictrium. J. Reprod. Pert. 19:169. Scliultz, R.H., M.L. Fahning and E.F. Graham. 1971. A chemical study of uterine fluid and blood serum of normal cows during the oestrous cycle. J. Reprod. Fert. 27:355. Seamark, R.F. and C. Lutwak-Mann. 19 72. Progestins in rabbit blastocysts. J. Reprod. Fert. 29:14 7. Sellner, R.G. and E.W. V/icke rsham. 1970. Effects of prostaglandins on steroidogenesis. J. Animi. Sci. 31:230. (Abstr . ) , Shemesh, M. , H.R. Lindner and N . Ayalon. 1971. Competitive protein binding assay of progesterone in bovine jugular venous plasma during the estrous cvcle. J. Reprod. Fert. 26:167.

PAGE 114

10 5 Shih, H.E., J. Kennedy and C Huggin.. 1940. Chemical ^composition of uterine secretions. .-jii. J. Physiol. 130:287 Shirai, B. , R. li^ulca and Y. NotaVe. 1972. Analysis of human uterine fluid protein. Pert. Steril. 25:522. Speroff, L. and P.W. Ramwell. 1970. Prostaglandin stimulation of in vitro progesterone synthesis, J. Clin. Endocrinof. 30:345. Spilmfin, C.H., G.E. Seidei, Jr., L.L. Larson, G.R. Vukman and R.M. Foote. ]9 73. Progesterone, 20S -hydroxypregn4-en-3-one, and luteinizing hormone levels in superovulated prepuberal and postpuberal cattle. Biol. Reprod. 9:116. Squire, G.D., F.W. Bazer and P. A. Murray, Jr. 1972. Electro phoretic patterns of porcine uterine protein secretions during tlie estrous cycle. Eiol. Reprod. 7:321. Stabenfeldt, G.tl. , L.L. E^'/ing and L.E. McDonald. 1969. _ Peripheral plasma progesterone levels during the bovine oestrous cycle. J. Reprod. Pert. 19:433. Stevens, K.R., H.D. Hafs and A.G. Hunter. 1964. Immunochemical and electrophoretic properties of oestrous _ rabbit uterine fluid proteins obtained by uterine ligation. J. Reprod. Pert. 8:319. Tarkov;ski , A.K. 1961. Mouse chimaeras developed from fused eggs. Nature, Lend. 190:857. Tervit, H.R., D.G. IMiittingham and L.E. A. Rowson. 1972. Successful culture in vitro of sheep and cattle ova. J. Reprod. Pert. 30:4"9 3. Tuft, H.P. and B.G. Boving. 19 70. The forces involved in 'water uptake by tlie rabbit blastocyst. J. Exp. Zool. 174:165. Urzua, M.A., R. Stambough, G. Plickinger and L. Mastroianni, Jr. 19 70. Uterus and oviduct fluid protein patterns _ in the rabbit before and after ovulation. Pert. Steril. 21: 860. Warren, J.E., Jr., H.W. Hawk and W.F. Williams. 1971. ; Effects of homogenized embryos on lUD-induced luteal regression in the ewe. J. Anim. Sci. 32:496. Warren, M.R. 19 38. Observations on the uterine fluid of the rat. Am. J. Physiol. 122:602.

PAGE 115

106 Wettemann, R.P,, H.D. Hafs , L.A. Edgerton and L.V. Sv/anson. 1972. Estradiol and progesterone in blood serum during the bovine estrous cycle. J. .Anim. Sci. 54:1020. Wiitson, G.L. and F.A. Murray. 1974. Cell cultureof raammaiian endometriun; and synthesis of blastokin^n in vitro, Science 183:668. KTiitten, W.K, 1956. Cultujie of tubal mouse ova. '.ature, Lond. 177:96. IVhitten, Vv'.K. 1957. Culture of tubal ova. Natur-j , Lond. 179:1031. l^Tiitten, W.K. 19 70. Nutrient requirements for tlie culture of prei nip] ant atiou embryos in_ vitro . Adv. Biosci. 6:129, IMiitten, W.K. and J.D. Biggers. 196 S. Complete development in vi_tro_ of the pre implant ation stages of the mouse in a sim-ple che.niically defined medium. J. Reprod. Pert. 17:399. Whittingham, D.G. 1968. Developm.ent of zygotes in cultured mouse oviducts. I. The effect of varying oviductal conditions. J. Exp. Zocl. 169:391. William.s , W.F., J.O. Johnston, M. Lauterbach and B. Pagan. 196 7. Luteclytic effect of a bovine uterine powder on the coi'pora lutea, follicular development, a]\d progesterone synthesis of t"ne pseudopregnant rabbit ovary. J. Dairy Sci . 50:555. Wilson, L. , Jr., R.L. Butcher and E.K. Inskeep. 1972. Prostaglandin F2a i^^ bovine uterus during early pregnancy. Biol. Reprod. 7:105. (Abstr.). Wiltbank, J.N. and L.E. Casida. 1956. Alteration of ovarian activity by hysterectom.y . J. .Anim.. Sci. 15:154. Winters, L.M., \U\\\ Green and R.E. Comstock. 1942. Prenatal develoDi.ient of the bovine. Minn. Agr. Exn. Sta. , Tech. Bull. 151. Wu , D.H. and W.M. Allen. 1959. Maintenance of pregnancy in castrated rabbits by 17alpha-hydroxy-progesterone caproate and by progesterone. Pert. Steril. 10:439. Yasukawa, J.J. and R.K. Meyer. 19 66. Effect of progesterone and oestrone on the pre-implantation and implantation stages of em.brvo development in the rat. J. Reprod. Pert. 11:245.

PAGE 116

BIOGRAPHICAL SKETCH Albert Carter Mills, III, son of Albert C, Jr., and Kathleen R. Mills, was born in Zachary, Louisiana, on June 7, 194 3. He attended Zachary High School and was graduated in May, ]961. In September, 1961, he entered Louisiana State University and received a B.S. degree in zoology in May, 1966. In September, 1966, he began study in the Department of Aniiiial Science as a nonTnatriculating student. The author entered Louisiana State University graduate school in February, 196 8, and worked as a Graduate Research Assistant until he received an M.S. degree from the Department of ibiimal Science in August, 1969, \v'ith a major in physiology of reproduction and a miinor in dairy science. In September, 1969, he entered the University of Florida graduate school and has since worked to^^'ard a Ph.D. in reproductive physiology in the Animal Science Departiiient as a Graduate Research Assistant. From March to Septemiber, 1974, the author was employed by Ova I of Glasgow, Montana, during which timic he worked on the perfection of nonsurgical embryo transplant techniques in the bovine. Since October, 1974, he has been in the employ of Montroy Ranch Industries, Inc., of Big Timber, Montana. The intention of Mr. Mills and this company is to set up a commercial nonsurgical embryo transplant unit which 107

PAGE 117

108 will, in addition to transplanting enbryos, conduct research in the area of bovine embryo transfer. AlberL Carter Mills, III, is married to the former Dorothy Watts Mason and is the stepfather of Holly Morrow Mason and Jennifer Dorr It Mason. He is the father of Albert Carter Mills, IV. The author is a member of the American Society of Mimal Science,

PAGE 118

I certify that 1 have read this study and that in my opinion it conforms to acceptable standards oi scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of I'octor of Philosophy. / Alvin C. Warnick , Chairman Professor of Animal Science I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ij. 0. 'Fuller W. Bazer Associate Professor of Animal Science I certify that 1 liave read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for tlie degree of Doctor of Philosophy. Donald E. Franke Associate Professor of Animal Science 1 certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. iL.-.J-^ r^^ J} Ci^xJ\Ji' Donald H. Barron Professor of Obstetrics and Gynecology

PAGE 119

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Paul T. Cardeilhac \.^,uC-^ Associate Professor of Veterinary Science This dissertation was submitted to the Graduate Faculty of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. June , 19 75 /3/S' /X*^^^
PAGE 120

UNIVERSITY OF FLORIDA 3 1262 08666 275 5