INTERRELATIONSHIPS OF CERTAIN THERMAL AND ENDOCRINE PHENOMENA
AND REPRODUCTIVE FUNCTION IN THE FEMALE BOVINE
FRANCIS CHARLES GWAZDAUSKAS
A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL. OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
The author is sincerely grateful to Dr. W. W. Thatcher, Chairman
of the Supervisory Committee, for his guidance, assistance, encourage-
ment, patience and friendship during the study.
Gratitude is expressed to Dr. C. J. Wilcox for his invaluable
assistance with statistical analyses and preparation of this manuscript.
The author is grateful to Dr. R. M. Abrams for the close association
and assistance throughout this endeavor. A word of thanks is due Drs.
D. H>-Barron, F. W. Bazer, D. Caton and H. H. Head for suggestions and
moreover for their assistance in projects and the authors' increase in
knowledge as members of the Supervisory Committee.
A special thanks goes out to Dr. R. B. Becker for his encourage-
ment and friendship throughout the entire study. The writer is indebted
to Drs. P. S. Kalra, C. A. Kiddy and M. J. Paape for their assistance
during different phases of the experiments. Thanks are given to Mr.
J. P. Boggs, Mr. A. L. Green and Mr. J. E. Lindsey for their help in the
barn and with cattle handling; to Mr. M. Casey, Mrs. D. Clark, Mrs. L.
Owens and Miss N. Baldwin for laboratory assistance and Miss L.
Buzzerd for clerical assistance.
Gratitude is expressed to R. W. Adkinson, J. R. Chenault, H.
Roman, L. C. Fernandez, R. Eley, J. M. Knight, E. Muljono, E. G. Benya,
L. W. Whitlow, S. Chakriyarat and J. L. Kratz who as fellow graduate
- ii -
students assisted technically and encouraged the author academically.
The author wishes to express his gratitude and appreciation to
his wife, Judy, for her constant understanding and encouragement during
the course of his studies.
- iii -
TABLE OF CONTENTS
LIST OF TABLES vi
LIST OF FIGURES vii
SECTION I 3
REVIEW OF LITERATURE 3
Influences of Thermal Stress on Reproductive
Hormone Relationships of Uterine Blood Flow and
SECTION II 21
HORMONAL PATTERNS DURING HEAT STRESS FROM PGF2a
INJECTION THROUGH ESTRUS AND OVULATION AND FOLLOWING
ADRENAL SIMULATION BY ACTH IN HEIFERS 21
Materials and Methods 23
Results and Discussion 27
SECTION III 60
EXPERIMENT 1: THERMAL CHANGES OF THE BOVINE UTERUS
FOLLOWING ADMINISTRATION OF ESTRADIOL-178 60
- iv -
Table of Contents (continued):
Materials and Methods
Thermocouple Preparation and Calibration
Surgical Techniques and Experimental Protocol
Results and Discussion
EXPERIMENT-2: THERMAL CHANGES OF THE BOVINE UTERUS
FOLLOWING PGF2a INJECTION THROUGH ESTRUS AND OVULATION
Materials and Methods
Results and Discussion
SUMMARY AND CONCLUSIONS
LIST OF REFERENCES
LIST OF TABLES
Table Page No.
1 Physiological parameters of heifers in environmental 28
chambers at 21.3 C and 32.0 C.
2 Simple correlations between hormone measurements. '37
3 Physical characteristics of plasma in heifers at 49
21.3 C and 32.0 C.
4 Overall least squares analyses of variance for 96
hormones in heifers at 21.3 C and 32.0 C.
5 Plasma progestins (ng/ml) following PGF2a injection. 97
6 Plasma estradiol (pg/ml) following PGF2a injection. 98
7 Plasma estrone (pg/ml) following PGF2a injection. 99
8 Plasma LH (ng/ml) following PGF2a injection. 100
9 Plasma prolactin (ng/ml) following PGF2a injection. 101
10 Plasma corticoids (ng/ml) following PGF2a injection. 102
11 Plasma corticoids (ng/ml) prior to and following 103
200 IU ACTH.
12 Plasma progestins (ng/ml) prior to ACTH injection. 104
13 Simple correlations between hormones and temperatures. 105
14 Analysis of variance for aortic and uterine temperatures. 106
15 Hormonal and temperature measurements for G665 (G) and 107
- vi -
LIST OF FIGURES
1 Evaluation of thermal stress on transitory hormonal
changes in the bovine during the period of
luteal regression, estrus and ovulation:
2 Sequential changes in
at 21.3 C or 32.0
the LH peak.
3 Sequential changes in
at 21.3 C or 32.0
the LH peak.
4 Sequential changes in
at 21.3 C or 32.0
5 Sequential changes in
21.3 C or 32.0 C.
6 Sequential changes in
at 21.3 C or 32.0
plasma progestins in heifers
C synchronized to the time of
plasma estradiol in heifers
C synchronized to the time of
plasma estrone in heifers
C synchronized to the time of
plasma LH in heifers at
plasma prolactin in heifers
C synchronized to the time of
7 Sequential changes in plasma corticoids synchronized
to the time of the LH peak using pooled means of
heifers at 21.3 C and 32.0 C.
8 Transitory changes in plasma corticoids following
injection of 200 IU ACTH in heifers at 21.3 C
and 32.0 C.
9 Uterine and aortic temperature prior to and following
IV injection of 12 ml sterile physiological saline.
10 Uterine and aortic temperature prior to and following
IV injection of 3 mg estradiol-17B.
11 ATuterus-aorta prior to and following either 12 ml
saline or 3 mg estradiol-17B.
List of Figures (continued):
Figure Page No.
12 Uterine and aortic temperature prior to and after 70
injection of estradiol-17B from continuous
13 Uterine and aortic temperatures prior to and following 75
PGF2a injections in G665.
14 Uterine and aortic temperatures prior to and following 76
PGF2a injections in JN15.
15 Changes in ATua following PGF2a injections. 77
16 Uterine and aortic temperatures, LH and estradiol in 82
G665 and air temperatures.
17 Uterine and aortic temperatures, LH and estradiol in 83
JN15 and air temperatures.
18 Circadian uterine, aortic and air temperature changes. 85
19 Changes in AT associated with endogenous LH and 87
- viii -
Abstract of Dissertation Presented to the Graduate Council
of the University of Florida in Partial Fulfillment of the Requirei nts
for the Degree of Doctor of Philosophy
INTERRELATIONSHIPS OF CERTAIN THERMAL AND ENDOCRINE PHENOMENA
AND REPRODUCTIVE FUNCTION IN THE FEMALE BOVINE
Francis Charles Gwazdauskas
Chairman: W. W. Thatcher
Major Department: Animal Science
Ten normally cycling Holstein heifers were assigned to one of two
environmental treatment groups (21.3 C, 59% RH or 32.0 C, 67% RH). PGF2a
was used to cause luteal regression and synchronize estrus. Least-squares
analyses were conducted to characterize treatment, animal and within-
animal time trends in plasma progestins, estradiol, estrone, LH,
prolactin and corticoids.
Environmental treatment (32.0 C) evoked a 1.49 C increase in
rectal temperature and a 3.59 C increase in skin temperatures. Length
of estrus was shorter (P<.10) for the 32.0 C heifers. Two of four
heifers at 21.3 C inseminated were pregnant at 40 days compared to none
of five at 32.0 C.
Average progestin concentration between treatments were not differ-
ent (P>.10; .53 ng/ml at 21.3 C compared to .65 ng/ml at 32.0 C). Mean
estradiol concentrations were significantly (P<.10) lower in 32.0 C
heifers (3.45 pg/ml compared to 2.96 pg/ml). There was a significant
elevation (P<.05) of estrone due to heat stress (1.55 pg/ml compared to
1.85). No significant differences (P>.10) were found in mean LH con-
centrations between heifers at 21.3 C or 32.0 C. Preovulatory peak LH
- ix -
concentrations were 32.2 and 33.2 ng/ml plasma, respectively. All
animals had a preovulatory LH surge, suggesting that hyperthermia did
not prevent the triggering mechanism for LH release. Mean prolactin
(14.51 ng/ml at 21.3 C compared to 14.78 ng/ml at 32.0 C) and corticoid
(8.01 ng/ml at 21.3 C compared to 7.76 ng/ml at 32.0 C) concentrations
were not different between temperature treatments (P>.10).
In an attempt to determine if plasma dilution may have occurred,
total protein concentration and osmolality were measured. There was
no difference (P>.10) in total protein concentration or osmolality
between treatment groups. The affinity (K ) of cortisol for CBG was not
different between treatments (P>.10); however, the binding capacity of
CBG for cortisol was reduced (P<.05) in the 32.0 C heifers.
Results of this experiment showed only subtle thermal effects on
estradiol and estrone plasma concentrations and no effects on LH, pro-
gestins, corticoids and prolactin. Apart from possible hormonal involve-
ment with duration of estrus, heat stress did not appear to affect the
hormonal mileau in peripheral plasma associated with corpus luteum re-
gression, follicle growth and ovulation.
Eight days following ovulation in the last heifer, 200 IU ACTH was
injected, IV, into the 10 heifers. The 32.0 C heifers responded with
significantly lower (P<.10) corticoid concentrations. The 6th order re-
gression response curves were not parallel (P<.01) suggesting that the
hot group response was earlier to reach a peak (75 compared to 105 min.),
had a lower magnitude (73.5 compared to 100.2 ng/ml corticoids) and was
of shorter duration (4 compared to 5 hr.).
Because the first experiment did not specifically consider environ-
mental and hormonal effects on uterine temperature it was necessary to
document possible estrogen induced uterine thermal changes. In the second
experiment thermocouples were placed into the uterine serosa and saphenous
artery of four dairy heifers. Injection of 3 mg estradiol-178 caused a
.25 C decrease (P<.01) in the difference between uterine and aortic
temperature (ATu-a) by 2.5 hr. postinjection. In contrast, there was no
significant change (P>.10) in the ATu-a after injection of saline.
The final experiment was an attempt to document and evaluate changes in
uterine temperature during the period of luteal regression, follicle
growth and ovulation induced by PGF2a under conditions of a mild heat
stress. Thermocouples were placed into the uterine serosa and aortic
blood vessel of four dairy cattle. PGF2a caused an immediate drop in
uterine and aortic temperatures, and a decrease in the AT of almost
.4 C at 45 min. postinjection. The two cows, in which thermocouples
remained operational for the duration of the study, had monophasic daily
uterine and aortic temperature rhythms. However, both temperatures lagged
about 6 hr. behind air temperature changes. Uterine temperatures reached
40 C for periods of up to 6 hr. Failure to detect an association between
ATa and hormonal measurements may have been due to a time lag association.
Not until the preovulatory surge of LH was there an appreciable rise in
ATua (P<.01), and this occurred at a time when estradiol was decreasing.
The mild environmental heat stress may have contributed to the high
uterine and aortic blood temperatures.
- xi -
Reduced reproductive efficiency occurs during the hot seasons of
the year in many parts of the United States. Lowered conception rate
due to heat stress occurs over a prolonged period of the year in
Florida and represents a major production problem to dairymen. A 12
year study in the University of Florida dairy herd revealed a con-
ception rate per service of less than 40% (Gwazdauskas, Thatcher and
Wilcox, 1975). Economically, poor reproductive performance under con-
ditions of thermal stress decreases heifer replacement availability
and long term milk production and increases calving interval and
Before systems for reproductive management can be developed to
counter these adverse effects of heat stress, several fundamental
questions must be answered. Among these are:
1. How does a standard heat stress alter hormonal and phy-
siological responses during the normal estrous cycle? The objective
of the first experiment (Section II) was to characterize hormonal
changes progestinss, estradiol, estrone, LH, prolactin and corticoids),
rectal temperature, plasma protein concentration and osmolality and
plasma cortisol binding capacity (CBC) in heifers subjected to a
standard heat stress (21.3 compared to 32.0 C). In addition, plasma
- 1 -
- 2 -
corticoids in response to ACTH were measured to evaluate possible
thermal stress effects on adrenal responsiveness.
2. What are the factors influencing uterine temperature?
Can estradiol, which is known to have a marked effect on uterine blood
flow and metabolism, alter uterine temperature? What are the changes
in uterine temperature during the period of luteal regression, follicle
growth and ovulation under conditions of mild heat stress? In Section
III a series of experiments were designed in an attempt to answer these
Prostaglandin F2a (PGF2 ) causes luteal regression in the bovine
and has enabled the researcher to use it efficiently to control endo-
crine and physiological changes near the time of estrus and ovulation.
Such a compound maximizes use of experimental facilities over short
periods of time without adverse chronic alterations of normal bovine
physiology. In answering questions one and two above, PGF2a was used
to synchronize the hormonal events associated with corpus luteum re-
gression, estrus and ovulation.
REVIEW OF LITERATURE
Stress is defined as a condition harmful to an organism, which
results from inability of the organism to maintain a constant internal
environment (Taber, 1961). Factors involved in altering homeostasis
include trauma, surgical operations, restraint, extreme cold or heat,
intense solar radiation, social stress due to peck order, nutritional
stress and internal stress caused by pathogens or toxins (Hafez, 1968;
Guyton, 1966). The purpose of the initial review section is to report
on the effects of thermal stress on reproductive performance with major
emphasis on hormonal or endocrine aspects.
Influence of Thermal Stress on Reproductive Performance
Hot environments may exert their depressive effect on fertility
via the gonads, accessory sex glands, uterine environment, gametes or
endocrine system (Hafez, 1959; Ulberg and Burfening, 1967). Reproductive
behavior has been shown to be altered by heat stress, in that estrous
duration was shorter (Branton et al., 1957; Gangwar, Dranton and Evans,
1965; Hall et al., 1959), there was an increased frequency of quiet
ovulations (Labhsetwar et al., 1963) and anestrus (Bond and McDowell,
1972) and a reduction in estrous intensity (Gangw:ar, Branton and Evans,
- 4 -
High temperatures exert direct effects on fertilized ova grown
in vitro (Alliston et al., 1965). Fertilized ova grown through first
cell division at 40 C in vitro had a lower rate of embryo survival than
those grown at 38 C when returned to synchronized pseudopregnant recipient
rabbits. There were no morphological.differences between ova in dif-
ferent media. As the period of culture at 40 C was delayed to second
cell division, differences in post-implantation death losses disappeared.
An environment of 32.2 C and 65% Relative Humidity (RH) did not inhibit
estrus or alter ovulation rate in sheep (Alliston and Ulberg, 1961).
However, an increase in embryo death was detected when embryos (2-32
cell stage) were transferred from donor ewes kept at 32.2 C to recipient
ewes at 21.1 C ambient temperature. Embryo survival was highest when
both donor and recipient ewes were maintained at 21.1 C, indicating that
damage to the early embryo was most likely to occur in uteri of ewes
kept at 32.2 C. Heat stressing one or both parents of a mouse embryo
affected the rate of thymidine, uridine and guanine incorporation into
nucleic acids during pre-implantation development, which may lead to
altered DNA and RNA synthesis and subsequent embryonic mortality (Sheean,
Durrant and Ulberg, 1974).
Due to limitations of facilities and methodology, most investi-
gations of the effects of heat stress on hormonal balance and their
relationships to reproductive performance have monitored only one or
two hormonal responses. Therefore pooling results from different
laboratories by combining data from different animals within and between
species can be misleading when an effort is made to develop a hypothesis
on how thermal stress affects reproduction. Stott, Thomas and Glenn
- 5 -
(1967) found progesterone to be elevated on the day of estrus in
thermally stressed cows. Heifers maintained at 32 C and 21 C for
72 hr. beginning at the onset of estrus had conception rates of 0 and
48%, respectively, according to a report by Dunlap and Vincent (1971).
Associated with the decreased fertility was an elevated plasma progestin
concentration of only .42 ng/ml plasma (Mills et al., 1972). In con-
trast, Stott and Wiersma (1973) reported depressed plasma progestin
levels during chronic heat stress in the bovine.
Evidence of detrimental progesterone effects on embryo cleavage
stages has been reported by Dickman (1970). Fertilized ova were
transferred on day 4 of pregnancy to pseudopregnant rats which had
been ovariectomized on day 2. When transfer was preceded by 2, 3, 4,
5 or 6 days of progesterone treatment in pseudopregnant rats, 49, 38,
13, 2.5 and 2% of the transferred morula developed into fetuses.
However, when blastocysts were transferred, there was a 62.5% fetal
survival. Overstimulation with progesterone apparently interfered
with embryonic development. Johnsson et al. (1974) reported a 60 to
75% reduction in fertility in ewes receiving a single injection of
progesterone on days 0, 1, 2 or 3 or daily injections on days 1 to 4.
The progesterone given before day 4 may have affected embryo transport
through the oviduct or directly altered it and therefore inhibited or
abolished its ability to cope with the 'luteolysin' and prevent corpus
luteum regression during pregnancy.
Progesterone, superimposed on estradiol administration in ewes,
caused a prompt decrease in uterine blood flow (Greiss and Anderson,
1970). Extreme or prolonged limitation of blood flow to the vicinity
- 6 -
of the embryo can result in fetal death. Blockage of blood supply to
the uterus one day post coitum in the mouse was most detrimental to
implantation rate (Senger et al., 1967). These observations may have
been the result of the uterine coagulation procedure since necrosis of
the tissue was detected (F. W. Bazer, personal communication). However,
coagulation of blood vessels to one uterine horn resulted in 51% fewer
embryos migrating to the uterus by 4 days after mating in mice. At 10
days post mating there were fewer live fetuses on the coagulated side
compared to the control side (57% compared to 73%). Therefore, reduced
blood flow may be responsible for failure of embryo transport to the
uterus and increased fetal death rate (Bazer, Ulberg and LeMunyan, 1969).
Thus, if heat stress increased plasma progesterone levels, an altered
blood flow may be a factor associated with reduced fertility.
Heat stress has been shown to cause elevated plasma corticoid
levels in the bovine within 4 hr. of exposure which suggests increased
adrenal activity (Christison et al., 1970). Other workers reported
that plasma corticoids decreased during chronic heat stress in cattle
(Alvarez and Johnson, 1973; Christison and Johnson, 1972; Rhynes and
Ewing, 1973) and that corticoid turnover rates decreased (Christison
and Johnson, 1972). However, chronic hyperthermia resulted in elevated
epinephrine and norepinephrine, though corticoids were depressed, which
suggested a decreased sensitivity to physiological actions of
catecholamines (Alvarez and Johnson, 1973). Shayanfar (1973) compared
adrenal responsiveness to adrenocorticotropin (ACTH) in cows exposed
to ambient temperatures of greater or less than 21.1 C. At ambient
temperatures above 21.1 C, plasma corticoid response to ACTH was slower,
- 7 -
peak levels were lower and the response was of shorter duration. Yousef
and Johnson (1967) reported a 30 to 40% increase in heat production fol-
lowing injections of hydrocortisone acetate to cattle at 35 C ambient
temperature. Thus, during prolonged thermal stress, depressed plasma
corticoids and lowered adrenal responsiveness to ACTH may be indicative
of altered adrenal function.
Madan and Johnson (1971) have reported that the preovulatory peak
of plasma LH and basal plasma LH concentrations were lower in heifers
maintained at 33.5 C and 55% RH compared to those at 18.2 C and 55% RH.
These results support the hypothesis that thermal stress may alter
secretion or metabolism of various hormones associated with reproductive
function. However, Riggs, Alliston and Wilson (1974) detected a breed
difference in response of the pre-ovulatory surge of LH to heat stress
in gilts. Heat stressed pigs of the Duroc breed had no difference in
magnitude of the preovulatory plasma LH surge compared to controls,
whereas pigs of the Hampshire breed had a three to sixfold increase in
the preovulatory peak of LH compared to their controls. It appears
that species and breeds may,therefore,respond differently to thermal
stress and that inferences among species and breeds must be reviewed
Koprowski and Tucker (1973), Schams and Reinhart (1974) and
Thatcher (1974) found elevated peripheral plasma prolactin concentrations
during the hot months of the year, suggesting that photoperiod and
temperature modulate prolactin release. At a constant day length,
calves exposed to 27 C had significantly higher plasma prolactin con-
centrations than those exposed to 10 C, whereas at 27 C prolactin levels
- 8 -
were only slightly higher than at 21 C (Wetteman and Tucker, 1974).
However, Karg and Shams (1974) found a positive correlation between
day length and basal prolactin concentrations in male and female cattle.
Relkin (1972) demonstrated that changes in light:dark ratios for rats
altered pituitary content and plasma concentrations of prolactin. It
would appear that this question has yet to be resolved in cattle, i.e.
whether photoperiod or temperature is the primary factor affecting
plasma prolactin concentrations.
These various studies suggest that thermal stress may alter
circulating plasma concentrations of certain pituitary, adrenal and
ovarian hormones. Such excesses or deficiencies of these hormones may
influence certain reproductive phenomena and account for lowered
Environmental factors play an important role in bovine fertility.
Seasonal depressions in conception rate due to heat stress effects on
the male can be eliminated through A.I. (artificial insemination) in
which semen from bulls can be collected and frozen during cooler times
of the year. Under these circumstances Stott (1961) still found a
seasonal depression in breeding efficiency of cows which paralleled
high climatic temperatures in Arizona and California. Thus, this
experiment indicated that altered reproductive efficiency in the female
was the major contributor to summer depression of fertility.
In Florida, a year-long study was conducted to relate climatic,
rectal and uterine temperatures, plasma corticoid and progesterone
concentrations, breed, service number, time of service, sire and age to
conception rate (Gwazdauskas, Thatcher and Wilcox, 1973). Significant
- 9 -
effects due to environmental temperature on the day after insemination,
rectal and uterine temperatures at insemination, sire and days post-
partum were detected on conception rate. Deleting environmental
temperature from the statistical model revealed significant effects of
uterine temperature the day after insemination on fertility. Relationships
between uterine temperatures at insemination or the-day after insemination
with fertility were intriguing. Uterine temperature the day after in-
semination appeared to be positively associated with environmental
temperature on that day. Their inverse associations with fertility
may reflect direct detrimental thermal effects on early cleavage and
development of the embryo. In contrast, the association of uterine
temperature at insemination with fertility might be related to certain
physiological (uterine blood flow and vaginal thermal conductance) and
hormonal changes occurring at estrus that may be associated with proper
timing of insemination to achieve maximal fertility.
In a subsequent study with 12 years of data, effects on conception
rate of age of cow, inseminator, service sire, month, year, breed and
21 climatological variables were evaluated (Gwazdauskas, Wilcox and
Thatcher, 1975). Age, inseminator, sire and breed had significant
effects on conception. Maximum temperature the day after insemination,
rainfall the day of insemination, minimum temperature the day of insem-
ination, solar radiation the day of insemination and minimum temperature
the day after insemination were the five highest ranking climatological
variables associated with fertility. The most potent environmental
variable, maximum temperature the day after insemination, had a
significant curvilinear relationship with fertility. As maximum
- 10 -
temperature increased from 21.1 to 35 C, conception rates declined from
40 to 31%. Month effects were found to have a significant relationship
with fertility when climatological measurements were deleted from
statistical models. This agreed with most previous research (Stott,
1961; Hafez, 1959). In no case were month effects significant when
climatological measurements were included in the model, suggesting that
month effects may have represented climatological factors to a greater
degree than nutritional and management factors. This work clearly
showed the importance of ambient environmental conditions at the time
of insemination and fertilization on bovine fertility.
Alterations of peripheral plasma estrogen concentrations during
periods of thermal stress have not been reported previously. Therefore,
speculation as to estrogenic effects on uterine blood flow and estrogen
participation in pre-ovulatory LH release cannot be reviewed here.
Physiological and endocrine factors controlling thermal properties of
the uterus at estrus and ovulation need clarification. In addition,
effects of stressful ambient temperatures on these'factors and uterine
temperature need further study as they relate to fertility.
Prostaglandins (PG), unsaturated 20 carbon fatty acids containing
a cyclopentane ring and two alipahtic side chains, were first dis-
covered in extracts of human and sheep seminal vesicles during the
1930's. The first report of their activity before identification was
shown when fresh semen was placed into the human uterus and caused the
- 11 -
uterus.to contract or relax (Kurzok and Lieb, 1930). The luteolytic
effects of PGF2a were reviewed extensively by Inskeep (1973),and
effects in cattle well documented by Hafs et al. (1974)-and Chenault
(1973). Injection of PGF2a-Tham Salt in the bovine does not drastically
alter the normal sequential hormonal patterns leading to estrus and
ovulation. In addition, fertility of cattle to the PGF2a induced
ovulation is apparently the same as in a normal spontaneous ovulation
of control cattle (Lauderdale et al., 1974). Therefore, PGF2a can be
used as an experimental tool to synchronize estrus and investigate
physiological and hormonal changes under conditions the researcher
wishes to impose. This would be beneficial in large animal research
where animal numbers may be few and biological events (the estrous cycle
and pregnancy) of long duration. The intention of this portion of the
review on prostaglandins is to recapitulate various effects of prosta-
glandins on the circulatory system and reproductive tract. Such
knowledge is essential in evaluating effects of PGF2 in the following
The role of the autonomic nervous system in controlling uterine
contractility and blood flow has been discussed by Shabanah et al.
(1964). The parasympathetic system generates uterine contractions and
causes vasodilation. The excitatory (a) action of the catecholamines
is manifested chiefly on the circular fibers whereas the inhibitory
(W) action influences the whole myometrium. Acetylcholine causes
vasodilation, especially of smaller blood vessels (Koelle, 1970).
Epinephrine is a vasopressor. Vasoconstriction occurs markedly in the
venous system, as well as smaller arterioles and precapillary sphincters.
- 12 -
Norepinephrine increases peripheral vascular resistance due to veno-
constriction (Innes and Nickerson, 1970). Estrogens govern the
parasympathetic acetylcholinee) activity and are responsible for the
basic contractile mechanism of the uterus (Shabanah et al., 1964),
whereas progesterone influences the sympathetic activity (epinephrine
and norepinephrine). Morris (1967) reviewed the sympathetic vaso-
constrictor action on the uterine vascular bed. Epinephrine and nore-
pinephrine reportedly cause a decrease in uterine blood flow with a
concomitant increase in arterial pressure, suggesting increased
vascular resistance in the uterus. Isoproteranol also acts as a
vasoconstrictor. These vasoconstrictor actions appeared to be due to
increased vascular resistance because myometrial tension changes were
Clegg (1966) reported that prostaglandins produce two types of
effects on smooth muscle. They produce direct short-lived actions such
as stimulation of the isolated uterus or relaxation of the isolated
tracheal chain preparation. Alternately, they potentiated long-term
effects of other stimulants when given in low doses. An example of an
indirect long-term effect of prostaglandins (PGF and PGE series) is
depression of responses of various isolated smooth muscle preparations
to sympathomimetic substances (epinephrine, norepinephrine, phenylephrine
Different classes of prostaglandins have various effects on smooth
muscle and blood pressure and have been reviewed by Bergstrom et al.
(1968). PGE's and PGF2a cause contraction of uterine myometrium in rats
and guinea pigs. However, in humans, the PGE's decrease tonus, frequency
- 13 -
and amplitude of spontaneous contractions of the myometrium. There is
an increase in sensitivity of the rat uterus to PGF2a following estrogen
treatment (Anggard and Bergstrom, 1963). PGF2a also has been shown to
stimulate and increase tone of the rabbit fallopian tube in vivo
(Bergstrom et al., 1968; Horton and Main, 1965), whereas PGE1 causes
relaxation. Isolated strips of human myometrium have a regular motility
pattern. This motility in the nonpregnant myometrium was inhibited by
PG's A, B and E; however, PG's Fla and F2 stimulated contractions. The
sensitivity of the myometrium was highest late in the menstrual cycle
and during pregnancy. PGF2. also increased motility of the human oviduct
in vivo. Intra-uterine application of PGF2a or intravenous infusion
increased the motility of the non-pregnant human uterus (Eliasson, 1973).
Following intramuscular injection of 10 or 20 mg PGF2a in non-
pregnant women, no cardiovascular changes were observed but there was
pain at the injection site and increased uterine activity within
minutes. The uterine contractility lasted 2 to 3 hr. (Karim et al.,
1971). Within 1 to 6 hr. after vaginal insertion of PGE2 or PGF2,
10 to 12 women had menstrual-like uterine bleeding. This bleeding
was preceded by a marked increase in uterine contractions which started
within 10 min., peaked between 60 to 90 min., and lasted about 4 hr.
PGE and PGF2a induced uterine activity that was similar to that recorded
for the non-pregnant uterus during the time of the menstrual flow.
Contractions measured between 50 to 200 mm Hg and occurred every 1 to 2
min. (Karim, 1971).
In non-pregnant dogs, PGE1 infused into the uterine artery reduced
perfusion pressure. The dilator effect of PGE1 was seen at doses as
- 14 -
little as 20 pg/ml blood. Such potent vasodilatory effects of PGE1 were
not seen in pregnant dogs near term even with large doses. PGF2, on the
other hand, had little effect on vascular smooth muscle in dogs, but
potentiated responses to sympathetic nerve stimulation, occasionally
in parallel with increased responses to norepinephrine. PGF2a appears
to work primarily on nerve terminals in the dog uterus since there was
a greater effect on neurogenically induced vasoconstrictor responses
than to responses of norepinephrine itself (Clark et al., 1972).
In a review by Brody (1973) PGF2a was reported to influence effector
response to sympathetic nerve stimulation. Vasoconstrictor action in
cutaneous and muscle vessels was facilitated by PGF2a without any change
in responsiveness of the vessels to norepinephrine, suggesting that PGF2a
facilitated liberation of the adrenergic transmitter. This specificity
was not found in venous smooth muscle when PGF2a facilitated responses
to both sympathetic nerve stimulation and to norepinephrine. Thus,
PGF2a venoconstrictor action was dependent upon integrity of sympathetic
No changes in cholinergic vasodilator nerves were noted in the
presence of prostaglandins (Brody and Kadowitz, 1974). Responses of
uterine vessels to norepinephrine were potentiated at PGF2a concentra-
tions which had no effect on uterine vascular resistance.
Recently, Ryan et al. (1974) showed that, in the dog,PGE1
redistributed the blood flow from the myometrium to the endometrium.
Therefore, PGE1 maybeavasodilator intermediate in an estrogen induced
uterine hyperemic response. To test this hypothesis estrogen was in-
jected into rats causing a visible intense hyperemia and a doubling of
- 15 -
uterine blood volume. In comparison, rats pre-treated with indomethacin,
a prostaglandin inhibitor, failed to show a large increase in uterine
blood volume. In conflict with the observation that PGF2a was a vaso-
constrictor was the finding of elevated uterine PGF content following
estrogen treatment which could be inhibited by indomethacin pre-
treatment. Except for this latter observation, the PGF series appears
to be associated with vasoconstrictor actions and reduced blood flow
to the uterus.
The cardiovascular actions of PGF2a also are complicated because
of quantitative species variation. Anggard and Bergstrom (1963) reported
that intravenous injection of PGF2a into cats caused increased right
ventricular pressure and a decreased systemic blood pressure. Intra-
arterial injections into muscles caused increased blood flow through
that area, i.e. vasodilation. PGF2a perfused into rabbit hindquarters
also caused tissue vasodilation. Horton and Main (1965) reported that
PGF2a or PGE1 injected intravenously in rabbits caused a fall in arterial
blood pressure. A review by Bergstrom et al. (1968) contrasts these re-
sults with the pressor action of prostaglandins in the rat, dog and spinal
chick. In dogs the pressor action of PGF2a is accompanied by an
increase in cardiac output and right atrial pressure, but the calculated
peripheral resistance was unchanged. It appeared as if there were a
decrease in venous capacitance because when a pressure stabilizer was
put into the venous side, it caused a shift of blood into the stabilizer
reservoir. Ducharme et al. (1968) reported similar results and also
found that, in the dog, PGF2a had little effect on'femoral arterial
pressure or small artery pressure but caused an increase in small vein
- 16 -
pressure when administered to an innervated limb. Abolishing the
sympathetic chain to the limb eliminated the venoconstrictor activity
of PGF2a. They found no real change in myocardial contractility. Thus
the pressor action was due to an increased venous return.
Horton (1969) found PGF to be weakly dilatory on arterioles.
In some species (rat and dog) they act as a venoconstrictor, thus in-
creasing venous return.and cardiac output. Neither PGE1 or PGFla
injected close-arterially released catecholamines from the adrenals
of anesthetized cats, but PGE1 did so in dogs. PGF2. injected intra-
arterially caused no constant change in blood pressure or in baroreceptor
discharge frequency. Moreover, intravenous injections caused a transient
rise in arterial pressure which was associated with an increase in
baroreceptor discharge. It appeared that the increased discharge fre-
quency was secondary to the pressure rise because in animals where the
blood pressure fell slightly, so did discharge frequency. PGF2a in-
jections into the carotid artery resulted in a variable response on
chemoreceptor discharges. Intravenously injected PGF2. caused a small
increase, decrease or no change at all in blood pressure (McQueen and
Belmonte, 1974). Therefore, the authors suggested that direct action
was on pressure changes not by way of baroreceptors. These actions
may be related to the rapid disappearance of prostaglandins as only
5 to 10% of the injected PGF2a was detected 1 min. later and negligible
PGF2a was found at 90 sec. (Raz, 1972). Also, more than 95% of injected
prostaglandins were removed during one circulation through the pulmonary
vascular bed (Ferreira and Vane, 1967).
Various investigators observed actions of prostaglandins on
- 17 -
respiratory smooth muscle. Main (1964) has shown that PG's E1, E2, E3
and Fla relaxed tracheal muscle in vitro in rabbit, guinea pig (also
Puglisi, 1972), ferret, pig, sheep, cat and monkey preparations. Except
fn the cat, they decreased lung resistance to inflation in vivo (also
Anggard and Bergstrom, 1963). PGF2a has similar biological activity to
PGFla, so these observations should hold for its actions. This con-
clusion was confirmed in a cat-trachea preparation by Horton and Main
(1965), in which PGF2a inhibited acetylcholine produced contractions.
In the dog the action of PGF2a was a reduction in dynamic lung compliance
and alveolar ventilation (Horton, 1969).
Investigations on systemic actions of prostaglandin in the bovine
are very limited to date. Lewis and Eyre (1972) reported that PGE1
and E2 lowered systemic blood pressure in calves, but PGF2a caused a
pressor response. Furthermore, pulmonary arterial pressure and abdominal
venous pressure were raised by the three substances. PGF2a caused con-
traction of the pulmonary artery and vein and produced an increase in
heart rate. Also noted was an increase in respiratory volume produced
by PGF2 Anderson et al. (1972) also concurred that PGF2a increased
pulmonary arterial pressure, but they found a drop in cardiac output
and essentially no change in femoral arterial pressure, left ventricle
and diastolic pressure, heart rate, blood gases and pH.
There is a definite need for more study on actions of prostaglandins
to determine their roles in physiological functions in the bovine.
Additional work is needed because of the contradictory results obtained
among and within species,
- 18 -
Hormone Relationships of Uterine Blood Flow and Temperature
The uterus responds to cyclic hormonal changes during the estrous
or menstrual cycles. Blood levels of estrogen and progesterone are in-
volved with this phenomena. In the human, the first half of the cycle
is associated with rapid growth of the uterine vascular elements and is
under estrogenic control. This is a period of tissue repair and pro-
liferation. The latter half of the cycle is characterized by glandular
secretary activity and elaboration of vascular elements under the con-
trol of estrogen and progesterone (Reynolds, 1949).
One of the principal characteristics of the uterus following
estrogen administration is its bright red color. The degree of redness
suggests a high level of oxygen saturation of the blood and there is
a high rate of blood flow. In the presence of an active corpus luteum
(progestational influence), the uterus is bluish in color. Oxygen
consumption is low and blood flow is sluggish. Oxytocin causes intense
muscular spasms within the uterus without affecting the rate of blood
flow, whereas vasopressin causes relaxation of uterine musculature but
a constriction of its vasculature (Reynolds, 1949).
Uterine hyperemia following injection of estrogen has been
estimated in ewes by direct collection of uterine venous blood (Huckabee
et al., 1970), by flow meters (Greiss and Anderson, 1970; Rosenfeld et
al., 1973) and microspheres (Rosenfeld et al., 1973). Endogenous estrogens
produced during the estrous cycle appear to have similar effects on uterine
blood flow. Patterns of change in plasma estradiol concentrations
(Scaramuzzi, Caldwell and Moor, 1970) are very similar to records of
- 19 -
uterine blood flow changes in the ewe (Greiss and Anderson, 1970;
Huckabee et al., 1968, 1970). Specificity of estrogen actions on
uterine blood flow have been shown by local injection of estrogen into
one uterine horn artery. An increase in blood flow was measured only
in that uterine artery (Resnik et al., 1974).
Progesterone injected into ovariectomized ewes did not alter
uterine blood flow, whereas progesterone superimposed on estradiol in-
jections caused a decrease in uterine blood flow rates (Greiss and
Anderson, 1970). Estrogens did not appear to affect systemic blood-
pressure (Huckabee bt al., 1968, 1970; Resnik et al., 1974), but caused
a fall in the coefficient of oxygen utilization [(AV)02 X 100] in the
uterus. However, due to the higher uterine blood flow there was
essentially no change in oxygen uptake of the uterus (Huckabee et al.,
1968, 1970). Thus a dissociation between uterine metabolic rate and the
rate of blood flow might be reflected in temperature differences between
the uterus and aortic blood. In sheep a decrease in the temperature
difference between the uterine cavity and aortic blood provided a con-
venient method for monitoring increased uterine blood flow changes
following estrogen injection. A rise in blood flow resulted in a
lowered uterine temperature (Abrams et al., 1970a, 1971).. The actions
of estrogen to lower uterine temperature may be mediated through its inter-
action with acetylcholine to cause vasodilation (Shabanah et al., 1964)
or through the release of uterine histamine which was shown to be
involved in a rapid onset of hyperemia and water imbibition '( Jensen
and DeSombre, 1972). Lowering the rate of uterine heat production is
unlikely because of the many metabolic activities induced by estrogens
- 20 -
(Talwar and Segal, 1971; Jensen and DeSombre, 1972).
In cattle, plasma estrogens increasedprior to estrus and declined
precipitously during estrus (Chenault et al., 1973; Henricks, Dickey
and Hill, 1971). These changes may have distinct thermal effects on
the uterus. Greiss and Anderson (1969) reported increased uterine blood
flow associated with the onset of estrus in sheep, which could cause a
drop in uterine temperature (Abrams et al., 1970a,1971; Caton et al.,
1974) and thus be related to an optimal time for insemination to achieve
maximal fertility. However, the thermal response of the bovine uterus
to estrogen has not been documented.
Although several hormonal changes due to hyperthermia have been
documented in the bovine there is a sparcity of results related to a
multiplicity of hormonal responses to a controlled thermal stress in
which such sources of variation due to breed, age, animal and time
responses are evaluated. Uterine blood flow and temperature relation-
ships have been reported in sheep in response to estrogen injections,
but have not been reported in the bovine. Also, changes in utetine
temperature during phases of the estrous cycle under conditions of mild
heat stress have not been reported.
HORMONAL PATTERNS DURING HEAT STRESS FROM PGF2a INJECTION
THROUGH ESTRUS AND OVULATION AND FOLLOWING ADRENAL
STIMULATION BY ACTH IN HEIFERS
Lowered conception rate due to heat stress occurs over a pro-
longed period of the year in tropical and subtropical climates. Before
systems for reproductive management can be developed to counter these
adverse effects of heat stress, a more complete understanding of
endocrine and physiological changes within the same animals must be
made. We need to know how a standard heat stress alters hormonal and
physiological responses during the normal estrous cycle, and determine
if chronic heat stress alters adrenal responsiveness to an IV injection
Due to limitations of facilities and methodology, most inves-
tigations of heat stress effects on hormonal balance and their
relationships to reproductive performance have monitored only one
or two hormonal responses. Therefore, pooling results from different
laboratories and from different animals within and between species
can be misleading when an effort is made to develop a hypothesis on
how thermal stress affects reproduction.
- 21 -
- 22 -
Reduced fertility in hot environments was associated with elevated
body temperature (Dunlap and Vincent, 1971; Gwazdauskas, Thatcher and
Wilcox, 1973). Hot climates may exert their depressive effects on
fertility by acting on the gonads, uterine environment, endocrine
system or gametes (Hafez, 1959; Ulberg and.Burfening, 1967). Seasonal
Infertility has been attributed primarily to the bovine female (Stott,
1961). Studies on hormonal alterations due to thermal stress have
failed to document interrelationships of more than two different hormones
in the same animals. -Plasma progestin changes have been documented by
Mills et al. (1972), Abilay and Johnson (1973) and Abilay, Johnson and
Seif (1973); changes in plasma corticoid levels have been reported by
Lee, Roussel and Beatty (1973), Christison and Johnson (1972), Abilay
and Johnson (1973), Shayanfar (1973) and Miller and Alliston (1974a).
Plasma LH changes have been reported by Madan and Johnson (1971) and
Miller and Alliston (1973) and seasonal changes in prolactin have been
detected by Koprowski and Tucker (1973), Schams and Reinhart (1974) and
Thatcher (1974). Such hormonal alterations may be causative agents
contributing to suppressed estrous manifestation and depressed fertility
under hot climatic conditions.
There are essentially no studies designed to test specific effects
of heat stress environments on a multiplicity of hormonal responses with-
in the same animal. Such a study is needed in which variations due to
breed, age, animal (among and within) and hormonal interrelationships
are considered in evaluating thermal stress effects.
Objectives of this study were to characterize changes in peripheral
plasma concentrations of LH, progestins, estradiol, estrone, prolactin
- 23 -
and cor.ticoids after an intramuscular (IM) injection of PGF2-Tham Salt
(PGF2a) under controlled environmental temperatures (21.3 C compared to
32.0 C), and to determine if chronic heat stress alters adrenal responsive-
ness to an intravenous (IV) injection of ACTH (200 IU).
Materials and Methods
Ten normally cycling Holstein heifers at the USDA, Agricultural
Research Center, Beltsville, Maryland, were assigned alternately, based
on age, to one of two treatment groups (figure 1). All heifers were
placed in one of two environmental chambers at 21.3 C, 59% RH for 2
weeks. On day 9 of this adaptation period, 8 of 10 heifers in the luteal
phase of the estrous cycle received 30 mg PGF2 a (IM) to cause luteal
regression. This injection allowed all heifers to be in the luteal
phase of the cycle when PGF2a was injected 12 days later. PGF2a
effectively regresses the bovine corpus luteum and synchronizes estrus.
Lauderdale et al. (1973, 1974), Louis et al. (1974) and Chenault et al.
(1974) reported that fertility at the synchronized estrus, and induced
hormonal changes resulting in estrus and ovulation appeared normal in
PGF2, treated cattle. Thus, it was felt that PGF2a treatment could be
utilized as a tool to best control reproductive status of the heifers
and maximize efficient use of the chambers.
On day 14, the environment of one chamber was adjusted to 32.0 C,
67.2% RH. On day 20, all heifers were fitted with indwelling polyvinyl
jugular catheters (V-7; Bolab Inc., Derry, N.H.) and PGF2a (30 mg, IM)
apGF2a-Tham Salt was graciously supplied by Dr. J. W. Lauderdale, Upjohn
Co., Kalamazoo, Michigan.
- 24 -
at 21.3 C
5 Heifers (21.3 C)
5 Heifers (32.0 C)
sample every 6 hr.
30 mg -- 6 hr. sample
4 hr. samples
Day 23 ovulation
Evaluation of thermal stress on transitory hormonal changes
in the bovine during the period of luteal regression, estrus
and ovulation: Experimental design.
- 26 -
was given on day 21. Such treatment would allow for monitoring of
hormonal responses associated with corpus luteum regression, follicle
growth and ovulation under two different environments (21.3 C compared
to 32.0 C).
Blood samples (50 ml) were collected from jugular catheters at
-18, -12 and 0 hr. pre-PGF2a (day 21), at 6 hr. intervals for 48 hr.
post-injection, every 4 hr. thereafter until ovulation, and then twice
daily until the last heifer ovulated (figure 1). All blood was col-
lected into heparinized syringes, placed immediately into an ice bath,
centrifuged at 12,000 g for 10 min. at 4 C, and plasma stored at -20 C
until analyzed for progestins, LH, estradiol, estrone, corticoids, pro-
lactin, protein concentration, osmolarity and cortisol binding capacity.
Beginning 48 hr.post-PGF2a injection, animals were checked visually
for estrus at 4 hr. intervals. Heifers were artificially inseminated 12
hr. after onset of estrus, and ovulation determined by rectal palpation
of an ovarian ovulatory crypt. Palpations were performed at 4 hr.
intervals following cessation of estrus.
Chamber temperatures and relative humidities were recorded con-
tinuously (Honeywell Recorder, Washington, Pa.), and temperature of
each chamber verified daily with a tele-thermometer [Model 46 TUC,
Yellow Springs Instrument Co., Inc. (YSI), Yellow Springs, Ohio; air
temperature probe T 2620 (YSI)]. Rectal temperatures were monitored
daily (tele-thermometer probe T 2600, YSI). Skin temperatures taken
on the shoulder, rump and approximately 5 to 8 cm lateral to the vulva
(surface temperature probe T 2630, YSI) also were monitored daily during
the serial blood collection period. Thermister probes were calibrated
- 26 -
against a Bureau of Standards Certified Thermometer in a well-stirred,
insulated water bath held at 35 to 40 C. Data collected from the various
probes were corrected for constant temperature differences above and
below the certified thermometer readings.
Plasma samples were pooled within heifers (after the drop of the
preovulatory LH peak to basal levels and having less than 3 pg/ml
estradiol) to determine total protein concentration (Lowry method),
osmolality, (freezing point depression, Fiske Osmometer, Model G-61,
Fiske Ass., Inc., Bethel, Conn.), cortisol binding capacity and cortisol
association constants (Pegg and Keane, 1969; Shayanfar, 1973) for each
In the second phase of the trial adrenal responsiveness to ACTH
was tested. Eight days after the last heifer ovulated, all heifers
received 200 IU ACTH (Porcine ACTH, Sigma Chemical Co., St. Louis, Mo.).
Blood samples (50 ml) were collected from indwelling jugular catheters
at -2, -1, 0 hr. pre-injection, 15, 30, 45, 60 min. and hourly thereafter
up to 12 hr. postinjection.
LH and prolactin were assayed in plasma at two dilutions using
the double antibody radioimmunoassay (RIA) of Niswender et al. (1969).
Guinea pig antibovine LH serum (Oxender, Hafs and Edgerton, 1972) was
supplied by Dr. H. D. Hafs of Michigan State University and revalidated
with NIH-LH-B7 for measuring plasma LH in our laboratory (Troconiz,
1973). Guinea pig anti-bovine prolactin serum (Koprowski and
Tucker, 1971) was donated by Dr. H. A. Tucker of Michigan State
University and revalidated for measuring plasma prolactin with NIH-P-B3
prolactin (Chakriyarat, 1974 personal communication). Plasma progestins,
- 27 -
estradiol and estrone were measured by RIA procedures described by
Abraham et al. (1971) and Hotchkiss et al. (1971), respectively.
The antiprogesterone antibody was a gift from Dr. K. Kirton of the
Upjohn Co., and the estrogen antibody was donated by Dr. V. L. Estergreen
of Washington State University. Extraction, purification and quantita-
tive procedures were validated in our laboratory by Chenault et al.
(1973, 1974, 1975). Plasma corticoids were extracted, isolated and
quantified by competitive protein binding (Gwazdauskas, Thatcher and
Wilcox, 1972, 1973). The only modification was use of a .2 ml dextran
coated charcoal suspension (100 mg Dextran, Type 60 C, Sigma Chemical
Co., St. Louis, Mo.; 1 gm Norit A, Sigma Chemical Co. and 100 ml deionized
water) instead of 80 mg florisil for adsorption of free steroid in the
competitive protein binding assay.
An extensive series of least-squares analyses was conducted to
characterize treatment, animal and within-animal time trends in plasma
LH, progestins, estradiol, estrone, prolactin and corticoids. Other
response variables were analyzed by analysis of variance.
Results and Discussion
Averages and standard deviations for chamber conditions, rectal
and skin temperatures, and events associated with estrus are shown in
table 1. Based on Christison and Johnson's (1972) criteria for a
moderate heat stress condition (rectal temperature increase of .5 C)
climatic conditions of our study exerted a greater than moderate heat
stress since rectal temperatures of heifers in the 32.0 C chamber were
- 28 -
Table 1. Physiological parameters of heifers in environmental
chambers at 21.3 C and 32.0 C.
PGF2 to.LR peak
PGF2a to ovulation
a(4 + SD)
c(n=26), d(n=23), e(n=5), f(n=4)
- 29 -
1.49 C greater than heifers in the 21.3 C chamber. Skin temperatures
also were significantly (P<.01) elevated in the hot chamber. Visual
appraisal of the data showed there was no tendency for skin or rectal
temperatures to decline during chronic exposure to the heat stress of
our experiment which would have suggested adaptation. Indices of
physiological response to PGF2a (duration of time between PGF2a injection
and the LH peak and time between PGF2a injection and ovulation) were not
different between treatments (P>.10). The interval from the LH peak to
ovulation was approximately 24 hr. for both groups (24 hr. in 21.3 C;
23.2 hr. in 32.0 C). This interval is similar to that reported by
Chenault et al. (1973, 1975). Arije, Wiltbank and Hopwood (1974), and
Christenson, Echternkamp and Laster (1974) for unsynchronized animals
and the PGF2a induced interval reported by Chenault et al. (1974) and
Hafs et al. (1974). If the trend for the heat stress group to have
shorter intervals from PGF2a injection to LH peak and ovulation is real
(table 1), we may have failed to detect differences due to small numbers
(n=5 each) and appreciable variation. In this study, if hyperthermia
affected endocrine-physiological interactions, it did not appear to
alter time between the LH peak and ovulation.
Length of estrus was significantly shorter (P<.0.) for the heat
stressed heifers and was comparable to the 14 hr. duration of estrus re-
ported by Gangwar, Branton and Evans (1965) for Holstein heifers in hot
natural summer climatic conditions of Louisiana. Two of four heifers
inseminated in the 21.3 C chamber were pregnant at 40 days compared to
none of 5 in the 32.0 C chamber. Though there were small numbers of
animals inseminated in this trial, the percentage of successful pregnancies
- 30 -
was comparable to that of the Dunlap and Vincent (1971) environmental
chamber study. This related well to observations that thermal stress
in this study interfered with the overall reproductive process in
heifers. Thus, the environmental condition did affect body temperature,
duration of estrus and probably overall fertility. Whether hormonal re-
sponse under these conditions varied was. of utmost interest.
Pre-PGF2a injection plasma samples were analyzed by analysis of
variance to detect possible differences in progestins, estradiol, estrone,
LH, prolactin and corticoids due to temperature, sampling time, temper-
ature X sampling time interaction and animals within temperature treatment.
Progestins (X = 3.21 ng/ml, 21.3 C; 7 = 3.16 ng/ml, 32.0 C) were not in-
fluenced by temperature or sampling time. However,a significant (P<.05)
temperature by time interaction suggested different progestin concentra-
tions at different times of sampling in each treatment chamber. Plasma
progestins appeared to decline in the heat stressed group with pro-
gressive sampling, whereas in the cool group they did not change
(Appendix, table 12). Also, there was significant (P<.01) variability
in progestin concentrations among heifers in each chamber. This would
suggest that there is considerable variation in progestin secretion from
heifers during the luteal phase of the cycle. Animal variability in
pre-injection plasma estradiol concentrations (X = 3.02 pg/ml, 21.3 C;
2.32 pg/ml, 32.0 C) was significant (P<.01). Estrone (X = 3.04 pg/ml,
21.3 C; X = 3.25 pg/ml, 32.0 C), LH (X = .53 ng/ml, 21.3 C; X = 182 ng/ml,
32.0 C) and prolactin (X = 12.73 ng/ml, 21.3 C; X = 15.32 ng/ml, 32.0 C)
pre-PGF2a plasma concentrations were not influenced by temperature, time
of sampling, temperature X time interaction or animals within temperature
treatment (P>.10). Variability in plasma corticoids was significant
due to sampling time (P<.01) and treatment X sample time interaction
(P<.01). However, there was no difference due to main effects of
temperature (9.0 ng/ml, 21.3 C; 9.55 ng/ml, 32.0 C).
A logical physiological reference to analyze the data is the
preovulatory surge of LH. All animals had a preovulatory LH surge, and
each hormone was analysed initially to determine time relationships
with the LH surge. Least squares statistical models were selected
based on tests of significance of higher order terms (time) in the
regression analyses and visual appraisal of the graphs.
Figure 2 shows the progestin responses at 21.3 C and 32.0 C.
Data for the regression analyses have been synchronized to the LH peak
for analysis. The statistical model included treatment, heifer within
treatment and time trends (Appendix, table 4). The progestin time trend
for heifers at 21.3 C was best described by the equation Y (progestin,
ng/ml) = 1.533 + 1.160X .344X2 + .034X3 .0014X4 + .00002X5 (P<.01)
where X = .1 hr., whereas the time trend for the 32.0 C heifers showed
a significant (P<.G1) curvilinear relationship best described by a third
order equation: Y = 7.923 1.123X + .061X2 .0010X3
Tests for heterogeneity of regression were significant (P<.01),
suggesting that the 5th order regression curves for each treatment were
not parallel. This observation implies that there was a different time
response to treatments. We feel that this difference was probably due
to heifers in the 21.3 C chamber having their LH peak approximately
22 hr. later from the start of blood sampling than the heifers main-
tained at 32.0 C (table 1). Therefore, on the average cool heifers had
-96 -72 --48 -24 0 24 43 72 97
SEQUENTIAL CHANGES IN PLASMA PROGESTINS IN HEIFERS AT
SYNCHRONIZED TO THE .TIE OF THE LH PEAK.
21,3 C OR 32,0 C
-21.- C (.2= .5 2)
---- 2.0 C ( .)
- 33 -
a longer plateau of low progestins prior to the LH peak. PGF2a caused
a drop in progestin concentration by 18 hr. after injection in all
heifers (Appendix, table 5). Apparently, factors controlling the pre-
ovulatory release of LH in cool heifers were slightly delayed (~ 24 hr.)
due either to treatment effects or chance. Thus, due to a shorter interval
between PGF2a and the LH peak for the 32.0 C heifers, one would expect
higher progestins because of shorter trough duration. Also, the number
of observations earlier than -120 hr. was very small. Significant
(P<.01) among heifer differences were detected within treatments, but
differences among treatments were not significant [(P>:10) Appendix,
table 4). This finding is in direct conflict with Stott, Thomas and
Glen (1967), who reported elevated progesterone on day of estrus in
thermally stressed cows. However, Stott and Wiersma (1973) more recent-
ly reported a depression in plasma progestins due to chronic thermal
The progestin RIA has a sensitivity of 25 pg or .025 ng/ml plasma.
Coefficients of Variation (C.V.) for progestins after accounting for
variability due to treatment, heifer in treatment and time trends were
115% (cool) and 99% (hot). Acute increases of 2.5 (Gwazdauskas,
Thatcher and Wilcox, 1972) and 1.5 ng progesterone per ml plasma (Wagner,
Strohbehn and Harris, 1972) for a period of 2 hr. were detected follow-
ing 200 IU and 100 IU ACTH injections, respectively. If thermal stress
had elicited an adrenal release of progesterone we would have been able
to detect it. Mills et al. (1972) detected a significant elevation of
only .47 ng/ml progestins in heifers thermally stressed for 72 hr. at
the onset of estrus. Therefore, with overall progestin levels of
- 34 -
.147 + .095 ng/ml (X + SD) the day of estrus and day after estrus and
no differences due to temperature, we cannot conclude that chronic
heat stress caused adrenal hyperprogesterone response. Our findings
support Miller and Alliston (1974b), who found no difference in plasma
progesterone during the bovine estrous cycle when twice daily measure-
ments were made in control (17-21 C) or heat stress (21-34 C)
environments. Progestins do not appear to be elevated during the
period between luteal regression and ovulation.
Data for regression analyses of estradiol were separated into
two periods and independently analyzed to characterize time trends.
The two periods were from -144 hr. to time 6f the LH peak (including
LH peak time) and from the LH peak to +96 hr. (also including the LH
peak time). Pre and post LH peak time trends for estradiol were best
described by Y (estradiol, pg/ml) = 4.53 .482X .045X + .0069X
(P<.05) and ?post = 45860.38 12165.711X + 1285.829X2 67.674X3 +
1.773X4 .019X5 (P<.05) for the cool heifers and Y = 5.67 1.279X
+ .092X2 (P<.01) and Yp = 490.168 73.503X + 3.647X2 .060X3
(P<.01) for the hot heifers (figure 3). Plasma estradiol was depressed
(P<,10) in the 32.0 C chamber (peak estradiol:10.4 pg/ml plasma for
21.3 C heifers compared to 7.2 pg/ml plasma for 32.0 C heifers; Appendix,
tables 4 and 6).
Lower plasma estradiol may have contributed to the shorter periods
of estrus seen in the 32.0 C heifers. However, these lower concentra-
tions of estradiol were adequate to elicit estrous behavior and LH
release causing a subsequent ovulation. The lower estradiol may reflect
altered production, secretion, clearance or receptor binding under
* -.144 -120 -96
-24 0 24
SEQUENTIAL CHANGES IN PLASfMA ESTRADIOL IN HEIFERS AT 21.3 C OR 32.0 C
SYNCHRONIZED TO THE TIME 01- THE LH PEAK.
,,1.,_,:,, rbYILTO~C~-~II-C ly- II.-I---tUMUI(P---~-r-~U~-P-? --1
- 36 -
conditions of hyperthermia in cattle. Significant among heifer vari-
ability (P<.0l) was detected in the 32.0 C heifers (Appendix, table 4).
Whether slightly altered plasma estradiol would affect such factors as
uterine and oviductal blood flow, temperature of the reproductive tract,
tract motility, gamete and embryo transport that may contribute to poor
fertility under heat stress is not known. Results of the present study
reveal only a subtle effect of heat stress on plasma estradiol.
Unlike the time responses of estradiol, estrone showed no apparent
association with onset of estrus or LH peak when data were synchronized
to the time of the LH peak (r-0, estrone:estradiol; table 2). Hansel
(1971) reported an estrone peak between days 13 to 16 of the bovine
estrous cycle and suggested that this elevation in estrone may be related
to corpus luteum regression. Therefore, data were analyzed from the
time of the PGF2o injection (figure 4; Appendix, tables 4 and 7). Time
responses were best characterized by Y (estrone, pg/ml) = 2.671 + 1.488X
.763X2 + .120X3 .0077X4 + .00018X5 (P<.01) for the 21.3 C group and
^2 3 4 5
S= 2.298 + 1.828X .781X + .145X3 .0072X + .00017X (P<.05) for
the 32.0 C heifers. The significant elevation (P<.05) of estrone
(Appendix, tables 4 and 7) due to heat stress is best seen following
PGF2. through 72 hr. (figure 4). There was no evidence that these
curves were not parallel (5th order, P>;10) suggesting that in both
treatments estrone followed the same decline post-PGF2 However, estrone
levels were higher through 72 hr. in heat stressed heifers. The slight
rise in estrone prior to PGF2a injection to +12 hr. may be related to a
luteolytic action as suggested by Hansel (1971), although heifers were
only day 9 of the estrous cycle at time of PGF2c injection.
- 37 -
Table 2. Simple correlations between
LH -.14* .45**
S"\ 2 1.3 C -- R2 .39
32.00--- -,R2 .27
;IE P PG F2c<
-18 0 24 48 72 96 120 144
FIGURE 4, -SEQUENTIAL CHANGES IN PLASMA ESTRONE IN HEIFERS AT 21,3 C OR 32.0 C
SYNCHRONIZED TO THE TI'E OF PGF2, INJECTION,
- 39 -
Chenault et al. (1974) and Henricks et al. (1974) reported
estrone to vary greatly within and among animals after PGF2a injections.
The C.V.'s for estrone in this study after accounting for heifer and
time variability were 65.3 and 74.0% for cool and hot groups, respectively.
Plasma estrone concentrations were less than 5 pg/ml and agree with the
work of Echternkamp and Hansel (1973). However, they reported that
estrone was slightly elevated at estrus in one cow. This observation
was difficult to support since there was no statistical analysis or re-
port of estrone variability.
Because LH increases above basal levels for only about 10 hr.
(Chenault et al., 1975), LH data were separated into four time periods
to analyze, independently, both basal and preovulatory peak concentra-
tions. These periods were: (1) from -148 to 8 hr. prior to the LH
peak (-8 hr.); (2) -8 hr. to the peak of LH; (3) LH peak to +8 hr.
and (4) +8 hr. to +96 hr. (figure 5). Average LH concentrations for
period (1) were 1.80 + 1.21 ng/ml plasma (X + SD; n=102) for the cool
heifers (21.1 C) compared to 1.75 + 1.27 ng/ml (n=74) for the heat
stressed heifers (32.0 C). A linear increase (P<.01) in LH occurred
between -8 hr. to the LH peak (peak LH 32.19 ng/ml for cool heifers
compared to 33.17 ng/ml for hot heifers), and then dropped linearly
(P<.01) to basal levels by +8 hr. (.43 ng/ml cool compared to .49
ng/ml hot). Basal levels were defined as any LH concentration within
three standard deviations of the mean for all samples (n=169) up to
-8 hr. (1.75 + 1.22 ng/ml). During the initial period, two heifers in
the cool chamber had sporadic peaks (>3 S.D.) of LH that occurred
between -72 and -12 hr. prior to the LH peak (Appendix, table 8).
il 21.3 C ---
-~3 -72 -403 -~
0 24 40 72
SEQUENTIAL CHANGES IN PLASMA LH IN HEIFERS AT EITHER 21,3 C OR 32,0 C,
- 41 -
The preovulatory surge of LH remained above basal levels for 10.4 and
9.6 hr. for the 21.3 C and 32.0 C groups, respectively. LH concentra-
tions in this study agree with those of Henricks, Dickey and Niswender
(1970) and Snook, Saatman and Hansel (1971). Unlike results of Madan
and Johnson (1971) and Miller and Alliston (1973), we found no
significant difference in LH levels in response to heat stress (Appendix,
table 4). Our contradictory results under conditions of thermal stress
may be due to frequency of sampling (twice daily; Miller and Alliston,
1973), sensitivity of experimental design in which among animal variability
was considered in the present study, duration of the LH peak (~ 10 hr.)
or breed differences (Madan and Johnson, 1971). Riggs, Alliston and
Wilson (1974) detected a difference in the preovulatory LH surge during
heat stress between Hampshire and Duroc gilts. Such differences may
exist between the study of Madan and Johnson (1971) in which Guernsey
cattle were used and in our study where only Holstein heifers were
All animals had a preovulatory LH surge, suggesting that hyper-
thermia did not prevent the triggering mechanism for LH release.
Although estradiol levels in peripheral plasma were slightly depressed
in the 32.0 C heifers, it does not appear that these lowered estradiol
concentrations altered LH release.
Plasma prolactin was analyzed initially in the same manner as
estradiol (pre- and post peak of LH). However, there was no change in
prolactin associated with estrus or the LH peak as previously reported
by Swanson and Hafs (1971). Absence of any association between prolactin
and estrus is supported by work of Hoffman et al. (1974) in which an
- 42 -
inhibitor of prolactin secretion caused no estrous cycle disorders.
Also, Wetteman and Hafs (1973) were unable to find elevated prolactin
on the day of estrus.
Since there were no detectable changes in prolactin associated
with the LH peak, data were analyzed further relative to time of PGF2,
injections. Hafs et al. (1974) reported that prolactin increased
immediately following PGF2a injection (within 1 hr. and lasting for
4 hr.). However, an increase in our study was not detected because
the first blood samples were not taken until 6 hr. after injection. No
differences were detected between the 21.3 C and 32.0 C treatment groups
(P>.10; Appendix, tables 4 and 9). Time trends of prolactin for the
21.3 C and 32.0 C treatment groups were described by the following
equations: Y (prolactin, ng/ml) = 14.50 1.707X + .371X2 .018X3 and
Y = 14.28 + 2.738X .733X2 + .066X3 .002X4 (3?.0 C) (figure 6). The
4th order time curves were not parallel (P<.005). Prolactin C.V. after
accounting for heifers and the above time equations was 49.4% (21.3 C)
and 26.3% (32.0 C). During the initial 42 hr. (-18 to 24 hr.), the
prolactin response in the cool chamber appeared to decline. This
observation may be due to a lowering of stress-induced prolactin secretion
with more sampling (Tucker, 1971). Apparently heifers in the 32.0 C
chamber could not adjust to sampling as quickly since prolactin increased
and remained elevated until 24 hr. after PGF2 However, this is
questionable since trends are very subtle and the curves account for
little of the variability (figure 6).
An increase in plasma prolactin due to heat stress was antici-
pated in the present study based on reports of Koprowski and Tucker (1973),
21.3 C R2 .07
32.0 ---- ,R2= .lI
0 24 48 72 96
SEQUENTIAL CHANGES IN PLASMA PROLACTIN IN HEIFERS AT 21,3 C OR 32,0 C
SYNCHRONIZED TO THE TIME OF PGF2, INJECTION,
;------- --- ----- -ra~F~ur~arr~-a90a~~a~i~oy-- ----~-.-~~rm
- 44 -
Schams and Reinhart (1974) and Thatcher (1974), in which seasonal
changes of plasma prolactin were detected (high during summer). Wetteman
and Tucker (1974), using twice daily sampling, detected only slight
differences (P<.10) in serum prolactin in 3 mo. old calves exposed either
to 21 or 27 C temperatures under a 12 hr. per day light regime. The
induced release of prolactin following injection of thyrotropin releasing
hormone (TRI) was twice as great in calves exposed to 27 C as calves
at 10 C. Furthermore, they suggested that these results, which are
opposite those seen in lactating cows following the milking stimulus,
may be due to differences in anterior pituitary responsiveness of 3 m6.
old calves at different temperatures.
In our study, no differences in average plasma prolactin concentra-
tions were detected between heifers at 21.3 C or 32.0 C (figure 6;
Appendix, tables 4 and 9). Thus at a constant 14 hr. light 10 hr.
dark regime an environmental temperature of 32.0 C caused no increase
in prolactin compared to controls at 21.3 C. Perhaps other factors
control the seasonal increase in prolactin previously reported (Koprowski
and Tucker, 1973; Schams and Reinhart, 1974; Thatcher, 1974). Karg and
Schams (1974) reported a positive correlation of day length and basal
prolactin levels in cattle. Relkin (1972) showed that changes in light:
dark ratios for rats altered pituitary prolactin content and plasma
prolactin concentrations. This effect is seen after only 4 to 8 hr. of
exposure to different lighting regimes. Photoperiod may be a factor
influencing pituitary prolactin secretion. Under Florida conditions
seasonal temperature changes also are correlated with increasing periods
of day length (Gwazdauskas, unpublished observations). Thus, temperature
- 45 -
and photoperiod effects are confounded in evaluating seasonal effects
on plasma prolactin concentrations. Under controlled environmental
conditions of the present study, temperature seemed unimportant in
eliciting a major change in prolactin secretion.
Figure 7 shows the plasma corticoid response when data were
synchronized to the LH peak. Statistical analysis revealed no dif-
ferences (P>.10) between treatment means (Appendix, tables 4 and 10).
Furthermore, we were unable to detect any individual treatment time
trends after looking-at regressions up to the 5th order. After account-
ing for corticoid variability due to treatment, heifers within
treatment and time trends (up to the 5th order), there was a 65% C.V.
for plasma corticoid concentrations. Our data, with blood samples
taken at 4 hr. intervals at least 2 days prior to estrus, do not sup-
port Miller and Alliston's (1974a)finding of increased corticoids
early the day of estrus (twice daily sampling). Nor does it support
a report which showed lower plasma corticoids in dairy cows during
summer months in Arizona (Stott and Wiersma, 1973). However, Arizona
climatic conditions of high temperature and low humidity may be dif-
ferent from our study with high humidity and high temperature. Our
results show numerous episodic peaks during the day. Wagner and
Oxenreider (1972) also reported episodic peaks of plasma corticoids
when measured at 30 min. intervals throughout the day. They also noted
diurnal corticoid variation, but we were unable to detect any time of
day differences (P>.10) when data were analysed at 4 hr. intervals.
Due to large variability in plasma corticoid levels, a large treatment
difference in corticoid concentrations would be needed to detect a
POOLED MEANS -
-96 -72 -48 -24 0 24 48 72 96
SEQUENTIAL CHANGES IN PLASMA CORTICOIDS SYNCHRONIZED TO THE
LH PEAK USING POOLED MEANS OF HEIFERS AT 21,3 C AND 32,0 C,
TIME OF THE
- 47 -
significant difference. We failed to detect any differences in plasma
corticoid associated with estrus, ovulation or heat stress when
monitoring plasma concentrations at 4 hr. intervals.
When time was removed from the model and each hormone considered
as a dependent variable, plasma LH had a negative association with
progestins (r = -.14, P<.05; table 2). Progestins also were negatively
related to estradiol (r = -.14, P<.05). In the overall model LH was
positively related to estradiol (r = .45, P<.01). These observations
are consistent with findings of Chenault et al. (1975) and support
their hypothesis that progestins may be inhibiting estradiol biosynthesis
and LH release. The significant relationship between plasma corticoids
and prolactin (r = .22, P<.01) may be related to the stress response of
both hormones (Gwazdauskas, Thatcher and Wilcox, 1972; Koprowski and
Increases in plasma and blood volumes due to heat stress have
been reported in cattle (Bianca, 1965), and chronic exposure to heat
resulted in decreased hematocrit. Changes in plasma electrolyte
concentrations due to thermal stress are reported to be slight (T. N.
Wegner, personal communication). Cattle in tropical areas had a
higher body water content in summer than in winter months (Thompson,
1973). Such factors as plasma volume and dilution may influence
interpretation of hormonal responses to a controlled heat stress. For
example, a difference in plasma corticoid concentration was not detected
in our study. However, a heat stress induced increase may have been
undetectable due to a possible plasma dilution response of these heifers.
Conversely, a decrease in estradiol concentration may have occurred
- 48 -
due to plasma dilution and not necessarily decreased secretion. Such
criticisms apply to all other hormonal responses in this study. Thus,
it was of interest to measure plasma total protein concentration and
plasma osmolality to determine if a possible plasma dilution had occurred.
Plasma samples from 8 to 96 hr.. after the LH peak and having less
than 3 pg/ml estradiol were pooled within heifer for evaluation. There
was no difference (P>.10) in total protein concentration or plasma
osmolality between heifers at 21.3 C and 32.0 C (table 3). These re-
sults suggest that no appreciable plasma dilution had occurred.
However, we have no measurement of total plasma volume of heifers for
A Corticosteroid Binding Globulin (CBG) of plasma has been re-
ported for various species (Seal and Doe, 1965) and also is present in
the bovine (Lindner, 1964). Such a protein acts as a corticoid carrier
molecule through the blood. Although total plasma protein concentration
(table 3) did not vary between treatments, certain alterations of pro-
tein composition may have occurred. Although plasma corticoid
concentrations did not differ between treatments, their potential
biological effectiveness would be appreciably altered if the concentra-
tion of plasma CBG differed.
Utilizing the procedure of Pegg and Keane (1969), the association
constant (Ka) and cortisol binding capacity of CBG were determined on
pooled samples (within heifer) of each experimental heifer. The
association constant did not vary due to treatment (P>.10; table 3).
An average experimental Ka of 1.86 X 10' M-1 was indicative of a protein
with an intermediate affinity for the cortisol ligand. It is a protein
- 49 -
Table 3. Physical characteristics of plasma
at 21.3 C and 32.0 C.
Cortisol Binding Capacity
(Ka X 107 M-1)
259.65 + 11.65
6.7 + 1.2
118.93 + 48.02
75.50 + 9.05
268.70 + 10.41
5.9 + 1.0
55.86 + 11.19*
2.20 + .69
a (I + SD)
- 50 -
with an association constant higher than a low affinity protein such as
human serum albumin (Ka = 1 X 10' M-1) but lower than the Ka for a
tissue receptor protein such as the cortisol mammary gland receptor
(Ka = 5 X 108 M-1; Tucker, Larson and Gorski, 1971). It was not expect-
ed that thermal stress would alter the physical chemical properties of
the CBG protein (Ka) but perhaps may alter the amount (capacity) of CBG
per ml of plasma. Indeed there was a significant difference (P<.05) in
cortisol binding capacity (ng/ml) between treatments (table 3). Thus
under experimental conditions for quantifying cortisol binding capacity
at 4 C, plasma of heat stressed heifers had a 53% lower capacity to
bind cortisol. This suggested that under such conditions, plasma from
hyperthermic heifers contained a decreased concentration of CBG.
Therefore, under environmental temperatures of 32.0 C at a body tempera-
ture of 40.24 C both the concentration of plasma CBG and cortisol bound
CBG (product) would be less if the rate constant for the forward reaction
was not different at an elevated body temperature of 1.5 C (table 1).
With this reasoning, heifers exposed to a thermal stress characteristic
of our experiment would have a greater percent free cortisol compared
to CBG bound cortisol at a constant total cortisol concentration. As
previously described, total plasma corticoid concentrations did not
vary between treatments (6.7 compared to 5.9 ng/ml; table 3).
Under other stressful conditions a lowered corticoid binding
capacity has been reported for various species. In human burn patients
a slightly lowered cortisol binding capacity has been reported (Mortensen
et al., 1972), in which decreased capacity was inversely related to burn
area. By analogy, lactation can be considered a stress in the sense
- 51 -
that reproductive efficiency is lower during this period. Lactation
inhibits the onset of estrous cycling in rats nursing 6 or 12 pups
compared to post-parturient rats which do not lactate (Tucker and
Thatcher, 1968). Early weaning of calves from their dams increased the
occurrence of estrus and increased pregnancy rates in beef and dairy
cattle (Laster, Glimp and Gregory, 1973). Troconiz (1973) has reviewed
the cystic ovary condition in dairy cattle. High milk producing cows
had a greater incidence of cystic ovaries and therefore a greater fre-
quency of reproductive problems. In rats nursing 12 pups, CBG activity
was lower in comparison to rats nursing only four pups (Westphal, 1970).
Such a nursing intensity will delay occurrence of normal estrous cycles
(Tucker and Thatcher, 1968). Thus under conditions of lactational
stress (relative to reproductive performance) CBG activity was depressed,
The liver is the reported source of CBG (Guyton, 1966) and the
thyroid gland is reported to exert a controlling influence on CBG
activity. Gala and Westphal (1966) showed that TSH stimulated CBG
activity in hypophysectomized rats and was primarily responsible for
regulation of CBG levels. In cattle under conditions of high environ-
mental temperatures, thyroid activity was depressed (Johnson and Yousef,
1966). If the hormonal control of CBG production is grossly comparable
between rats and cattle, then lower CBG binding capacity of heifers
detected in our study would be expected.
Hypothyroid patients have a slower turnover of cortisol. Both
bound and free steroid fraction disappearance rates were slower than
normal or hyperthyroid patients (Beisel et al., 1964). In our study
the amount of free hormone would have a greater biological role in the
- 52 -
heat stress group due to the lower binding capacity. This may result in
a lower level of ACTH secretion due to greater negative feedback inhibi-
tion. In the bovine, corticoid turnover rates were depressed during
chronic heat stress (Christison and Johnson, 1972). This also suggests
a longer biological life for the circulating corticoid allowing a greater
ACTH negative feedback since less corticoid is also bound to transcortin.
However, the amount of free cortisol in our study, estimated by extra-
polation at 4 C, was not different (P>.10) between the 21.3 C heifers
(1.39 ng/ml) and 32.0 C heifers (1.43 ng/ml).
Clarification is needed in this area as to the physiological role
of bound and free steroids because of conflicting reports between species
and stress situations. Our finding that corticoids were not elevated
during chronic heat stress, irrespective of the binding capacities, may
be advantageous to the cow in that heat production has been shown to
increase 30 to 40% at 35 C when hydrocortisone acetate was administered
(Yousef and Johnson, 1967).
In the second phase of the experiment, 8 days following ovulation
in the last heifer, 200 IU ACTH was injected, IV, into 10 heifers. The
ACTH was given while heifers were in the luteal phase of the estrous
cycle or at a time when a progesterone increases in peripheral plasma
due to ACTH injection (Gwazdauskas, Thatcher and Wilcox, 1972; Wagner,
Strohbehn and Harris, 1972) may not have a detrimental effect on the
developing embryo (Johnsson et al., 1974). Figure 8 shows the corticoid
response curves following ACTH injection. The 32.0 C group responded
with significantly lower (P<.10) corticoid concentrations. The 6th
order regression curves were not parallel (P<.01) suggesting that the
-2 -I 0 I 2 3 4 5 6 7 8 9 10 II 12
C OR 32,0
CORTICOIDS FOLLOWING INJECTION
OF 200 IU ACTH
21.3Co ---,R2P .67
- 54 -
hot group response was earlier to reach a peak (75 min. compared to
105 min.), had a lower magnitude (73.5 compared to 100.2 ng/ml corticoid)
and was of shorter duration (4 hr. compared to 5 hr.; Appendix, table 11).
This response is comparable to that reported by Shayanfar (1973) in
which lactating cows exposed to environmental temperatures above 21.1 C
responded the same way. The cool heifer response was best described by:
Y (corticoids, ng/ml) = -521.387 + 4648.999X 12253.643X2 + 13609.684X3
- 5942.539X4 + 4.326X5 + 464.541X6 (P<.01), whereas the hot heifer re-
sponse was best characterized by Y = -613.342 + 6525.433X 23576.092X
+ 41551.725X3 38756.632X4 + 18359.263X5 3477.023X6 (P<.01).
The apparent reduced ability of the adrenal to secrete and/or
synthesize corticoids following ACTH stimulation during heat stress may
be related to a chronic lower level of endogenous ACTH secretion. The
lower plasma cortisol binding capacity in the 32.0 C heifers may provide
a greater amount of free corticoid to exert a feedback inhibition on
endogenous ACTH secretion. In addition, there is also a lower level of
corticoid turnover and secretion during chronic heat stress (Christison
and Johnson, 1972). As a result the degree of chronic endogenous ACTH
secretion maybe less, causing a reduction of responsive adreno-cortical
tissue. These conditions may result in a lower adrenal corticoid increase
in response to a pharmacological challenge with ACTH. A reduced level
of adrenal function during heat stress would be advantageous to the animal
calorigenically. Corticoid secretion did not appear to be higher in the
heat stressed group since resting corticoid levels were not greater than
controls. However, a possible decreased adrenal secretion rate was not
reflected by a lower plasma corticoid concentration. It was not until
- 55 -
a response to ACTH was evaluated that adrenocortical function appeared
to be depressed.
To determine the significance of this apparent reduction in
adrenal responsiveness to ACTH due to hyperthermia plasma ACTH levels
need to be determined in the bovine under different physiological stress
situations. Other possibilities include determining effects of stress
on ACTH receptors, more definitive studies on corticoid-CBG binding
properties in relation to thermal stress and possible steroidogenic-
enzyme alterations in the adrenal.
Pre-ACTH plasma progestin and corticoid concentrations were test-
ed to detect differences in levels due to heat stress and pregnancy
status (Appendix, tables 11 and 12). The analyses include temperature,
pregnancy status and heifers nested in temperature-pregnancy status.
There were no statistically significant differences (P>.10) either in
hormone concentrations due to temperature or pregnancy status. However,
significant among animal variability (P<.05) was found in progestin
levels. These results agree with the pre-PGF2a treatment hormonal
values in the first phase of this experiment (Page 30). Our observa-
tions conflict with a summer seasonal depression in corticoid and
progestin concentrations reported by Stott and Wiersma (1973). The
present study also did not confirm their finding of higher progestins
in fertile cows on day 15 of pregnancy or the estrous cycle. This
period of corpus luteum function is comparable to our study (heifers
were between estrous cycle days 9-13).
In summary, environmental treatment of 32.0 C evoked a 1.49 C
increase in rectal temperature and a 3 to 4 C increase in skin
- 56 -
temperatures. The time durations between PGF2a injection and the LH
peak and the period between PGF2a and ovulation were not different
(P>.10) between treatments. Length of estrus was shorter (P<.10) for
the heat stressed heifers. Two of four heifers inseminated in the 21.3
C chamber were pregnant at 40 days compared to none of five in the 32.0
C chamber. Thus, the environmental condition did affect body temperature,
duration of estrus and overall fertility.
Preinjection plasma samples showed no differences (P>.10) in
any of the hormonal measurements due to the main effect of temperature.
Average progestin concentration between treatments was not different
(P>.10). However the 5th order response curves were not parallel (P<.01)
indicating a different time response between treatments. Progestin
concentrations declined in a similar manner in both groups following
PGF2a injection. Heifers in the 21.3 C group, on the average, had a
LH surge about 24 hr. later than heifers in the 32.0 C group. This
24 hr. time lag would account for the difference in time responses when
data were synchronized to time of LH peak. Mean estradiol concentra-
tions were significantly (P<.10) lower in the heat stressed heifers.
The lower plasma estradiol may have contributed to the shorter estrous
periods seen in the 32.0 C heifers. However, these lower concentrations
of estradiol were adequate enough to elicit estrous behavior and trigger
LH release causing a subsequent ovulation.
Estrone showed no apparent association with the onset of estrus
or LH peak when the data were synchronized to the time of the LH peak.
There was a significant elevation (P<.05) of estrone due to heat stress
but there was no evidence that estrone time trends following PGF2a were
- 57 -
not parallel (P>.10) suggesting that in both treatments estrone follow-
ed a similar decline postinjection. No significant differences (P>.10)
were found in mean LH concentrations between heifers at 21.3 C or 32.0
C. Preovulatory peak LH concentrations were 32.2 ng/ml and 33.2 ng/ml
plasma for the 21.3 C and 32.0 C heifers, respectively. All animals
had a preovulatory LH surge, suggesting that hyperthermia did not pre-
vent the triggering mechanism for LH release.
There was no change in prolactin associated with estrus or the
LH peak, therefore prolactin was analyzed relative to time of PGF2a
injection. Mean prolactin concentrations were not different between
treatments (P>.10). The 4th order time curves were not parallel (P<.005).
Heifers in the 21.3 C chamber had a decline in plasma prolactin after
the initial sampling as compared to increased prolactin concentrations
in the 32.0 C heifers during this early blood sampling period. The
summer seasonal increase in plasma prolactin reported by various
researchers may be more related to photoperiod effects. There was no
difference (P>.10) between treatment means in plasma corticoid concentra-
tions. Furthermore, we were unable to detect any individual treatment
time trends after looking at regressions up to the 5th order. Plasma
corticoid C.V. was 65% after accounting for variability due to treat-
ment, heifers within treatment and time trends up to the 5th order.
In an attempt to determine if plasma dilution may have occurred,
total protein concentration and osmolality were measured. There was no
difference (P>.10) in total protein concentration or osmolality between
treatment groups. However, no measurement of total plasma volume was
made. Cortisol binding capacity of CBG and its association constants
- 58 -
(Ka) were determined. The affinity (Ka) of cortisol for CBG was not
different between treatments (P>.10); however,the binding capacity
of CBG for cortisol was significantly (P<.05) reduced in the 32.0 C
heifers. This observation suggested that under experimental conditions
(4 C) for determining the binding capacity of cortisol, the hyperthermic
heifers may have had a decreased concentration of CBG.
ACTH (200 IU) was injected, IV, into 10 heifers. The 32.0 C
heifers responded with a significantly lower (P<.10) corticoid concen-
tration. The 6th order regression response curves were not parallel
(P<.01) suggesting that the hot group response was earlier to reach a
peak (75 min. compared to 105 min.), had a lower magnitude (73.5 compared
to 100.2 ng/ml corticoids) and was of shorter duration (4 hr. compared
to 5 hr.). Adrenal responsiveness was significantly less in heifers
maintained at 32 C.
Results of this experiment show only subtle thermal effects on
plasma concentrations of estradiol and estrone and no effects on LH,
progestins, corticoids and prolactin. Apart from possible hormonal
involvement with duration of estrus, heat stress does not appear to af-
fect the hormonal milieu associated with corpus luteum regression,
follicle growth and ovulation. The significance of possible lowered
adrenal response in hot environments may be related to a state of
lowered heat production. Since corticoids are known to be calorigenic
(Yousef and Johnson, 1967) a lowered adrenal responsiveness in hyper-
thermic heifers might be physiologically advantageous.
The experiment described in this section has not specifically
considered the possible environmental and hormonal effects on uterine
temperature. It was of prime importance to characterize uterine
thermal changes during the period of luteal regression, follicle growth
and ovulation under conditions of a mild heat stress, and to document
possible estrogen induced uterine thermal changes.
EXPERIMENT I: THERMAL CHANGES OF THE BOVINE UTERUS FOLLOWING
ADMINISTRATION OF ESTRADIOL-17B
The first experiment (Section II) indicated that a thermal stress
increased body temperature, suppressed fertility and caused a slight de-
crease in endogenous estradiol secretion. Furthermore, we reported
previously that uterine temperatures both on day of and day after
insemination were inversely related to fertility (Gwazdauskas, Thatcher
and Wilcox, 1973). This directly indicated that temperature of the
uterus was closely associated with fertility.
Other factors in addition to environmental temperature may in-
fluence uterine temperatures. For example estrogen administration was
shown to increase uterine blood flow in sheep (Huckabee et al., 1970;
Greiss and Anderson, 1970; Rosenfeld et al., 1973; and Resnik et al.,
1974). This uterine hyperemia may have caused heat to be dissipated
from the uterus, thus cooling the uterine cavity (Abrams.et al., 1970a).
Uterine blood flow changes in sheep were monitored following estrogen
injections by looking at differences in temperature between the uterus
and aorta. A rise in blood flow rate resulted in a lower uterine
temperature (Abrams et al.,o1970a).
- 60 -
- 61 -
Although uterine temperature and estrogen relationships have
been foundin sheep, this phenomena has not been examined in the bovine.
Objectives of this study were to determine if uterine-aortic temperature
differences exist in the bovine, and if such differences change following
injection of Estradiol-178.
Materials and Methods
Thermocouple Preparation and Calibration
Lengths of 36 gauge, nylon coated, copper constantan wire
(Revere Corp., Wallingford, Conn.) were pulled through polyvinyl
tubing (V5-V7; Bolab Inc., Derry, N. H.) for measurements of uterine
and blood temperatures. The terminal thermojunctions to be placed in
the saphenous artery then were pulled through a larger polyvinyl tube
(V-12) for additional support. The ends of all thermojunctions were
heat-sealed in the polyvinyl by pushing them through a siliconized,
narrowbore glass tubing which was being heated on a soldering iron.
After sealing, ends were coated with liquid tygon (U. S. Stoneware Co.).
Stranded, untinned copper extension wires (Leads and Northrup, #27-32-36,
Philadelphia, Pa.) were soldered to divided copper wires leading to the
thermojunctions. All extension wires led either to a millivolt
potentiometer (#8686, Leads and Northrup, Philadelphia, Pa.; limits
of error of recording system + .075 C) or to a strip chart recorder
(Hewlett-Packard, M 7100B; limits of error of recording system j+ .03 C).
Most, but not all of the potentials from the aortic-ice water thermo-
couple were suppressed by known amounts before being amplified and
- 62 -
Calibration of the thermocouples was made routinely by use of a
Bureau of Standards Certified Thermometer in a well-stirred, insulated
water bath held at intervals between 36 to 40 C. The thermocouple
readings were 0.05 to 0.075 C above the certified thermometer reading,
so all data collected were corrected for these constants.
Surgical Techniques and Experimental Protocol
Four 2-year-old heifers with histories of regular estrous cycles
were used in these experiments. Prior to surgery, heifers were placed
on a 48 to 72 hr. feed and water fast. Heifers were anesthetized with
2 to 4 g sodium thiopental (Abbott Laboratories, North Chicago, Ill.)
dissolved in saline (2 g/20 ml) while standing and restrained. They
were placed onto a portable operating table, tracheotomized and maintained
under surgical anesthesia with methoxyfluorane (Pitman-Moore, Washington
Crossing, N. J.). After removal of hair, the abdominal and inguinal
regions were scrubbed thoroughly with germicidal soap and rinsed with
A 15 cm longitudinal midventral incision was made through the
abdominal wall at the cranial margin of the mammary gland. A sharpened
stainless steel cannula was carried into the abdominal cavity through
this midventral incision and pressed through the abdominal wall in the
flank area. All thermojunctions and approximately 2.5 m of extension
wires were drawn through the cannula leaving the remainder of the 3 m
of extension wire and connectors coiled up in a canvas pack. The
cannula was removed from the abdominal cavity by sliding it over the
thermojunctions and withdrawing it through the midline incision. The
- 63 -
pack subsequently was attached to the flank with one or two stainless
steel pins passed through a flap of skin.
The uterus was elevated so that the junction of the uterine horns
with the uterine body could be visualized. Using small scissors and
straight forceps, a 3 to 4 cm tunnel was made under the serosa in the
medial aspect of one uterine horn about 1 cm from the bifurcation. A
thermojunction was inserted into this tunnel and tied in place with
000 silk thread. The extension wires were secured by two to three
additional ties through the serosa along the uterine horn.
Thermojunctions for aortic blood temperatures were routed through
the midventral incision, tunneled under the skin to the inguinal area
where the saphenous artery was exposed. These thermojunctions then
were inserted into the saphenous artery, passed 70 to 75 cm upward to
the abdominal aorta and extension wires fixed with silk suture at the
point of entry into the vessel. Incisions were closed in layers.
Thermocouple placements were confirmed prior to their surgical removal
7 to 10 days after completion of the experiment.
Twenty-four hr. prior to intravenous (IV) injection either of
3 mg Estradiol-17 (Progynon-Schering Corp., Bloomfield, N. J.) or 12
ml of .9% sterile saline, heifers were fitted with polyvinyl catheters
(V-7) by jugular venipuncture. Catheters were'filled with heparin
solution (15 U/ml of .9% saline), capped with a brad and the external
catheter placed in an adhesive tape pouch glued to the neck with branding
cement (Electro Cote Co., Minneapolis, Minn.).
On the day of injection heifers were placed in a stanchion barn
on rubber comfort mats at least 2 hr. prior to recording temperatures.
- 64 -
Each of the four heifers received an estradiol injection, and two of the
heifers also received two saline injections each. Thus there were a
total of four estradiol and four saline experiments. All heifers received
treatment during the luteal phase of the cycle. Recordings were made
from the millivolt potentiometer at 15-min. intervals beginning 1 hr.
prior to IV injection either of Estradiol-17g or saline and ending 6 hr.
after the injections. A pre-experimental control period of 1 hr. was used
to determine a steady state level of uterine temperature. Repeatability
of triplicate measurements at each time was 0.92 for aortic temperature
(mV) with a C.V. of 0.07% (n=66). Repeatability and C.V. for
ATuterus-aorta (V) were 0.99 and 6.44%, respectively.
Initially, an additional thermocouple was placed in the uterine
lumen as well as in the uterine serosa. Prior to and following estrogen
injection the temperatures at both reference points were identical. To
avoid any possible complication due to presence of an intrauterine object,
all subsequent animals were fitted only with a uterine serosa thermo-
junction. In a separate experiment, temperatures were recorded
continuously before and after an injection of Estradiol-173. The major
statistical technique to analyse time changes was least squares as de-
scribed by Harvey (1960). Statistical models were selected based on
tests of significance of the higher order terms in the regression
analyses and visual appraisal of the graphs.
Results and Discussion
The uterine and aortic temperature response following intravenous
injection of 12 ml.saline is shown in figure 9. The slight increase in
OC 12 ml
38.2- o0 o --- o
0[ o Y = 38.13 +- .0'223X
38.0 AORTA A
A A '
37. 8 A .
37.8- A A ^A A A
Y-= 37.80 + .0210X
o, A = actual means
S. 0 I 2 3 4 5 6
FIGURE 9, UTERINE AND AORTIC TEMPERATURE PRIOR TO AND FOLLOWING IV INJECTION OF 12 ML
- 66 -
both mean temperatures (~ 0.2 C) which occurred during the 7 h experi-
ment may be related to the normal rhythmic rise in body temperature in
cattle during the day (Bianca, 1968). The greater variability in both
temperatures 4 to 6 hr. after saline injection could have been because
of blood temperature changes induced by some restlessness due to long
confinement in the stanchions. In spite of these changes in uterine
and aortic temperatures, temperature differences between the two were
quite stable during the experiment, indicating that the ratio between
uterine heat production and uterine heat loss had remained unchanged.
Relationships of time (X) and uterine (Y ) and aortic (Y ) temperatures
are shown in figure 9. There was no evidence of curvilinearity; fitting
the two equations accounted for 36 and 37% of the within-heifer vari-
ability in Yu and Ya, respectively. There was no evidence that the
two slopes were not parallel, which suggested that the saline vehicle
had no depressive effect either on uterine or aortic temperature.
Effects of Estradiol-17 on uterine and aortic temperatures are
illustrated in figure 10. The initial fall in uterine temperature of
slightly more than 0.3 C compares favorably with the response noted
previously in sheep (Abrams et al., 1970a). The slight rise in uterine
temperature between 4 to 6 hr. post injection was undoubtedly due to
the rise in blood temperatures as noted in control experiments.
Uterine changes were curvilinear (P<.01) as indicated by the equation
(R = 0.18). A significant quadratic equation (P<.01; R = 0.04) best
describes the aortic temperature response. Why aortic temperature fell
initially is not known. Increased respiratory evaporative heat loss or
sweating may have been responsible. Estrogens are known to be potent
39.06 -.2783X + .0304X
S, A = actual means
Y = 38.53 -
.120X + .0157X2
-1i 0 I 2
UTERINE AND AORTIC TEMPERATURE PRIOR TO AND FOLLOWING IV INJECTION OF 3 MG
- 68 -
vasodilators of skin blood vessels (Reynolds and Foster, 1940), and to
the extent that heat loss was promoted by this increased skin blood flow,
a lowered temperature may result. One may propose that estrogens could
have had a subtle effect of the thermoregulatory "set point" (Hammel
et al., 1963) which resulted in activating one or more heat loss
mechanisms. However, the decrease in aortic temperature was only about
When the difference in temperature between uterine serosa and
aorta was examined the result of Estradiol-17P administration was obvious
(figure 11). The decrease in ATuterusaorta (AT ) of 0.25 C was de-
scribed by a highly significant (P<.01) curvilinear trend over time.
The ATa began to plateau at approximately 2.5 hr. post estrogen
injection and remained depressed for the duration of the recording period,
although both uterine and aortic temperature started to rise 4 to 5 hr.
post-injection. There was no significant change (P>.10) in ATua follow-
ing saline injection.
Figure 12 is a plot in 30 sec. intervals taken from a continuous
recording of temperatures of one heifer prior to and following Estradiol-
178 injection. In this animal the estrogen effect on uterine temperature
was noted within 1 hr. Rapid oscillations in temperature of the uterine
tracing were considerably dampened by the heat capacity of the uterine
A consistent finding in the estrogen experiments was the decrease
in the temperature difference between the uterus, as represented by the
subserosal temperature and the blood df the abdominal aorta. Such a
decrease in AT could occur as a result of a lowered rate of uterine
y =.34 + .0038X
y = .68 -.309X
or A =actual means
R = .78
- -. ......---.-. S, .4
FIGURE 11, ATuterus-aorta
PRIOR TO AND FOLLOWING EITHER 12 [iL SALINE OR 3 MG
38.8 T, RUS
3 mg (IV)
37. 8 I
-15 0 15 30 45 60 75 90 105 115
FIGURE 12. UTERINE AND AORTIC TEMPERATURE PRIOR TO AND AFTER INJECTION OF ESTRADIOL-17$
-FROM CONTINUOUS RECORDING,
- 71 -
heat production, a possibility which appears remote in view of the many
cellular metabolic activities induced by estrogens (Talwar and Segal,
1971). A more reasonable explanation for the lowered AT is the
augmented rate of heat loss resulting from the marked estrogen induced
elevation in uterine blood flow. Endogenous estrogens released during
the estrous cycle in ewes are known to be associated with elevated
uterine blood flow rate (Greiss and Anderson, 1970) and increased
vaginal blood flow as inferred from a significant rise in vaginal thermal
conductance in cattle (Abrams et al., 1973). Thus, there is reason to
believe that comparable cyclic, blood flow-induced changes in temperature
of the reproductive tract may occur during the estrous cycle in the
bovine. High uterine temperatures at the time of artificial insemination
are associated with diminished fertility (Gwazdauskas, Thatcher and
Wilcox, 1973). Elevated environmental temperature is thought to
suppress fertility by acting directly on the developing embryo and/or
through altering maternal endocrine function (Vincent, 1972).
Findings in the first experiment indicated that plasma estradiol
of heat stressed heifers was lower during the pre-estrous period.
Results of the present study indicate that a pharmacological injection
of Estradiol-170 can significantly decrease uterine temperatures. In
the final experiment, attempts were made to evaluate changes in uterine
temperature during the period of luteal regression (decreasing progesterone),
follicle growth (increasing estradiol) and ovulation under conditions of
a mild heat stress.
EXPERIMENT 2: THERMAL CHANGES IN THE BOVINE UTERUS FOLLOWING
PGF2a INJECTION THROUGH ESTRUS AND OVULATION
The first experiment (Section II) indicated that a thermal stress
increased body temperature, suppressed fertility and caused a slight
decrease in endogenous estradiol secretion. Next, an effect of exogenous
Estradiol-17B on uterine temperature was documented. In this final
experiment, estrus was synchronized by PGF2a and an attempt was made to
evaluate changes in uterine temperature and aortic blood temperature
with plasma estradiol and LH under conditions of mild heat stress. Such
an experiment would closely mimic responses of animals under normal
field conditions and provide additional insight into factors controlling
uterine temperature under conditions of poor reproductive efficiency.
Materials and Methods
Thermocouple preparation and calibration were the same as described
in the previous experiment with the exception that all thermocouples were
made in triplicate for each location. During the experiment, extension
wires led to a recording potentiometer (9835 A, D-C Microvolt Amplifier
and Speedomax G, Model S6000 Recorder, Leeds and Northrup, Philadelphia,
Pa.; limits of error of the recording system + .0125 C). Surgical
- 72 -
- 73 -
techniques were identical except cattle were anesthetized with 3 g
sodium thiamylal (Surital-Park Davis, Detroit, Michigan) dissolved
in saline (3 g/20 ml) and were maintained under surgical anesthesia
with halothane (Fluothane-Ayerst Laboratories, Inc., New York, N. Y.).
Blood samples were collected prior to PGF2a injection (0 hr.), at
6 hr. intervals for 48 hr. and every 4 hr. until 24 hr. after visual
detection of estrus. Measurements of LH and estradiol were by methods
previously cited. Three cycling first lactation dairy cows between 60
to 90 days postpartum and one cycling heifer were given 30 mg PGF2a-
Tham Salt (IM). All animals were between days 9 to 15 of the estrous
cycle at the time of injection. Each animal had a functional corpus
luteum at the time of surgery, 4 to 5 days earlier. At the time of
PGF2a injection each animal maintained a uterine-aortic temperature
difference (AT u) greater than .3 C during the previous 2 days and a
palpable corpus luteum. A second injection of PGF2a (10 mg) was given
to 3 of the 4 cows at 21 hr. after the first injection. This was done
to insure complete luteal regression. Cow aortic temperatures and AT
were monitored continually from 5 hr. prior to the initial PGF2a
injection until 24 hr. after the detection of estrus. Twice daily,
recordings were temporarily interrupted for 90 min. (0800 and 2000 hr.)
for estrous checks and exercise. Temperatures in the heifer were re-
corded continually for 15 min. prior to PGF2a injection until 6 hr.
Rectal palpations were made 2 days postinjection to confirm
corpus luteum regression, and again approximately 24 hr. after visual
appraisal of estrous behavior to detect occurrence of ovulation.
- 74 -
Thermocouple placement was verified 4 days after estrus by surgical
examination of the reproductive tract. At ovariectomy confirmation
of luteal regression, ovulation and new corpus luteum formation was
verified by direction.
Results and Discussion
Regression of the corpus luteum, as determined by rectal palpation,
occurred in all four animals. The three cows, at the time of ovariectomy,
had newly formed corpora lutea near the area where the old corpus luteum
had regressed. Two of the cows were detected in estrus while the thermo-
couples remained functional. The usual life span of thermocouples was
about 2 weeks. However, due to mechanical failure the thermocouples in
one cow lasted only 7 days, and the heifer was not detected in estrus
during this period. At the time of ovariectomy and recovery of thermo-
couples, confirmation of ovulation was made on the basis of a newly
formed corpus luteum.
The immediate effects of PGF2a on uterine and aortic temperatures
of two cows and ATu-a of all four animals are shown in figures 13, 14
and 15. To simplify and assimilate the continuous recordings, points
at 15 min. intervals were taken to describe the data. Following the 30
mg injection of PGF2,, the ATu-a dropped .4 C (P<.01) from approximately
.54 C to .16 C at 45 min. postinjection (figure 15). A similar drop
(P=.10) in ATua occurred following the 10 mg PGF2a injection. However,
the magnitude of the decline was only about .15 C which occurred 30 min.
postinjection. The lower ATa before injection of PGF2. (10 mg) and
!, ,\,, ,.
25.0 -- -
I I t I i
---I I I I a A
UTERINE AND AORTIC
2200 0200 0600 1000 1400 1800
HOURS OF DAY
TEMPERATURES PRIOR TO AND FOLLOWING PGF2, INJECTIONS
t s t t I t I I I I
I I/ i
I I 1 t I I
1 t I t I
HOURS OF CAY
FIGURE 14, UTERINE AND AORTIC TEMPERATURES PRIOR TO AND FOLLOWING PGF2. INJECTIONS
_~ I I __ ___ __ _____
s 1_4_ 1 1 I
( 5 -
o- MG PG5c (M=4)
- --- 10 NG PGF (Nt3)
FIGURE 15. CHANGES IN ATu-a
FOLLOWING PGF2 INJECTIONS.
- 78 -
smaller decline may be related both to time and hormonal status after
first injection and also dose of PGF2 These observations were not
anticipated because various researchers (Bergstrom et al., 1968; Brody
and Kodowitz, 1974; Clark et al., (1972) have reported a vasoconstrictor
effect of PGF2 A vasoconstrictor action would tend to decrease blood
flow through the uterus and therefore elevate the ATua. The marked
drop in blood temperature might be attributed to an increased respiratory
evaporative heat loss. Indeed, an increased respiratory rate was de-
tected shortly after PGF2a injection but not quantified. Sweating is
negligible in cattle, so in order for heat to be eliminated by way of
respiratory evaporative heat loss there has to be a tremendous increase
in lung ventilation (Brody, 1945). Also, Lewis and Eyre (1972) report-
ed increased respiratory volume following PGF2a administration to
calves. However, if aortic temperature did fall, a decrease in the
ATua would not occur unless there was selective PGF2a action on the
uterus to increase heat loss or decrease heat production.
If one uses the thermal balance equation, Q=FcAT, then theoretical
heat production, Q, and uterine blood flow, F, can be calculated based
on the ATu-a changes.
Q = rate of uterine heat production (cal/gm tissue-min.)
F = rate of uterine blood flow (gm blood/gm tissue-min.),
density of blood taken as 1 gm/ml
c = specific heat of blood (.87 cal/gm blood-C)
ATa = temperature difference between the uterus and
aortic blood (C); (adapted from Abrams et al., 1970b).
aortic blood (C); (adapted from Abrams et al., 1971b).
- 79 -
The major assumption is that all heat loss from uterine tissue
is by way of the uterine veins. Therefore, in order to calculate Q,
uterine blood flow during the luteal phase of the estrous cycle needs
to be obtained. Assuming no species differences, then.we can use for
cattle the value of 119 ml blood/kg-min. for uterine blood flow in
sheep (Huckabee et al., 1968).
Theoretically, at ATua = .55 C just prior to PGF2a injection
in the present experiment:
Q = FcAT
Q = .119 gm blood/gm tissue-min. X .87 cal/gm blood-C X .55 C =
By contrast, the uterine heat production rate, Q, calculated 45 min.
post-PGF2a injection when ATu-a = .16 C was determined to be:
Q = .119 X .87 X .16 =
This theoretical calculation would suggest a 3.4 fold decrease in
uterine heat production in response to PGF2a.
Lowered ATua in response to PGF2a could also be explained by an
increase in heat loss. One mechanism of heat loss would be an increase
in uterine blood flow. Be rearranging the equation, the theoretical
blood flow, before and after PGF2a, can be calculated based on oxygen
consumption data for a 350 kg Jersey cow (oxygen consumption 3.26 ml/
kg-min.; Brody, 1945). Assuming no differences between oxygen con-
sumption (per kg) of various organs of the body (in sheep the oxygen
consumption of the total body as well as the uterus is approximately
5 ml/kg-min.), oxygen consumption of uterine tissue of 3.26 ml/kg-min.
- 80 -
multiplied by a calorific value of 4.8 cal/ml oxygen (based on an assumed
R.Q. of .8; Brody, 1945) would give a calculated heat production of
3.26 ml/kg-min. X 4.8 cal/ml = .0156 cal/gm-min.
Rearranging the original Equation:
then at ATua = .55 C (pre-PGF2a injection):
F .0156 cal/gm-min.
.87 cal/gm blood-C X .55 C
32.6 ml blood/kg-min.
and at ATua = .16 C (post-PGF2a injection):
.87 X .16
112 ml blood/kg-min.
The 3.4 fold increase in theoretical uterine blood flow calculated
above would be comparable to changes reported by various researchers
(Huckabee et al., 1970; Abrams et al., 1970a) following estrogen in-
jections. However, the time course of maximum PGF2a response (45.min.)
was of shorter duration and had a more rapid onset than an estrogen in-
duced decrease in ATa (Section III, Experiment 1). These observations
indicate the need for more definitive experiments in the bovine to pin-
point the cause of the lowered ATu-a in response to PGF2 A basic
question is whether there is an increase in uterine blood flow or a
- 81 -
decrease in heat production.
Figures 13, 14, 16 and 17 show individual cow aortic and uterine
temperature changes pre- and post PGF2a injection (figures 13 and 14)
and at the time of the LH peak (figures 16 and 17). Both cows appeared
to exhibit circadian changes in aortic and uterine temperatures. The
range of aortic temperatures (37.9 to 41.0 C) within the two cows is
comparable to body temperatures reported by Bligh and Harthoorn (1965)
in African cattle. They reported that maximum and minimum body tempera-
tures were closely associated with sunset and sunrise, respectively.
In the latter study thermistors were implanted 8 cm into the dorsal
caudal neck region to record deep body temperature. Aortic temperature
patterns in our study showed that maximum daily deep body temperature
occurred close to midnight, whereas minimal body temperature occurred
between 0800 and 1200 hr. Although cows were turned out to exercise
at 0800 and 2000 hr., their aortic temperatures returned to pre-
turnout baselines within 2 hr. after their return to the barn. Also,
there appears to be a 4 to 6 hr. lag behind barn air temperature in
maximum and minimum body temperature.
The uterine and aortic temperatures were highly correlated
(Appendix, table 13) but no correlation between AT and either uterine
or aortic temperature was detected. This might suggest that there was
no change in uterine blood flow and uterine heat production. However,
these correlations were based on data throughout the entire experiment
and any possible increases in ATua at higher body temperatures (figures
16 and 17) may have been undetected statistically. Visual appraisal of
figures 16 and 17 do show a widening of the AT at maximum daily
- 82 -
i rd t,'dU I
2400 1200 2400 1200 2400 120
HOURS OF DAY
0 2400 1200
FIGURE 16. UTERINE AND AORTIC TEMPERATURES, LH AND
ESTRADIOL IN G665 AND AIR TEMPERATURES,
I --* \ %
- 83 -
/ 1I- / /
I -" aorta \
l \ r I
1200 2400 1200 2400 1200
HOURS OF DAY
UTERINE AND AORTIC TEMPERATURES, LH AND ESTRADIOL
IN JN15 AND AIR TEMPERATURES,
- 84 -
uterine and aortic temperatures.
Figure 18 shows the significant (P<.01) curvilinear (2nd order)
time trends for uterine and aortic temperatures when data were pooled
across days for each time of blood sampling. The data representing
each individual sampling point is the average of individual 15 min.
points + 2 or 3 hr. from time of the blood sample. Also only tempera-
tures preceding turn out of cows (0800 and 2000 hr.) were used in
obtaining an average for the 0800 and 2000 hr. blood sampling times.
The air temperature plot is comprised of average values across the
individual days. Uterine temperature and aortic temperature were not
influenced by barn air temperature (P>.10; Appendix, table 13). The
uterine temperature trend throughout the day was best described by
Y(uterine temperature, C) = 40.04 .143X + .007X (P<.01) where X:= hr.,
whereas aortic temperature was best characterized by Y(aortic tempera-
ture, C) = 39.50 .125X + .006X2 (P<.O0; Appendix, table 14). The
body temperature lag of about 6 hr. behind air temperature (1600 hr.-
peak air temperature compared to 2400 hr. body temperature peak) is
best seen in figure 18.
Possible explanations for the time delay could be:
1) Thermal inertia of the cow. For example, ambient temperature
will increase more rapidly than body temperature because of the mass and
heat capacity of the large mammal.
2) Inherent circadian rhythm of body temperature which may be
independent of environmental temperature and cued to external events
such as light-dark cycles, feeding regimen, presence of barn personnel
and other factors which were uncontrolled in this experiment.
T R2 = .35
39.5 Tuterus, R2 35
Taortic R2 = .45
0400 0800 1200 1600 2000 '2400
HOURS OF DAY
FIGURE 18. CIRCADIAN UTERINE, AORTIC AND AIR TEMPERATURE CHANGES.
- 86 -
Correlated responses between concurrent temperature measurements
(uterine temperature, aortic temperature and air temperature) are low
due to the time delay phenomena.
Of interest in this experiment is the observation that uterine
temperatures during the day reached 40 C for periods of up to 6 hr. as
ambient temperature fluctuated near 30 C (figures 16 and 17). Tempera-
tures of this magnitude (40 C) are damaging to embryo development at
the 1 to 4 cell stages's (Alliston et al., 1965). In JN15 (figure 17)
this increased uterine temperature was at the time when artificial
insemination would normally have been performed.
We failed to detect an association between concurrent measurements
of AT with estradiol or LH (Table 13). Based upon Experiment 1 there
was a 2.5 hr. delay between injection of a pharmacological dose of
estradiol and the minimum ATu-a From PGF2a injection through the
LH surge (figure 19) estradiol concentrations fluctuated considerably
as did the AT (figures 16 and 17). Not until the massive LH dis-
charge was there an appreciable rise in ATa at a time when estradiol
was decreased (figure 19).
Findings of this experiment indicate that uterine and aortic
temperatures followed a daily circadian rhythm, and,because of a time
lag in these temperatures behind air temperatures, correlations between
body temperatures and ambient temperature were negligible. Failure to
detect an association between AT and hormonal measurements may be
due to the time lag, also, The mild heat stress (which by definition
occurs in cattle anytime ambient temperatures exceed 30 C), to which
these cows were subjected to, may have contributed to the high uterine
and aortic blood temperatures. Uterine temperatures periodically
< R2 = .76
I i ler
SI I H AD EST L
Fo 1% I
4 )S % -
-84 -72 -60 -48 -36 -24 -12 0 +12 +24
FIGURE 19. CHANGES IN ATu-a ASSOCIATED WITH ENDOGENOUS LH AND ESTRADIOL CONCENTRATIONS.
exceeded 40 C (Appendix, table 15). Since survival rate of fertilized
ova exposed to 40 C for 3 hr. is seriously decreased, these observations
may be of some practical significance in improving reproductive per-
formance in a hot climate.
SUMMARY AND CONCLUSIONS
Ten normally cycling Holstein heifers at the USDA, Agricultural
Research Center, Beltsville, Maryland, were assigned to one of two en-
vironmental treatment groups (21.3 C, 59% RH or 32.0 C, 67% RH).
PGF2a-Tham Salt (PGF2c) was used to cause corpus luteum regression and
synchronize estrus. Blood samples were collected prior to PGF2a
injection and at 6 or 4 hr. intervals following injection through
ovulation. Plasma samples were analysed to determine concentrations of
progestins, estradiol, estrone, LH, prolactin, corticoids, total protein
concentration, osmolality, cortisol binding capacity and cortisol
association constants. In the second phase of this first experiment
adrenal responsiveness to ACTH (200 IU) was tested by quantification'of
corticoid concentrations in plasma prior to and up to 12 hr. following
injection of ACTH. Least-squares analyses were conducted to characterize
treatment, animal and within-animal time trends in plasma progestins, estra-
diol, estrone, LH, prolactin and corticoids. Other response variables were
analyzed by analysis of variance.
Environmental treatment of 32.0 C evoked a 1.49 C increase in
rectal temperature and a 3.59 C increase in skin temperatures. Time
durations between PGF2x injection to LH peak and ovulation were not
different (P>.10) between treatments. Length of estrus was shorter
- 89 -