The role of light and melatonin on the reproductive endocrinology of anestrus, cyclic and early pregnant mares

MISSING IMAGE

Material Information

Title:
The role of light and melatonin on the reproductive endocrinology of anestrus, cyclic and early pregnant mares
Physical Description:
xv, 353 leaves : ill. ; 29 cm.
Language:
English
Creator:
Cleaver, Brian Douglas, 1966-
Publication Date:

Subjects

Subjects / Keywords:
Horses -- Reproduction   ( lcsh )
Animal Science thesis, Ph. D
Dissertations, Academic -- Animal Science -- UF
Genre:
government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1996.
Bibliography:
Includes bibliographical references (leaves 318-348).
Statement of Responsibility:
by Brian Douglas Cleaver.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002117440
oclc - 35777918
notis - AKV7156
System ID:
AA00009008:00001

Full Text












THE ROLE OF LIGHT AND MELATONIN ON THE REPRODUCTIVE
ENDOCRINOLOGY OF ANESTRUS, CYCLIC AND EARLY PREGNANT MARES













BY

BRIAN DOUGLAS CLEAVER


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




UNIVERSITY OF FLORIDA


1996














ACKNOWLEDGEMENTS


The author wishes to express his deepest appreciation to his major advisor, Dr. Dan C.

Sharp, and the members of his graduate committee, Dr. William C. Buhi, Dr. Michael J.

Fields, Dr. Louis J. Guillette and Dr. William W. Thatcher, for their guidance and direction in

the research and preparation of this manuscript. The author wishes to express his appreciation

to the Farm Crew and Staff of the Horse Research Center for their assistance in the

maintenance and care of the ponies. Special thanks go to Michael Porter for his friendship

and unique outlook on the way things should be. Further thanks goes to Morgan Peltier and

Gill Robinson for their assistance in the collection of data; to Dr. Phillip Fields for his

collaboration on histology; and Drs. Inho Choi and Frank Simmens for their collaboration on

concepts aromatase measurements. Very special thanks to Connie Nicklin for her friendship,

support and love during the preparation of this manuscript. My deepest thanks and love to my

parents, Donald and Carolyn, for their endless support and encouragement which allowed me

to pursue all my dreams toward their fulfillment. And special thanks to all the "girls" at the

HRC, past and present, without whose involvement, if not always complete cooperation, this

work could never have been accomplished.














TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS................................................................................................ ii

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

LIST OF FIG URES......................................................... .................................................. ix

A B STR A CT ......................................................................................... ............................... xiv

CHAPTER

I INTRODUCTION ................................................................. 1

II REVIEW OF LITERATURE...................................................... 4

Seasonal Breeding.................................................................. 4
D aylength................................................................................. 5
Photoperiodism and Biological Rhythms............................... 5
Photoperiodism in Mares................................. ....................... 9
The Anatomical Photoperiodic Pathway .......... ............... 13
The Structure of the Pineal Gland..................................... 16
Melatonin: Synthesis and Secretion...................................... 21
Melatonin Receptors .......................................................... 26
The Reproductive Hormones................................................ 30
Ovarian Steroid Synthesis in Mammals............. ......... 30
Steroid Binding Globulins............................................ 33
Ovarian Steroid Biosynthesis and
Secretion in M ares.................................. ................ 33
Gonadotropin Synthesis and Secretion
in M am m als................................................................ 36
Effects of GnRH on Gonadotropins.................................... 37
The Seasonality of Gonadotropin Secretion in Mares......... 48
Early Pregnancy in Mares ....................................................... 49
The Influence of Melatonin on Reproduction................... 51

III LH SECRETION IN ANESTROUS MARES EXPOSED
TO ARTIFICIALLY LENGTHENED PHOTOPERIOD
AND TREATED WITH ESTRADIOL.................................... 59










Materials and Methods............................................................. 61
Experimental Design...................................................... 61
Blood Collection and Handling.................................. 62
Light Treatment............................................................ 62
Estradiol Treatment..................................................... 65
GnRH Challenge.......................................................... 66
Ovarian Activity.......................................................... 66
LH A ssays.................................................................... 66
Statistical Analysis...................................................... 67
R esults..................................................... ................................. 68
Ovarian Activity........................................................ 68
Hormone Concentrations........................................... 68
Discussion..................................... ................. ........... 73

IV EFFECT OF ESTRADIOL TREATMENT ON
LH SECRETION DURING LATE ANESTRUS:
AN INVESTIGATION OF THE EXISTENCE
OF A PHOTIC GATE.............................................................. 83

Materials and Methods............................................. ......... 84
Experimental Design...................................................... 84
Blood Sample Collection ............................................. 85
Hormone Assays (LH, E2 and P4)........................... 88
Ovarian Follicular Activity......................................... 89
Statistical Analysis......................................................... 89
R esults.......................................................... ....... ................ 90
Ovarian Activity .......................................................... 90
Efficacy of E2 Treatment............................................... 93
LH Concentrations............................................ ....... 93
D discussion ................................. .. ....... ................................ 99

V THE EFFECTS OF MELATONIN ADMINISTRATION DURING
PROESTRUS, ESTRUS OR DIESTRUS PLASMA LH,
ESTRADIOL, PROGESTERONE AND FOLLICULAR
AND LUTEAL GROWTH...................................................... 108

Materials and Methods........................... ......................... 110
Animal Handling............................................................ 110
Experimental Designs..................................................... 111
Blood Sample Handling ........................................... 113
Measurement of Plasma Melatonin,
LH, E2 and P4 Concentrations.................................. 113
Analysis of Episodic Secretion in Study #1 116
Statistical Analyses......................................................... 116
R esults...................................................................................... 119
Study #1: Melatonin Treatment at Proestrus............... 119


CHAPTER


page







CHAPTER page

Study #2: Melatonin Treatment
at Proestrus and Estrus............................................ 158
Study #3: Effects of Melatonin Treatment
at D iestrus............................................................... 175
D iscussion............................................................................. 189

VI THE EFFECT OF MELATONIN ADMINISTRATION AT
ESTRUS ON FOLLICULAR, LUTEAL AND
EMBRYONIC GROWTH AND SELECT
ENDOCRINE FACTORS OF PREGNANT
PONY M ARES......................................................................... 193

M materials and M ethods......................................................... 195
Anim al Handling......................................................... 195
Experimental Designs................................................. 195
Blood Sample Handling.............................................. 197
Measurement of Plasma LH, eCG, and
P4 Concentrations.................................................... 197
Statistical Analyses...................................................... 198
R esults................................................................................... 199
Study #1: Melatonin on E2 and P4
through Day 16 of Pregnancy................................. 199
Study #2: Melatonin Treatment at Estrus
then at Day 35 to 45 of Pregnancy........................ 205
D iscussion............................................................................. 220

VII THE in vitro AND in vivo EFFECTS OF
MELATONIN ON in vitro ESTRADIOL
BIOSYNTHESIS BY FOLLICLE WALL AND
CONCEPTUS MEMBRANES................................................. 225

M materials and M ethods......................................................... 227
Anim al Handling......................................................... 227
Experimental Designs.................................................. 227
E2, P4 and DNA Assays............................................. 233
Statistical Analyses..................................................... 234
R esults................................................................................... 236
Study #1: Effects of in vitro Melatonin
on Follicle Wall E2 Biosynthesis........................... 236
Study #2: Effects of in vitro Melatonin
on Conceptus Membrane E2 Biosynthesis 241
D iscussion.......................................................................... ... 241

VIII MELATONIN ADMINISTRATION TO
HYSTERECTOMIZED PONY MARES
AT ESTRUS: EFFECTS OF MORNING VERSUS
EVENING ADMINISTRATION............................................. 249








CHAPTER Page

M materials and M ethods............................................................ 251
Anim al Handling......................................................... 251
Experimental Designs.................................................. 251
Measurement of LH and P4 Concentrations 252
Statistical Analyses...................................................... 253
R esults................................................................................... 253
Study #1: Effects of Melatonin Treatment
during the Morning.................................................. 253

Study #2: Effects of Melatonin Treatment
during the Afternoon............................................... 256
A comparison of A.M.
versus P.M. melatonin treatment........................... 263
D iscussion............................................................................. 268


IX ANALYSIS OF LH, MELATONIN AND STEROID
HORMONE CONCENTRATIONS IN PLASMA AND
FOLLICLE FLUID COLLECTED AT EITHER
NOON OR MIDNIGHT FROM
PONY MARES AT ESTRUS.................................................. 271

Materials and Methods......................................................... 273
Anim al Handling......................................................... 273
Measurement of Plasma and Follicle
Fluid Hormone Concentrations............................... 274
Statistical Analysis...................................................... 275
R esults................................................................................... 276
Plasma Hormone Concentrations.............................. 276
Follicle Fluid Hormone Concentrations.................. 276
Follicle Size, Follicle Fluid Recovery
and Time of Aspirations....................................... 285
Comparisons of A4 Versus A5
Biosynthetic Pathways............................... ................. 285
Discussion...................................................................... .... 285

X THE EFFECTS OF MELATONIN TREATMENT AT ESTRUS
ON THE MORPHOLOGY OF LUTEAL CELLS COLLECTED
FROM PREGNANT PONY MARES.................................. 289

Materials and Methods....................................................... 290
Anim al Handling............................................................ 290
Experimental Design ...................................... .......... 290
Statistical Analysis......................................................... 291
R results .......... ........................................................... ........... 291
Discussion................................................... ....................... 296








CHAPTER

XI


page

SUMMARY AND CONCLUSIONS....................................... 299

LIST OF REFERENCES......................................................... 318


APPENDICES
A. HEPARIN FORMULATIONS FOR BLOOD TUBES........... 349

B. MELATONIN FORMULATIONS FOR INJECTIONS ......... 350

C. MEDIA FORMULATIONS FOR INCUBATIONS OF FOLLICLE
WALL AND CONCEPTS SEGMENTS INCUBATIONS...... 351

BIOGRAPHICAL SKETCH........................................................................................... 353














LIST OF TABLES


TABLE PAGE


III-I. Mean LH Concentrations at Periods 1, 2, 3, 4 and 5
for all G roups..................................................................................... 69

IV-I. Mean LH Concentrations for Groups 1, 2, 3 and 4
in Frequent Samples Collected During Periods 1, 2,
3, 4 and Prior to GnRH Administration at Period 5........................ 97

V-I. Mean Number of Episodes of LH, E2 and P4 Secretion in
Samples Collected at Proestrus, Estrus and Diestrus....................... 127

VII-I. Mean E2 Concentrations in Media From Incubations of
Conceptus Segments from Control and Melatonin
Treated M ares.................................................................................... 242














LIST OF FIGURES


FIGURE Page

III-I Experimental Design Illustration.......................................................... 64

III-II Mean LH Concentrations During Period 5 in Group
1, 2, 3 and 4 M ares........................................................................... 72

III-III Mean Pre-, Post- and Maximum Plasma LH Response to
the GnRH Challenge Administered at the Conclusion of
Period 5.............................................................................................. 75

III-IV Mean Plasma LH Response of All Experimental Groups to
the GnRH Challenge at the Conclusion of Period 5........................ 77

IV-I Experimental Design Illustration......................................................... 87

IV-II Average and Maximum Follicular Size
in Group 1, 2, 3 and 4 Mares.......................................................... 92

IV-Inl LH Concentrations in Daily Blood Samples for Group
1, 2, 3 and 4 M ares........................................................................... 95

IV-IV Mean LH Response to the GnRH Challenge Administered
at Period 5......................................................................................... 101

IV-V Mean Pre-, Post- and Maximum Response to the GnRH
Challenge Administered at Period 5................................................. 103

V-I Mean Melatonin Concentrations in Control and
Melatonin Treated Mares During Proestrus..................................... 121

V-II Mean LH Concentrations in 10-Minute Interval Samples
Collected at Proestrus........................................................................ 124

V-Ill Mean LH Concentrations During the Four, One-Minute
Interval Samples Collected at Proestrus........................................... 126

V-IV Mean E2 Concentrations in 10-Minute Interval Samples
Collected at Proestrus........................................................................ 129

ix








FIGURE PAGE

V-V Mean E2 Concentrations During the Four, One-Minute
Interval Samples Collected at Proestrus........................................... 131

V-VI Mean LH Concentrations in 10-Minute Interval Samples
Collected at Estrus............................................................................. 134

V-VII Mean LH Concentrations During the Four, One-Minute
Interval Samples Collected at Estrus................................................ 136

V-Vy I Mean E2 Concentrations in 10-Minute Interval Samples
Collected at Estrus............................................................................. 138

V-IX Mean E2 Concentrations During the Four, One-Minute
Interval Samples Collected at Estrus................................................ 140

V-X Mean LH Concentrations in 10-Minute Interval Samples
Collected at D iestrus ....................................................................... 143

V-XI Mean LH Concentrations During the Four, One-Minute
Interval Samples Collected at Diestrus............................................. 145

V-XII Mean E2 Concentrations in 10-Minute Interval Samples
Collected at Diestrus......................................................................... 147

V-XIII Mean E2 Concentrations During the Four, One-Minute
Interval Samples Collected at Diestrus............................................. 149

V-XIV Mean P4 Concentrations in 10-Minute Interval Samples
Collected at Diestrus......................................................................... 151

V-XV Mean LH Concentrations at Proestrus, Estrus and
D iestrus............................................................. .. ............................... 153

V-XVI Mean E2 Concentrations at Proestrus, Estrus and
D iestrus.............................................................................................. 155

V-XVII Mean P4 Concentrations at Proestrus, Estrus and
D iestrus.............................................................. ........................... 157

V-XVIII Mean LH Concentrations During the 10 hour Sample
Collection of Mares Treated at Proestrus......................................... 160

V-XIX Mean E2 Concentrations During the 10 hour Sample
Collection of Mares Treated at Proestrus ...................................... 162

V-XX Mean LH Concentrations During the 10 hour Sample
Collection of Mares Treated at Estrus........................... ................ 165

x








FIGURE page

V-XXI Mean E2 Concentrations During the 10 hour Sample Collection of
M ares Treated at Estrus........................................................................ 167

V-XXII Mean LH Concentrations in Daily Samples Collected
in Mares Treated at Both Proestrus and Estrus................................ 169

V-XXIII Mean E2 Concentrations in Daily Samples Collected
in Mares Treated at Both Proestrus and Estrus................................ 171

V-XXIV Mean P4 Concentrations in Daily Samples Collected
in Mares Treated Both at Proestrus and Estrus................................ 174

V-XXV Mean Follicle Diameter in Control and Melatonin Treated Mares....... 177

V-XXVI Mean P4 Concentrations During Frequent Sampling at
D iestrus.............................................................................................. 169

V-XXVII Mean LH Concentrations in Daily Samples Collected From
Control and Mares Treated With Melatonin at Diestrus.................. 171

V-XXVIII Mean E2 Concentrations in Daily Samples Collected From
Control and Mares Treated with Melatonin at Diestrus.................. 173

V-XXIX Mean P4 Concentrations in Daily Samples Collected From
Control and Mares Treated with Melatonin at Diestrus.................. 176

V-XXX Mean Maximum Follicle Diameter and CL Diameter
in Control and Mares Treated with Melatonin
at Day 8, Postovulation..................................................................... 188

VI-I Mean E2 Concentrations in Pregnant Mares During
the First 15 Days of Pregnancy........................................................ 201

VI-II Mean P4 Concentrations in Pregnant Mares During
the First 15 Days of Pregnancy........................................................ 203

VI-III Mean Conceptus Diameter From Day 10 to Day 15,
Postovulation..................................................................................... 207

VI-IV Mean LH and/or eCG Concentrations Between Days 0
and 60, Postovulation......................................................................... 209

VI-V Mean P4 Concentrations in Control and Melatonin
Treated Mares During the First 60 Days of Pregnancy................... 211

VI-VI Mean Conceptus Diameter Between Day 10 and Day 20,
Postovulation in Control and Melatonin Treated
M ares .......................... ......................................... ......................... 2 14
xi







FIGURE oage

VI-VII Average and Maximum Follicle Size in Control
and Melatonin Treated Mares Following
the Onset of Daily Melatonin Injections
at Day +35, Postovulation............................................................... 217

VI-VIII Mean Date of Formation of the First Accessory
Corpus Luteum in Control and Melatonin Treated
M ares................................................................................................ ... 205

VII-I An Illustration of How Culture Plates Were Prepared
for in vitro Experiments................................................................... 231

VII-II Mean E2 Concentrations in Culture Media of
Follicle Wall Segments Incubated with 0 nM,
1 nM, 10 nM or 100 nM Melatonin in the
Absence of Supplemental Testosterone............................................ 238

VII-Ill Mean E2 Concentrations in Culture Media of
Follicle Wall Segments Incubated with 0 nM,
1 nM, 10 nM or 100 nM Melatonin with
Supplemental Testosterone................................................................ 240

VII-IV Mean E2 Concentrations in Media Collected from
Incubations of Segments of Conceptus from
Control and Melatonin Treated Mares.............................................. 244

VIII-I Mean LH Concentrations in A.M. Control and
Melatonin Treated HYX Mares........................................................ 255

ViII-II Mean P4 Concentrations in A.M. Control and
Melatonin Treated HYX Mares ........................................................ 258

VIII-III Mean LH Concentrations in P.M. Control and
Melatonin Treated HYX Mares........................................................ 260

VIII-IV Mean P4 Concentrations in P.M. Control and
Melatonin Treated HYX Mares........................................................ 262

VIII-V A Comparison of Mean LH Concentrations in Both
A.M. and P.M. Control and Melatonin Treated
H Y X M ares........................................................................................ 265

VIII-VI A Comparison of Mean P4 Concentrations in Both
A.M. and P.M. Control and Melatonin Treated
H Y X M ares........................................................................................ 267







FIGURE page

IX-I Mean Plasma and Follicle Fluid Melatonin
Concentrations in Samples Collected at Either
Noon or M idnight.............................................................................. 278

IX-II Mean Plasma LH and E2 Concentrations in
Samples Collected at Either Noon or Midnight............................... 280

IX-III Mean LH, DHEA, A and E2 Concentrations in Follicle
Fluid Collected at Either Noon or Midnight.................................... 282

IX-IV Mean P4, 17aP4 and T Concentrations in Follicle
Fluid Collected at Either Noon or Midnight.................................... 284

X-I Mean Nuclear Score of Luteal Cells From Control
Mares or Mares Which Were Treated with Melatonin
at Estrus............................................................................................. 293

X-II Representative Electron Micrographs (x 20,000)
of a Nuclear Score 1 (Oval Nucleus)
and Score 3 (highly pleomorphic Nucleus)..................................... 295

XI-I The Hypothetical Mechanism for the Differential
Effects of E2 Treatment on LH Secretion in
Anestrous M ares................................................................................ 302

XI-II The Hypothesized Existence of a Photic Gate of
LH Sensitivity to Estradiol Treatment.............................................. 305

XI-III A Diagrammatic Representation of the Effects of
Melatonin Treatment on the
Hypothalamic-Pituitary-Ovarian Axis............................................. 308

XI-IV The Postulated Effects of Melatonin on
Follicular Steroidogenesis................................................................ 311

XI-V The Postulated Effects of Melatonin on the
Luteinization Processes by Which Granulosa
Cells Become Luteal Cells............................................................... 314














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


THE ROLE OF LIGHT AND MELATONIN ON THE REPRODUCTIVE
ENDOCRINOLOGY OF ANESTRUS, CYCLIC AND EARLY PREGNANT MARES

By

Brian Douglas Cleaver

August, 1996

Chairman: Dr. Daniel C. Sharp
Major Department: Animal Science


There were two major focuses of this research: 1) the interaction of light and estradiol

(E2) on the regulation of luteinizing hormone (LH) secretion in anestrous mares; and 2) the

effects of melatonin administration on LH and ovarian steroidogenesis in cyclic, pregnant and

hysterectomized (HYX) mares.

The first study demonstrated that anestrous mares exposed to ambient daylength

(December) have minimal LH secretion. Furthermore, treatment of mares exposed to ambient

photoperiod with E2 did not affect LH secretion while treatment of anestrous mares with 2.5

hours supplemental light and E2 daily resulted in elevated LH concentrations in the peripheral

circulation and enhanced responsiveness to a gonadotropin-releasing hormone (GnRH)

challenge. Exposure to 2.5 hours of supplemental light in the absence of E2 treatment had no

effect on LH secretion or responsiveness to exogenous GnRH.








Beginning on December 21, mares were exposed to ambient daylength and were

administered E2 treatment beginning on December 21, January 4 or January 18. It was

hypothesized that all mares exposed to E2, regardless of the time at which treatment began,

would respond with enhanced LH production once daylength was adequate. Only mares which

began E2 treatment on Jan. 4 responded. LH secretion in all other

groups remained basal.

Investigations of the effects of melatonin on ovarian steroidogenesis revealed that

treatment of non-pregnant, pregnant or HYX mare with melatonin resulted in an acute decline

in E2 biosynthesis in vivo and a direct suppressive effect of melatonin on follicle wall

segments, in vitro. Treatment with melatonin resulted in the formation of a corpus luteum

which secreted 1.5 to 2 times more P4 compared to controls.

Embryos observed in mares which had been treated with melatonin tended to grow

faster than control embryos. Culture data showed that incubation of concepts segments from

embryos from mares treated with melatonin secreted higher concentrations of E2, in vitro.

Follicular fluid collected from preovulatory follicles at noon or midnight showed that

melatonin was higher at midnight. Follicular fluid concentrations of P4, 17atOH P4 and

testosterone also differed between noon and midnight. This supports the contention that

melatonin is present at the site of ovarian steroidogenesis and could be associated with

alterations in steroid hormone levels in vivo.














CHAPTER I
INTRODUCTION


Mares (Equus caballus), like other mammals in temperate climates, evolved

mechanisms to entrain annual breeding cycles to the annual changes in daylength. This

ensured that normal gestation of 340 days would result in birth when temperature and nutrition

were optimal for survival of mother and offspring. The mares annual reproductive cycle

consists of four distinct phases: the breeding season; an autumnal transition into anestrus;

anestrus; and vernal transition back into the breeding season (Sharp, 1980).

The breeding season occurs from late spring through summer. During this time, the

mare is polyestrous with cycles of approximately 21 days in length occurring until either

pregnancy occurs or decreasing daylength initiates the transition into anestrus (Hafez, 1980).

The autumnal transition into anestrus (acyclicity) usually begins during early fall (September

through October) and is characterized by the development of a preovulatory size follicle which

fails to ovulate or the persistence of an established corpus luteum (Ginther, 1992). During

anestrus, GnRH synthesis and secretion (Strauss et al., 1979b; Hart et al., 1984; Sharp and

Grubaugh, 1987) are greatly reduced, gonadotropin secretion is absent (Freedman et al., 1979)

and ovarian activity is minimal (Ginther, 1992). In the weeks following the winter solstice,

GnRH secretion begins to increase (Silvia et al., 1986) and FSH secretion recommences

(Garcia et al., 1979). The elevation in FSH reinitiates ovarian follicular activity (Sharp et al.,

1987) and signifies the onset of the vernal transition period. An average of 3.7 0.9 follicles

greater than 30 mm diameter develop and regress prior to the first ovulation of the year (Davis








2

et al., 1986). Renewed follicular estradiol synthesis (Davis and Sharp, 1986) followed by

renewed pituitary LH synthesis (Sherman et al., 1992) results in the first ovulation, and the

onset of the breeding season. Furthermore, administration of estradiol during late anestrus

resulted in elevated LH secretion (Sharp et al., 1991).

Published data indicate that at least two control mechanisms regulate LH secretion in

mares: an ovarian-independent mechanism, which most likely involves photoperiodic

regulation of hypothalamic function (Garcia et al., 1979; Sharp et al., 1988), and an ovarian-

dependent mechanism, which involves gonadal steroid feedback on both the hypothalamus and

pituitary (Sharp et al., 1991; Ginther, 1992).

Responsiveness to changes in photoperiod is commonly used as a management tool

to advance the onset of the breeding season. Administration of as little as 2.5 hours of

supplemental light beginning in December resulted in a significant advance in the time of the

onset of the breeding season (Sharp, 1980). The mares ability to respond to changes in

daylength involves the pineal gland and its major secretary product melatonin. Pinealectomy

resulted in a significant delay in the onset of the breeding season (Grubaugh et al., 1982) and

administration of melatonin at the same time as supplemental light in the winter blocked

photostimulation of advanced onset of the breeding season (Cheves and Sharp, 1985).

Furthermore, suppression of melatonin either by exposure to constant light (Cleaver et al.,

1991) or by pinealectomy (Cleaver et al., 1992) resulted in elevated GnRH and LH

concentrations during anestrus while chronic treatment with melatonin via implant during the

summer resulted in suppression of GnRH (Strauss et al., 1978) and disruption of entrainment

to photoperiod (Peltier and Sharp, 1994).

Recent information has indicated that melatonin can alter steroidogenesis directly at the

level of the gonads (Wallace et al., 1988; Brzezinski et al., 1992) independent of its effects at








3

the level of the central nervous system. Thus, melatonin, and possibly other photoperiod

mediated mechanisms, may be involved in the ovarian-dependent regulation of reproductive

function in mammals as well.

Clearly, the mechanisms of ovarian-independent and ovarian-dependent regulation of

reproductive function are not mutually exclusive in many mammals. The following studies

were carried out to investigate the interactions between ovarian-independent mediators such as

photoperiod and melatonin and ovarian-dependent mediators such as estradiol in mares.














CHAPTER II
REVIEW OF LITERATURE


Seasonal Breeding

Mammals that evolved in temperate climates experienced strong selective pressures to

give birth when temperature and food supply were optimal. Since gestation lengths vary

greatly among mammalian species, ovulation and fertilization must occur at different times of

the year. For example, the short gestation length of rodents (= 20 days)(Magnussum and

LeMaire, 1981) enables them to undergo sexual recrudescence during periods of optimal

temperature and nutrition. However, generally, as mammals increase in mass, the required

length of gestation also increases, necessitating that sexual recrudescence and onset of the

breeding season must occur at different times of the year (Bittman, 1984).

As with other seasonal breeders, mares evolved in environmental conditions that

required that offspring be born when food and temperature are optimal for survival. Large

mammals like horses have long gestation periods (= 340 days)(Hafez, 1980) thus, conception

must occur during the period from late spring to early summer. For this reason, the mare is

defined as a long day, seasonal breeder (Sharp, 1980).

Despite the differences in gestation length and timing of the breeding period, all

seasonally breeding species share a common requirement; that is, they must synchronize their

reproductive physiology with some common, reliable environmental characteristic (Farner and

Gwinner, 1980). The environmental characteristic utilized by a vast majority of temperate

mammalian species, studied to date, is daylength.








5

Daylength

By definition, a solar day on Earth is the time interval between two successive noons

(Webster's 3rd Int. dictionary, 1971). Since the Earth travels in an elliptical orbit around the

Sun, the actual length of a solar day varies slightly. However, for all practical purposes, a

solar day is considered to be 24 hours and 4 minutes in length (Webster's 3rd Int. Dictionary,

1971).

The rotation of the Earth on its axis as it orbits the Sun creates the natural day-night

cycle to which most organisms are exposed. This cycle of day and night is important for

many reasons. Almost all energy entering the biotic environment is derived from the Sun

(Saunders, 1977). Green plants utilize the Sun's light/energy to produce oxygen, water and

edible tissue. These three materials provide the basis for most food chains upon which

mammalian species rely. Therefore, mammals have evolved their biological patterns to take

advantage of conditions which favor optimal plant/energy availability.

The length of day (light) and night (darkness) varies continually throughout the year.

The rotation of the Earth, tilted on its axis, when combined with its revolution around the Sun,

is the defining pretext for the sidereal year (365 days, 6 hours, 9 minutes, 10

seconds)(Webster's 3rd Int. Dictionary, 1971). The presentation of the Earth to the Sun differs

during the sidereal year creating differing periods of day and night. These relative

"daylengths" when combined with the angle at which sunlight strikes a given place on the

Earth provides the basis for the seasons.



Photoperiodism and Biological Rhythms

Throughout their evolution, mammals have developed inherent rhythmicities the

periods of which match those of the cycles of day and night (circadian, period = 24 hours),








6

lunation (circalunar, period = 29 days), tide flow (circatidal, period = 12.4 or 24.8 hours) and

the seasons (circannual, period = one year)(Saunders, 1977). Of greatest relevance to the

present discussion are those rhythms dealing with the circadian and circannual periods.

In order for a biological rhythm to be defined as a circadian rhythm, it must meet

three main criteria. First, it must have a natural period of approximately one solar day (24

hours) in the absence of all environmental stimuli. Second, it must maintain this natural

period accurately for long periods of time under constant conditions. Third, it must be

entrainable to environmental stimuli (Weaver and Reppert, 1988).

The ability of an organism to modify its biological functions based on changing

environmental cues is defined as entrainability. In a natural 24-hour cycle of light and dark,

the circadian system of an organism becomes entrained to that of the environment. In such a

system the entrainer, light, is defined as the Zeitgeber. By achieving steady-state entrainment

to the Zeitgeber, an organism can partition its activities temporally, and thereby perform

behavioral and physiological activities at the most appropriate time of day (Underwood, 1984).

There are four basic ways by which an animal can discriminate changes in daylength:

(1) measurement of absolute duration of light, (2) measurement of absolute duration of

darkness, (3) measurement of the light:dark ratio, and (4) comparison of the incidence of light

with an endogenous rhythm of photosensitivity (Bittman, 1984).

The hour-glass hypothesis states that photoperiod discrimination is dependent on some

biochemical process that is restricted to one phase or the other (light or dark) with the end

product providing the measure of the duration of that phase (Lees, 1966). The coincidence

theory states that external photoperiod is compared to an internal rhythm of light sensitivity.

Within the total photoperiod there exists a 'light-sensitive' and a 'light-insensitive' period.








7

Light coincident with the light insensitive phase is interpreted as a short day whereas light

coincident with the light sensitive period is interpreted as a long day (BUnning, 1936).

The most commonly studied of annual cycles are those related to reproduction. In

general, responsiveness to photoperiod varies at different times of the year. For example, the

Syrian hamster undergoes gonadal regression upon exposure to short days in the autumn, but

gonadal activity resumes in the spring. Thus, in the autumn, hamsters are said to be

photosensitive, whereas in spring, they become photorefractory (Reiter, 1991). These

conditions can be reproduced in the laboratory by artificial variations in photoperiod.

Hamsters that have been raised under a long day environment undergo gonadal regression after

6-10 weeks exposure to short days (<12 hours light), regardless of the time of the year they

are transferred. However, after 20-24 weeks exposure to short days, hamsters become

photorefractory and the gonads undergo complete recrudescence (Elliot and Goldman, 1981).

To reinstate the photosensitive state, refractory animals must be exposed to long days for

approximately 10 weeks. After such exposure, hamsters are once again capable of undergoing

gonadal regression in response to short days (Stetson et al., 1977).

The resonance paradigm has been utilized to study the Golden hamster's

responsiveness to photoperiods in which light exposure was held constant (6 hours) and dark

phases of varying length. Male hamsters were exposed to photoperiods of 24 hr (6:18 L:D),

36 hr (6:30), 48 hr (6:42) and 60 hr (6:54). Photosensitive males exposed to either 6:18 or

6:42 underwent testicular regression whereas testicular regression persisted in photorefractory

males. Photosensitive males exposed to 6:30 or 6:54 interpreted these cycles as long days and

gonadal function remained intact whereas in photorefractory hamsters, these cycles terminated

the gonadal regression (Stetson et al., 1976).








8

Thus, Golden hamsters were capable of discriminating daylength in these experiments

even though the length of light exposure did not vary. Hamsters do not simply measure

absolute length of light exposure. The results also demonstrate that hamsters do not measure

absolute length of darkness, as the longest dark phase (54 hours) was interpreted as a long day.

These results provide support for the coincidence theory.

Other studies have shown that the photosensitive phase described by the coincidence

theory is restricted to the latter half of the animals' circadian day (i.e., hours 12 [noon]

through 24 [midnight]). Light present, even briefly, during this phase is interpreted as a long

day. The absence of light during this period is interpreted as a short day (Elliot, 1976).

In a short-day breeder such as the domestic ewe, a similar phenomena exists. Ewes

are reproductively competent during short days (autumn and winter) but are reproductively

quiescent anestruss) during the long days of spring and summer (Legan and Winnans, 1981).

Sheep kept indoors under artificial photoperiods varying between long days (16:8 L:D) and

short days (8:16 L:D) at 90-day intervals experienced two complete breeding cycles (two

breeding seasons, two anestrus periods) during a 365-day period (Legan and Karsch, 1983).

Application of the resonance paradigm to sheep has shown that response to

photoperiod was opposite to that of hamsters (Almeida and Lincoln, 1982). The rams in these

experiments interpreted the photoperiods in the same way hamsters did (6:18 and 6:42 short

days; 6:30 and 6:54 as long days). However, the reproductive response in sheep to short days

was one of gonadal stimulation.

Exposure to constant photoperiods for extended periods of time significantly alters

reproductive patterns in some mammals. Reproductive patterns in ewes maintained under

constant equatorial photoperiods (12:12 L:D) eventually became completely temporally

dissociated from control ewes exposed to normal seasonal changes in daylength (Jackson et al.,








9

1990). Experiments that utilized a 'symmetric skeleton' photoperiod procedure demonstrated

that gonadally regressed hamsters exposed to short days given one 250-millisecond pulse of

light 14 hours after the onset of darkness responded with recrudescence of the testis (Ellis and

Follett, 1983). Immature hamsters exposed to constant light have been shown to reach puberty

much later than those under normal photoperiods (Brackmann and Hoffman, 1977). Constant

light has also been shown to disrupt circadian behavioral activity severely in rodents

(Chesworth et al., 1987). Finally, exposure to constant darkness resulted in 'free run' in

nocturnal activity in rodents (Cassone et al., 1986) and eventual reproductive asynchrony when

compared with animals under normal photoperiods (Reiter, 1980).



Photoperiodism in Mares

Physiologically, the annual breeding cycle of the mare is divided into four phases:

anestrus, transition into the breeding season (vernal transition), the breeding season and

transition back to anestrus (Sharp, 1980). Anestrus is the period of reproductive quiescence

and is characterized by sexual disinterest and ovarian atrophy. The annual distribution of

ovulation in both horse and pony mares has been shown to be highly correlated with seasonal

changes in photoperiod. The occurrence of ovulations in horse mares (Van Niekerk and Van

Heerden, 1972) and pony mares (Ginther, 1974) in both the Northern and Southern

hemispheres is highest during the long days of late spring and early summer and lowest during

the short days of winter. If the criteria of less than 25% of mares ovulating is used to define

anestrus, horse mares are anestrus from December through February and pony mares from

November through March (Sharp, 1980).

The vernal transition period most probably begins at the winter solstice when

daylength begins to increase. However, ovarian changes characteristic of this period do not








10

become evident until some weeks later. During this period, ovarian follicular development

recommences; however, an average of 3.7 0.9 preovulatory like follicles (_> 30 mm) develop

prior to the first ovulation of the year (Davis et al., 1987). These transitional follicles have

been shown to be steroidogenically incompetent as they do not produce significant amounts of

either androgens or estrogens (Davis and Sharp, 1991). It is theorized that vernal transition

ends and ovulation occurs when a follicle develops the ability to complete the steroidogenic

pathway to estradiol. Once the first ovulation of the year occurs, the breeding season has

begun.

During the breeding season, the mare is polyestrous with cycles of approximately 21

days in length (Ginther, 1979) occurring until either pregnancy occurs or decreasing daylength

initiates the transition into anestrus. During these approximate 21-day cycles, ovulation leads

to formation of a corpus luteum (CL) which secretes progesterone which, among other

functions, supports a possible pregnancy (Hafez, 1980). If pregnancy does not occur, or if

maternal recognition of pregnancy fails, the CL regresses (McDowell, 1986).

Little is understood about the transition into anestrus. After the last ovulation of the

year, a preovulatory type follicle develops but fails to ovulate (Snyder et al., 1978). During

this last episode of follicular development, LH secretion fails to increase to the levels observed

during normal breeding cycles. The mechanisms which prevent the normal preovulatory

increase in LH are unknown, but most likely represent the basis of anestrus physiology.

Burkhardt (1947) first described the manipulation of the breeding cycle of mares using

changes in daylength. This study demonstrated that artificially lengthening the day at a rate

twice that of the ambient increase in photoperiod beginning on January 1, advanced the onset

of the breeding season dramatically. These results have since been confirmed by many others

(Sharp, 1980).








11

How changes in daylength influence the mare are not clear. The existence of a

window of photosensitivity, as described by the coincidence theory, has been investigated in

the mare. Sharp (1980) reasoned that daylength varies seasonally only at the time of sunrise

and sunset and not during other times of the day. Accordingly, they reasoned that if such a

window of photosensitivity exists, it would be near either dusk or dawn.

To test this hypothesis, mares were supplemented with an extra 2.5 hours of light after

dusk, prior to dawn, or both. Mares that received supplemental light in the morning entered

the breeding season at the same time as control mares exposed to natural photoperiods. Those

mares supplemented with light after dusk, or at both dusk and dawn entered the breeding

season significantly earlier than normal mares. Such results suggest that a window of

photosensitivity may exist around the time of dusk (Sharp, 1980).

However, such results are controversial. Palmer and Driancourt (1981) also

investigated the existence of a window of photosensitivity. In a series of night-interruption

experiments, supplemental lighting was given to anestrous mares to simulate daylengths of

from 8 hours of light to 20 hours of light. From their results, Palmer and Driancourt (1981)

reported that light present 9.5 to 10.5 hours after dusk (morning) stimulated an early onset of

the breeding season. Malinowski and co-workers (1985) came to the same conclusion in their

experiments in which mares given a photoperiod of 10L:8D:2L:4D entered the breeding season

earlier than mares under natural photoperiods. Thus, according to this group, a window of

photosensitivity exists closer to dawn than dusk.

Furthermore, the occurrence of a photosensitive phase in the early morning hours, as

suggested by Palmer and Driancourt (1981), would mandate the development of a

photosensitive phase that would never be exposed to light under natural conditions. As was

previously stated, daylength changes at both dusk and dawn under natural conditions (Sharp,








12

1980). Thus, the Palmer experiments were biased toward detecting only AM photosensitivity.

Additionally, it is likely that interrupting night with light resets the mare's endogenous timing

system. Thus, the existence of a photosensitive period at dusk remains a possibility.

The success of constant daily photoperiods demonstrated that the natural, gradual

increase in photoperiod is not required to initiate the onset of the breeding season.

Interestingly, exposure of mares to constant light has been shown to be less effective in

advancing the onset of the breeding season than exposure to 16:8 L:D (Kooistra and Ginther,

1975). It should be recognized, however, that the mare does not simply need to enter the

breeding season per se, but to enter the breeding season at the appropriate time to optimize

survival of the foal. Under such conditions, the gradual lengthening of photoperiod in the

spring may be critical in timing the first ovulation of the year.

The role of changing daylength in the termination of the breeding season has also been

investigated. Results demonstrate that mares induced to an early onset of the breeding season

by exposure to stimulatory photoperiods cease to ovulate (enter anestrus) at the same time as

normal mares (Kooistra and Ginther, 1975). Thus, termination of the ovulatory system appears

to be a function of decreasing daylength. When mares were exposed to a stimulatory

photoperiod (16:8 L:D) continuously through the breeding season, they entered anestrus

significantly later than normal mares. Such mares also entered the next breeding season earlier

than normal mares (Ginther, 1979). These results are in line with what is known about Golden

hamsters and ewes both of which eventually become refractive to stimulatory photoperiods

(Bittman, 1984).

From the present information in mammals, including mares, the following was

inferred: First, mammals have biological functions which oscillate with rhythms which are

circadian (or about 24 hours in length), circannual (or about one year in length) or both.








13

Second, these biological functions are entrained, or fine-tuned, by daily and seasonal changes

in daylength. Finally, seasonal breeders, both short day breeders and long day breeders, utilize

these changes to determine when they should reproduce. In the following sections, the

mechanisms by which the cycles of rodents, ruminants, and specifically mares, are controlled

will be illustrated to demonstrate how circannual organization may occur.



The Anatomical Photoperiodic Pathway

The mechanisms by which changes in daylength reach the central nervous system have

been investigated in a number of mammalian species. These investigations provide a fairly

complete understanding of the anatomical pathway involved in photoreception.

Unlike many nonmammalian vertebrates, mammals lack extra-retinal photoreceptive

capability and must rely entirely on the eye. Bilateral enucleation leads to a dramatic

regression of the reproductive organs in Golden hamsters (Reiter, 1968) and in rats (Wurtman

et al., 1964). Light striking the photoreceptive rod cells of the retina causes neural

transduction signals to fire toward the retinal ganglia (Takahashi et al., 1984). The retinal

ganglia project their neural tracts to the optic chiasma. Bilateral sectioning of the optic nerves

or the ablation of the optic chiasma of rats and hamsters have the same effect as enucleation

(Moore, 1978a).

Research has identified a direct, independent projection from the retina to the

suprachiasmatic nuclei (SCN). Elimination of visual sight by transaction of the primary and

accessory optic tracts leaving this retino-hypothalamic tract (RHT) intact maintained the ability

of both rats and Golden hamsters to respond to changes in daylength (Moore and Lenn, 1972).

Attempts to section the RHT selectively without destroying the SCN have proven difficult.








14

For this reason, experiments investigating this portion of the anatomical pathway have centered

on the SCN.

There have been no critical studies of the effects of unilateral or bilateral enucleation

in mares with regard to the annual breeding rhythm. However, the existence of a RHT in

mares has been demonstrated. Sharp and co-workers (1984) found that horseradish peroxidase

injected into the aqueous humor of the eye was found in the contralateral SCN. These results

suggest the existence of an retinohypothalamic tract in this species as found in others (Moore,

1978b).

Lesions of the SCN eliminated the ability of golden hamsters to discern changes in

photoperiod in terms of reproductive response and eliminated the animals ability to sustain

circadian behavioral rhythms (Rusak, 1977). Work with ewes has shown that lesions of the

SCN disrupt normal seasonal breeding patterns (Przkop and Domanski, 1980). Entrainment of

the rat circadian system by changes in photoperiod has also been shown to depend on an intact

SCN (Cassone et al., 1986).

Efferent projections from the SCN have been shown to terminate directly in the

paraventricular nuclei (PVN) of rats (Swanson and Sawenko, 1983) and the hamster (Don

Carlos and Finkelstein, 1982). Electrolytic lesions of the PVN in hamsters blocked testicular

regression upon exposure to short days (Bittman et al., 1983).

Projections from the PVN to the medial forebrain bundle (MFB) have been identified

(Swanson, 1977) and lesions of the projections or the MFB blocked pineal melatonin synthesis

in rats (Moore and Klein, 1974). Lesions to the MFB area have also been shown to block

short-photoperiod-induced gonadal regression in golden hamsters (Morin et al., 1987).

Neural projections from the MFB to the dorsal spinal cord and preganglionic neurons,

which terminate at the superior cervical ganglia (SCG), have been described (Don Carlos and








15

Finkelstein, 1982). The synapse into the SCG represents a departure of the anatomical

photoperiodic pathway from the CNS into the autonomic nervous system (Cardinali et al.,

1981). The SCG provide sympathetic innervation to the pineal gland, the cephalic blood

vessels, the choroid plexus and the eye (Cardinali et al., 1981). Removal of the SCG or

transaction of the preganglionic trunks prevents nocturnal increases in pineal melatonin

secretion (Moore, 1978a). Additionally, bilateral superior cervical ganglionectomy (SCGx)

eliminates the ability of rats and hamsters (Reiter and Hester, 1966) and sheep (Lincoln and

Almeida, 1982) to respond to changes in photoperiod.

There have been no investigations in equids of any of the anatomical sites between the

SCN and the superior cervical ganglia (SCG) shown above to be involved in the transfer of

photic input in other species. However, involvement of the SCG in horses is well established.

Sharp and co-workers (1979) demonstrated that bilateral SCGx altered seasonal breeding

patterns. Mares which were SCGx entered the first post-surgery breeding season at the same

time as normal mares. However, SCGx mares entered the second post-surgery breeding season

significantly later (> 2 months) than normal, control mares.

Postganglionic fibers from the SCG innervate the pineal gland directly via the internal

carotid nerve and the nervii conarii (Ariens-Kappers, 1960). These postganglionic sympathetic

axons terminate in the perivascular spaces which surround the pinealocyte (Moore, 1978b).

While definitive studies of the innervation of the equine pineal have not been carried out, the

empirical data presented above suggest that the SCG innervate the equine pineal via the

internal carotid nerve and the nervii conarii as well.








16

The Structure of the Pineal Gland

The pineal gland was first described more than 2300 years ago by the anatomist

Herophilus. Herophilus thought that the pineal was a valve through which the flow of

memories was controlled. The Greek anatomist Galen considered the pineal to be a gland

which was involved in the control of fluid flow in the brain (Preslock, 1984). The renowned

French philosopher Rene Descartes has provided probably the most 'enlightened' view of the

pineal by declaring it to be the seat of the soul (Kitay and Altschule, 1954).

Physicians of the 17th and 18th centuries linked the gland to various psychiatric

diseases, but Gibson was the first to advance the idea that the pineal is an organ, which, he

said controlled the separation of lymph from blood (Reichlin, 1981). Ogle in England (1899)

and Heubner in Germany (1899) independently recognized a link between pineal tumors and

precocious puberty in children. However, after this time, little work was done in this area

until, in 1959, Lemer and co-workers isolated melatonin from bovine pineal glands. Almost

all information now known about pineal anatomy, physiology and biochemistry has been

described since Lemer's work was published.

The pineal is a reddish-grey structure that occupies the depression between the superior

colliculi at the posterior border of the corpus callosum and the third ventricle in most non-

rodent mammals (Preslock, 1984). The rodent pineal is attached to the undersurface of the

confluence of the venous sinuses overlying the brain (Reiter, 1980). Non-rodent species

exhibit a small ependymal lined recess of the third ventricle that extends into a short stalk

connecting the pineal to the diencephalic roof (Preslock, 1984). The major blood supply is

provided by branches of the posterior choroidal arteries and innervation arises from the

superior cervical ganglia (Arendt, 1988).








17

The pineals from different mammals are heterogeneous in shape and size. The pineal

of man is piriform in shape, weighing from 50 to 150 mg, whereas that of the rat weighs only

1 mg or less (Arendt, 1988). There is some evidence for an increase in pineal size relative to

body weight with increasing distance from the equator (i.e., increasing need for

seasonality)(Krstic, 1986).

Birds and reptiles have pineals that appear to have mixed photoreceptive and secretary

function. However, the pineal in mammals is thought to be entirely secretary in nature.

Typically, the mammalian pineal consists of pinealocytes, which are essentially modified

photoreceptive cells (Pevat, 1981).

The pineal is composed of two cell types: pinealocytes and neurological 'astrocyte-

like' cells. The pinealocytes are polyhedral cells with elongated dendrite-like processes which

intermingle with the processes of neighboring cells. These processes often terminate on the

endothelium of capillaries (Wolfe, 1965). Pinealocyte nuclei are irregular and slightly

heterochromatic. They exhibit well developed endoplasmic reticulum and Golgi bodies with

dense, membrane-bound vesicles prominent in the perivascular expansions (Preslock, 1984).

The neurological cells resemble astrocytes and appear to play a supportive role to the

pinealocytes (Bloom and Fawcett, 1975).

The presence of 'synaptic ribbons' in the pinealocyte of the rat have been described

(Vollrath and Howe, 1976). The relative number of ribbons during light is low but increases

4-fold during darkness. Vollrath and Howe also described 'synaptic spherules', which may

vary in a manner similar to the ribbons. The presence of granulated vesicles in pinealocytes of

the Golden (Syrian) hamster has also been reported (Krasovich and Benson, 1982). Such

vesicles exhibit no circadian rhythm under long photoperiods (14:10 L:D), but exhibit marked

circadian rhythmicity under short photoperiods (10:14 L:D), with the number of vesicles in the








18

dark phase being significantly less. These cellular structures and their changes with

photoperiod have been shown to be closely correlated with melatonin-forming activity (Rudeen

et al., 1975).

Surgical removal of the pineal gland (pinealectomy, PNX) has profound effects on the

reproductive activity of seasonally breeding mammals. Early work in rats demonstrated that

PNX increased ovarian weight in females and seminal vesicle and prostatic weight in males

and induced premature vaginal opening in prepuberal rats (Motta et al., 1967). However,

hamsters (Syrian, Turkish and Djungarian) have been the most intensively studied of all

mammalian species.

Exposure of Syrian hamsters (Mesocricetus auratus) to short (inhibitory) photoperiod

initiated profound and immediate gonadal regression (Reiter, 1974). PNX completely blocked

short photoperiod-induced regression when PNX occurred prior to short photoperiod exposure

(Reiter, 1974; 1980). PNX after short-day induced gonadal regression has begun leads to

restoration of reproductive function (Matt and Stetson, 1980).

In Turkish hamsters, (Mesocricetus brandti), also long-day breeders, PNX induced

gonadal regression in long photoperiods and prevented regression during exposure to short

photoperiods (Carter et al., 1982). Additionally, PNX Syrian hamsters remained

reproductively active under both long and short photoperiods whereas PNX Turkish hamsters

underwent involution of the gonads regardless of photoperiod (Goldman, 1984).

Djungarian hamsters (Phodopus sungorus) exhibited a responsive pattern intermediate

between Syrian and Turkish hamsters. As with Syrian hamsters, PNX in Djungarian hamsters

during long photoperiod blocked the influence of short photoperiod on reproduction (i.e.

gonadal regression). In hamsters, testicular regression induced by exposure to short

photoperiod was rapidly reversed upon transfer back to long photoperiods. This effect was








19

blocked by PNX just prior to transfer from short to long photoperiod (Hoffman, 1981b). PNX

in male Djungarian hamsters caused recrudescence 9 to 12 weeks after regression regardless of

photoperiod whereas pineal-intact males needed only 6 weeks to achieve recrudescence when

transferred from short to long photoperiod (Goldman et al., 1982).

Other rodent species require an intact pineal in order to respond to changes in

photoperiod. Removal of the pineal in voles (Charlton et al., 1976) and White-footed mice

(Goldman, 1984) eliminated short-photoperiod-induced testicular regression. Non-rodent

species also require an intact pineal. Pinealectomy (PNX) in ferrets, members of the carnivore

family, blocked the ability of long photoperiod to accelerate onset of seasonal estrus (Herbert,

1981). PNX also influenced delayed embryonic implantation events in the Western Spotted

Skunk (Meade, 1972) and the Tammer's wallaby (Renfree et al., 1982).

Early studies involving PNX in sheep were apparently without effect, leading to the

erroneous conclusion that sheep were unique. Roche and co-workers (1970) pinealectomized

ewes without apparent effect on the timing of the breeding season. Later experiments

conducted using both rams (Kennaway et al., 1981) and ewes (Matthews et al., 1981) agreed,

leading to the conclusion that the pineal was not essential for the interpretation of

environmental seasonal changes in this species. However, these early experiments were

carried out under natural photoperiodic conditions and PNX subjects had exposure to pineal

intact animals.

Natural seasonal cycles can be disrupted by exposure to artificial (i.e. short

photoperiod exposure during summer) photoperiods (Sykes and Cole, 1944) and removal of

the pineal eliminated such effects (Bittman et al., 1983). In sheep kept indoors under artificial

photoperiod, it has been shown that alternation between artificial long days (16:8 light:dark)

and artificial short days (8:16 light dark) at 90-day intervals produced two breeding seasons








20

and two anestrus seasons in a single year (Legan and Karsch, 1983). Ovarian-intact ewes

PNX in early spring did not respond to the 90-day shifts as pineal-intact ewes did (Bittman et

al., 1983). These results provided evidence for an endogenous annual reproductive cycle that

relies on the pineal to interpret changes in daylength.

The equine pineal is a reddish-grey organ that weighs between 50 and 100 mg

(Cleaver, B.D. et al., unpublished observations). As in the ewe, the equine pineal occupies the

depression between the superior colliculi at the posterior border of the corpus callosum and the

third ventricle (Preslock, 1984) and the major blood supply is through the posterior choroidal

arteries (Kappers, 1960).

As in other species, PNX has been shown to have profound effects on reproduction in

equids. Grubaugh and co-workers (1982) were the first (and to date, only) group to

successfully PNX mares. Their studies examined the effect of PNX on seasonal reproductive

patterns and the ability of PNX mares to respond to photostimulation. As with

ganglionectomy, PNX mares enter the breeding season at the same time as intact mares under

ambient photoperiod the first post-surgical year, but entered the second year significantly later

than intact mares. Furthermore, PNX mares exposed to supplemental lighting (2.5 hours after

dusk) did not undergo any advance in the date of first ovulation.

Why PNX and SCGx did not delay the first, post-surgery breeding season is not

known. One explanation is that the annual breeding cycle is controlled by endogenous timing

systems that are in turn, entrained to changes in photoperiod. The ability of both PNX and

SCGx mares to enter and exit post-surgical breeding seasons would support this. Such a

timing system would be sufficient to maintain cycle integrity with intact mares the first year

thereafter, the cycle would "free run" with no entrainment possible. However, the definitive

conclusions await the identification and characterization of such endogenous timing systems.








21

The one conclusion which can be drawn from this work is that an intact photoperiodic-neural

pathway is essential to the ability of the mare to respond to changes in photoperiod.

Taken together, the effects of pineal removal on both long-and-short day breeders

convincingly demonstrate the critical role of this gland in photoperiodism. The mechanisms

by which the pineal transduces changes in photoperiod to influence changes in reproductive

state have been shown to be endocrine in nature. One hormone which is now widely accepted

as an endocrine link between the pineal and its target tissues is melatonin.



Melatonin: Synthesis and Secretion

Melatonin is one of the major secretary products of the pineal gland and was first

characterized by Lerner and co-workers (1959) from bovine pineal glands. The hormone

derives its name from the role it plays in amphibian melanocyte aggregation during darkness

(Bagnara, 1960). The association of melatonin with the light-dark cycles of many species of

vertebrates indicates that it may serve as a neuroendocrinee transducer' or photic information

from the external environment to the tissues of the body (Wurtman et al., 1968; Axelrod,

1974; Gem, 1982).

Melatonin is synthesized and secreted by the pineal primarily during periods of

darkness in all vertebrate species studied. This rhythm of synthesis is accomplished by

pinealocytes following the uptake of tryptophan (Cassone, 1990). Tryptophan enters the brain

through a neutral amino acid transport system, but the method by which it enters the

pinealocyte has not been widely investigated (Sugden, 1979). The concentration of tryptophan

in the pineal is high but, unlike brain levels, is not correlated to serum tryptophan levels which

are known to vary with feeding (Sugden, 1979; Young and Anderson, 1982).








22

Serotonin (5-hydroxytryptamine)(5-HT) is synthesized from tryptophan by the action

of two enzymes. The first, tryptophan hydroxylase, transfers a hydroxyl group to the 5-

position of the indole ring to yield 5-hydroxytryptophan (5-HTP). The second enzyme,

aromatic amino acid decarboxylase, removes the side-chain carboxyl group to form serotonin

(Young and Anderson, 1982). Tryptophan hydroxylase activity in the pineal is extremely high

and the pineal is the source used to purify this enzyme for laboratory use (Nukiwa et al.,

1974). Several groups have independently detected a nocturnal elevation in this enzyme's

activity with night-time activity approximately two-fold higher than daytime activity (Shibuya

et al., 1978; Sitaram and Lees, 1978). Stimulation of tryptophan hydroxylase activity by

norepinephrine (NE) in rat pineals has been demonstrated establishing a link to adrenergic

control via the superior cervical ganglia (Shein et al., 1971).

As stated previously, the product of enzymatic conversion of tryptophan is serotonin.

The concentration of 5-HT in the pineal gland is higher than in any other body tissue

(Wiechmann, 1986). While evidence now exists that 5-HT of pineal origin may play a

significant role as a secretary product on its own (Garrick et al., 1983), most 5-HT serves as a

precursor for melatonin synthesis (Reiter, 1991). The enzyme responsible for the conversion

of 5-HT to N-acetylserotonin is serotonin-N-acetyltransferase (SNAT).

Almost all information known regarding SNAT activity is due to studies on the rat

pineal (Buda and Klein, 1978). During the day, in vivo and in vitro, SNAT activity in the

pineal is very low. Activity is markedly increased by NE release into the pineal perivascular

space at night from sympathetic nerve endings terminating in the gland. NE acts on the

pinealocyte membrane to stimulate B-adrenoreceptors (Auerbach et al., 1981). Radioligand

binding studies have also identified B-adrenoreceptors in the sheep (Foldes et al., 1983) and the

hamster (Craft et al., 1985). B-adrenoreceptor stimulation activates the adenylate cyclase-








23

cAMP second messenger system. In rats, B-stimulation leads to increased cAMP

concentrations to levels 60-fold higher than baseline. The elevation in cAMP is short-lived,

however, as concentrations of this second messenger return to baseline concentrations within

10 minutes (Klein et al., 1978).

Current evidence suggests that cAMP elevation by NE involves the a-adrenoreceptor

sub-groups as well. The first evidence that a-adrenoreceptors were present in the pineal came

from studies illustrating that NE initiated an increase in phosphotidylinositol (PI) turnover as is

the case in other tissues (Smith, 1979). These a-adrenoreceptors have been since identified in

the rat (Sugden and Klein, 1984), and the sheep (Sugden et al., 1985). Activation of rat

pinealocyte aX-adrenoreceptors does not, in itself, increase cAMP, SNAT and melatonin

synthesis. However, B-adrenergic stimulation of cAMP and SNAT is greatly potentiated by

simultaneous activation of the cx,-adrenoreceptor (Vanecek et al., 1985). Furthermore, the

mechanism underlying a1-amplification has been shown to involve a Ca2-activated protein

kinase C (Sugden et al., 1985).

N-acetylserotonin is converted to melatonin by the enzyme hydroxy-indole-O-methyl

transferase (HIOMT)(Buda and Klein, 1978). Evidence suggests that HIOMT concentrations

in the pineal remain constituitively high throughout the day lending support to the suggestion

that the availability of SNAT is the rate-limiting step in the biosynthesis of melatonin (Reiter,

1991).

The existence of specific mechanisms governing the release of melatonin into

circulation is unclear. Due to its high lipophilic nature, melatonin readily exits from the

pinealocyte. For this reason, it has been hypothesized that melatonin secretion is an automatic

consequence of its synthesis (Reiter, 1991). Some preliminary data suggest that blockade of








24

calcium channels, within the pinealocytes, may impede melatonin secretion (Morton et al.,

1989) but the potential role calcium may play in the secretion of melatonin is untested.

In all species studied to date, pineal secretion is higher at night than during the day.

As melatonin is thought to diffuse directly from the pinealocyte into adjacent capillaries,

changes in blood melatonin levels are thought to be a direct result of changes in melatonin

synthesis (Reiter, 1986). The pineal is the dominant source of the nocturnal increase in

circulating melatonin. Evidence lies in the fact that pinealectomy or bilateral superior cervical

ganglionectomy prevented the nighttime rise in plasma melatonin (Kennaway et al., 1977;

Arendt et al., 1980; Lewy et al., 1980; Vaughn and Reiter, 1986).

There is evidence that melatonin is released in a pulsatile manner from the pineal

(Reiter et al., 1987). While the most evident episodes of melatonin secretion came from

samples taken from the veins adjacent the pineal (Colombo et al., 1987), melatonin pulses

could still be identified in jugular blood samples (Cozzi et al., 1988). Any role pulsatile

melatonin secretion may play in its effects on target tissues is unknown.

The bulk of circulating melatonin is bound to albumin (Pardridge and Mietus, 1980).

Melatonin's half-life in the blood is short (= 40 minutes)(Kopin et al., 1961) and upwards of

90% of circulating melatonin is cleared by the liver during a single pass (Pardridge and

Mietus, 1980). Such data provide strong support for possible importance of episodic or

pulsatile patterns of melatonin secretion and its effects on target tissues. However, putative

target tissues for melatonin await identification.

Biochemical studies of the synthesis of melatonin in equids have been limited.

HIOMT is the enzyme responsible for the conversion of N-acetylserotonin into melatonin.

Wesson and co-workers (1977) investigated the correlation between "melatonin forming

activity" (which they describe as pineal HIOMT content) based on radioligand incorporation








25

and the percentage of mares ovulating throughout the year. They found that the melatonin

forming activity (and HIOMT activity) of pineals collected during the months of November,

December and January were much greater than during other months of the year. Furthermore,

the percentage of mares ovulating during this time period were minimal. The authors

concluded that the amount of HIOMT activity within the pineal was highest during winter.

However, studies which have been done in other species have found that N-

acetyltransferase (NAT) activity and not HIOMT activity displays circadian and circannual

variations (Reiter, 1991). Therefore, in the absence of definitive evidence, it is likely that

Wesson and co-workers were not measuring changes in HIOMT activity, but NAT activity.

Melatonin secretion in the mare has been shown to occur in a nyctohemeral pattern

with mean daytime secretion from 0 to 30 pg/ml and nighttime secretion from 100 to 400

pg/ml (Cleaver, B.D. and Grubaugh, W.R. unpublished observations). The pineal gland is the

major source of circulating melatonin in equids as PNX significantly reduced plasma melatonin

and eliminates the nocturnal increase observed in pineal intact mares (Grubaugh et al., 1982).

Pineal sympathectomy by SCGx also eliminates the nocturnal increase in melatonin secretion

but overall melatonin secretion tends to be higher than in PNX mares (Sharp, D.C. et al.,

1979).

Melatonin secretion in equids has also been shown to vary with changes in daylength.

A comparison of average melatonin secretion by month for a 12 month period clearly shows

that as daylength becomes shorter (winter), mean melatonin levels measured increase.

Conversely, when daylength is long (summer), mean melatonin is lower (Sharp et al.,

unpublished results).

Exposure of mares to constant darkness has been shown to severely disrupt the

nyctohemeral pattern of melatonin secretion observed under light:dark progressions. Melatonin








26
secretion initially appeared to "free run", similar to other photoperiodic mammals, in the days

immediately after entering constant darkness. However, secretion was completely disrupted

(no clear patterns) by the end of the experimental period indicating apparent absence or failure

of endogenous rhythms in melatonin secretion in this species (Kilmer et al., 1982).

Melatonin secretion in mares has been shown to increase rapidly after sunset. Sharp

and Grubaugh (1983) have also detected the presence of episodic events in melatonin secretion

in pony mares. The possible function of such episodes is unclear at this time and awaits

further study. It should be pointed out that there is minimal support for theories concerning

the importance of episodic melatonin secretion in relaying the melatonin message to target

tissues in other photoperiodic mammals (Reiter, 1991).



Melatonin Receptors

Most evidence suggests that melatonin acts through receptors located in the brain

(Glass and Lynch, 1981; Hastings et al., 1988). The identification of high affinity melatonin

receptors has relied on three specific areas of biochemical activity: the inhibition of dopamine

release from the rabbit retina, the inhibition of cAMP production in hypothalamic and

hypophyseal structures and in vitro autoradiographic localization utilizing 2-[125l]iodomelatonin.

Each of these areas will be discussed in detail below.

The understanding that melatonin is a very potent inhibitor of dopamine release from

the rabbit retina in vitro led to the first pharmacological characterization of a functional

receptor for melatonin (Dubocovich, 1983). In both rabbit and chicken retinas, melatonin

selectively inhibited the calcium-dependent release of [3H]-dopamine with pharmacological and

functional characteristics of a specific melatonin receptor (Dubocovich, 1988). Melatonin's

effects were concentration dependent with an IC50 value of 40 pM reported for the rabbit retina








27

(Dubocovich, 1985). Serotonin, a-adrenergic dopamine and opiate receptor antagonists, all of

which are known to influence dopamine release, had no affect on melatonin's inhibition of

dopamine release (Krause and Dubocovich, 1991). This eliminated the possibility that

melatonin was working through either opiate, a-adrenergic or serotonin receptors.

The potency with which melatonin inhibited [3H]-dopamine release from the retina

depended on extracellular concentration of calcium (Dubocovich and Hensler, 1985).

Furthermore, any factor which decreases neurotransmitter release (i.e reduction in extracellular

Ca2 concentration or depolarization) significantly enhanced the ability of melatonin to inhibit

the release process. This suggests that melatonin may act to decrease the availability of

calcium (Dubocovich, 1990).

The involvement of cyclic AMP (cAMP) in a biologic effect of melatonin was first

suggested twenty years ago in a study of amphibian skin (Abe et al., 1969). Melatonin not

only inhibited melanocyte-stimulating hormone's (MSH) ability to darken skin pigmentation, it

also inhibited the rise in cAMP. Melatonin has also been shown to inhibit cAMP production

stimulated by forskolin in the pars tuberalis of Djungarian hamsters (Carlson et al., 1989),

sheep (Morgan et al., 1989), neonatal rats (Vanachek and Vollrath, 1989) and White-footed

mice (Weaver et al., 1990).

Melatonin has no affect on basal cAMP levels or phosphodiesterase activity suggesting

that decreased cAMP may be due to inhibition of adenylate cyclase activity (Carlson et al.,

1989). However, studies using other cellular preparations indicate that melatonin's actions on

cAMP may be indirect (Morgan et al., 1989; Benitez-King et al., 1990). Evidence exists that

the high-affinity melatonin receptor can also inhibit cyclic GMP (cGMP) production with

potency equal to that of cAMP inhibition (Vanachek and Vollrath, 1989). The effects of

melatonin on both cAMP and cGMP was blocked by pretreatment with the pertussis toxin








28

suggesting that a G-protein may be involved in melatonin receptor function (Vanachek and

Vollrath, 1989).

Both dopamine and cyclic nucleotide inhibition assay systems have done much to

enhance our understanding of the melatonin receptor. However, the discovery that melatonin

could be iodinated at the 2-position has proven to be the most useful tool to date in studying

the location and function of melatonin receptors.

Due to rapid metabolism of radioligands in vivo, in vitro test systems were favored

(Morgan and Williams, 1989). 2-[125I]iodomelatonin was shown in 1987 to be a useful

compound for membrane homogenate binding studies of melatonin receptors (Dubocovich and

Takahashi, 1987) as well as autoradiography (Vanecek et al., 1987). The results from both

studies correlated well with known functional data from the rabbit retina studies (Dubocovich,

1983).

Binding sites for 2-[125l]iodomelatonin with affinity in the picomolar range have been

characterized in a variety of mammalian tissues. These include the pars tuberalis/median

eminence region of sheep (Morgan et al., 1989), rats (Vanecek et al., 1987) and hamsters

(Carlson et al., 1989); the SCN of rats (Laitinen and Saavedra, 1990) and humans (Reppert et

al., 1988); the thalamic PVN of mice (Fang et al., 1990); and in whole brain homogenates in

chickens (Siuciak et al., 1990).

Specific binding sites have been characterized by saturation analysis of both membrane

homogenates and autoradiographic samples. The binding has been shown to be reversible and

saturable indicating a single class of binding site (Dubocovich and Takahashi, 1987).

However, recent studies have suggested that melatonin receptors may exist in two affinity

states (Carlson et al., 1989). Study in this area has advanced the concept that a high affinity








29

binding site (defined as ML-1) and lower affinity binding site (defined as ML-2) exist

(Dubocovich, 1990). ML-1 receptors are thought to be involved in all known melatonin-

induced effects at known target tissues (i.e. dopamine release inhibition, cAMP inhibition, etc.)

whereas ML-2 receptors have no known function at the present time (Krause and Dubocovich,

1991).

ML-1 linkage to a G-protein has been investigated for possible involvement in

melatonin-receptor function. Micromolar concentrations of guanine nucleotides specifically

inhibited 2-[125I]iodomelatonin binding in a dose-dependent manner in a number of tissues

(Dubocovich et al., 1990). GTP and GDP were found to be equally effective in inhibiting

binding whereas ATP, ADP and their analogues had no effect on binding. Evidence appears

to indicate the existence of an inhibitory (G,) type complex (Carlson et al., 1989). G-proteins

are known to be involved in many of the intracellular processes which are also affected by

melatonin. For these reasons, a melatonin receptor G protein relationship is highly likely.

Investigation of the target tissues for melatonin in equids is extremely new. In fact,

the author has only identified one investigation (Stankov et al., 1991) in this area. The results

of this investigation will be the basis of this discussion.

Melatonin binding was characterized in the horse brain by using an in vitro binding

technique with 2-[125I]iodomelatonin as the labelled ligand. Utilizing crude membrane

preparations, binding was shown to be rapid, stable, saturable, reversible, of high affinity (KI =

10 pM) and low capacity (Bm between I and 20 fmol/mg protein). As in other mammalian

species, the pars tuberalis of the horse exhibited the highest specific binding of all tissue areas

measured. Specific binding was also detected in the median eminence, the SCN, the preoptic

area (POA), lateral hypothalamus, olfactory bulb, frontal cortex, occipital cortex, lateral

geniculate nucleus, dentate nucleus and the area postrema.








30

The importance of these initial results can not be understated. The existence of

specific binding sites in the hypothalamic areas known to be associated with biological time-

keeping (the SCN) and reproduction (POA, median eminence and pars tuberalis) in other

mammalian species provides evidence for how melatonin functions in circadian and circannual

events in Equids.

In summary, high affinity melatonin receptors have been identified, characterized and

localized to specific areas of the hypothalamus and pituitary gland. The presence of such

receptors in discrete areas known also to be involved in reproductive activity may prove to be

very important.



The Reproductive Hormones



Ovarian Steroid Synthesis in Mammals

The fundamental structures of ovarian steroid biosynthesis are the preovulatory follicle

and the corpus luteum. Structurally, the preovulatory follicle in mares is composed of a layer

of fibroblast cells thecaa extema), a vascularized layer of steroidogenic cells thecaa intema), an

acellular basement membrane, multiple layers of granulosa cells and a fluid filled antrum

(Baird, 1984). The granulosa cells occur in two general subtypes: the mural granulosa, which

are steroidogenic, and the cumulus granulosa cells which support and surround the oocyte

(Byskov, 1979). The basement membrane isolates the granulosa cells and the oocyte from

direct vascular support, necessitating that all metabolic precursors, nutrients and

endocrine/paracrine regulatory factors must first pass through the theca intema cell layer

(Baird, 1984).








31

The corpus luteum is made up of fibroblast cells, and large and small luteal cells of

other mammalian species are derived from both the granulosa and the theca intema cells

(Niswender and Nett, 1994). The large cells range in size from 20 pm to 40 pm in diameter

(Enders, 1973) and often appear polyhedral with light staining cytoplasm and large, centrally

located nucleus. The small luteal cells, have a diameter of less than 22 pm and appear

spindle-shaped with darkly staining cytoplasm, large lipid droplets and an irregularly shaped

nucleus with cytoplasmic inclusions (Enders, 1973).

Ovarian steroid hormones, both of follicular and luteal origin, are biochemically

classified as progestins, androgens or estrogens. The basic chemical structure of each is

composed of a ring complex of three cyclohexane rings and a cyclopentane ring referred to as

the perhydrocyclopentano-phenanthrene nucleus, or more simply, as the steroid nucleus (Kellie,

1984). All steroid hormones are related to or derivatives of four basic parent compounds,

cholestrane, pregnane, androstane and estrane (Gore-Langton and Armstrong, 1988).

The biochemical synthesis of the steroids begins with cholesterol derived from either:

active uptake from blood in the form of lipoproteins; cholesterol liberated from the plasma

membrane, cholesterol esters or lipid droplets stored within the cell; or cholesterol synthesized

de novo within the ovarian cells (Strauss et al., 1981). Cholesterol is converted to

pregnenolone by the cleavage of the C-20,22 bond by the enzyme P450-side chain cleavage

(P450 SCC)(Miller, 1988). This enzyme complex is localized on the matrix side of the inner

mitochondrial membrane (Farkash et al., 1986).

Pregnenolone is converted by the enzyme complex A5-3B-hydroxysteroid

dehydrogenase:A"-isomerase (3B-HSD) into progesterone (Miller, 1988). The 3B-HSD is

localized on smooth endoplasmic reticulum (a.k.a. microsomes)(Hall, 1984). Pregnenolone can

also be retained by the P450 SCC complex and converted via and unstable intermediate to








32

dehydroepiandrosterone (DHEA) by P450-17a-hydroxylase:C17.20-lyase enzyme complex

(Miller, 1988). DHEA is converted to androstenedione by the enzyme complex 3B-HSD:A54-

isomerase (Miller, 1988). Selective steroid metabolism from pregnenolone to androstenedione

via DHEA is defined as the A5 pathway. Progesterone conversion to 17a-hydoxyprogesterone

(17aP4) then on to androstenedione is catalyzed by the P450-17ca-hydroxylase:lyase enzyme

complex as well (Miller, 1988). Metabolism of pregnenolone to androstenedione by this

pathway is referred to as the A4 steroid pathway.

Androstenedione is converted either to testosterone by the 17-ketosteroid reductase

enzyme or to estrone by the P450-aromatase enzyme (Miller, 1988). Testosterone is converted

into estradiol by the P450-aromatase enzyme and estrone and estradiol are believed to rapidly

interchange under influence of the enzyme 17B-hydroxysteroid dehydrogenase (Miller, 1988).

In ovarian preovulatory follicles, steroid biosynthesis/metabolism is carried out as

described by the "two cell" theory. This theory indicates that under influence of LH, theca

intema cells will synthesize and secrete the androgens, androstenedione and testosterone (Gore-

Langton and Armstrong, 1988). The mural granulosa cells, under the influence of LH,

synthesize and secrete progesterone for use by the theca intema cells for androgen biosynthesis

and will take up androgens produced by the theca internal cells to synthesize and secrete

estradiol (Gore-Langton and Armstrong, 1988).

The primary steroid secretary product of the corpus luteum is progesterone (Niswender

and Nett, 1994). Once again, two cell types are present in the mature CL: the small luteal

cell, and the large luteal cell. The large cell appears to be the predominant progesterone

synthesizing cell in ruminants while the small luteal cell appears to be a "resting" sub-type that

can undergo conversion to a large luteal cell under the influence of LH (Niswender and Nett,








33

1994). Indeed, most LH receptors (>80 %) are found on small luteal cells (Niswender et al.,

1985).

LH has been shown to be stimulatory to progesterone production in small luteal cells

(Niswender and Nett, 1994). The intracellular process by which LH regulated progesterone

synthesis has been shown to involve receptor activation of protein kinase A (PKA) via G

proteins and cAMP (Anhar and Menon, 1975). Protein kinases have been demonstrated to

influence nuclear events, transcriptional events, gene expression and protein synthesis including

steroidogenic enzymes and cholesterol binding proteins (Jungmann and Hunzicker-Dunn,

1978). The model indicates that PKA stimulates synthesis of steroidogenic enzymes; activates

cholesterol esterase, mitochondrial cholesterol transport regulatory mechanisms and LDL

uptake from circulation; and steroid transport from mitochondria to the smooth endoplasmic

reticulum (Niswender and Nett, 1994).



Steroid Binding Globulins

Steroid hormones are not stored, but are secreted as they are synthesized indicating

that increased secretion is indicative of increased synthesis. Following secretion into

circulation, steroid hormones are bound to transport glycoproteins which are made in the liver

(Gill, 1987). These plasma binding proteins provide a reservoir of hormone, protected from

metabolism and renal clearance. Thus, the mechanisms which regulate levels of binding

globulins ultimately regulates circulating concentrations of steroid hormones.



Ovarian Steroid Biosynthesis and Secretion in Mares

The preovulatory follicle in mares is composed of a layer of fibroblast cells thecaa

extema), a vascularized layer of steroidogenic cells thecaa intema), an acellular basement








34

membrane, multiple layers of granulosa cells and a fluid filled antrum (Baird, 1984). The

granulosa cells occur in two general subtypes: the mural granulosa, which are steroidogenic,

and the cumulus granulosa cells which support and surround the oocyte (Byskov, 1979). The

basement membrane isolates the granulosa cells and the oocyte from direct vascular support,

necessitating passage of all non-granulosa synthesized metabolic precursors, nutrients and

endocrine/paracrine regulatory factors through the theca intema cell layer (Baird, 1984).

Steroidogenesis in the preovulatory equine follicle occurs in the following manner.

Mural granulosa cells in mares have been shown to contain receptors for FSH (Fay and

Douglas, 1987), 38-hydroxysteroid dehydrogenase (converts pregnenolone to

progesterone)(Hay et al., 1975) and P450 aromatase (converts testosterone to estradiol)(Tucker

et al., 1991). The theca intema cells appear to be the primary site of androgen synthesis.

Both 17a-hydroxylase:C17-20-lyase (conversion of P4 to 17oaOH P4 then to androstenedione)

and the 178-hydroxysteroid dehydrogenase (conversion of androstenedione to testosterone)

enzyme complexes have been localized in theca intema cells (Miller, 1988; Sirois et al., 1991).

Thus, the "two-cell" hypothesis is mares indicates that mural granulosa cells synthesize P4,

which diffuses into theca intema cells and is converted into androgen precursor (testosterone).

Testosterone then diffuses back across the basement membrane and into the mural granulosa

cells where it is converted to E2 (Sirois et al., 1991).

The primary secretary product of the equine preovulatory follicle is estradiol.

Estradiol concentrations in circulation begin to increase 6 to 8 days prior to ovulation, reach

maximum concentrations in circulation approximately 2 days prior to ovulation and declined

by the time of ovulation (Ginther, 1992). Concentrations of estradiol decline to low baseline

levels within 2 days after ovulation (Ginther, 1992). The primary functions of estradiol in

mares are discussed in later in relevant sections.








35
Unlike other mammalian species, ovulation in mares occurs only at the ovulation fossa.

As a result, the entire structure of the corpus luteum (CL) is contained within the stroma of the

ovary (Mossman and Duke, 1973). This is in contrast to other species such as the cow or the

gilt where ovulation occurs outward at the site of follicle development and the luteal gland

itself often extends prominently off the surface of the ovary (Niswender and Nett, 1994).

When mature, the equine CL is flesh-colored with a trabeculated appearance owing to the

collapse of the large preovulatory follicle. A central cavity is often present and the CL often

extends into the ovulation fossa itself (Ginther, 1992).

The following information is based on histological analysis reported by Van Niekerk

and co-workers (1975) and Levine and co-workers (1979). The secretary cells of the equine

CL are derived exclusively from granulosa cells. Thecal cells can be observed to degenerate

just prior to ovulation and are not visible within 24 hours post-ovulation. Granulosa cells are

approximately 10 pm at ovulation and enlarge to around 15 pm within 24 hours after ovulation

and contain a vesiculated nucleus and cytoplasmic vacuoles. The luteinization process appears

to be complete by day 3, postovulation, but luteal cells continue to hypertrophy until around

day 9, postovulation, when they achieve an average diameter of around 37.5 pm. Maximum

progesterone, the major secretary product of the equine CL, secretary activity usually occurs

by this time. In addition to the large, light-staining luteal cells, approximately 15 % of all

luteal cells are small cells which are thought to represent a resting stage that can undergo

conversion to large cells. LH is required for maintenance of luteal function following

ovulation in mares. Treatment with antisera against LH (Pineda et al., 1973), or administration

of either hCG or equine pituitary extracts in vivo, or to luteal cell cultures in vitro, increased

progesterone concentrations and/or extended the lifespan of the CL (Kelly et al., 1988).








36

Furthermore, receptors for LH have been identified on equine luteal cells (Roser et al.,

1982). However, little has been reported regarding the intracellular effects of LH binding to

its receptor on luteal cells in mares. The functions of progesterone in the mare are discussed

later in relation to specific tissue sites of action.



Gonadotropin Synthesis and Secretion in Mammals

Mammalian gonadotropins are members of a family of glycoprotein hormones that

share many structural similarities. The family also includes the pituitary hormone TSH and

the placentally derived chorionic gonadotropins. Structurally, these hormones consist of a

common a subunit covalently bound to a unique B subunit. Individually, the subunits have no

known biological activity, thus, the formation of a heterodimer (a-Bx) is essential for activity

(Gharib et al., 1990).

The a subunit, within a species, has an identical amino acid sequence in all three

members of the hormone family (LH, FSH, and TSH). The B subunits share considerable

amino acid homology with one another, indicating that most likely they evolved from a

common precursor. There is also a certain amount of structural and amino acid sequence

homology between a and 8 subunits, suggesting that they too share a common ancestral gene

(Stewart and Stewart, 1977).

Both subunits are glycosylated at specific residues and contain multiple intramolecular

cross-links. The subunits achieve their tertiary structures by the formation of internal disulfide

bonds (5 in the a and 6 in the 1). The location of cysteine residues is highly conserved

among the hormones as well. When the 8 subunits are aligned by their 12 cysteine residues,

regions of amino acid similarity and variability become apparent. The similar regions are








37

believed to be involved in a subunit binding and variable regions involved in receptor-

hormone recognition (Pierce and Parsons, 1981).

The gonadotropin subunits are encoded by separate genes localized on different

chromosomes (Naylor et al., 1983). The structure of the genes for LHB and FSHB are very

similar but the a gene is organized somewhat differently. The ca gene in humans (Fiddes and

Goodman, 1981), cows (Goodwin et al., 1983), mice (Chin et al., 1981) and rats (Burnside et

al., 1988) has been characterized. In each species, a single gene encodes the a subunit and is

composed of 4 exons and 3 introns. The positions of the introns are highly conserved in all

species. There is considerable variation in subunit size among species (8 to 16.5 kbases). An

a mRNA of 730 to 800 nucleotides is produced in all species. There is a 24 amino acid

leader peptide followed by a 96 amino acid mature protein in all species studied to date (Chin

et al., 1983).

The human (Talmadge et al., 1983), rat (Jameson et al., 1984) and bovine (Maurer,

1989) genes encoding the LHB subunit are much smaller (= 1.5 kB) than those encoding the a

subunit. There are 3 exons and 2 introns which encode a mature protein 121 amino acids in

length (Fiddes and Goodman, 1980). The rat (Gharib et al., 1989), bovine (Kim et al., 1988)

and human (Watkins et al., 1987) genes encoding the FSH8 subunit are highly conserved

among species. Human FSHB genes differ from the bovine and the rat in that 4 mRNA

species are transcribed (Jameson et al., 1988) whereas the other two species transcribe a single

1.7 kB mRNA.



Effects of GnRH on the Gonadotropins

The frequency and amplitude of GnRH pulses can affect the amounts of gonadotropins

secreted. Current information on the role of GnRH on LH and FSH biosynthesis is








38
conflicting. Some studies have shown that GnRH induces changes in the glycosylation of LH

but does not increase the biosynthesis of the protein itself (Lui et al., 1976). Other studies

showed that actinomycin D (a transcriptional blocker) inhibited GnRH-induced LH synthesis

and release. This suggests that mRNA synthesis is required for the induction of LH synthesis

by GnRH (Lui and Jackson, 1978). If GnRH is added after cultured pituitary cells have been

exposed to estradiol, an increase in LH synthesis occurs (Ramey et al., 1987). These results

suggest that GnRH added alone may down-regulate its own receptors and estradiol is required

for induction of new receptor synthesis/recycling.

Further studies utilized pituitary stalk transected in vivo models to study the effects of

GnRH on gonadotropin subunit synthesis. Within one month of transaction in ewes,

concentrations of a and LHB mRNAs were undetectable (Hammemik and Nett, 1987).

However, if GnRH was supplemented in a pulsatile fashion, mRNA levels returned to normal

concentrations (Hammemik and Nett, 1987). The pattern of GnRH pulse replacement has been

shown to be highly important to subunit concentration response. The GnRH pulse amplitudes

which produced the highest levels of a and LHB mRNAs were 75 and 25 ng/pulse,

respectively. It was also found that GnRH administered at 7.8 to 8 minute intervals increased

a, had little effect on LHB and had no effect at all on FSHB mRNA levels. Thirty minute

pulse intervals increased the levels of all three gonadotropin subunit mRNA levels while 120

minute pulses increased only FSHB mRNA (Haisenleder et al., 1988).

Thus, there is evidence that GnRH regulates gonadotropin subunit biosynthesis in vivo

systems, however, the data from in vitro studies are conflicting. Failure to detect changes in

subunit levels may be due to experimental artifact, not the absence of physiological response.

Studies on the stimulation of LH secretion in response to GnRH initially implicated

cAMP as the second messenger (Adams et al., 1979). Subsequent work has proven this to be








39
incorrect. Conn and co-workers (1979) conclusively demonstrated that LH release in vitro can

be accomplished when the cAMP system was blocked. Current understanding shows that

calcium functions as second messenger in acute releases of LH in response to GnRH. The

elimination or chelation of extracellular calcium inhibited GnRH-induced LH secretion

(Hopkins and Walker, 1978). When the eliminated calcium was reintroduced, normal LH

secretion resumed (Bates and Conn, 1984).

Additional evidence implicating calcium's role in this process was demonstrated by the

fact that ionophores of calcium or liposomes loaded with calcium administered to

gonadotropes resulted in LH secretion at levels comparable with those seen after treatment

with GnRH (Marian and Conn, 1979). GnRH has also been shown to stimulate an increase in

intracellular calcium (Hopkins and Walker, 1978) and calcium channel agonists (i.e.

maitotoxin) stimulated LH release while channel blocking agents (i.e. verapamil) blocked

stimulated LH release (Conn et al., 1983; 1987).

Such evidence, when combined, strongly implicated calcium as the second messenger

for the GnRH-receptor system. However, final proof comes from the discovery that the

calmodulin kinase and phosphotidylinositol systems, systems known to involve calcium, are

also involved in GnRH-induced LH secretion.

GnRH and GnRH agonists have been shown to cause an increase in the turnover of

polyphosphoinositides (Andrews and Conn, 1986) which results in a stimulation in intracellular

calcium mobilization. Calcium and diacylgycerol (DAG), which is also derived from the

phosphoinositide system, are known activators of protein kinase C (PKC). Gonadotropes

release increased amounts of LH in response to treatment with synthetic DAG (Conn et al.,

1985) or PKC activators such as phorbol esters (Harris et al., 1985). Some conflicting

evidence shows that PKC-depleted cells maintain their ability to release LH in response to








40

GnRH binding (Lewis et al., 1989). Therefore, while GnRH binding to its gonadotrope

receptor does lead to an increase in intracellular concentrations of known PKC activators (i.e.

calcium and DAG), PKC itself does not appear to be involved in GnRH-induced LH release.

An alternative calcium sensitive system is the calmodulin system. Calmodulin is a

known activator of many regulatory enzymes and cytoskeletal proteins (Means and Dedman,

1980; Chalfouleas et al., 1982). The administration of GnRH has been shown to cause

redistribution of calmodulin in rats and has also been shown to associate directly with GnRH

occupied receptors (Conn et al., 1981). The administration of calmodulin inhibitors (i.e.

pimozide) can inhibit the GnRH-induced LH release as well (Conn et al., 1984). These data

provide strong evidence for a role of calmodulin in this system.

The best described mechanism linking GnRH and LH secretion is that of the LH pulse

generator. This is an undefined neuroendocrine mechanism which gives rise to the pulsatile

mode of LH secretion in rodents (Kalra and Kalra, 1983), primates (Knobil, 1989) and

ruminants (Karsch et al., 1984). The ewe will be used to describe this system.

Ewes are short day seasonal breeders with the breeding season occurring in the Fall

and Winter and an anestrus period during late Spring and Summer (Karsch et al., 1980).

During the estrous cycle, the frequency of LH pulses is modulated primarily by progesterone

which is believed to act at the level of the brain to prolong the interval between bouts of

GnRH release. In contrast, the amplitude of LH pulses is limited by estradiol which is

believed to act primarily at the pituitary gland to diminish response to each pulse of GnRH.

The system is believed to function in the following manner.

During the luteal phase elevated progesterone inhibits pulse generator function creating

low GnRH frequency. Additionally, basal estradiol secretion limits the amplitude of LH pulses

by acting at the level of the gonadotrope to limit responsiveness to GnRH and may also








41

decrease LH synthesis (Goodman and Karsch., 1983). The resulting pattern of low-frequency

LH pulses does not provide a sufficient stimulus to support the final stages of follicular

maturation. This, in turn, prevents the large increase in estradiol necessary to initiate the

preovulatory LH surge.

Luteolysis results in a precipitous fall in circulating progesterone, thus eliminating the

suppression of the pulse generator. As a result, GnRH and LH pulse frequency begin to

accelerate. Presumably, without progesterone to oppose its action, estradiol enhances pulse

frequency even further. However, the presence of estradiol also maintains the low-amplitude

of the LH pulses. These resulting low-amplitude, high-frequency LH pulses initiate an

increase in circulating LH which supports final follicular maturation. This results in the

preovulatory rise in estradiol and later, the ovulatory surge of LH.

Following ovulation, progesterone levels increase again as the corpus luteum becomes

established. This returns the system to a status where LH and GnRH pulse frequency becomes

low. The accompanying decline in circulating estradiol levels initiates a return to a higher

pulse amplitude characteristic of diestrus (Goodman and Karsch, 1980a).

Results from experiments in which rats were hypophysectomized and the pituitary

transplanted into the renal capsule failed to establish any strict correlation between LH, FSH

and GnRH secretion. However, basal FSH secretion did rise after treatment with GnRH and

pretreatment immunoneutralization of GnRH eliminated this rise (Strobl and Levine, 1988).

The removal of feedback signals from the gonads via, gonadectomy, results in

increased gonadotropin secretion. In rats, FSH increases rapidly in both sexes, whereas LH

rises rapidly in males, but only after a 4- to 7 day lag in the female (Spitzbarth et al., 1988).

Replacement of the sex-appropriate steroid can prevent the post-gonadectomy LH increase.

Testosterone administration also restores FSH to precastration levels in male rats (Summerville








42
and Schwartz, 1981) and monkeys (Plant et al., 1978), and the addition of follicular fluid

suppresses FSH secretion even further (Summerville and Schwartz, 1981). OVX females

respond to estrogen administration with a decline in FSH secretion, but intact levels can be

obtained only by the addition of follicular fluid (Cambell and Schwartz, 1979). Pulse

amplitude and frequency of LH in OVX rats remained suppressed for several days after

removal of estrogen implants, but FSH pulsatility in estrogen treated animals was

indistinguishable from untreated (no estrogen) animals (Luderer and Schwartz, 1989).

Gonadotropin subunit mRNA levels increased gradually following gonadectomy. In

female rats, LHB levels increased for 7 days after OVX to reach levels which were 10 to 20

fold higher than intact rats. a subunit mRNA levels also increased following OVX and FSHB

mRNA were higher by 3 days post-OVX, reaching levels 3 to 10 fold higher than intact rats

by 21 days (Gharib et al., 1987).

Estrogen replacement to OVX rats suppressed all three subunits (a, LHB and FSHB)

mRNA levels within 4 hours of treatment, reducing them to precastration levels within 24

hours (Shupnik et al., 1988). Testosterone or DHT treatment for 7 days suppressed LHB and

a mRNA levels to intact levels in intact male rats, but had no effect on FSHB mRNA levels

(Gharib et al., 1988). However, in pituitary cell cultures, androgen treatment increased FSHB

message and did not alter either a or LH8 mRNA levels (Gharib et al., 1990).

Positive steroidal feedback from the gonads has long been known to be crucial in the

development of the preovulatory gonadotropin surge (Mahesh, 1985). The primary surge is

triggered by high estradiol levels on diestrus and proestrus. Estrogen given to OVX rats

during afternoon LH surges for 3 to 5 days (Legan et al., 1975) and progesterone given at

noon to estrogen primed OVX rats advanced the time of the GnRH (and LH) surge (Caligaris








43
et al., 1971). However, progesterone administration at noon also abolished the daily, afternoon

LH surges observed in estrogen treated rats (Freeman et al., 1976).

Recently, a vast amount of information regarding the role of ovarian peptides, growth

factors and interactions with the immune system on the control of FSH secretion has surfaced.

While beyond the scope of this review, these factors have been reviewed extensively in recent

years (Gore-Langton and Armstrong, 1988). The isolation and characterization of these factors

have greatly increased our understanding of how these gonadotropins are regulated.

GnRH synthesis and secretion in mares from the hypothalamus is obligatory for the

synthesis and secretion of LH (Robinson et al., 1994; Porter et al., 1995). As in other

mammals, GnRH is a decapeptide and has been immunolocalized in nerve terminals in the

area of the median eminence near the portal vessels and in adjacent hypothalamic areas (Dees

et al., 1981). Positive staining neurons were also found in the area of the SCN and arcuate

nuclei as well as in the periventricular nuclei (Dees et al., 1981).

The manner of GnRH is secretion is controversial in equids. Studies utilizing push-

pull perfusion technique showed that GnRH secretion was minimal during anestrus, increased

during vernal transition and was maximum during the breeding season. However, there were

no significant differences between periods of the annual cycle with regards to either amplitude

or frequency of secretary episodes, and the occurrence of episodes was not consistent with

patterns observed in mammals with a "pulse generator" (Sharp and Grubaugh, 1987).

In other studies, the cavernous sinus which drains the pituitary was catheterized and

direct venous drainage was collected. Using this technique, Alexander and Irvine (1987)

described a close correlation between pulsatile releases of GnRH and pulsatile releases of LH

and FSH. Additionally, > 90 % of the detected GnRH pulses elicited releases of LH and FSH.

Unfortunately, this study can only be described as observational as most sampling days were








44

represented by only one mare (no basis for comparison) and sampling intervals were limited (4

to 6 hours).

Another study using this technique detected episodic releases of GnRH every 12 to 24

hours with each episode lasting from 2 to 4 hours (Pantke et al., 1991). These results are

more in line with the push-pull perfusion data discussed earlier. Thus, when compared to the

current dogma accepted for other mammals, it is appears unlikely that a pulse generator system

similar to that found in rats and sheep exists in equids.

Secretion of GnRH in the horse has been shown to vary with season. Strauss and

co-workers (1979) were the first to demonstrate that both total hypothalamic content as well as

distribution within the hypothalamus varied seasonally. Total GnRH content was highest

during the summer (breeding season), lowest during the winter anestruss), and intermediate

during the spring (vernal transition). The areas within the hypothalamus which showed highest

activity corresponded to the median eminence, the arcuate nuclei, the preoptic area and the

SCN.

Hart and co-workers (1984) also showed that GnRH content within the POA,

hypothalamic body and median eminence were highest during the breeding season and lowest

during anestrus. Additionally, this study found that GnRH receptor content of pituitaries from

different seasons did not vary. From this information, it could be surmised that during

anestrus, reduced GnRH synthesis and not decreased pituitary sensitivity to GnRH is the basis

for anestrus.

As with other mammalian species, the equine gonadotropins consist of structurally

identical at subunits and unique 8 subunits. Equine LH has a molecular weight of = 33,500

(Sherwood and McShan, 1977) and contains a high proportion of carbohydrate (27 %) and a








45

large compliment of sialic acid residues (6 %). Equine FSH has a molecular weight of around

33,200, is about 24.2 % carbohydrate and 6.8 % sialic acid (Sherwood and McShan, 1977).

Gonadotropin levels in the plasma change seasonally in mares (Sharp, 1980). In

winter when ovarian activity is low, circulating levels of LH are low while during the breeding

season, LH varies with stage of cycle. Relatively low LH levels (1.0 to 1.5 ng/ml) are found

during mid-diestrus; levels increase toward the end of diestrus and the beginning of estrus (1.5

to 2.0 ng/ml); and increases to maximal levels at or near the time of ovulation (maximum -

5.0 to 6.0 ng/ml). A distinctive feature of the ovulatory LH surge in equids is its long

duration. While this event occurs over a course of hours in rats and ewes, in horses and

ponies it lasts as much as 4 to 5 days (Whitmore et al., 1973).

The annual secretion of FSH has been shown to decline in the winter (Garcia and

Ginther, 1976) or to not vary at all throughout the year (Turner et al., 1979). During anestrus,

FSH secretion into the circulation is low (= 10.0 ng/ml) while during the breeding season,

plasma levels vary with stage of cycle. Maximum FSH secretion during the estrous cycle

occurs in two waves. The first occurs following ovulation and the second during late diestrus

(Miller et al., 1977). Such increases in secretion are thought to be related to the induction of

follicular growth (first wave) and the selection and maturation of the preovulatory follicle

(second wave). Plasma levels begin to decline near the onset of estrus (Ginther, 1979).

FSH secretion during the vernal transition period has also been investigated in more

detail. Hines and co-workers (1991) examined FSH pulsatility during the vernal transition

period. This study found that the decline in FSH secretion as vernal transition progressed was

due to a decline in FSH pulse amplitude. They theorized that such a decline could be due to

the increasing presence of gonadal peptides (inhibin) and steroids as the first ovulation

approached.








46

Much of the research into the secretion of the gonadotropins has utilized exogenous

GnRH. Treatment of mares with exogenous GnRH results in increased LH and FSH secretion

in the mare (Ginther and Wentworth, 1974) while active immunization of mares to GnRH

produced a state similar to anestrus (Garza et al., 1986). Conversely, LH and FSH response to

GnRH have been shown to vary with season. The release of LH in response to GnRH during

transition was not significant until about 21 days prior to the breeding season (Silvia et al.,

1986; Davis et al., 1987). However, FSH levels in response to GnRH declined as vernal

transition progressed toward the breeding season (Silvia et al., 1987).

A single injection of GnRH to anestrus mares induced an increased in serum FSH but

not LH (Evans and Irvine, 1976) and three injections per day were able to cause ovulation in

anestrus mares (Minoia and Mastronardi, 1987). Supplementation with a pulsatile pattern of

exogenous GnRH (50 to 250 pg/hour) to anestrus mares increased LH secretion and hastened

the onset of the breeding season (Johnson, 1986). However, other studies have shown that

constant release implants were also an effective means of hastening the onset of the breeding

season in anestrus mares (Hyland et al., 1987; Allen et al., 1987).

The manner in which GnRH affects the secretion of the gonadotropins in equids has

been the subject of a great deal of investigation. As with GnRH secretion, controversy exists

as to the pattern in which these two hormones are released.

The episodic nature of gonadotropin release in equids has been studied for some time.

Evans and co-workers (1979) described what they called secretaryy bursts" in the periovulatory

period. However, the criteria for determination of a 'peak' were not given in the study; the

number of mares utilized in the study was low (n=3); and the reported frequency of occurrence

in each mare was highly variable.








47

In a series of papers by Fitzgerald and co-workers, the changes in LH pulsatility in the

mare throughout the year were addressed. The first study demonstrated that LH pulses could

be detected only when LH levels were low anestruss) or were increasing (vernal transition).

These workers concluded that LH pulses were no longer detectable during periods when LH

concentrations are high (the breeding season) (Fitzgerald et al., 1983). A second study

examined the presence of LH pulses in post-partum mares. In this case, LH pulses could be

detected only during the luteal phase (1 to 3/hour), not in the pre-ovulatory state (Fitzgerald et

al., 1985). A third study investigated LH secretion during the vernal transition period in mares

which were exposed to an artificial, stimulatory (long day) photoperiod. This study showed

that exposure to light led to increased LH pulse frequency (0 to 1 pulse/12 hours to 2.5 to 3.0

pulses/12 hours) and decreased LH pulse amplitude (1.51 ng/ml to 0.56 ng/ml)(Fitzgerald et

al., 1987).

Thompson and co-workers (1987) compiled what is probably the most comprehensive

study of gonadotropin pulsatility in this species. In this study, mares which were: 1) ovarian

intact, anestrus in winter; 2) ovarian intact, diestrus in summer, 3) ovarian intact, estrus in

summer, and 4) ovariectomized (OVX) in summer were examined. Relatively high frequency

peaks of short duration were observed only in OVX mares. Low frequency peaks of long

duration (- 8 hours) were observed in both OVX and intact mares during both summer and

winter (5 vs 3 in summer and 4 vs 0 in winter, OVX vs ovarian intact, respectively). Estrous

mares exhibited high, variable LH concentrations with low stable FSH concentrations. OVX

mares, on average, exhibited more LH and FSH peaks/12 hour period than intact mares.

Discrepancies observed in this type of analysis are not unusual. Thompson and co-

workers (1987) found that duplicate assay of each hormone was necessary to measure actual

secretary events efficiently. As much as half the events observed in one assay disappeared








48

when the results of 2 or more assays were combined. For this reason, as many as 50 % of the

secretary episodes described in studies in which results from one assay were analyzed may be

incorrect.

The Seasonality of Gonadotropin Secretion in Mares

One of the unique features of the annual reproductive cycle of mares is that, unlike

ewes, seasonal changes in LH and FSH (and presumably GnRH) are independent of the ovary

and its secretions. Mean levels of both gonadotropins in OVX mares increase to maximal

levels during summer and decline to minimal levels during winter (Garcia and Ginther, 1976).

Thus, in the mare, seasonal onset of the breeding season appears to be dependent on changing

photoperiod, or on an endogenous timing mechanism entrained to photoperiod.

The influence of ovarian steroids on GnRH and gonadotropin secretion has been

investigated in equids. Early studies demonstrated that co-administration of estradiol and

GnRH resulted in increased LH secretion compared with GnRH alone (Garcia and Ginther,

1975). Other studies demonstrated that estradiol benzoate injection reduced the normal

diestrus increase in FSH secretion (Thompson et al., 1983). Estradiol treatment by itself has

been shown to decrease LH secretion while co-treatment with GnRH resulted in prolonged

elevation of LH release (Vivrette and Irvine, 1979). From these results, it appears that the

endogenous preovulatory increase in circulating estradiol may be important to the induction of

the extended preovulatory increase in LH.

Testosterone has also been shown to affect FSH secretion. Thompson and co-workers

(1983) found that administration of either testosterone propionate or dihydrotestosterone

resulted in enhanced FSH secretion following exogenous GnRH administration. Immunization

of mares to androgens resulted in elevated LH secretion during estrus but did not alter FSH

secretion. The immunization against androgens did not alter estrogen secretion in this study so








49

any effects on LH were not mediated through estrogens (Thompson et al., 1987). From these

results they concluded that during estrus, androgens may also be involved in the regulation of

LH secretion.

During the vernal transition period, treatment with estradiol resulted in elevated

pituitary LH content and peripheral secretion. However, administration of progesterone, or

progesterone with estradiol had no effect and decreased LH secretion, respectively.

Furthermore, when estradiol-treated mares were administered exogenous GnRH, the magnitude

of the secretary response was no larger than that observed in control mares. Thus, increased

sensitivity to GnRH (i.e. increased receptor number) can not account for the observed response

to estradiol. However, none of the treatments had any effect on FSH secretion (Sharp et al.,

1991).



Early Pregnancy in Mares

Gestation in mares is, on average, 340 days (Hafez, 1980). However, the relevance to

the present subject relates only to the events surrounding the first 60 days of pregnancy.

Following fertilization, the equine concepts enters the uterus on day 5-6 (Oguri and Tsutsumi,

1975). Unlike other mammalian embryos, the equine concepts remains spherical and

migrates throughout the uterine lumen (Ginther, 1983). Maternal recognition of pregnancy

(MRP), which occurs between day 14 and 16 (Sharp et al., 1989), requires this uterine

migration (Leith and Ginther, 1985). If MRP is successful, the CL will be maintained, P4

secretion will be maintained, and pregnancy will continue.

LH is believed to be the primary luteotrophic hormone in mares during early

pregnancy with regards to maintaining P4 secretion. Active immunization of mares against

gonadotropins induced luteal regression in diestrous mares (Pineda et al., 1973). The








50

maintenance of P4 secretion is essential to the maintenance of pregnancy. Many of the

proteins secreted into the uterine lumen require P4 for their synthesis and secretion. That P4

is obligatory was shown by Hinricks and co-workers (1985) in which P4 administration to

OVX mares was sufficient to maintain pregnancy. Total uterine luminal protein concentrations

increase from ovulation through day 14, postovulation in both cyclic and pregnant mares (Zavy

et al., 1979; McDowell et al., 1990). Following day 14, protein content within the uterus

continues to increase in pregnant mares whereas non-pregnant mares protein content drops

dramatically following luteal regression and the decline in P4 secretion (Sharp, 1994).

Progesterone also has been shown to enhance the growth and development of the

concepts. Administration of exogenous P4 to pony mares resulted in conceptuses of greater

diameter and mobility (Weithenaur et al., 1990). The concepts continues to migrate until

around day 16 to 17 when size prevents further movement (Ginther, 1983). Upon fixation,

concepts membrane expands to fill the uterine lumen. The capsule ruptures around day 22

(Enders and Lui, 1991) and two types of trophoblast become recognizable: the placental

trophoblast and the chorionic girdle (Allen et al., 1973).

The chorionic girdle reaches maturity by day 35, postovulation, and begins to invade

the surrounding uterine epithelium on around day 36 to 37 (Allen et al., 1973). Invading

trophoblast cells are surrounded by a wall of lymphocytes and develop into mature endometrial

cup cells (Allen et al., 1973). The major secretary product of the endometrial cups, equine

chorionic gonadotropin (eCG), first appears in the maternal circulation around day 40,

postovulation (Cole et al., 1931). Beginning around day 40, postovulation, accessory corpora

lutea begin to form (Amoroso et al., 1948) and plasma P4 concentrations begin to increase

(Urwin and Allen, 1982). Due to its inherent LH-like activity in mares (Bousfield et al.,








51

1987), eCGs primary function appears to be the support of the primary CL and the formation

and support of the accessory CLs (Allen et al., 1973).

The predomination of eCG over LH in circulation was estimated to begin at around

day 36, postovulation, after which concentrations of eCG in circulation increased rapidly

throughout the collection interval (Allen, 1969).



The Influence of Melatonin on Reproduction

Photoperiodic control of reproduction is largely a result of alterations in gonadotropin

secretion. Both long and short day breeding mammals respond to inhibitory photoperiods with

decreased FSH and LH secretion and synthesis (Davis and Meyer, 1972; Turek et al., 1976;

Lincoln et al., 1982). The effects on gonadotropin secretion may be mediated by changes in

GnRH secretion from the hypothalamus (Lincoln and Short, 1980) or through an alteration in

hypothalamic-pituitary responsiveness to the feedback effects of steroids (Tamarkin et al.,

1976; Legan et al., 1977).

The first report of an effect of melatonin on the reproductive system was by Wurtman

and co-workers (1963). This study reported that daily injections of low doses of melatonin

into maturing female rats retarded ovarian growth. Subsequent studies using adult male rats

demonstrated that melatonin injections decreased testicular weight, prevented spermatogenesis

and decreased serum testosterone if rats had first been rendered anosmic (olfactory

bulbectomy)(Nelson and Zucker., 1974).

Golden hamsters are extremely sensitive to the effects of melatonin. Melatonin (10 or

25 pg) induced gonadal regression in adult male hamsters if given in the afternoon (6 to 14

hours after lights on) in animals exposed to long day photoperiods while injections given in

the morning had no effect (Tamarkin et al., 1976). Timed melatonin injections replicated the








52

influence of short days on White-footed mice (Glass and Lynch, 1982) and Grasshopper mice

(Frost and Zucker, 1983). Such melatonin injections regressed the testes, arrested estrous

cycles and heightened the feedback potency of testosterone upon LH secretion (Sisk and

Turek, 1982).

Melatonin has also been shown to block the ability of short day photoperiod (<12.5

hours) to induce gonadal regression in hamsters (Reiter et al., 1974) when administered

continuously via subcutaneous implants. Such implants also blocked the ability of afternoon

melatonin injections to induce gonadal regression (Reiter et al., 1977). Reiter (1982) has

proposed that the continuous supply of melatonin provided by such implants "down-regulates"

all melatonin receptors, thus rendering tissues containing melatonin receptors insensitive to

further melatonin treatment. Under such circumstances, morning injections of melatonin

would be ineffective in influencing gonadal function because all receptors would be down

regulated from night time melatonin secretion. However, this hypothesis may not be entirely

accurate as pituitary glands collected in the morning and afternoon and superfused with

melatonin exhibited reduced LH and FSH secretion (Reiter, 1991).

In female hamsters, the reproductive cyclic state is characterized by a continuation of

the daily LH and FSH surges (Seegal and Goldman, 1975). Some evidence suggests that

melatonin does not change LH secretion but does affect FSH synthesis and secretion. Female

hamsters responded to exogenous melatonin with gonadal suppression, cessation of estrus

cyclicity and discontinuation of ovulation (Trakulrungsi et al., 1979). White-footed mice

which were given melatonin implants also exhibited gonadal regression regardless of the

photoperiod to which they were exposed (Petterborg and Reiter, 1982).

Pinealectomized hamsters responded to melatonin differently from pineal-intact

hamsters. PNX animals showed little or no response to single daily injections of melatonin








53

even when the injection occurred in the same phase at which intact animals responded

(Tamarkin et al., 1976). PNX hamsters, however, responded to three injections of melatonin

administered at 6 hour intervals. In this paradigm, the time-of-day of melatonin treatment did

not appear to be important, suggesting that the phase of sensitivity to melatonin in pineal-intact

hamsters may be related to pineal induced rhythms (Goldman et al., 1979).

Current evidence shows that photostimulation of hamsters leads to increased FSH

secretion and testicular weight prior to any detectable increases in LH secretion (Simpson et

al., 1982). Significant changes in serum FSH levels were observed within 3 days and

significant testicular growth within 5 days of photostimulation. However, significant increases

in serum LH levels were not detected for 21 days (Hoffman, 1981).

FSH administration to hamsters maintained on short photoperiod also led to increased

testicular development while administration of LH had no effect (Luderer and Schwartz, 1991).

Such effects of FSH administration were completely abolished if porcine follicular fluid

(inhibin and follistatin source) were administered concurrently (Milette et al., 1988). Thus,

FSH appears to play a critical role during the early stages of spontaneous testicular

recrudescence.

The development of an RIA for melatonin by Rollag and Niswender (1976)

dramatically increased the experimental possibilities available. After that time, research

demonstrated that melatonin secretion in all seasonally reproducing mammals closely paralleled

the duration of darkness. With the characterization of endogenous melatonin secretion came

the ability to substitute a unique rhythm and observe the response.

Programmed subcutaneous infusions of melatonin were administered on a daily basis

to juvenile male hamsters. By using infusions, it was possible to vary (1) the circadian phase

of melatonin administration; (2) the daily duration of melatonin exposure; (3) the amplitude of








54

the melatonin 'peak'; and (4) the total amount of melatonin systematically. In the first

experiment, daily doses as low as 10 ng/day infused over a 12 hour time for 12 consecutive

days resulted in complete testicular regression in PNX Djungarian hamsters raised under long

days. Other experiments demonstrated that 12 hours of melatonin treatment resulted in

testicular regression regardless of the time of day the infusions were given. Infusions of

shorter duration (i.e. 4 or 6 hours) were ineffective in inducing regression (Yellon et al., 1980).

Thus it appeared that the duration of 'peak' secretion of melatonin and not the time of day

melatonin administration was most important.

Hamsters raised from birth under short day photoperiod displayed significant

reductions in testis growth. Testicular activity was stimulated by transferring animals to long

day photoperiods. However, PNX completely blocked the long day growth induction. The

stimulatory effects of long days could also be mimicked in the PNX animals by programmed

infusions of melatonin lasting from 4 to 6 hours (Hoffman and Kuderling, 1975). Infusions

from 8 to 12 hours resulted in continued testicular regression (Carter and Goldman, 1982).

The nocturnal rise in melatonin secretion may be as critical to melatonin's role in

sheep. Constant-release melatonin implants blocked the ability of photoperiod to alter

testicular function in intact rams (Lincoln, 1982). Timed supplementation of melatonin via

feeding or injection in intact ewes temporarily shifted the onset of the breeding season.

Afternoon melatonin supplementation during long days produced an exogenous peak in

advance of endogenous nightly rises in melatonin. Such a supplementation pattern created the

semblance of a short day photoperiod and thus, stimulated the onset of gonadal activity

(recrudescence) (Arendt et al., 1982; Nett and Niswender, 1982).

Ewes which had been pinealectomized and given melatonin infusions representative of

short day photoperiods could be manipulated to enter the breeding season earlier, even when








55

exposed to long day inhibitory photoperiods (Bittman et al., 1982). Those given infusion

patterns representative of long day photoperiods remained reproductively inactive (Bittman et

al., 1984).

The extent, timing and relative importance of specific photoperiodically driven events

within annual reproductive cycles varies among breeds and species (Evans and Robinson,

1980). The photoperiodic modulation of the secretion of LH and FSH is critical to the annual

reproductive rhythms of both long and short day breeders (Bemdtson and Desjardins, 1974).

Studies have demonstrated that melatonin implants administered into the anterior

hypothalamic area (AHA) and suprachiasmatic nuclei (SCN) of White footed mice resulted in

gonadal regression as well (Glass and Lynch, 1982). Additionally, afternoon injections of

melatonin into the SCN also resulted in gonadal regression while morning injections had no

effect (Glass and Lynch, 1982). These data suggest that melatonin acts at the level of the

hypothalamus.

The pattern of GnRH release in both hamsters and sheep is episodic in nature. A

putative mechanism collectively referred to as the pulse generator is responsible for regulating

the number, frequency and amplitude of GnRH pulses into the portal circulation. Changes in

photoperiod have been shown to affect the pulse generator with stimulatory photoperiods

enhancing its activity and inhibitory photoperiods inhibiting its activity (Stetson and Watson-

Whitmyre, 1984). These alterations result in the seasonal changes in the reproductive state of

the animal.

Daylength has been shown to regulate the frequency of GnRH secretion (Carmel et al.,

1976). During summer, LH pulses in male sheep occurred infrequently however, as the

breeding season approached, LH pulse frequency increased (Katanogle et al., 1974). In female

sheep, daylength modulates follicular growth and E2 (Goodman and Karsch, 1980) and in








56

female hamsters, inhibitory photoperiods were associated with high amplitude, daily discharges

of LH and FSH (Seegal and Goldman, 1975). The primary consequence of exposure to short

days and melatonin injections in female Golden hamsters was the loss of estrous cyclicity and

the loss of the LH and FSH surge (Seegal and Goodman, 1975). Short day exposure also

increases the sensitivity of the gonadotropes to exogenous GnRH in the presence or absence of

the ovaries. This may be simply the result of the availability of more GnRH receptors on the

gonadotropes as endogenous GnRH secretion is low at this time. Short day exposure also

abolished the positive feedback effects of estradiol gonadotropin release (Stetson and Watson-

Whitmyer, 1984).

In sheep, experiments were carried out to determine the ability of melatonin and

photoperiod to regulate reproductive activity. PNX ewes were given nightly melatonin

infusions mimicking a long day (16:8 L:D) and intact ewes were exposed to long days. After

90 days, melatonin patterns and photoperiod were switched to short day patterns (8:16 L:D).

In each case, both photoperiod and short day melatonin patterns were able to induce

reproductive recrudescence out of season. The melatonin pattern resulted in a pulsatile LH

pattern of secretion indistinguishable from that seen in intact ewes (Bittman et al., 1983).

The responsiveness of PNX ewes to GnRH administration demonstrated the ability of

exogenous melatonin to mimic photoperiod (Bittman, 1984). As with hamsters, ewes given

melatonin patterns representing inhibitory photoperiods remained sensitive to GnRH

stimulation. This demonstrated that photoperiod had little effect on sensitivity of pituitaries to

GnRH in these species. Therefore, seasonal changes in episodic release of LH and FSH must

result from seasonal changes in the synthesis and pulsatile release of GnRH from the

hypothalamus (Bittman, 1984).








57

Recently, information has been forthcoming indicating that melatonin could have a

direct role in the regulation of steroidogenesis at the level of the preovulatory follicle or corpus

luteum. Administration of melatonin advanced the onset of the breeding season, stimulated an

earlier onset of puberty in sheep (Wallace et al., 1988). Furthermore, this same investigation

found that progesterone concentrations in the first diestrus after ovulation were significantly

higher in sheep treated with melatonin (Wallace et al., 1988). Melatonin has also been shown

to elevate basal P4 production by bovine and human granulosa cells in vitro (Webley and

Luck, 1987) or in hCG stimulated human or baboon granulosa cells (Brzezinski et al., 1992).

Melatonin has been identified in the follicle fluid of humans (Brzezinski et al., 1987), rats

(Penny et al., 1987) and baboons (Brezezinski et al., 1992). Additionally, melatonin receptors

have been reported in the ovaries of hamster, rat and humans (Cohen et al., 1978).

Investigations into the effects of melatonin on the reproductive system of the mare

have been limited. The first known investigation of the effects of exogenous melatonin

administration to mares was done by Strauss and co-workers (1979). In this study, OVX pony

mares were given melatonin in subcutaneous beeswax implants during the summer (breeding

season in intact mares). Such implants significantly reduced hypothalamic GnRH content as

compared with control mares. Thus, high doses of melatonin, delivered chronically, appear to

be antagonistic to reproduction.

Other studies also place melatonin in the role of an antagonist to reproduction. As

was presented previously, exposure of anestrus mares to 2.5 hours of artificial light after dusk

advanced the onset of the breeding season. However, when exogenous melatonin was

administered during the period of photostimulation (2.5 hours of melatonin at dusk), the onset

of the breeding season was no longer advanced (Cheves and Sharp, 1985).








58

In an effort to complement their investigation into the existence of a photosensitive

phase, Guillaume and Palmer (1991) investigated the effects of exogenous melatonin

administration on the date of first ovulation (onset of the breeding season) in anestrus mares.

In this study, mares were photostimulated with either a 14.5:9.5 L:D and evening melatonin or

17.5:6.5 L:D and morning melatonin. Evening, but not morning melatonin blocked the

stimulatory effects of long days during anestrus. They concluded that the photosensitive phase

described in their earlier investigations (Palmer et al., 1981) may be sensitive to light even in

the presence of melatonin.

While no clear data exist in this species, results from other photoperiodic mammals

suggest that the existence of a light sensitive but melatonin insensitive phase is unlikely. If

melatonin is the endocrine signal by which changes in daylength are encoded in the mare,

melatonin sensitivity, not light sensitivity, would be critical to timing circadian and circannual

reproductive behavior.

In conclusion, it has been demonstrated that seasonally breeding mammals, including

mares, utilize changes in daylength to time their reproductive cycles to produce offspring at

the most appropriate time of the year. The pattern of daylength which stimulates the onset of

the breeding season depends on the gestation length of the mammal. Such changes in

daylength are transduced into an endocrine language, melatonin, by the pineal gland. Removal

of the pineal or alterations in the melatonin pattern have drastic effects on the annual breeding

cycle. Melatonin receptors have been identified in close proximity to the hypothalamic areas

associated with reproduction and changes in photoperiod or melatonin pattern can alter the

secretion of the reproductive hormones. Furthermore, there is increasing evidence that

melatonin could be involved in the regulation of gonadal function independent of

hypothalamic-pituitary mechanisms.














CHAPTER III
LH SECRETION IN ANESTROUS MARES EXPOSED TO ARTIFICIALLY LENGTHENED
PHOTOPERIOD AND TREATED WITH ESTRADIOL


During anestrus (November through January, Northern Hemisphere)(Sharp, 1980),

pituitary LH synthesis (Sherman et al., 1992), storage (Hart et al., 1984) and secretion (Garcia

and Ginther, 1976) are minimal. Conversely, pituitary content of follicle stimulating hormone

(FSH) was not different during anestrus or the breeding season (Hart et al., 1984)(Freedman et

al., 1979). The relative absence of FSH secretion during anestrus likely reflects low

hypothalamic content of gonadotropin-releasing hormone (GnRH)(Strauss et al., 1979; Hart et

al., 1984) and secretion (Sharp and Grubaugh, 1987). Shortly after daylength begins to

increase, hypothalamic GnRH content also begins to increase (Silvia et al., 1986).

Hypothalamic GnRH secretion has not been measured during January but was shown to be

higher by early February (the beginning of vernal transition) compared to December

(anestrus)(Sharp and Grubaugh, 1987). The time of increased GnRH secretion correlates with

the initial increase in FSH secretion (Freedman et al., 1979). LH secretion, however, remains

low until just prior to the first ovulation of the year (Freedman et al., 1979; Davis et al.,

1986). These data indicate that the mechanisms involved in seasonal regulation of the

synthesis and secretion of FSH and LH are different.

Burkhardt (1947) first showed that treatment with an artificially extended daylength

during anestrus and early vernal transition advanced the onset of the breeding season. Sharp

(1980) demonstrated that the addition of as little as 2.5 hours of supplemental light at sunset

beginning in December advanced the onset of the breeding season. However, elimination of








60

the ability of mares to perceive changes in daylength either by pineal denervation (superior

cervical ganglionectomy; SCGx)(Sharp et al., 1979) or pinealectomy (PNX)(Grubaugh et al.,

1982) blocked photostimulated advance of the onset of the breeding season in anestrous mares.

Furthermore, mares pinealectomized prior to the onset of deep anestrus (October-November)

were shown to have higher plasma LH content during anestrus (December) compared with

pineal-intact controls (Cleaver et al., 1992) and both SCGx and PNX resulted in a significant

delay in the onset of the breeding season (Sharp et al., 1979; Grubaugh et al., 1982).

Garcia and co-workers (1979) demonstrated that both LH and FSH content increased

during late spring and summer in OVX mares and declined during the winter indicating that

gonadotropin secretion in the mare occurred independent of ovarian factors. Recent work in

this laboratory showed that exposure of anestrous OVX mares to 2 weeks of a stimulatory

photoperiod stimulated GnRH secretion and FSH secretion (Sharp et al., 1988). Similarly,

exposure of anestrous OVX mares to 2 weeks of constant light (Cleaver et al., 1991) resulted

in increased hypothalamic GnRH content and LH and FSH secretion. The correlation between

changes in hypothalamic-pituitary function and changes in daylength have fostered the belief

that entrainment to photoperiod is a major regulator of the seasonal patterns of LH and FSH

synthesis and secretion.

A possible role of estradiol 178 (E2) in the regulation of LH secretion during sexual

recrudescence has been investigated. Treatment of ovariectomized (OVX) mares with estradiol

during late anestrous/early vernal transition (early February) (Garcia and Ginther, 1978; Sharp

et al., 1991) elicited an increase in plasma LH. In ovarian-intact mares, the first ovulation of

the year is preceded by the reinitiation of ovarian follicular E2 synthesis and secretion and is

temporally associated with the first major increase in LH secretion (Davis et al., 1986; Tucker

et al., 1993). Thus, while E2 is not required for seasonal changes in gonadotropin synthesis








61

and secretion in mares, experimental evidence suggests that it may be involved in the

reinitiation of LH synthesis and secretion prior to the onset of the breeding season in ovarian

intact mares.

The current study was designed to investigate the role of increasing daylength and E2

in seasonal regulation of LH synthesis and secretion in mares. During the short days of

December, anestrous mares provide a sensitive model due to an absence of ovarian factors

such as E2 and limited pituitary LH synthesis (Sherman et al., 1992) and secretion (Freedman

et al., 1979). Thus, any alterations in LH synthesis or secretion observed can be attributed to

the effect of treatments during the experimental period. Our specific hypothesis was that the

positive effects of E2 on LH synthesis (as measured by pituitary stores released in response to

exogenous GnRH) and secretion (as measured by circulating concentrations) would be manifest

only after exposure of mares to a stimulatory photoperiod.



Materials and Methods



Experimental Design

Fourteen mature, anestrous pony mares were selected based on the following criteria:

1) average ovarian follicle size of less than 10 mm diameter with no follicle greater than 10

mm diameter, and 2) 3 consecutive weeks of plasma progesterone less then 1 ng/ml.

Progesterone was quantified in plasma by radioimmunoassay (RIA)(Diagnostic Products

Corporation, Coat-a-Count assay kit). The assay was validated for use in mares in our

laboratory (Davis et al., 1986). The sensitivity of this assay was 0.1 ng/ml and the inter- and

intra-assay coefficients of variation were 5.65 % and 6.37 %, respectively. All animals were

maintained on pasture and were supplemented with hay throughout the experimental period.








62

Mares were divided randomly into four groups: Group 1 (ambient photoperiod + control

injections; n=3); Group 2 (ambient photoperiod + E2 injections; n=4); Group 3 (supplemental

light + control injections; n=3); and Group 4 (supplemental light + E2

injections; n=4)(figure III-I). The experiment extended from November 23 through December

21 (28 days).



Blood Collection and Handling

Blood samples were collected frequently from all mares on November 23 (Period 1),

November 30 (Period 2), December 7 (Period 3), December 14 (Period 4) and December 21

(Period 5) (experimental days 0, +7, +14, +21 and +28, respectively)(figure III-I). Blood

samples collected on experimental day 0 were to establish pretreatment plasma hormone

content for all mares in all experimental groups. At each sampling period, jugular vein

catheters were inserted aseptically and blood samples were collected at 30 minute intervals for

8 hours (09:00 through 17:00). Harvested blood was collected and placed into heparinized

tubes and stored on ice for transport to the laboratory. Samples were centrifuged at 1500 g for

15 minutes and plasma was collected then stored frozen at -20 C until assayed for hormone

content.



Light Treatment

Group 1 and 2 mares were exposed to ambient photoperiod throughout the

experimental period. Ambient daylength at the onset of the experiment was 10 hours 30

minutes light and 13 hours 30 minutes of dark and at the end of the experiment was 10 hours

14 minutes of light and 13 hours 46 minutes of dark (Data supplied by the University of

Florida Department of Astronomy). Group 3 and 4 mares were exposed to ambient












0










tf) 40
79













s R









Cu.-~ -
VCA
*S g OO




|l|
u 27 .25oug





m










0 0 3 "
En





a a so o


a s
i~r.





















c, o
5 0 Cu?



















g 2 L
96 e .
pi3!a~













x
c'i

CMj
w
Q.
0
L-


4-'




0
C-



o
0
clJ)


0
C.'

0
L-

0
5mz


0
x
cij
Cj
CL


0
L-
0


0
x
X


0
Q-

oL
0
(L.
0


LO

0
.)





.a


co
CL
CO
"0

0



0
L.


0

0


co
CM




C\J
+


N,














[1r
0>


+



N



+

0


0
C
-< 5


(0
a

,,i






0.
x

m'
X
tU








65

photoperiod during Period 1 then received 2.5 hours of supplemental light daily beginning on

November 24 (experimental day 1) and continuing throughout the experimental period (figure

IH-I). Lights were turned on at 17:00 to maintain light exposure levels then remained on for

2.5 hours after sunset (as determined by U.F. Department of Astronomy information for

Gainesville, Florida). This treatment regime, when administered beginning on December 1,

has been shown to advance the onset of the breeding season (Sharp, 1980). The addition of

2.5 hours of supplemental light to ambient daylength on experimental day +1 (November 24)

created a photoperiod of 13:00/11:00 L:D, equivalent to the ambient photoperiod on April 20

at 296' north latitude. To facilitate administration of the light treatment, Group 3 and 4

mares were isolated in a light paddock every day prior to sunset. Supplemental light was

delivered by halogen lamps on poles 20 feet in height evenly spaced around the paddock

creating uniform illumination of greater than 100 lux. At the completion of the daily light

treatment, mares were released onto pasture.



Estradiol Treatment

Beginning on December 8 (experimental day +15), Group 1 and 3 mares received

twice daily injections intramuscularr) (approximately 12 hours apart) of sesame oil with 10 %

benzyl alcohol, by volume, (3.0 ml total volume per injection) between 08:00 and 09:00 and

again between 20:00 and 21:00. Group 2 and 4 mares received twice daily injections

intramuscularr) of sesame oil + E2 (5.0 mg E2 in 3.0 ml sesame oil with 10 % benzyl alcohol)

during the same time periods as Group 1 and 3 mares (figure III-I). This E2 treatment has

been shown by our lab to be effective in stimulating increased pituitary LH content and

secretion (Sharp et al., 1991). Blood samples collected during Period 4 were assayed for E2

by RIA (Diagnostic Products Corporation, double-antibody assay kit) to establish efficacy of








66

the treatment regime during the current experiment. The sensitivity of this assay was 5.0

pg/ml and the intra assay coefficient of variation was 1.02 %. Injection treatments were

continued daily through the end of the experiment (December 21)(14 days of injections).



GnRH Challenge

To quantify pituitary LH response, all mares were administered a GnRH challenge at

the completion of Period 5 blood sample collection. Samples collected during the final 2

hours of Period 5 (15:00 through 17:00) were utilized to establish plasma LH content prior to

GnRH challenge. Immediately after collection of the blood sample at 17:00, a single 100 pg

bolus of GnRH (Sigma, U.S.A.) in 1.0 ml sterile saline was administered directly through the

jugular vein catheter. Blood samples were collected at 15 minute intervals for an additional 4

hours (17:15 through 21:00) to evaluate pituitary LH response to the GnRH challenge.



Ovarian Activity

Ovarian follicular activity was monitored at weekly intervals by rectal palpation.



LH Assays

Plasma LH was analyzed using an RIA validated in our laboratory for use in the mare

(Sharp et al., 1991). The primary antisera used was an anti-ECG antiserum provided by Dr.

Donald Thompson Jr. (Louisiana State University) in conjunction with purified equine LH

donated by Dr. Harold Papkoff (University of California at San Francisco) for radioiodination

and standard preparations. The sensitivity of the LH assay was 1.2 ng/ml and the intra- and

interassay coefficients of variation were 9.6 and 10.2 %, respectively.










Statistical Analysis

Data were analyzed by least-squares analysis of variance to evaluate differences in

mean plasma LH at each sampling period and after the GnRH challenge at the end of Period

5. The main effects of group, mare within group, period, blood sample and the interaction

group with period, mare within group with period, sample with group, sample with period,

sample with mare within group and the residual error were included in the model. Plasma E2

concentrations at period 4 were analyzed by least squares analysis of variance with the main

effects of group, mares within group, blood sample, and the interactions of group with blood

sample and the residual error were included in the model. Furthermore, mean pre-GnRH

challenge, post-GnRH challenge and maximum LH response to the GnRH challenge were

analyzed by LSANOVA with the main effects of group, mare within group, challenge phase,

sample and the interactions group by challenge phase, mare within groups by challenge phase,

group by sample, mare with group by sample, challenge phase by sample, group by challenge

phase by sample and the residual error were included in the model. Average and maximum

follicle sizes were analyzed by LSANOVA and the model included the main effects of group,

mare within group, week and the interactions group by week and the residual error.

Orthogonal contrast analysis was utilized to test for differences among groups, among

periods, groups at each period (group by period) and the pre-, post- and maximum mean LH

concentrations in response to the GnRH challenge. Potential differences in time trends among

groups at each of period and during the GnRH challenge were examined through tests for

homogeneity of regression with models being tested at the highest significant order of

regression.








68

Results



Ovarian Activity

Analysis of weekly ultrasound measurements of ovarian follicular activity showed no

significant increase in follicle number (2.4 follicles/group/week) or average follicle diameter

(8.5 mm/group/week) over the experimental period.



Hormone Concentrations

Period 1. Period 1 (pretreatment) orthogonal contrasts of mean plasma LH content did

not differ among experimental groups (table III-I). Tests for homogeneity of regression

revealed that the LH response of all mares in all groups were best described as linear and did

not change (zero slope) over the 8 hour period (data not shown). Data were analyzed at the

first order.



Period 2 and 3. After exposure to 7 (Period 2) or 14 (Period 3) days of 2.5 hours of

supplemental light, orthogonal contrast analysis indicated that mean plasma LH content in

Group 3 or Group 4 mares were not different from the mean plasma LH content of Group 1

and 2 mares (table III-I). Furthermore, mean plasma LH content of all experimental groups

did not differ compared to pretreatment (Period 1) mean plasma LH content (table III-I). Tests

for homogeneity of regression revealed that plasma LH concentrations were linear in all groups

and did not change (zero slope) throughout period 2 (November 30) and period 3 (December

7)(data not shown). Data were analyzed at the first order.















Table III-I. Mean LH Concentrations at Periods 1, 2, 3, 4 and 5 for all Groups. Values are
LSMeans S.E.M.


I Period 1
(Nov.23)


Period 2
(Nov.30)


LH (ng/ml)

Group 1 Group 2 Group 3 Group 4
Ambient Ambient Ambient Ambient
1.91 1.14 3.70 0.99 1.81 1.14 3.25 0.99


Group 1 Group 2 Group 3 Group 4
Ambient Ambient + Light + Light


2.29 0.86


3.43 0.74


1.96 1.05


3.17 0.74


Period 3 3.55 0.79 2.94 0.72 3.80 0.97 2.90 0.69
(Dec. 7)

Group 1 Group 2 Group 3 Group 4
Ambient Ambient + Light + Light
+ Control + E2 + Control + E2
Period 4 2.28 0.37 1.67 0.32 1.39 0.45 2.67 0.32
(Dec.14)
Period 5 2.42 0.66 2.39 0.57 2.42 0.81 3.96 0.57
(Dec.21) ____








70

Period 4. Plasma E2 concentrations were analyzed in blood samples collected after 7

days (Period 4) of twice daily control (sesame oil; experimental groups I and 3) or E2

injections (experimental groups 2 and 4). Mean plasma E2 content in control groups I and 3

was 7.0 15 pg/ml. Mean plasma E2 concentrations, 44.9 10.7 pg/ml, in experimental

groups 2 and 4 were significantly higher (p= 0.0001) than mares receiving control injections.

The mean plasma E2 concentrations were equivalent to plasma E2 concentrations observed

during the E2 surge preceding ovulation in estrous mares (Ginther, 1979).

Orthogonal contrast analysis of mean LH concentrations at period 4 revealed that

neither exposure to 21 days of supplemental light (Groups 3 and 4) nor treatment with E2

twice daily for 7 days (Groups 2 and 4) had any effect on mean plasma LH concentrations

(table III-I). Mean plasma LH concentrations among mares were not different. Tests for

homogeneity of regression indicated that curves were linear and did not change (zero slope)

over the 8 hour sample collection interval (data not shown). Data were analyzed at the first

order.



Period 5. Orthogonal contrast analysis indicated that mean plasma LH concentrations

in Group 4 mares during period 5 (28 days of supplemental light, 14 days of E2 treatment)

tended to be higher (p=0.061) than mean plasma LH concentrations of the other experimental

groups (table III-I). Tests for homogeneity of regression revealed that curves representing LH

concentrations in all groups were linear, but tended to be higher (p=0.06) in Group 4 mares,

but did not change over the pre-GnRH period (zero slope)(figure III-II). Data were analyzed

at the first order.

Orthogonal contrast analysis indicated that mean post-GnRH challenge LH response of

Group 4 mares were higher (p=0.037) than the mean post-GnRH challenge response of other






























Figure III-II. Mean LH Concentrations During Period 5 in Groups 1, 2, 3 and 4 Mares. LH
concentrations are least squares means S.E.M. Values are LSmeans S.E.M. Groups are
defined in the legend at the top of the figure. Groups 1 and 2 mares were exposed to ambient
photoperiod while groups 3 and 4 were exposed to ambient photoperiod plus 2.5 hours of
supplemental light in the evening. Group 1 and 3 mares received control injections whiles
Groups 2 and 4 mares received twice daily E2 injections on experimental days 15 through 28.
Regression analysis found that LH concentrations during all sampling periods (including period
5) were best described as linear with no change in plasma LH content observed over the 8
hour collection intervals. Mean LH concentrations in Group 4 mares tended to be higher
(p=0.061) compared to LH concentrations in other groups.












Group
Group


1 0 Group 3
2 Group 4


SI


~JD


09:00 11:00 13:00 15:00 17:00
09:00 11:00 13:00 15:00 17:00








73

experimental groups (figure III-IH). The maximum mean plasma LH concentrations observed

after the GnRH challenge at the conclusion of Period 5 was greater (p= 0.013) in Group 4

mares (28 days of supplemental light, 14 days of E2 treatment) compared with the other

experimental groups (figure Ill-III). Tests for homogeneity of regression indicated that the LH

responses to the GnRH challenge in Groups 1, 2 and 3 mares were linear with no differences

in LH concentrations observed over the 2 hour pre-challenge or 4 hour post-challenge

collection periods (zero slope)(figure III-IV). The LH response curve of Group 4

mares was not parallel to other groups, was best described as quadratic and was significantly

different (p=0.007) from the other experimental groups (figure III-IV). Data were analyzed at

the second order.



Discussion



The mechanisms involved in the regulation of seasonal changes in LH secretion are

largely unknown. Clearly, these mechanisms involve the annual changes in photoperiod and to

some extent, ovarian steroid hormone feedback. However, the manner in which these two

phenomena influence LH secretion appears to be independent as OVX mares exhibit clear

annual changes in both LH and FSH secretion (Garcia et al., 1979) in the absence of ovaries.

The spontaneous rise in gonadotropin secretion normally occurs in May/June in OVX mares

exposed to ambient photoperiod (Garcia et al., 1979). However, exposure to 16 hours of light

daily resulted in elevation of LH by early March (Garcia et al., 1979). Thus, changes in

daylength are an important component in the annual regulation of LH secretion.

While the effect of exposure to supplemental light on hypothalamic GnRH content or

secretion was not measured in the current study, previous evidence indicated that changes in



























Figure III-Ill. Mean Pre-, Post- and Maximum Plasma LH Responses to the GnRH Challenge
Administered at the Conclusion of Period 5. Group I and 2 mares were exposed to ambient
photoperiod throughout the experiment. Group 1 mares received control injections and Group
2 mares received E2 injections on experimental days 15 through 28. Group 3 and 4 mares
were exposed to 2.5 hours of supplemental light for 28 days. Group 3 mares received control
injections and Group 4 received E2 injections of experimental days 15 through 28. Bars are
LSMeans S.E.M. and are defined in the figure legend above. Mean pre-GnRH challenge
plasma LH content represents the final 5 samples (15:00 through 17:00) collected during
Period 5. Mean post-GnRH challenge plasma LH content represents the 4 hours of 15 minute
interval samples collected after the GnRH challenge. Maximum mean plasma LH response to
the GnRH challenge represents the highest mean plasma LH value observed following the
GnRH challenge. The single (*) denotes that Group 4 mares post-GnRH challenge plasma LH
content was significantly higher (p=0.037) than the post-GnRH challenge plasma LH content
of the other experimental groups. The double asterisk (**) denotes that the Group 4 maximum
mean plasma LH value in response to the GnRH challenge was significant higher (p=0.013)
than the maximum mean plasma LH value of the other experimental groups.









20 -
0
10 -


__ Group 2 8
20-

10- o


Group 3

20 )

10- co


Group 4 **
20 -

10- T






























Figure III-IV. Mean plasma LH Response of All Experimental Groups to the GnRH challenge
at the Conclusion of Period 5. The mean plasma LH response for each experimental group is
defined by the figure legend. Group 1 and 2 mares were exposed to ambient photoperiod
throughout the experiment. Group 1 mares received control injections and Group 2 mares
received E2 injections on experimental days 15 through 28. Group 3 and 4 mares were
exposed to 2.5 hours of supplemental light for 28 days. Group 3 mares received control
injections and Group 4 received E2 injections of experimental days 15 through 28. Values are
LSMeans S.E.M. of the interaction group by sample. Mean plasma LH response curves for
Groups 1, 2 and 3 were linear and did not differ over the 2 hour pre-challenge and 4 hour
post-challenge interval. The mean plasma LH response of Group 4 mares was quadratic and
was different (p=0.007) from the mean plasma LH response of the other experimental groups.












Group 1
Group 2


0 K--
15:00


-.-.4. ~4


I I I


0 Group 3 v
* Group 4 v


17:00 19:00


GnRH
Challenge


20





10


biD








78
daylength are involved in the seasonal regulation of GnRH. Prior to the onset of increasing

daylength (i.e. prior to December 21), hypothalamic GnRH content was significantly lower

compared to content during other seasons (Strauss et al., 1979; Silvia et al., 1986). Silvia and

co-workers (1986) indicated that an increase in hypothalamic GnRH content was observed

after 19 days exposure to a naturally increasing daylength (on January 9). Thus, it appears

that the mechanisms which regulate hypothalamic GnRH are very responsive to even small

increases in daylength.

GnRH secretion rate during early February was significantly higher than GnRH

secretion rate during anestrus (December) (Sharp and Grubaugh, 1987). Furthermore, exposure

of anestrous mares to 2 weeks of a stimulatory photoperiod (16:8 L:D) during December

resulted in increased hypothalamic GnRH secretion (Sharp et al., 1988). Thus, GnRH

secretion was increased when measured after six weeks exposure to increasing ambient

daylength (an increase of only 33 minutes in daylength at the experimental location; U.F.

Department of Astronomy data) or after two weeks exposure to an abrupt and dramatic

increase in daylength (from 10:15 light to 16:00 hours of light). Therefore, it is reasonable to

speculate that 28 days exposure to an abrupt increase in daylength (from 10:15 light to 13:00

light) would result in increased hypothalamic GnRH content and secretion in the current study.

In terms of circadian biology, the proposed reinitiation of GnRH secretion in response

to supplemental light (Group 4) represents an excellent example of light functioning as a

"zeitgeber". Recall that a zeitgeber is a factor which coordinates other factors to function in a

specific temporal manner (see Chapter II). Thus, supplemental light is the zeitgeber which

activates the synthesis and secretion of GnRH, presumably through the neuroendocrine

mechanisms through which GnRH production is regulated, which in turn, is then available to

stimulate renewed LH synthesis.








79

GnRH secretion is essential for the synthesis and secretion of LH. Addition of GnRH

to cultures of rat pituitary cells increased the incorporation of radiolabeled amino acids into

LHB subunits (indicative of de novo protein synthesis)(Lui and Jackson, 1978). Furthermore,

the down-regulation of GnRH receptors with constant treatment with high concentrations of

GnRH (Lalloz et al., 1988a) or elimination of endogenous GnRH with antisera or antagonists

(Lalloz et al., 1988b) results in an acute decline in LH subunit mRNA levels. Furthermore,

hypothalamus-Pituitary disconnection (HPD) in ewes resulted in the complete disappearance of

the a and LHB subunits mRNAs within one week (Hamemik and Nett, 1988). Thus, GnRH is

essential for the expression of the a and LHB subunit genes and synthesis and secretion of the

LH apoprotein.

Pituitary stalk transaction resulted in a rapid decrease in plasma LH content (Porter et

al., 1994). Furthermore, stalk-transection also resulted in a dramatic decline in pituitary

mRNA content for the a and LHB subunits (Robinson et al., 1994). However, if GnRH was

replaced via intravenous infusion, LH secretion (Porter et al., 1994) and pituitary mRNA

content (Robinson et al., 1994) were maintained at levels not different from pituitary-stalk

intact mares. These results were supported by Sherman and co-workers (1992) who indicated

that pituitary mRNA for the LHB subunit was nearly undetectable during anestrus (December)

when endogenous GnRH levels are low, but increased during the time of the onset of the

breeding season (late April) when GnRH secretion is elevated (Sharp and Grubaugh, 1987).

Thus, the predominant explanation for the absence of LH synthesis and secretion during

anestrus is the relative absence of hypothalamic GnRH secretion.

In the current study, mares exposed to ambient, winter photoperiod had low plasma

LH content throughout the experimental period. Additionally, this low anestrous plasma LH

content was not influenced by 28 days of treatment with a supplemental light regime known to








80

be effective in advancing the onset of the breeding season if continued (Sharp, 1980). Thus,

GnRH secretion, in-and-of itself, may not be sufficient to stimulate increased LH synthesis and

secretion during anestrus. This contention is supported by evidence showing that although

GnRH secretion was shown to be elevated by early February (Sharp and Grubaugh, 1987), LH

synthesis and secretion did not increase until just prior to the onset of the breeding season

(April or May)(Freedman et al., 1979; Davis et al., 1986; Sherman et al., 1992; Peltier and

Sharp, 1994).

Ovarian steroids, specifically E2, have been shown to be stimulatory to LH synthesis

and secretion. Both the a and LH3 subunits mRNA levels increased during the preovulatory

E2 surge in ewes (Landefeld, 1985) and when E2 was added to cultures of pituitary cells

(Shupnik et al., 1989). Treatment of primates (Knobil, 1974) and rats (Legan and Karsch,

1975) with E2 caused an initial suppression followed by a massive release of LH.

Furthermore, treatment of mice genetically lacking proper GnRH production with E2 resulted

in an increase in a, but not LHB subunit mRNA (Mason et al., 1986).

The potential role of ovarian steroid feedback on the annual secretion of the

gonadotropins in mares is most evident during the end of sexual recrudescence in vernal

transition. It is unlikely that steroid feedback contributes to increased GnRH or FSH during

early vernal transition as estradiol and progesterone levels in the circulation are minimal at that

time (Oxender et al., 1977; Garcia and Ginther, 1978; Davis et al., 1987). However, later in

vernal transition, LH secretion just prior to the first ovulation of the year is preceded by

significant E2 production by the follicle destined to be the first ovulatory follicle of the year

(Davis et al., 1987; Tucker et al., 1993).

Data presented in this study indicate that the effect of E2 on reinitiation of LH

secretion requires exposure to increased photoperiod. Administration of E2 to Group 1 mares








81

(exposed to December, short day photoperiod) had no effect on plasma LH whereas

administration of E2 to OVX mares during February using the same E2 treatment protocol

resulted in increased plasma LH content (Sharp et al., 1991). However, in the current study,

the combination of exposure to a stimulatory photoperiod and E2 treatment tended to enhance

LH secretion (Group 4 mares). This supports the contention that exposure to a stimulatory

photoperiod is required to reinitiate GnRH synthesis and release and allow for the positive

feedback effects of E2 on LH secretion.

Silvia and co-workers (1986) demonstrated that pituitary GnRH receptor concentration

did not differ in pituitaries collected during anestrus, early or late vernal transition or estrus.

This suggests that in anestrus or early vernal transition GnRH receptors on the pituitary may

not be limiting. In support of that idea, Sharp and co-workers (1991) reported that pituitary

responsiveness to GnRH did not increase following E2 administration. Therefore, the LH

release observed in Group 4 mares (exposed to photostimulation and treated with E2)

following the GnRH challenge most likely reflected a reactivation of LH synthesis and not an

alteration in pituitary responsiveness. Furthermore, in a study of LH response to varying doses

of GnRH in anestrous mares, there was no difference in LH response following treatment with

25, 50, 100 or 400 pg of GnRH, respectively (Davis and Sharp, unpublished results).

Therefore, the dose of GnRH used in the present study (100 pg) should have been adequate to

elicit an LH response. Although, the possibility that E2 increased pituitary responsiveness to

GnRH cannot be ruled out, as GnRH receptors were not quantified in the current study, failure

of E2 treated, short photoperiod-exposed mares (Group 2) to respond to the GnRH challenge

argues against this.

From these results, the following hypothesis can be suggested. Exposure to

supplemental light stimulates increased hypothalamic GnRH synthesis and secretion. GnRH








82

stimulated increased a and LHB subunit synthesis. Treatment with E2 synergized with

hypothalamic GnRH to increase LH production and release. The lack of LH response in mares

exposed to supplemental light but not treated with E2 and mares treated with E2 but exposed

to ambient daylength indicates that a minimum period of exposure to a stimulatory

photoperiod is required to reactivate the hypothalamo-pituitary axis in anestrous mares. The

data from the current study demonstrated that exposure to a stimulatory photoperiod was

required to enable E2 treatment to stimulate an increase in plasma LH content in anestrous

mares. Thus, anestrous mares provide an excellent research model for the elucidation of the

mechanisms involved in seasonal regulation of LH production in the mare.














CHAPTER IV
EFFECT OF ESTRADIOL TREATMENT ON LH SECRETION DURING LATE
ANESTRUS: AN INVESTIGATION OF THE EXISTENCE OF A PHOTIC GATE



Mares exposed to ambient photoperiod have an anestrous period characterized by low

hypothalamic GnRH content (Strauss et al., 1979; Hart et al., 1984) and secretion (Sharp and

Grubaugh, 1987) and an absence of LH synthesis (Sherman et al., 1992) and secretion (Garcia

and Ginther, 1976). Sexual recrudescence is characterized by increased hypothalamic GnRH

(Silvia et al, 1986) within weeks after the onset of increasing daylength (i.e. the Winter

Solstice) and a corresponding increase in FSH release from pituitary stores (Freedman et al.,

1979). Pituitary LH synthesis (Sherman et al., 1992) and secretion (Freedman et al., 1979;

Davis et al., 1986) does not increase until just prior to the first ovulation of the year (early

April in horse mares and early May in pony mares)(Sharp, 1980). This increase in LH

production and release corresponds to the onset of follicular estradiol (E2) production (Davis et

al., 1987; Tucker et al., 1993).

The specific role E2 plays in the reinitiation of LH production is, as yet, unclear, but

could involve both hypothalamic and pituitary effects. Results from Chapter III indicated that

anestrous mares exposed to ambient daylength during December and treated with E2 did not

exhibit increased LH production. However, ovariectomized (OVX) mares under ambient

photoperiod and treated with E2 in early February display both increased LH synthesis and

secretion (Sharp et al., 1991). But E2 is not essential for LH production as OVX does not

prevent seasonal changes in both LH and FSH secretion (Garcia et al, 1979). Thus, it appears








84

that both ovarian independent (e.g. photoperiod regulated changes in GnRH production) and

ovarian-dependent (e.g. follicular E2 production) are involved in the reinitiation of LH

production during sexual recrudescence in the mare. Furthermore, it appears that a critical

period of time must exist somewhere between December 21 (the end of the experiment in

Chapter III) and February 5 (the beginning of the Sharp et al., 1991) when ambient

photoperiod is sufficient to allow mares to respond to E2 treatment During this period of

time, the dominant zeitgeber, light, stimulates the reinitiation of GnRH production which in

cooperation with E2, will stimulate a reinitiation in LH production.

The current investigation was carried out to test the hypothesis that a "photic gate"

exists during January under ambient photoperiod before which the mare will be insensitive to

E2 treatment (with regards to LH production) and after which, E2 treatment will initiate

increased LH production.



Materials and Methods



Experimental Design

Sixteen mature, anestrous pony mares were selected based on the following criteria: 1)

average ovarian follicle size of less than 10 mm diameter with no follicle greater than 10 mm

diameter; and 2) 3 consecutive weeks of plasma progesterone less then 1 ng/ml. Progesterone

was quantified in plasma by radioimmunoassay (RIA)(Diagnostic Products Corporation, Coat-

a-Count assay kit). This assay has been validated for use in the mare in our laboratory (Davis

et al., 1986). The sensitivity of this assay was 0.1 ng/ml and the inter- and intra-assay

coefficients of variation were 5.79 % and 7.46 %, respectively.








85

All animals transported from the Horse Research Center to pasture in Gainesville on

December 6 and were allowed 2 weeks to acclimate to the new environment. Throughout the

study mares were maintained on pasture and were given hay, water and mineral supplement ad

libitum as well as a commercial grain feed once daily. Animals were exposed to natural

photoperiod throughout the investigation.

Mares were randomly assigned to four groups. Group 1 (n=4) were control mares and

received no treatment throughout the study. Group 2 mares (n=4) received twice daily E2

injections (see Chapter III for details of E2 injection solution and injection protocol) beginning

on evening of experimental day 0 (December 21) and continuing throughout the end of the

study (experimental day 56, February 15). Group 3 mares (n=4) received twice daily E2

injections beginning on experimental day 14 (January 4) and continued through the end of the

study. Group 4 mares (n=4) received twice daily E2 injections beginning on experimental day

28 (January 18) and continued through the end of the study (figure IV-I). All injections were

intramuscular and the site of injection on each mares body was rotated daily to minimize

trauma (e.g. morning injection day 1 in left neck; evening injection day 1 in left rump;

morning injection day 2 in right rump; and evening injection day 2 in right neck).



Blood Sample Collection

Daily blood samples were collected from Group 1 mares throughout the investigation

(day 0 through day 56). Blood samples were collected daily from Groups 2, 3 and 4 mares

only during E2 injections but were collected every three days (Groups 3 and 4) prior to onset

of treatment (January 4 and January 18, respectively)(figure IV-1). Samples were collected by

jugular venipuncture (10 ml Vacutainer brand with sodium heparin, Becton-Dickinson, U.S.A.)

and stored on ice until transported to the laboratory. Blood was centrifuged at 1500 x g for 15