Department of Clinical Veterinary Sciences Faculty of Veterinary Medicine

University of Helsinki Finland

Department of Veterinary Clinical Sciences

Faculty of Veterinary Science University of Sydney

Australia

Endocrinological basis of seasonal infertility in pigs

Anssi Tast

ACADEMIC DISSERTATION

To be presented, with the permission of

the Faculty of Veterinary Medicine, University of Helsinki, for public criticism in Auditorium Maximum,

Hämeentie 57, Helsinki, on September 9th, 2002 at 12 noon

HELSINKI 2002

ISBN 952-91-5024-5ISBN (PDF) 952-10-0655-2

Helsinki 2002Yliopistopaino

1

CONTENTS

ABSTRACT 3

LIST OF ORIGINAL PAPERS 6

ABBREVIATIONS 7

1. INTRODUCTION AND REVIEW OF LITERATURE 9

1.1. INTRODUCTION 9 1.2. MANIFESTATIONS OF SEASONAL INFERTILITY 10

1.2.1. Reduced farrowing rate 10 1.2.2. Delayed puberty 11 1.2.3. Prolonged weaning to oestrus interval 12

1.3. ENVIRONMENTAL FACTORS CONTRIBUTING TO SEASONAL INFERTILITY 12 1.3.1. Photoperiod 12 1.3.2. Temperature 13 1.3.3. Feeding level 14 1.3.4. Other environmental factors 15

1.4. PATTERN OF MELATONIN SECRETION SYNCHRONIZES BREEDING WITH

SEASON 16 1.4.1. Circadian rhythm in melatonin secretion 16 1.4.2. Pineal secretion of melatonin 18 1.4.3. The role of melatonin in regulation of hypothalamo – pituitary – gonadal axis 18 1.4.4. The melatonin pattern in the pig 20 1.4.5. Effects of light intensity on melatonin 22

1.5. RECOGNITION OF PREGNANCY 23 1.5.1. PGF2α secretion during the oestrous cycle 24 1.5.2. Embryonic signals 24 1.5.3. Oestrogen signals alter PGF2α secretion during pregnancy 27 1.5.4. Conceptus secretory proteins (CSP) 28

1.6. MATERNAL RESPONSE 29 1.6.1. Formation of CL and luteotrophic role of LH 29 1.6.2. Progesterone in early pregnancy 31

2. THE AIMS OF THE STUDY 33

3. MATERIALS AND METHODS 34

3.1. ANIMALS 34 3.2. BLOOD COLLECTION 35

3.2.1. Jugular vein catheters (I, II, III, V) 35 3.2.2. Single blood samples (IV, V) 35 3.2.3. Saphenous arterial catheters (I) 36

3.3. GNRH-IMMUNIZATION 37

2

3.3.1. Active immunization (V) 37 3.3.2. Passive immunization (V) 37

3.4. HORMONE AND ANTIBODY ASSAYS 37 3.4.1. Melatonin (I, II, III) 37 3.4.2. Progesterone (IV, V) 40 3.4.4. GnRH-antibody (V) 42

3.5. PREGNANCY TESTING 42 3.6. LIGHTING PROGRAMS 43 3.7. STATISTICAL ANALYSES 43

4. RESULTS 44

4.1. SEASONAL ALTERATIONS IN MELATONIN SECRETION (I) 44 4.2. EFFECTS OF ARTIFICIAL LIGHTING PROGRAMS ON MELATONIN SECRETION

(II, III) 45 4.3. EARLY DISRUPTION OF PREGNANCY (IV, V) 47

5. DISCUSSION 49

6. CONCLUSIONS 55

ACKNOWLEDGEMENTS 56

REFERENCES 58

3

ABSTRACT

Seasonal infertility of the domestic pig has different manifestations, but reduced farrowing rate has

been found to be of highest economic significance. In other seasonally breeding non-tropical

mammals, the seasonal effects on reproduction are mediated by the pineal hormone melatonin. In

the domestic pig, there has been considerable controversy about the existence and role of a

circadian melatonin pattern.

The aims of the study were to investigate the melatonin secretion of the wild boar and compare it

with that of the domestic pig. The effects of different lighting regimens and light intensities on the

circadian melatonin response of pig were also studied. The association between the seasonally

reduced farrowing rate and early disruption of pregnancy was clarified in a field trial. The

endocrinological mechanism of early disruption of pregnancy was investigated by active and

passive GnRH-immunization trials, the aim being to develop an experimental model of this

mechanism.

To determine seasonal alterations in melatonin profiles of the European wild boar and the domestic

pig, the wild boars and domestic gilts were sampled at two-hour intervals for 48 hours four times a

year. The wild boars were under the natural lighting and the gilts under the typical indoor piggery

light environment, where artificial lighting was provided for 12 hours a day and the sunlight

through the windows was not controlled. Both pigs expressed a clear seasonal variation in the

duration of melatonin secretion. There was no difference in lighting – melatonin transduction

between wild and domestic pigs.

To test the effect of different lighting regimens on the secretion pattern of melatonin an experiment

was set up, where young prepubertal boars in four different temperature- and lighting-controlled

climate rooms were exposed to short day or long day lighting conditions. Following a two-week

acclimatizing period, lighting regimens were changed to be opposite in each room. Blood samples

were collected at two-hour intervals for 48 hours spanning the changeover and assayed for

melatonin. The procedure was repeated three times so that the pigs ended up with the same lighting

with which they started. This experiment demonstrated existence of a circadian melatonin pattern

under both lighting regimens and in every animal sampled. Furthermore, the experiment showed

4

that pigs were able to respond to abrupt and extreme changes in lighting within two weeks in terms

of adjustment of the circadian melatonin pattern to a new lighting regime.

The effects of the light intensity during the photophase on the subsequent scotophase melatonin

response were also investigated. Three groups of four gilts were accommodated in three lighting-

and temperature-controlled climate rooms. In each room the light-dark cycle was 12 h : 12 h and the

light intensity was either 40 lx, 200 lx or 10000 lx during the photophase. Following a one-week

acclimatizing period, blood samples were collected at two-hour intervals for 24 hours and assayed

for melatonin. All the groups went through all the treatments (light intensities). This experiment

showed that pigs are able to recognize a light intensity as low as 40 lx. The experiment also

demonstrated no difference in the scotophase melatonin response to the different light intensities.

This experiment suggests that a light-dark cycle is important in generating the melatonin signal,

whereas the light intensity is of little importance.

The association of lowered farrowing rate with early disruption of pregnancies was studied in a

commercial 160-sow unit with a known history of seasonally lowered farrowing rate. Every

sow/gilt in production was included in the study for four months periods in the winter – spring and

in the summer – autumn. The pregnancies were followed for six weeks by serum progesterone

concentrations and ultrasound scanning. An increase in disruption of pregnancies was associated

with a seasonal decrease in the farrowing rate.

The mechanism of early disruption of pregnancy was studied by active and passive GnRH-

immunization. Sows were immunized for the first time on the day of farrowing and the second time

either on day 10 or on day 20 post mating. The passive immunization of gilts was carried out on day

12 post mating. None of the sows actively immunized on day 10 was detectably pregnant (real time

ultrasound) on day 18 post mating. The sows immunized for the second time on day 20 aborted with

a mean vaccination-abortion interval of 10 ± 1.5 days. The passive immunization on day 12 had a

similar effect on establishment of pregnancy as active immunization on day 10 post mating.

Progesterone samples showed that in both active immunization groups the regression of CL took

place within eight days following the second immunization. These results suggest that two different

mechanisms can cause disruption of pregnancy depending on the timing of the immunization. The

immunization on day 20 caused a regression of CL within eight days and abortion followed two

days later whereas the immunization on day 10 caused a failure in establishment of pregnancy

before a regression of CL took place. These findings suggest that seasonally decreased progesterone

5

secretion of CL, because of inadequate LH-support, may interfere with development of embryos

and their ability to produce oestrogen signals for recognition of pregnancy.

It is concluded that the basis of lighting-melatonin transduction in the pig is similar to other

mammals and that simple artificial lighting programs can manipulate the circadian rhythm of

melatonin secretion. The seasonally decreased farrowing rate can partly be explained by an

increased number of disrupted pregnancies. It is suggested that seasonal early disruption of

pregnancy is mediated through seasonally decreased progesterone concentrations, which interfere

with development of embryos. The seasonal decrease in progesterone may be caused by inadequate

LH-support for CL which results from changed melatonin secretion pattern that is inhibitory for

GnRH-secretion.

6

LIST OF ORIGINAL PAPERS

The present thesis comprises the following original papers. In the text, the papers are referred to by

their Roman numerals. They were reprinted by the kind permission of the journal concerned.

I) Tast, A., Hälli, O., Ahlström, S., Andersson, H., Love, R.J. and Peltoniemi, O.A.T.

Seasonal alterations in circadian melatonin rhythms of the European wild boar and

domestic gilt. J. Pineal Res. 2001, 30: 43-49.

II) Tast, A., Love, R.J., Evans, G., Telsfer, S., Giles, R., Nicholls, P., Voultsios, A. and

Kennaway, D.J. The pattern of melatonin secretion is rhythmic in the domestic pig and

responds rapidly to changes in daylength. J. Pineal Res. 2001, 31:294-300.

III) Tast, A., Love, R.J., Evans, G., Andersson, H., Peltoniemi, O.A.T. and Kennaway, D.J.

The photophase light intensity does not affect the scotophase melatonin response in the

domestic pig. Anim. Reprod. Sci. 2001, 65: 283-290.

IV) Tast, A., Peltoniemi, O.A.T., Virolainen, J.V. and Love, R.J. Early disruption of

pregnancy as a manifestation of seasonal infertility in pigs. Anim. Reprod. Sci. (in press)

V) Tast, A., Love, R.J., Clarke, I.J. and Evans, G. Effects of active and passive

gonadotrophin-releasing hormone immunization on recognition and establishment of

pregnancy in pigs. Reprod. Fertil. Dev. 2000, 12: 277-282.

7

ABBREVIATIONS

Ab antibody

BSA bovine serum albumin

bTP-1 bovine trophoplast protein-1

C18 active charcoal

CL corpus luteum

cm centimeter

CSP conceptus secretory proteins

CV coefficient of variation

EDP early disruption of pregnancy

EDTA ethylenediaminetetraacetic acid

EMAI Elizabeth McArthur Agricultural Institute

FSH follicle stimulating hormone

g grams

GnRH gonadotrophin releasing hormone

I125 radioactive iodine

IFN interferon

kg kilograms

L liters

LH luteinizing hormone

LPP long pseudo pregnancy

MB maximal binding

min minutes

MJ megajoules

ml milliliters

mm millimeters

NaCl natrium chloride

ng nanograms

NSB non specific binding

oTP-1 ovine trophoplast protein-1

pg picograms

PGF2α prostaglandin F2α

8

PGE prostaglandin E

registered trade mark

RBP retinol binding protein

RIA radioimmunoassay

rpm rounds per minute

SCN suprachiasmatic nucleus

SPP short pseudo pregnancy

µg micrograms

µl microliters

9

1. INTRODUCTION AND REVIEW OF LITERATURE

1.1. INTRODUCTION

The Western domestic pig breeds are derived from the European wild boar (Sus scrofa scrofa),

which has a breeding season in late autumn - early winter. Piglets of the wild boar are born in the

spring when climatic conditions and feed sources maximise the chances of survival. After giving

birth to piglets, a wild boar sow remains in anoestrus until the next winter. Initially this anoestrus is

associated with lactation, but after weaning the sow stays in anoestrus until the next breeding

season. The photoperiod is the primary factor determining the onset of the breeding season, while

availability of feed has a modifying effect. Changing photoperiod affects the pattern of melatonin

secretion from the pineal gland, and it is thought to be the most important factor that synchronizes

breeding with the preferred season.

Although the domestic pig is not a distinct seasonal breeder, it is clear that it has a tendency to a

lowered fertility in the autumn. Seasonal infertility appears in late summer and early autumn, same

time when the European wild boar experiences total anoestrus. A reduced farrowing rate, delayed

puberty in gilts and a prolonged weaning to oestrus interval manifest seasonal infertility. Gilts and

primiparous sows are most severely affected (Love, 1978; Peltoniemi et al., 1999b).

Seasonal effects on reproductive performance of the pig and its physiological and endocrinological

mechanisms are discussed in the present review of the literature. In the first section, different

manifestations of seasonal infertility are described. In addition, environmental factors, which

contribute to seasonal infertility, are briefly reviewed. In the second section, attention is paid to the

endocrinology of seasonal breeding and some particular features of melatonin secretion patterns in

pigs are pointed out. The third section discusses the effect of changing photoperiod on the

recognition and establishment of pregnancy.

10

1.2. MANIFESTATIONS OF SEASONAL INFERTILITY 1.2.1. Reduced farrowing rate

Farrowing rate is defined as the proportion of mated sows that farrow (Love, 1981). Repeated

matings/inseminations to the same oestrus are calculated as a single mating event. This may be

modified by removing sows that have died or been culled for some reason other than fertility

problems to give an adjusted farrowing rate.

Reduced farrowing rate is typical following matings in late summer and early autumn (Love et al.,

1978). In another study by Love (1981), more delayed returns to oestrus after mating were found

during summer-autumn than winter-spring. Using oestrone sulphate as an indicator of pregnancy, it

was found that 25-30 % of sows detected to be pregnant did not farrow in the autumn where as in

the spring only 4 % of sows detected to be pregnant did not farrow (Mattioli et al., 1987). These

studies provide strong evidence for the important role of early disruption of pregnancy in seasonal

infertility, and it appears to be important in reducing farrowing rate seasonally.

A reduction in the farrowing rate following matings in autumn has been revealed in many studies,

but the severity of the problem has been variable. Stork (1979) reported only a 3-5 % reduction in

farrowing rates caused by the season in Britain, and Lucia et al. (1994) found season to cause

approximately a 3 % reduction in farrowing rates in North American herds. In a retrospective study

on the seasonal effect on fertility, a 5-10 % reduction in the farrowing rate following matings from

August to October was found in Finland (Peltoniemi et al., 1999a). In an Australian study (Fig. 1.)

farrowing rates following autumn matings dropped down to 50 % in most severe cases and 10-15 %

reductions were commonly seen in this country (Love et al., 1993). In Italy, Enne et al. (1979)

reported that only 40 % of sows mated in the summer farrowed.

Variation in the severity of seasonal infertility is at least partly explained by the different

management and environmental factors (Hancock, 1988). It is also typical for seasonal infertility to

have a great variation between years, weeks, piggeries and even within the same piggery amongst

different groups of pigs (Love et al., 1978; 1993). The great variation in severity and an

unpredictable appearance of the problem make it difficult to control. Early disruption of pregnancy

causes irregular return to oestrus, as discussed earlier. The traditional method of detecting non-

pregnant sows by boar at three weeks after mating is unlikely to detect sows, which have had an

11

early disruption of the pregnancy. If any other pregnancy test methods are not used the risk of larger

numbers of sows being non-pregnant near the expected farrowing date increases following summer

and autumn matings.

F a r ro w in g ra te

6 46 66 87 07 27 47 67 88 08 28 48 6

s p r in g s u m m e r a u tu m n w in te r

N = 7 2 1 3

Figure 1: Seasonal effects on farrowing rates 1983-1991. Data from Love et al. (1993).

1.2.2. Delayed puberty

Most of the studies support the opinion that gilts reach puberty at an older age during the seasonal

infertility period compared with the rest of the year (Hughes, 1982; Peltoniemi et al., 1999a). In an

Australian study, 53 % of gilts reached puberty at 225 days’ age, when kept in short day lighting

conditions around the expected time of the puberty and isolated from boars, whereas only 13 % of

gilts reached puberty by that age when kept in long day lighting conditions and isolated from boars

(Paterson et al., 1991). The effect of season on puberty is dependent partly on other environmental

factors, especially on boar contact. However, the seasonal delay of puberty in gilts seems to remain

to some extent even under the influence of boar contact (Paterson et al., 1991). This seasonally

delayed puberty is typical for the European wild boar in its natural environment. After threshold

values in age and weight are achieved, the occurrence of puberty depends on season; if the right age

and weight are reached late in the spring, the attainment of the puberty will be delayed until the next

winter (Mauget, 1987).

In a Finnish study, (Peltoniemi et al., 1997a) puberty of domestic gilts was delayed by

approximately 10 days during the autumn compared with the rest of the year. This delayed

attainment of puberty has economic consequences in commercial piggeries, especially because it

12

occurs simultaneously with the period of reduced farrowing rate and prolonged weaning to oestrus

interval in sows.

1.2.3. Prolonged weaning to oestrus interval

Effects of seasonal infertility on the first oestrus after weaning seem to vary. Many authors have

reported severe adverse effects of season on weaning to oestrus interval. In these reports, weaning

to oestrus interval in primiparous sows was prolonged significantly during the period of seasonal

infertility (Hurtgen and Leman, 1980; Mattioli et al., 1987; Prunier et al., 1996). These sows

experienced a total anoestrus after weaning, analogous to the situation in the wild boar. In Finland,

a prolonged weaning to oestrus interval seems not to be a severe problem according to data

collected from the database of the Finnish Animal Breeding Association (Peltoniemi et al., 1999b).

There is not a clear explanation why prolonged weaning to oestrus interval is not such a severe

problem in Finland as in many other countries. One explanation could be the small average herd

size in Finland, where average number of sows per farm is only about 50. The oestrus detection

might be more efficient in small farms than in larger herds and, thus, weak oestrus signs might be

more easily recognized. It has to be taken into consideration that collecting data from this kind of

database may lead to an underestimation of some particular problem depending on the methods of

data collection and classification. However, in general, a prolonged weaning to oestrus interval is

still considered an important and economically significant manifestation of seasonal infertility.

1.3. ENVIRONMENTAL FACTORS CONTRIBUTING TO SEASONAL INFERTILITY

1.3.1. Photoperiod

Photoperiod synchronizes the beginning of the breeding season with the preferred season in most

non-tropical seasonally breeding mammals. For short day breeders such as the sheep, the long

duration of high nocturnal plasma melatonin concentrations leads to an increased frequency of

GnRH-pulses through complicated neural pathways and hormonal feed back mechanisms, which

stimulate the release of LH from the pituitary gland and thereby activates the gonads. In pigs,

effects of the photoperiod on melatonin and on subsequent hypothalamic activity are not as well

documented as in the sheep. Some authors have found extraordinary melatonin profiles like

13

increased melatonin concentration during the photophase in pigs (Peacock, 1991), and others seem

to think that the neuro-endocrinological basis of the pig is similar to sheep and that confusing

results in melatonin profiles may be due to technical difficulties in assays used (Paterson et al.,

1992). Some fundamental details in melatonin patterns in pigs under different lighting conditions

have remained unclear.

Confusing results of effects of photoperiod on reproduction (and how the photoperiodic information

is transduced into endogrinological form) in domestic pigs have led to the recommendation of a 16-

hours’ photophase all year round in many countries. For example, in Scandinavian countries where

pigs are kept indoors and lighting is controlled artificially, recommendation is to provide sows with

16-17 hours of light daily. However, the latest experiments have provided strong evidence that the

domestic pig still has a tendency towards seasonal breeding similar to European wild boar, which is

a distinct short day breeder. One could expect that short day lighting regimens might be beneficial

in commercial piggeries at least a couple of weeks before and after moving sows or gilts into a

breeding unit. The optimal lighting regimens might demand different lighting conditions in different

units but these programs are yet to be determined.

1.3.2. Temperature

It has been suggested that high ambient temperatures have importance in seasonal infertility

(Wetteman and Bazer, 1985). However, high temperatures do not correlate with the time of the year

when seasonal infertility is seen, namely late summer and early autumn. If the high temperature was

the main reason for seasonal infertility, one would expect to see reduced litter size rather than

disruption of the whole pregnancy. However, a reduced litter size is not a typical feature of seasonal

infertility (Love, 1978). Use of air conditioning and cooling showers has not solved the problem,

which speaks against high ambient temperature as the underlying factor in seasonal infertility

(Hurtgen and Leman, 1980). High ambient temperatures may, under some circumstances, have an

indirect adverse effect on fertility by reducing the voluntary feed intake of lactating sows leading to

energy imbalance and fertility problems (Prunier et al., 1996). Low ambient temperatures without

compensatory energy feeding in the autumn are more likely to have an adverse effect on fertility

than high temperatures. Low temperatures play an important role in the “autumn abortions

syndrome“ (Almond et al., 1985).

14

1.3.3. Feeding level

According to the conventional feeding recommendations pregnant sows have received a restricted

feeding level (2.0-2.2 kg/day, 13 MJ/kg diet) for the whole dry period. The rationale for restricted

feeding level after mating is based on findings, which have demonstrated less early embryonic

deaths taking place in primiparous sows when held under low feeding levels compared with high

feeding levels (Aherne and Kirkwood, 1985). The conventional feeding recommendations for sows

have been same throughout the year.

Effects of feeding level and the type of housing during early pregnancy (individual stalls vs. group

housing) have been studied in two large piggeries in Australia (Love et al., 1995). In group housing

system, the low feeding level (1.6-2.0 kg, 13 MJ/kg) increased the weaning to oestrus interval and

reduced farrowing rate (50 %) during summer and autumn (Table 1.). In the winter and spring, the

same level of feeding did not have any adverse effect on weaning to oestrus interval and the

farrowing rate remained high (> 85 %). Higher amount of feed (2.5 kg and over 3.6 kg) had a

positive effect on the farrowing rate in summer and autumn. Feeding level did not affect the litter

size. Seasonal infertility was not detected when sows were kept in individual stalls and fed 2.5 kg

per day. This is the best study available, because the comparison between groups was done in the

same piggery, whereas many other studies have compared housing systems and feeding levels

between piggeries, thereby confounding the data.

Table 1. Effect of type of housing and feeding level on farrowing rates (Love et al., 1995). Level of feeding: low = 1.6-2.0 kg/day, moderate = 2.5 kg/day, high > 3.6 kg/day (13 MJ/kg). These feeding levels were applied for the first four weeks of pregnancy. Thereafter a feeding level of 2.5-3.2 kg / day was given for the remainder of pregnancy.

Season Farrowing rate

Group housing Individual stalls

Feeding Low Moderate High Moderate

Summer-Autumn 50.0 69.0 74.1 84.0

Winter-Spring 87.0 - 87.5 87.6

15

It has been hypothesized that the protective effect of higher feeding against seasonal infertility

might be because of increased pituitary LH support to the CL. It has been documented that pituitary

LH-pulses are of lower amplitude and irregular under long day lighting, in other words during the

seasonal infertility period (Paterson et al., 1992; Peltoniemi et al., 1997b). Increased feeding level

might compensate for this seasonal decrease in LH-pulses. Peltoniemi et al. (1997c) studied the

effect of different energy levels and gut fill on LH-secretion in early pregnant gilts, and they found

that the gilts on the highest energy intake had higher LH-pulse amplitude compared with gilts on

lower energy intake. However, there was no effect of energy intake on plasma LH-pulse frequency,

mean concentration, area under the curve or mean nadir. They concluded that the protective effect

of higher feeding level against seasonal disruption of pregnancy appears to be mediated by a

mechanism other than an alteration in LH-secretion. In that study, the energy content of the feed

was increased by adding a fat supplement (soybean oil) to a commercial ration. It did not mimic the

situation in normal farm conditions, where the energy increase is achieved by increasing the total

amount of the basic commercial ration. Increasing the total energy intake through carbohydrates has

more effects on insulin-regulated glucose metabolism than increasing energy intake through fat.

The glucose metabolism, in particular, is known to affect the hypothalamic-pituitary-ovarian

function in the pig (Booth, 1990). More investigations of the mechanism are needed to determine

how higher feeding level in early pregnancy is able to protect against adverse effects of season on

establishment of pregnancy.

1.3.4. Other environmental factors

Pigs are known to have a very sensitive olfactory system. Pheromones secreted by the boar have

been documented to have a strong positive effect on oestrous signs in sows. It has also been shown

that oestrous sows have effects on other sows in the group leading to synchronization of oestrus in

the group (Pearce and Pearce, 1992). Similar synchronization has been reported in the European

wild boar sows (Delcroix et al., 1990). In some circumstances pheromones may have a negative

effect on oestrus, especially in young sows or gilts. In the summer and autumn, when sows appear

more sensitive to all kinds of adverse effects on fertility, negatively effecting pheromones may

abolish oestrus and even disrupt pregnancy. Too many animals in too small a pen can cause stress,

16

and sows or gilts that are of a lower social ranking may lose the pregnancy. In one study, mixing

older pregnant sows in the same pen with younger sows lowered fertility in the young sows (Wilson

and Love, 1990).

In many studies, group housing has led to lower fertility during the seasonal infertility period

compared with individual stalls or pens (Hurtgen and Leman, 1980; Love et al., 1995). This has at

least partly been explained by social stress related to competition for feed and a need to defend

social rank in a group housing situation. However, in the study carried out by Peltoniemi et al.

(1999a), where management and seasonal effects on fertility were examined in 1298 pig farms over

a four-year period in 1992-1996, a clear reduction in farrowing rate was evident during late summer

and early autumn despite 70.2 % of the sows being housed in individual stalls. This study did not

support the earlier studies where individual stalls have been found to protect against reduced

farrowing rate during the seasonal infertility period. It is obvious that some management strategies

essential for achieving good fertility in loose housing system differ from those conducted in

piggeries where animals are individually housed. However, if these differences in management,

especially in a feeding strategy during early pregnancy, are recognized, there is little evidence that

individual housing per se provides better fertility.

1.4. PATTERN OF MELATONIN SECRETION SYNCHRONIZES BREEDING WITH SEASON

1.4.1. Circadian rhythm in melatonin secretion

Seasonality in breeding is a characteristic of animals that inhabit the higher latitudes, where

variations in temperature and food availability are much greater than at lower latitudes. During

evolution, many mammals have opted to use changing photoperiod as the signal to provide

appropriate timing of the breeding season. Photoperiod is a very reliable signal of the season,

compared e.g., with the ambient temperature, which can vary considerably at a particular time of

the year. Light entering the eye stimulates retina and the pulses generated pass to the

suprachiasmatic nuclei (SCN) in the hypothalamus. This synchronizes endogenous melatonin

secretion of the pineal gland. Melatonin secretion increases in the dark and is inhibited by light. The

secretion rhythm is synchronized by ambient lighting conditions, but the melatonin secretion itself

17

is an endogenous circadian rhythm and this rhythm is generated by the SCN (Ebling and Hastings,

1992).

The pattern of circadian rhythm of melatonin secretion synchronizes the beginning of the breeding

season in seasonal breeders, irrespective of whether they are long day or short-day breeders (Turek

and Campbell, 1979). The photoperiod - pineal - hypothalamic pathway in rams has been reviewed

extensively by Lincoln (1992), and the principle features are discussed here. In short day breeders

such as the sheep, the long duration of high melatonin concentrations in the systemic blood

circulation causes an increase in gonadotrophin secretion, which leads to the onset of the breeding

season. The importance of the duration of melatonin secretion has been well documented in the

sheep. Daily 8-hour infusions of melatonin to pinealectomized sheep induce a long day response,

while 16-hour infusions induce a short day response. The daily pattern of melatonin secretion is a

cue in synchronizing the breeding season also in the hamster, which is a long day breeder. The

amount of melatonin, which induced gonadal regression, administered over 10 hours a day to

pinealectomized hamsters had an opposite effect if it was administered over a period of 4 hours

(Carter and Goldman, 1983). The function of the pineal gland is much more complicated than only

to give the information about absolute daylength or nightlength. It also has to provide information

to responsive organs if the days are getting longer or shorter. In this way a given organ can begin its

physiological adjustments in advance, so that it is able to change its function at the time required

(Reiter, 1991). The unusual melatonin profiles of the domestic pig under different lighting reported

in the literature have lead to suggestions that domestication has eliminated or attenuated the

circadian melatonin pattern of the pig (Green et al., 1996). However, this hypothesis has not been

proven.

A change in circadian melatonin secretion is not the only factor synchronizing the breeding season.

It is well known that, for example in the wild boar, the availability of feed affects the onset of the

breeding season. Social interactions are also known to be important in the regulation of seasonality.

The early studies in pinealectomized sheep failed to demonstrate the role of the pineal gland in

synchronizing seasonal breeding because the pinealectomized sheep were kept together with intact

sheep (Roche et al., 1970). Pinealectomized animals were able to respond to the social signals from

intact male and/or female sheep in the absence of photoperiodic information (Wayne et al., 1989).

Melatonin is an extraordinary hormone because its physiological effects do not depend simply on its

presence or absence or the total concentration secreted at a particular time. It seems that the pattern

18

of secretion over a longer period is more important, as discussed earlier. Another interesting feature

of melatonin is that it is able to cause different responses in different species, depending on whether

the animal is a long- or short day breeder. Many details of the action of melatonin are yet to be

clarified but it seems to be a key hormone in synchronizing the beginning of the breeding season

with preferred season along with some modifying factors.

1.4.2. Pineal secretion of melatonin

It has now been clearly documented that a pineal indoleamine melatonin is the substance that gives

the chemical indicator of darkness to the hypothalamus through a complex neural and hormonal

feed back mechanism. The pineal gland secretes quite a few different indoleamines and

neuropeptides, but the treatment of pinealectomized animals with the appropriate pattern of

melatonin induces seasonal effects on reproduction. It has been shown conclusively that melatonin

is the key hormone in interpreting photoperiods (Ebling and Hastings, 1992). SCN works as the

"inner clock", which produces the circadian rhythm of melatonin secretion. This endogenous

circadian rhythm is synchronized by light-dark cycles, and light also directly suppresses melatonin

secretion, provided that the light intensity is high enough or the duration of the light is long enough

(Aoki et al., 1998). The combination of these two factors ensures that nocturnal melatonin secretion

follows the nightlength (Ebling and Hastings, 1992).

1.4.3. The role of melatonin in regulation of hypothalamo – pituitary – gonadal axis

The role of melatonin in synchronizing the breeding season is dependent upon whether the animal is

a long day or a short day breeder. The regulating mechanism is complicated, since the basis of the

melatonin secretion seems to be the same, irrespective of whether the animal is a long day breeder,

a short day breeder or does not express seasonality in breeding. The interpretation of the melatonin

signal must be different in different species. In the short day breeders such as the sheep, the

breeding season is synchronized to the right time of the year by interpreting the nocturnal melatonin

secretion (together with secondary cues). When the duration of melatonin secretion is long enough,

it alters the pulsatile GnRH-secretion in hypothalamus, which in turn leads to increased pulsativity

(or higher amplitude pulses) in LH-secretion from the pituitary. However, the altered melatonin

secretion can only synchronize the timing of the breeding season within a certain timeframe,

19

because seasonal breeding follows a circannual rhythm. This means that even without any

photoperiodical cues reproductively active and passive (refractory) periods follow after certain

interval (for review see Ebling and Hasting, 1992; Lincoln, 1992).

The site of action of melatonin has been localised within the intracranial central nervous system

using active immunisation or systemic passive immunisation against melatonin (Arendt et al.,

1981), and comparing those results with intraventricular injection of an anti-melatonin serum

(Bonnefond et al., 1989). The melatonin antibodies in the systemic circulation have no access to

intracranial areas, and so systemic immunisation failed to have any effects on photoperiodic

responses, whereas the intraventricular injection of an anti-melatonin serum abolished the

photoperiodic response. The locations of melatonin receptors in the brain are different from those

exhibiting GnRH-neurones. The action of the melatonin is thought to be mediated by interneurones,

rather than being directly on GnRH-neurones. In the sheep, the neurotransmitters involved in

carrying the message have been thought to be catecholamines, endogenous opioids and glutamate

(for review see Ebling and Hastings, 1992; Lincoln, 1992).

The frequency of GnRH-controlled LH-pulses in the sheep is much lower in the non-breeding

season as compared with the breeding season (for review see Ebling and Hastings, 1992). There is

no difference in the number of immunocytochemically identified GnRH-neurons between the

breeding and the non-breeding season in the sheep (Lehman et al., 1986) and the amount of stored

GnRH in the hypothalamus is greater in the reproductively inactive state than in breeding season

(Ebling et al., 1987). Thus, inadequate synthesis of GnRH is not the reason for the inactive breeding

state.

There is also evidence for involvement of oestradiol in the control of GnRH-secretion. Karsch et al.

(1984) extensively reviewed the role of oestradiol in seasonal breeding. Some parts of the review

are discussed in this section. In this model GnRH-pulses are thought to be controlled by the GnRH-

pulse generator, the activity of which depends on inhibitory feedback of oestradiol. The change in

the photoperiod and in the melatonin secretion is thought to affect the sensitivity of the GnRH-pulse

genarator to the oestradiol feedback. In short day breeders, the long duration of melatonin secretion

reduces the negative feedback of oestradiol on the pulse generator and allows increased GnRH-

secretion, which leads to the commencement of the breeding season.

20

It is likely that the seasonality of reproduction in the pig is dependent on photoperiodism and

melatonin secretion, which in turn controls release of GnRH either through neurotransmitters and

activation/inhibition of particular neural pathways, as discussed earlier, or neurotransmitters control

the sensitivity of GnRH-pulse generator to oestradiol feedback. Most likely, both mechanisms are

involved in controlling GnRH-secretion (Robinson et al., 1985). The work by Peacock (1991)

demonstrated that the oestradiol feedback theory could not alone explain the seasonal effects on

reproduction. The photoperiod altered pituitary LH-secretion even if the oestradiol was absent

(ovariectomised gilts).

1.4.4. The melatonin pattern in the pig

The daily melatonin pattern is well established in many species. However, the melatonin profiles in

the pig have remained unclear, despite many studies. Results of the first experiments were

controversial, and it is impossible to draw any common conclusion from them. Most of the studies

were not able to demonstrate any changes between melatonin concentrations during the scotophase

and photophase (Minton et al., 1989; Diekman et al., 1992; Diekman and Green, 1997). Some

authors have reported a nocturnal rise in melatonin only in a small propotion of the animals studied

or only under particular lighting conditions (McConnell and Ellendorff, 1987; Griffith and Minton,

1992; Green et al., 1996; Bollinger et al., 1997). There are also some reports of an exceptional rise

in melatonin during illumination (Peacock, 1991). There is only one study, which has shown a

consistent nocturnal rise in melatonin in domestic pigs under different lighting regimens (Paterson

et al., 1992).

Minton et al. (1989) was not able to demonstrate a nocturnal rise in melatonin concentrations in

pubertal crossbred boars exposed to a 16-hour photoperiod in controlled climate rooms. It was

concluded that melatonin was not secreted in a rhythmic fashion under long photoperiods.

However, the basal melatonin concentrations detected in Minton’s study during the photophase

ranged between 10 and 50 pg/ml, which is a high basal concentration compared with other studies

(Paterson et al., 1992; Mack and Unshelm, 1997). Higher basal concentrations are probably due to a

crossreaction of the antiserum (Guildhay®) with some component(s) other than melatonin or non-

specific serum/plasma effects commonly observed in direct RIAs (Klupiec et al., 1997). Another

study carried out by Diekman et al. (1991) reported extremely high basal melatonin concentrations

in prepubertal gilts without any changes between the dark and light period. In that study,

21

crossreactivity (antisera R1055) was a possible explanation for the high basal concentrations and for

a lack of the circadian melatonin profile, even though the assay was an extraction assay. The study

in question employed chloroform extraction, and it is possible that chloroform also extracted the

substance(s) causing the crossreaction. Furthermore, in Diekman’s study, or at least in some parts of

it, one reason for the lack of nocturnal increase in melatonin might have been a relatively low light

intensity (80-100 lx) used during photophase. Mack and Unshelm (1997) have shown that 50 lx

intensity during illumination did not lead to an increase in melatonin levels during the scotophase

(measured from saliva). A similar problem with crossreactivity seems evident in other experiments

reported on another occasion by the same group (Diekman and Green, 1996). McConnell and

Ellendorf (1987) reported a nocturnal rise in melatonin in three sows out of four under a 12 L : 12 D

lighting regimen. In three sows during the dark phase, concentrations in individuals increased two-

to fivefold over the peak light phase values. When the lighting regimen was changed to 16 L : 8 D

or 8 L : 16 D, the melatonin surge was abolished. The assay was a direct RIA and lacked sensitivity.

The sensitivity of the assay was 16 pg / ml, which is higher than average nocturnal melatonin

concentrations in pig serum or plasma (Paterson et al., 1992). Green et al. (1996) found an increase

in nocturnal melatonin concentrations only in some of the animals and only under certain (close to

12 L : 12 D) lighting regimens. They used a direct RIA utilising an antiserum (Guildhay®) that led

to the high basal melatonin concentrations. Crossreactivity and insensitivity (5 pg/ml) most

obviously also affected the results published by Bollinger et al. (1997). They found a nocturnal rise

in melatonin in only 16 % of the gilts included in the experiment, although they detected a nocturnal

rise in 86.2 % of the ewes housed in the same room. They used an extraction assay but the assay

utilised an antiserum (Guildhay®), which has been shown to crossreact with some small molecular

weight compound(s) present in pig plasma or serum (Klupiec et al., 1997). Measuring melatonin in

the sheep serum does not require such a sensitive assay as does measuring melatonin in pig serum

because the scotophase melatonin levels are much higher in sheep than in pigs.

Crossreaction of the antiserum, non-specific plasma/serum effects in direct RIAs, inappropriate

extraction in extraction based RIAs and insensitivity of the assay are likely to be the most common

reasons for failure to reveal a melatonin profile in the pig. Klupiec et al. (1997) have shown that a

commonly used antiserum (Guildhay®) has some limitations in direct RIAs because of its

crossreactivity with some plasma substance(s) other than melatonin and a non-specific

plasma/serum effect. It seems to be questionable that direct RIAs are adequate for measuring

melatonin in pig plasma or serum because of relatively low melatonin concentrations present

demanding a high sensitivity of the assay. Furthermore, non-specific plasma/serum effect varies

22

considerably between samples from the same individual making correction of plasma/serum effect

difficult (Klupiec et al., 1997). The direct RIAs usually measure higher baseline melatonin

concentrations than extraction assays even if the assay is utilizing highly specific antiserum (G280).

This indicates that the non-specific plasma/serum effect does not exist only with Guildhay

antiserum but also to some extent with other antisera. However, it may be possible to measure

reliably pig melatonin also by employing a direct RIA, if correction of non-specific plasma/serum

effect is done properly and if the validation of the assay includes comparing results with an

extraction assay and/or mass-spectrometry, as shown by Paterson et al. (1992). It seems that an

extraction assay where a plasma or serum effect is eliminated by extraction and reconstituting of the

samples, together with a highly specific antiserum, is the most reliable RIA-method to analyze

melatonin in pigs.

1.4.5. Effects of light intensity on melatonin

The intensity of light has a decisive effect on melatonin secretion. If the intensity of light is not high

enough during the photophase, a rise in plasma melatonin will not take place during the dark phase.

In other words, the intensity of the light during the photophase has to be sufficient to suppress the

melatonin secretion, before there can be any nocturnal rise in melatonin. Griffith and Minton (1992)

showed that 113 lx during the photophase was not sufficiently intense to cause a rise in plasma

melatonin during the dark phase, whereas 1783 lx during the photophase led to a clear rise in

plasma melatonin during the dark phase. These results have to be interpreted with care because of

the difficulties in the assays discussed in the previous section. The exact limit of light intensity

required during the light period to cause high melatonin concentrations during the dark period in the

pig is not known, but some experiments have suggested that it might be somewhere between 200

and 300 lx (Love et al., 1993). If there is not enough light during the photophase, sows might be

unable to recognize the difference between night and day, which is known to be an essential part of

melatonin signals in regulating reproduction (Karch et al., 1988). In addition, it has been suggested

that pigs might be unable to react to rapid changes in lighting conditions (Paterson et al., 1992). It

seems evident that both naturally changing lighting conditions and insufficient artificial lighting

regimens can detrimentally affect reproductive performance of the pig.

Early studies in humans demonstrated that relatively high light intensities were needed to acutely

suppress melatonin secretion. Lewy et al. (1980) reported that 500 lx was not sufficient to suppress

23

melatonin secretion in humans, whereas 1500 lx caused an intermediate suppression and 2500 lx

caused a total suppression. However, Hashimoto et al. (1996) concluded that the threshold of the

light intensity to acutely suppress the nocturnal melatonin secretion in humans was somewhere

between 200 and 500 lx. Humans have been considered to be very insensitive to changes in lighting

conditions but the latest reports of the effect of the illumination on melatonin in humans have

demonstrated that humans are much more sensitive to light than was suspected initially. Boivin et

al. (1996) revealed that light intensity of 180 lx significantly shifts the human circadian pacemaker.

These values are quite similar to the light intensity needed in pigs to acutely suppress melatonin

secretion.

There seems to be a relationship between the light intensity and the illumination time in suppressing

melatonin secretion both in humans and in pigs. In indoor piggeries, controlled by artificial lights,

light intensity and the duration of photophase are the critical factors to consider when planning the

lighting regimens of the different units in piggeries. This is also one of the basic questions to be

considered when trying to prevent seasonal infertility problems in outdoor piggeries. Paterson and

Pearce (1990) showed that if gilts remained isolated from the boars, the mean age of the gilts at

puberty was 209 days when kept under short day lighting conditions and 214 days when kept in

long day regimens. Diekman et al. (1991) revealed that daily oral administration of 3 mg of

melatonin caused a reduction in age of puberty in gilts compared with control gilts. These studies

clearly demonstrated that the photoperiod and melatonin had an effect on reproduction in pigs.

However, details of the role of photoperiod (and melatonin) in controlling the reproductive

performance in pigs needs to be elucidated before different lighting regimens can be introduced to

commercial piggeries or melatonin supplementation can be used.

1.5. RECOGNITION OF PREGNANCY

The following part of the review will concentrate on the recognition and establishment of

pregnancy, since disruption of early pregnancy leading to a reduced farrowing rate seems to be the

most significant manifestation of seasonal infertility, as discussed in earlier sections. Events related

to the luteolysis in cyclic animals, embryos’ ability to prevent luteolysis and the maternal response

to the embryonic signals will be discussed in the following sections.

24

1.5.1. PGF2α secretion during the oestrous cycle

It was shown in the seventies that PGF2α is the hormone causing luteolysis in the sow during the

normal oestrous cycle. The role of PGF2α in controlling the oestrous cycle in the pig has been

reviewed extensively by Bazer et al (1982). The endometrium is the source of PGF2α since its

destruction or hysterectomy before day 14 of the oestrous cycle extends the function of CL beyond

30 days (Melampy and Anderson, 1968). In the cyclic sow, PGF2α secretion starts to increase after

day 12 of the oestrous cycle, and at about day 15 PGF2α concentration reaches its peak value when

measured in the utero-ovarian vein (Moeljono et al., 1977). PGF2α concentration falls to its basal

level between days 18 and 19 of the oestrous cycle. The corpora lutea of the sow appear to be

refractory to the luteolytic effects of PGF2α before day 11 of the cycle as demonstrated by Hallford

et al. (1975). They injected gilts intramusculary with PGF2α at 12-hour intervals on days 4 and 5 of

the cycle. The total dose of 80 mg of PGF2α did not affect oestrous cycle length. However, when the

same treatment protocol was used on days 12 and 13, the cycle length was reduced. Krzymowski et

al. (1976) infused a total of 2 mg of PGF-2α into the anterior uterine vein of sows over a 10-hour

period on either day 6, 8, 10, 12, 14 or 15 of the cycle. The PGF2α did not affect CL weight or

plasma progesterone levels when given on days 6, 8 or 10, but luteolysis was initiated in sows

treated on days 12, 14 or 15. These data clearly demonstrate that CL of the pig are refractory to the

luteolytic effects of PGF2α during the first eleven days of the oestrous cycle. Increased CL

sensitivity to PGF2α after day 11 has been associated with an increase in the number of PGF2α

receptors in the luteal cells (Gadsby et al., 1988). This increase corresponds to the time when LH

begins to dissociate from its membrane receptors (Henderson and McNatty, 1975) and CL become

dependent on a continual pituitary LH support (Anderson et al., 1967).

1.5.2. Embryonic signals

After fertilization has occurred in the oviduct of the sow, embryos move into the uterine horns at the

4-cell stage, about 60-72 hours after the onset of oestrus. Embryos reach the blastocyst stage by day

5 and hatching from the zona pellucida occurs between days 6 and 7. Thereafter, embryos grow to

2-6 mm diameter on day 10 and rapid elongation to a threadlike organism, 10 – 100 cm in length

(Anderson, 1978; Geisert and Yelich, 1997), takes place on days 12 – 16. These filamentous

25

organisms lie in a close contact with the endometrium and are able to give a local signal of their

presence by oestrogen secretion. Van der Meulen et al. (1988) has examined the critical time for the

presence of viable embryos in the uterus in the pig. They flushed blastocysts from the both uterine

horns of gilts on days 10, 11, 12 or 13 and observed effects on the interval to the subsequent oestrus

and progesterone profile. In mated non-pregnant control gilts flushing did not affect progesterone

profile or the cycle length. Flushing of the blastocysts on day 10 resulted in a cycle with a normal

progesterone profile and a normal length. However, flushing on days 11, 12 or 13 resulted in a

normal cycle or in maintenance of the CL for 3-13 days longer than a normal lifespan of CL (in

only one of 6 gilts flushed on day 11). The elevated progesterone levels and prolonged oestrous

cycles were 22.0 ± 1.2 (day 11), 24.8 ± 1.4 (day 12) and 26.3 ± 2.3 days (day 13). There was a large

variation in the stage of the development between embryos flushed on the same day. Only those

gilts in which the blastocysts were ≥ 8 mm or filamentous showed the prolonged oestrous cycle.

The authors concluded from these results that the first signal for maternal recognition of the

pregnancy is generated on day 12 and that blastocysts ≥ 8 mm or filamentous are required for

prolongation of CL function for three or more days. They also concluded that a second signal seems

necessary to maintain CL for the whole pregnancy. It is also an interesting finding that development

of embryos during the first twelve days of pregnancy varied considerably and it contributed to the

embryos’ capability to produce the signal. It has been well documented that endometrial secretions

(histotrophe), which are necessary to nourish embryos, are progesterone-induced. Abnormally low

progesterone concentrations in early pregnancy might interfere with the growth of embryos, and

hence their ability to signal their presence early enough. During the first twelve days of pregnancy

when CL are still working autonomously, subnormal progesterone concentrations may be related to

the feeding level provided to sows and especially to gilts. The energy level of feeding has effects on

progesterone concentrations through metabolic clearance of progesterone rather than on

progesterone secretion from CL (for review see Foxcroft, 1997). After day 12 when CL become

increasingly dependent upon pituitary LH (luteinizing hormone), inadequate LH secretion might

cause subnormal progesterone concentrations and failure of embryos to produce the second signal

needed for continuation of pregnancy. The absence of the second signal might play an important

role in the early disruptions of pregnancy seen typically during the seasonal infertility period, since

season has been shown to affect LH secretion (Paterson et al., 1992; Peltoniemi, 1994).

Several investigators have suggested that embryos give the signal, discussed earlier, by producing

oestrogen. Perry et al. (1973) first described the physiological role of oestrogen in the maintenance

26

of pregnancy in pigs and these results have been confirmed and expanded later by others (Gadsby et

al., 1980; Fischer et al., 1985). Conceptuses of the pig produce oestrogens between days 11 and 18

in a biphasic manner. Uterine concentration of oestrogen increases for the first time on day 11 or

12. At this same time conceptuses go through rapid elongation from round shaped blastocysts to

long filaments. This stage is followed by a decrease in uterine oestrogen content between days 13

and 14 before a new increase after day 14 (Stone and Seamark, 1985). An increase in uterine

content of oestrogen does not occur before day 15 in non-pregnant gilts or sows. Since the capacity

of endometrium to synthesize oestrogen in the early pregnancy is low, the uterine content of

oestrogen reflects the embryos’ oestrogen production (Fischer et al., 1985). It has been

demonstrated that pig blastocysts are very efficient in synthesizing oestrone and oestradiol from

different precursors including progesterone (Fischer at al., 1985). These “free oestrogens” produced

by embryos are most likely acting only locally in the endometrium, because they are conjugated to

biologically inactive sulphated forms in the endometrium as they move into systemic circulation

(Heap and Perry, 1974), but the possibility of the systemic effects can not be rejected.

Experiments where oestrogen has been injected into cyclic gilts have provided further evidence of

oestrogen being the signal that embryos produce to ensure prolonged CL function. Pusateri et al.

(1996a) carried out on experiment where they determined minimal requirement for exogenous

oestradiol-17β to induce either a short or long pseudopregnancy in cycling gilts. They carried out

five experiments where oestradiol-17β was injected into the cyclic gilts in different combinations

between days 11 and 25 post oestrus. The results demonstrated that short pseudo pregnancy (SPP)

can be induced in gilts by oestrogen treatment that is effective over a 2-day period between days 11

and 15 of the oestrous cycle. The most consistent induction occurred when the 2-day period

encompassed day 12 or 13, or both. The long pseudo pregnancy (LPP) was most effectively induced

by the injection regimen which begun on day 12 and continued with daily injections to between

days 17 and 19. None of the gilts that received daily injections from day 12 to day 16 exhibited

LPP. The duration of oestradiol treatment required to induce pseudopregnancy was clearly

demarcated between SPP and LPP, being 23-35 days in SPP and over 50 days in LPP. The authors

therefore suggested that maternal recognition of pregnancy might occur in two phases with

continuous exposure to oestradiol being required from day 12 to 17-19 of pregnancy. In this

experiment they used oestradiol-17β because of its shorter half-life compared with oestradiol

valerate or oestradiol benzoate. Intramuscular injection of oestradiol-17β at dosages of 4 to 17

µg/kg resulted in elevated serum oestradiol concentrations for 16 to 18 hours whereas peripheral

27

oestradiol remained elevated for at least 72 hours after a single injection of 5 mg oestradiol valerate.

After daily injections of 5 mg of oestradiol valerate for five days, serum oestradiol concentrations

remained above the preinjection baseline for five to seven days. Earlier studies done with oestradiol

valerate or benzoate did not provide accurate indications of the exact time when the oestrogen

signal has to be initiated to cause pseudopregnancy because of the long residual effects of these

compounds.

1.5.3. Oestrogen signals alter PGF2α secretion during pregnancy

There is common agreement that embryonic oestrogen signals are the factor causing the

maintenance of CL, but there is not unanimity of opinion on the mechanism of action. Several

mechanisms have been proposed and they include both systemic and local effects of oestrogen. It

appears that oestrogen is capable of stimulating progesterone secretion of CL possibly through an

increase in luteal LH (Garverick et al., 1982). However, the inability of embryos to maintain the

pregnancy when a large portion of the uterus is unoccupied or when the pregnancy appears only in

another horn of the uterus does not support the theory of systemic effects (see Dziuk, 1985). It has

also been proposed that oestrogens exert their effects by reducing endometrial release of PGF2α

(Guthrie and Rexroad, 1981). Although several studies have indicated that in vitro PGF2α

production of pregnant gilts’ endometrium or oestrogen treated cyclic gilts’ endometrium is

diminished compared with non-treated cyclic gilts’ endometrium, the luminal concentration of

PGF2α is greater on days 14-20 in oestrogen treated than in cyclic gilts (Frank et al., 1977). The

PGF2α concentrations are even greater in pregnant animals during this time (Zavy et al., 1980) that

might reflect embryonic prostaglandin production. The elevated PGF2α concentrations found in the

uterus of pregnant animals support the most widely accepted theory, proposed by Bazer and

Thatcher (1977), based on reorientation of endometrial PGF2α secretion. The secretion of PGF2α

during the mid and late luteal phase in the cyclic animals is mainly toward the uterine venous

drainage (endocrine direction) causing the regression of CL and allowing follicular development to

progress and oestrus to occur. In pregnant sows or gilts, oestrogens produced by the embryos on

days 12 and 18 alter the PGF2α secretion toward the uterine lumen (exocrine direction) without

reaching the systemic circulation, and the regression of CL does not take place. There are a number

of studies which have shown that PGF2α concentration in the utero-ovarian vein is higher in cyclic

animals between days 12 and 17 compared with pregnant or oestrogen treated animals (Frank et al.,

28

1977; Moeljono et al., 1977). It has also been shown that peripheral concentrations of PGF-

metabolites are higher in cyclic gilts than in pregnant gilts on days 12-13 and 15-17 reflecting the

reorientated PGF2α secretion into the uterine lumen in pregnant gilts (Shille et al., 1979).

The exact mechanism by which oestrogen is able to reorientate the PGF2α secretion is not known

but endometrial calcium release would appear to be involved. There is an increase in uterine

calcium concentration about twelve hours after the embryonic oestrogen signal (or after exogenous

oestrogen) and a re-uptake of calcium by the endometrium occurs about twelve hours after the

concentration of calcium in the uterine lumen reaches the maximum level. This oestrogen induced

endometrial calcium secretion and re-uptake has been associated with the reorientation of PGF2α

secretion. More evidence of calcium being involved in alteration of endometrial prostaglandin

release has been provided by Gross et al. (1990). This study demonstrated that treatment with

calcium ionophore shifts orientation of PGF2α secretion to the luminal side during in-vitro

perfusion. Reorientation of PGF2α secretion to the luminal side also involves interaction between

oestrogen and prolactin, since neither oestrogen nor prolactin alone causes the reorientation (Gross

et al., 1990). There is no increase in prolactin concentration in early pregnancy, but the number of

prolactin receptors increases between days 14 and 30 (Dehoff et al., 1984). There are also some

other factors, like embryonic prostaglandin-E (PGE) and conceptus secretory proteins (CSP), which

might have an effect on the reorientation of endometrial PGF2α secretion.

It appears that two embryonic oestrogen signals are needed between days 11 and 18 to extend the

function of CL until the end of the normal pregnancy period. Pusateri et al (1996b) demonstrated

that in all oestrogen treated cyclic gilts serum prostaglandin-metabolite concentrations peaked

always 4-6 days before the post-treatment oestrus, independent of the treatment response (SPP or

LPP). This indicates that the second signal is needed around pregnancy day 18 to ensure the

continuation of the endometrial PGF2α exocrine secretion pattern. It seems that no further signals

are needed after that, since two appropriately timed exogenous oestrogen signals are able to cause

the LPP of the normal pregnancy length (Pusateri et al., 1996b).

1.5.4. Conceptus secretory proteins (CSP)

In sheep and cattle, maternal recognition of pregnancy is thought to be initiated by the conceptus

secretory proteins, ovine trophoblast protein-1 (oTP-1) and bovine trophoplast protein-1 (bTP-1),

29

which are members of interferon-alpha (IFN-α) family. In these species, trophoplast proteins seem

to prevent the ability of oestrogen and oxytocin to stimulate endometrial PGF-2α secretion that

initiates the regression of CL. Cross and Roberts (1989) showed that IFN production by early

conceptuses occurs also in pig, between days 11 and 17, but the role of this IFN is unknown. A

large body of evidence suggests that in the pig the events related to prevention of luteolysis are

triggered by embryonic oestrogen secretion, as discussed earlier, and not by the IFN-secretion as in

sheep and cattle. However, it is likely that IFN plays an important role in establishment of

pregnancy also in the pig, perhaps independent on effects on luteal life span. Other potential roles of

conceptus IFN include regulation of the local maternal immune system, regulation of endometrial

secretions and induction of antiviral protection against infection (Cross and Roberts, 1989; La

Bonnardière, 1993). However, at this stage the potential roles of the IFN in recognition and

establishment of pregnancy in the pig are speculative and further investigations are needed.

1.6. MATERNAL RESPONSE

1.6.1. Formation of CL and luteotrophic role of LH

After ovulation the newly formed corpora lutea continue to develop autonomously independent of

pituitary support until day 15, when PGF2α secretion reaches its peak value, and if pregnancy has

not been established, regression of CL are initiated. Sammelwitz et al. (1961) showed that large

amounts of progesterone injected into pregnant pigs from the time of ovulation until days 10 to 13

of gestation did not prevent the formation of CL. Injections of progesterone beyond days 12 to 16 of

pregnancy resulted in the rapid and complete regression of CL. Large amounts of exogenous

progesterone are thought to cause a complete blockade of pituitary LH secretion through the

negative feed-back mechanism. Brinkley et al. (1964) reported similar results where formation of

CL after ovulation was not affected by the pituitary LH-blockade caused by exogenous

progesterone lasting from seven to ten days. These two experiments provide evidence that CL of

pigs are able to function autonomously until day 15, but after that the pituitary LH-support appears

to be necessary for continuation of pregnancy.

The work of Anderson et al. (1967) provides further evidence that establishment of pregnancy

needs firstly the reorientation of endometrial PGF2α secretion and secondly a certain requirement

for pituitary secretion of LH. In their study hypophysial stalk transection or hypohysectomy was

30

performed on pigs on the 1st day after oestrus or mating. Corpora lutea developed following the

hypophysial stalk transection, but by day 12 they were smaller in unmated stalk-sectioned animals

than those in mated stalk-sectioned, as well as being smaller than those in day 13 unmated or mated

control pigs. By day 20 CL regressed in both mated and unmated stalk-sectioned animals. Living

embryos were present in the uterus of mated stalk-sectioned pigs at days 12 and 16, but only

necrotic embryos were found at day 20. In hysterectomised animals with persisting CL, luteal

regression begun within four days after hypophysial stalk transection and was completed by day 12.

Oestrogen maintained the CL in intact control animals but not in stalk-sectioned pigs. Desiccated

porcine anterior pituitary or human chorionic gonadotrophin maintained CL in hypophysectomised

and hysterectomised pigs. Collectively these data confirm the autonomous function of the CL about

15 days following the ovulation and the essential role of pituitary LH in the maintenance of CL and

continuation of pregnancy after CL have lost their autonomous function.

The autonomous function of CL appears to disappear gradually at the same time as the PGF2α

secretion increases. Pituitary LH has to take a supporting role for CL and this support has to

continue until around pregnancy day 29 as demonstrated by GnRH-agonist implantation of gilts in

different stages of pregnancy (Peltoniemi, 1994). GnRH-agonist implants abolish the LH-surges

and if the implantation occurred on day 14 or 21 of pregnancy all the gilts aborted. However, if

implantation occurred on day 29, 50 % of the gilts were able to maintain the pregnancy. These

results indicate increasing independence of CL on pituitary LH when pregnancy progresses.

In the early stage of pregnancy, prolactin concentrations measured in pregnant gilt serum are very

low (Peltoniemi, 1994). However, prolactin is known to be the primary luteotrophin beyond day 60

of pregnancy in pigs (Szafranska and Tilton, 1993). It is not well understood how maintenance of

CL are ensured between days 30 and 60, when pituitary LH is no longer critical to the continuation

of pregnancy and before prolactin has taken a role of primary luteotrophin. This might be some kind

of transient period that is not extremely sensitive to the lack of either of these luteotrophins.

In conclusion, pituitary LH appears to be essential to the continuation of pregnancy between days

15 and 29. At that time, corpora lutea gradually lose the autonomous function and become

dependent on pituitary luteotrophin. After day 29, prolactin gradually takes its place as a primary

luteotrophin.

31

1.6.2. Progesterone in early pregnancy

Corpora lutea are the main source of progesterone in pigs throughout pregnancy, and functional CL

are needed for the entire length of pregnancy. The maintenance of CL is ensured by the prevention

of PGF-2α secretion into ovarian circulation and pituitary luteotrophic support as discussed earlier.

Progesterone induces the secretion of different kind of proteins like uteroferrin and retinol binding

proteins (RBP), which act by transporting vitamins and minerals to the developing conceptuses.

There is a wide consensus that progesterone is the key hormone inducing the secretion of these

proteins essential for development of conceptuses, but the effect of embryonic oestrogen on

regulation of this secretion is unsure. Trout et al. (1992) reported that while progesterone

concentration is sufficient to allow the secretion of RBP, oestrogens secreted by embryos have a

major influence on the amount and timing of RBP secretion. In contrast to this study, Vallet et al.

(1998) concluded from their experiments that conceptuses do not influence secretion of either total

protein or RBP during pregnancy and that the onset of secretion of these proteins is controlled by

progesterone alone. However, progesterone concentration seems to be critical during and after the

time CL begin to be dependent on pituitary LH. This might be because of the effect of progesterone

on secretion of nutrients to embryos, effect of progesterone and embryonic oestrogens together on

uterine secretion or effect of progesterone on embryos’ ability to produce an oestrogen signal to

prevent the luteolysis.

In normal pregnancy, there is even 30 % decrease in maternal progesterone around day 14 of

pregnancy, and this decrease has been proposed to be due to metabolism of conceptuses to produce

the oestrogen signal (see Bazer et al., 1982). Vallet et al. (1998) found that increase in progesterone

induced protein secretion always preceded the change in embryonic oestrogen secretion. They

proposed that the uterine protein secretion (induced by progesterone) might control conceptus

oestrogen secretion. Down-regulation of epithelial progesterone receptors in uterine epithelia occurs

after day 10 of gestation but receptors remain in deeper layers and might control endometrial

secretion throughout the pregnancy (Geisert et al., 1994). Even if all details of the role of

progesterone in early pregnancy are not clear it seems to be related to the recognition of the

pregnancy through the embryonic oestrogen signal. Progesterone may be the primary precursor of

oestrogen produced by embryos or progesterone may control embryos’ growth by protein secretion

and their ability to produce oestrogens when needed. However, the changes in maternal

progesterone concentrations at the same time as down-regulation of uterine epithelial progesterone

32

receptors and need for the embryonic oestrogen signal, are all part of critical events associated to

establishment of pregnancy.

33

2. THE AIMS OF THE STUDY

The primary aim of the study was to investigate the endocrinological basis of seasonal infertility

concentrating on the photoperiodic – melatonin transduction. The secondary aim of the work was to

study association between early disruption of pregnancy (EDP) and lowered farrowing rate in the

summer – autumn as well as the mechanism of EDP.

1) The first experiment where European wild boars (gilts and young boars) and domestic gilts

were studied in the same trial was aimed to give a basis to all the rest lighting experiments,

as there was a fundamental lack of information regarding the lighting – melatonin

transduction in the pig.

2) The other two lighting trials were aimed to give more detailed information of the ability of

the pig to respond to abrupt changes in lighting regimens and effects of the light phase light

intensity on the dark phase melatonin response.

3) The association between EDP and seasonally decreased farrowing rate was studied by a

field trial in a commercial herd having a constant history of seasonal infertility.

4) The endocrinological mechanism of the early disruption of pregnancy was investigated. The

aim was also to develop an experimental model of the mechanism for further studies. This

was done by active and passive GnRH-immunization trials.

34

3. MATERIALS AND METHODS

The experiments of the study were divided into two main categories; lighting – melatonin

transduction trials (papers I, II, III ) and trials concerning early disruption of pregnancy (papers IV,

V). This division was justified by the fact that seasonal changes in the circadian secretion pattern of

melatonin are behind the seasonally altered reproduction performance, whereas early disruption of

pregnancy appears to be the most important single manifestation of seasonal infertility in practise.

Both areas of the study were further divided into a field trial and strictly controlled laboratory

investigations.

Materials and methods that were used only in one of the studies or were similar in all the studies are

described here in details. Those materials and methods that varied between studies (number of

animals, management, feeding) are described in details in the paper concerned, and only general

guidelines are given in this section.

3.1. ANIMALS

The trial where melatonin secretion of wild boars was compared with that of domestic pigs (paper I)

used pure bred European wild boars (young boars and gilts) and crossbred (Yorkshire x Landrace)

non-pregnant gilts. The European wild boars were from a wild boar farm where the animals lived in

a semi natural environment. The population of the wild boars originated from animals captured

from the nature.

In the light intensity experiment (paper III), eight-month old Yorkshire-Landrace and Yorkshire-

Landrace-Duroc hybrid non-pregnant gilts were used.

Pre-pubertal male grower pigs were used in the study where ability of pigs to respond to abrupt

changes in lighting was investigated (paper II). The body weight of 40 – 45 kg was used as an

inclusion criterion.

The field trial investigating association between seasonally lowered farrowing rate and early

disruption of pregnancy (paper IV) used all the animals (gilts and sows) in production in a

35

commercial 160-sow unit. Most of the animals were hybrids (Yorkshire x Landrace) but

occasionally there were some purebred Yorkshire or Landrace gilts/sows.

The immunization experiments (paper V) used pregnant multiparous crossbred (Yorkshire x

Landrace) sows and pregnant crossbred (Yorkshire x Landrace) gilts.

3.2. BLOOD COLLECTION

3.2.1. Jugular vein catheters (I, II, III, V)

The frequent blood samples from domestic pigs were collected via jugular vein catheters inserted

non-surgically via an ear-vein (Peacock, 1991). Pigs were restrained using a soft rope snare and the

surface of the ear was cleaned and disinfected. A 14-gauge intra-venous catheter placement unit

(Jelco®, Johnson & Johnson, Chiyoda-Ku, Tokyo) was inserted into the ear-vein and approximately

50 cm of vinyl tube was threaded through the catheter into the vein and another 50 cm remained

externally. The vinyl tube (Dural Plastic and Engineering, Auburn, NSW) had an inner diameter of

1.0 mm and an outer diameter of 1.5 mm. A blunted 18-gauge needle hub was fitted into the end of

the vinyl tube to allow blood collection and a stopper was used in the needle hub to prevent a

backflow. To ensure the patency of the catheter, it was filled with sterile EDTA-saline (9 g NaCl

and 9 g di-potassium EDTA in 1 l of sterile water). A zippered pouch placed on the pig's neck was

used to store the external part of the vinyl tube and the stoppered needle hub. The catheter was

bandaged on the ear using elastic adhesive bandage (Elastoplast®, Smith and Nephew, Clayton,

Vic.) and the ear was then strapped to the neck to prevent the pig from dislodging the catheter. Prior

to collecting a blood sample, the first 2 ml drawn up was discarded to prevent contamination of the

sample with saline. The sample (approximately 8 ml) was drawn into a 10 ml syringe and

immediately emptied either into 10 ml plastic tubes containing 150 µl EDTA solution (250 mg/ml)

and 0.25 g spinning granules (Kwik Spin®, Disposable Products, Ridleyton, S.A.) or into plain

glass tubes.

3.2.2. Single blood samples (IV, V)

36

When single blood samples were needed, they were taken as a vein puncture of V. jugularis or V.

saphena medialis.

3.2.3. Saphenous arterial catheters (I)

The frequent blood samples of the wild boars were collected by the non-surgical method described

by Tast et al., 2000. Animals were sedated with a combination of medetomidin (Domitor® 1 mg /

ml, Orion-Farmos, Finland), zolazepam and tiletamine (Zoletil forte vet® 50 mg / ml + 50 mg / ml,

Virbac Laboratories, France). The dose of Domitor® was 0.025 mg / kg and of Zoletil® 5 (2.5 +

2.5) mg / kg. An air-pressure sedation gun with self-emptying shooting syringe and needle (both

from Dan Inject®, Denmark) was used to inject the sedatives intramuscularly. With the animal in

lateral recumbancy, the hair was cut from the medial surface of the hind leg (Musculus gracilis) and

the Arteria saphena medialis was palpated. The skin was washed with water and disinfected with

clorhexidin (Hibiscrub®, Zeneca Pharma, Norway). A small incision (2 mm) was made in the skin

with an injection needle. An Intraflon® 14 G catheter (Intraflon 2® 14 G, Vygon, France) was

inserted into the Arteria saphena medialis, and the inner metallic needle removed. A teflon tube

(Prevett, Finland, inner diameter 0.9 mm and outer diameter 1.5 mm.) was placed into the artery

through the Intraflon® 14 G catheter. Approximately 40 cm of this tube was inserted into the artery,

and the Intraflon® 14 G catheter was withdrawn. The Intraflon® 14 G catheter was attached by a

plastic clue to the outer end of the tube, where it supported the connection between the tube and the

19 G injection needle that had earlier been attached to the end of the tube. To ensure that the tube

was functioning properly, 10 ml of blood was aspirated, and 1 ml of diluted heparin (Heparin 5000

IE / ml®, Lövens, Denmark) was injected into the tube to prevent coagulation. The hair was cut

from the ventral side of the hind limb up to the back, and the tube was covered by tape

(Elastoplast® 6 cm, Beiersdorf AG, Spain). Interrupted skin sutures (Suturamid® 0, Ethicon,

Germany) at 5 to 10 cm intervals secured the tape to the skin .A blunted hypodermic needle hub (19

G) was fitted to the end of the tube, with a stopper to prevent back flow. The tape covered the

needle hub until the sampling begun. When sampling started, a small hole was made in the tape

allowing the 19 G injection needle to exit. The blood sample was drawn into a 10-ml syringe and

emptied immediately into a plain glass tube. The samples were refrigerated immediately and serum

was separated by centrifugation on the following day. The serum was stored at -20°C until analysis.

37

3.3. GNRH-IMMUNIZATION

3.3.1. Active immunization (V)

A commercially available vaccine (Improvac®, CSL Limited, Australia) was used to actively

immunize sows against GnRH. The first immunization was done on the day of farrowing injecting 2

ml of the vaccine subcutaneously. The booster was given 10 or 20 days after mating.

3.3.2. Passive immunization (V)

A number of cull sows and finisher pigs were immunized with Improvac® two or three times at

four-week intervals with the last vaccination given 14 days before culling to obtain GnRH-immune

pig serum. Blood was collected at slaughter into plastic containers and allowed to clot at room

temperature over night. On the next day, blood was centrifuged and serum harvested. Serum was

stored at -20ºC until used. Before use, serum was warmed to +38ºC. On day 12 post mating, the

gilts were infused intravenously with 1 L of pig serum containing GnRH-antibodies.

3.4. HORMONE AND ANTIBODY ASSAYS

3.4.1. Melatonin (I, II, III)

Melatonin concentrations of plasma/serum samples were determined by a commercially available

double-antibody radioimmunoassay (Bühlmann Laboratories AG, Switzerland) based on the

Kennaway G280 anti-melatonin antibody (Kennaway et al., 1982). Klupiec et al. (1997) have

demonstrated specificity of G280 antiserum for melatonin extracted from porcine plasma. The

sensitivity of the assay was 0.3 pg/ml.

As a modification for manufacturer’s instructions the ether extraction was used rather than column

extraction in the wild boar study (I). Before assay, controls and samples were extracted twice with

4.5 ml of diethyl ether. The tubes were shaken for 1 minute and put into a freezing bath. The

38

supernatant was decanted and the solvent was evaporated to dryness in a 37 ºC water bath. The

residue was dissolved in 1 ml of incubation buffer and proceeds with assay according to

manufacturer’s instructions with some minor modifications (described on the next page).

The samples of lighting program trials (II, III) were assayed in the Department of Obstetrics and

Gynaecology, University of Adelaide. This laboratory has originally provided the antiserum used in

the commercial Bühlmann Melatonin radioimmunoassay and the laboratory currently continues its

co-operation with Bühlmann Laboratories. The assay described in following paragraphs was

basically the same as the commercial one, but some components, which normally are included in

the commercial kit, were made in the laboratory following the manufacturer’s directions and with

the manufacturer’s permission. There were also some minor modifications in the extraction and

assay procedures. The components of the assay were as follows:

1. Extraction columns: The extraction columns (RK-MEL2, Bühlmann Laboratories,

Switzerland) containing charcoal (C18) were provided with the assay kit. All columns were

used up to five times if blockage of the columns did not occur earlier.

2. Melatonin buffer: Melatonin buffer is included to the commercial assay kit. Because of the

great number of samples analyzed the buffer was made up in the laboratory according to the

manufacturer’s directions, and contained di-Sodium Hydrogen Orthophosphate dihydrate

(14.2 g), Sodium Dihydrogen Orthophosphate (2.9 g), Sodium chloride (8.7 g), BSA (5 g)

and Sodium Azide (0.2 g) in one litre of distilled water.

3. Standards: One set of five vials, each containing different amount of melatonin, was

provided with the assay kit. The standards were reconstituted prior assaying adding 5 ml of

melatonin buffer into each vial, vortexed and incubated at room temperature for 1 hour and

vortexed again. After reconstitution, the standards contained 0.5, 1.5, 5, 15 and 50 pg/ml of

melatonin respectively.

4. Controls: One set of two vials containing 4.5 ml of human serum with lot-specific amounts

of melatonin (LOW and HIGH concentrations) were included in the assay kit. The quality

controls (LOW and HIGH) were included in the beginning and end of each assay to

calculate intra and inter assay coefficients of variation (CV).

39

5. Antiserum: The antiserum was made up in the laboratory adding 6.6 µl of stock antibody to

100-ml of melatonin buffer. In the commercial assay kit, antiserum is provided in ready-to-

use form.

6. Melatonin tracer: The tracer (125I γ-iodomelatonin) was diluted in the laboratory adjusting

total counts to be about 10 000 counts in 100 µl of the diluted tracer. The needed amount of

the tracer was approximately 75 µl in 100-ml of melatonin buffer. In the commercial assay

kit tracer is provided in ready to use form.

7. Second antibody: The solid phase bound anti-goat second antibody (RK-MEL2) was

provided in ready-to-use form.

Extraction and assay procedure:

All the samples were thawed the day before assay, stored in the refrigerator and taken to room

temperature one hour before extraction. Prior to extraction, samples were centrifuged for 10 min at

500 x g to remove any small clots in the plasma/serum that can block the extraction columns. The

extraction columns were inserted into the extraction tubes (Pyrex® 16x125 mm, Corning

Laboratory Science Company, USA) and conditioned two times adding 1 ml of methanol to each

tube and centrifuging for 1 min. at 200 x g. Columns were washed twice by adding 1 ml of water

and centrifuging for 1 min at 200 x g before loading columns with standards and samples. One

millilitre of melatonin buffer was added to columns representing NSB (non-specific binding) and

MB (maximum binding) tubes and 1 ml of standards and samples into respective columns and

centrifuged for 1 min at 200 x g. Columns were washed twice with 10 % methanol and once with

hexane centrifuging for 1 min at 500 x g between washings to remove everything but melatonin

from the columns. After washing, columns were transferred into borosilicate glass tubes (Kimble®

13x100 mm, ASG Inc., USA) and 1 ml of methanol was added into the columns and centrifuged for

1 min at 200 x g to remove the extracted melatonin from the columns. Methanol was evaporated

from the tubes using airflow and heating blocks. Once evaporated, the extracted samples were

reconstituted by adding 1 ml of melatonin buffer into the tubes, vortexed and allowed to equilibrate

30 min at room temperature. Once the reconstituted samples were equilibrated, the assay proceeded

by adding 500 µl extracted buffer into the NSB tubes, 400 µl extracted buffer into MB tubes and

400 µl extracted standards and samples into the respective tubes. Melatonin antibody (100 µl) and

40

tracer (100 µl) were added into each tube except into the NSB tubes, where only tracer was added.

Into the total count tubes 100 µl tracer was added. All the tubes were vortexed and incubated

overnight in a cold room. On the next day 25 µl of the second antibody was added into all the other

tubes but total counts, vortexed and incubated 30 min in a cold room. One millilitre of cold water

was added into all the tubes except the total counts and centrifuged for 15 min at 2000 x g. The

supernatant was decanted and radioactivity of the pellets was counted by γ-counter (Wallac®,

Turku, Finland).

Table 2. The cross-reactivities of the melatonin antiserum shown above were found at 50 % binding (Bühlman Laboratories, 1998).

C om p ound C rossreactivity

M elatonin 100 %

Seroton in < 0.001 %

5-H ydrox y-in doleacetic acid < 0.001 %

N -A setylseroton in 0.027 %

5-M eth oxytryptam in e 0 .003 %

5-M eth oxytryptoph an 0 .001 %

6-Sulfa toxym ela tonin 0 .002 %

5-M eth oxytryptoph ol 0 .001 %

3.4.2. Progesterone (IV, V)

Plasma progesterone concentrations were determined using a commercially available, solid-phase

progesterone RIA (Spectria®, Orion Diagnostica, Finland), which has previously been validated to

measure progesterone in pig plasma (Peltoniemi, 1994). In the assay, 50 µl of plasma sample and

500 µl of buffered I125-label were added to the antibody-coated tubes. The tubes were incubated for

two hours at room temperature and the supernatant was poured out. Radioactivity of the tubes was

41

counted by a γ-counter (Wallac®, LKB-Wallac, Turku, Finland). Sensitivity of the assay was 0.09

ng/ml.

Table 3. Cross-reactivity of the progesterone antiserum (Orion Diagnostica, 1999)

Compound Crossreactivity

progesterone 100 %

Pregnenolone 3.9 %

Corticosterone 0.9 %

5β-Dihydroprogesterone 0.75 %

11-Deoxycorticosterone 0.38 %

5α-Dihydroprogesterone 0.22 %

20β-Hydroxyprogesterone 0.045 %

17α-Hydroxyprogesterone 0.017 %

3.4.3. Luteinizing Hormone (V)

Plasma LH concentrations were determined using a direct double-antibody RIA as previously

reported (Niswender et al., 1970), with modifications reported by Peacock (1991). Purified porcine

LH (LER-786-3) was supplied Professor L.E. Reichert Jr. and radio-iodinated with I125 (Amersham

Australia Pty.Ltd.). In the iodination procedure, the Chloramine-T method described by Greenwood

et al. (1963) was used. After iodination, the label was stored at 20°C and used within six weeks.

0.01 M phosphate buffer with EDTA was used to dilute the label to give approximately 10 000

counts.

The antiserum used was rabbit anti-porcine LH (Niswender No. 566), donated by Professor G.

Niswender. It was frozen in a final dilution of 1:17 000 in the buffer. A solid-phase second antibody

coated cellulose suspension (Sac-Gel®, A-SAC1, Immunodiagnostics, Waverley, Australia) raised

in donkeys against rabbit serum was used to separate bound and unbound label. An assay buffer

42

(0.133 g NaH2PO4.2H2O, 1.299 g Na2HPO4, 9.000 g NaCl, 1.000 g BSA, 1.000 g NaN3 added to 1

L of distilled water) with additional 1 % Bovine Serum Albumine (BSA) was used to prepare the

standards. One hundred micro liters of the purified LH was added to 3.9 ml of the buffer. Two

milliliters of the previous standard was serially diluted into 2 ml of the buffer to obtain standards

from 16 to 0.125 ng/ml.

Two hundred microlitres of sample or standard were mixed with 600 µl of the buffer. Antiserum

(100 µl) was added and the tubes were incubated for 24 hours at 4°C. After the incubation, 100 µl

of the labeled LH was added and incubated for 48 hours at 4°C. The second antibody was then

added (100 µl), incubated for 30 min at room temperature and centrifuged for 8 min at 2000 rpm.

The supernatant was decanted and the tubes were counted on a gamma counter (Wallac®, LKB-

Wallac, Turku, Finland).

3.4.4. GnRH-antibody (V)

GnRH-antibodies from immunized animals and from the antiserum pool were analyzed using RIA

described by Carson et al., 1997. Iodinated GnRH (Amersham International Limited, Amersham,

England) was diluted to approximately 10 000 counts per minute per 100 µl of assay buffer (50 mM

sodium phosphate, 10 mM EDTA and 0.3 % BSA; pH7.2). One hundred microliters of diluted

samples were incubated at 4°C overnight in test tubes (75 x 12 mm; GS Ross Limited,

Macclesfield, England) with 100 µl of the labeled GnRH and 300 µl of assay buffer. Then 250 µl of

dextran (Pharmacia Fine Chemicals, Uppsala, Sweden) coated charcoal (Sigma) was added to

separate free from bound labeled GnRH. The tubes were centrifuged at 4°C for 15 min at 2000 rpm.

and the supernatant decanted and counted on a gamma counter (Wallac®, LKB-Wallac, Turku,

Finland). Titres were calculated by expressing binding of labeled GnRH minus non-specific bound

counts as the percentage of the total available labeled GnRH.

3.5. PREGNANCY TESTING

The sows and gilts were pregnancy tested trans-cutaneously (IV, V) using a real time ultrasound

(Hondex®, Honda Electronics Co. Ltd., Japan) with a linear probe (5 MHz).

43

3.6. LIGHTING PROGRAMS

The lighting experiments (II, III) were done in temperature and lighting controlled climate rooms,

except the wild boar trial (I), in which the wild boars were under the natural lighting and domestic

gilts under a typical light environment of an indoor piggery where the artificial lighting was

controlled but the natural light was allowed through the windows. In the climate rooms, the

scotophase light intensity was less than 7 lx (dim red lights) measured at pig eye-level. The light

intensity was measured using a digital light meter (Topcon® IM-2D, Japan) in the studies presented

in papers II and III. Another digital light meter (Gossen Mavolux digital®, Gossen, Germany) was

used in the wild boar trial (I).

3.7. STATISTICAL ANALYSES

Statistical analyses were carried out using Stata Intercooler, version 5.0 statistical package (Stata

Corporation, Texas, U.S.A).

An analysis of variance with repeated measures was used to study variation in melatonin and

progesterone profiles (I, III, IV, V). The method of linear mixed models for log-transformed data

was used to study effects of lighting programs on melatonin secretion (II). Fischer’s exact

probability test was used to test seasonal effect on rate of pregnancy disruption (IV). Where

appropriate, Student’s t-test was used for further analyses (I, IV, V).

44

4. RESULTS

The most important results are given in this section. For more detailed results, the reader is referred

to the original papers included to the end of the thesis.

4.1. SEASONAL ALTERATIONS IN MELATONIN SECRETION (I)

Both domestic pigs and European wild boars demonstrated a clear rhythm in serum melatonin

concentrations in every season. In all pigs, the serum melatonin concentrations increased rapidly

from values between < 0.3 pg/ml (the sensitivity of assay) and 1.5 pg/ml, up to values generally

between 10 and 25 pg/ml with the onset of darkness. In all seasons there was a considerable inter

individual variation in the amplitude of nocturnal melatonin increase but individual’s profile was

constant on consecutive days (Figure 2).

There was no significant difference in the darkphase (<100 lx) melatonin response, in terms of the

mean melatonin concentration or peak value, between the domestic gilts and wild boars or among

different seasons. In the wild boars, the duration of nocturnal melatonin secretion was in the spring,

summer, autumn and winter, 10, 6, 11 and 17 hours, respectively. In the domestic gilts, the duration

was, 9, 8, 12 and 11 hours, respectively. The differences in the duration of nocturnal increase in

melatonin between the summer and winter samplings were significant (P < 0.01) in both animal

groups (Figure 3).

Figure 2. Individual melatonin patterns of the wild boars and domestic gilts from the summer sampling.

Wild boars

0

10

20

30

40

13 17 21 1 5 9 13 17 21 1 5 9

Time (00 hrs)

pg/m

l

Domestic gilts

0

10

20

30

40

13 17 21 1 5 9 13 17 21 1 5 9

Time (00 hrs)

pg/m

l

45

Figure. 3. Mean (-sem) serum melatonin concentrations (�) for the wild boars and domestic gilts and light intensity ({) over 48 hours. The season and animal group (wild boar/domestic pig) are indicated in the figure.

4.2. EFFECTS OF ARTIFICIAL LIGHTING PROGRAMS ON MELATONIN SECRETION

(II, III)

There were no significant effects of lighting regimen (16 L : 8 D vs. 8 L : 16 D) on scotophase mid-

interval mean melatonin concentrations (II). When the lighting regimen was changed from long day

to short day, there was an immediate change in the pattern of melatonin profile toward alignment

with the new lighting conditions. When the lighting regimen was changed from short day to long

day there was only a small change in the melatonin profile (Figure 4). At 1800, 2000 and 0600

hours when lights were on mean melatonin concentration continued to significantly exceed the

basal level (P<0.001, 0.001, 0.01 respectively) but at 0800 hours the mean was at basal level. After

one week there was further progression towards a long day melatonin pattern. After two weeks,

melatonin profiles were adjusted to a new light environment excepting of some variation in the first

and the last sample of the scotophase.

Dom es tic (S um m er)

05

1 01 52 02 53 03 5

1 3 1 7 2 1 1 5 9 1 3 1 7 2 1 1 5 9

Tim e (00 hrs )

pg

/ml

0

1 0 0

2 0 0

3 0 0

4 0 0

lux

W ild B oar (S um m er)

05

1 01 52 02 53 03 5

1 3 1 7 2 1 1 5 9 1 3 1 7 2 1 1 5 9

Tim e (00 hrs )

pg/m

l

0

1 00 00

2 00 00

3 00 00

4 00 00

lux

W ild boar (W inter)

05

1 01 52 02 53 03 5

1 9 2 3 3 7 1 1 1 5 1 9 2 3 3 7 1 1 1 5

Tim e (00 hrs )

pg/m

l

0

1 0 0 0

2 0 0 0

3 0 0 0

4 0 0 0

lux

Domestic (Winter)

05

101520253035

19 23 3 7 11 15 19 23 3 7 11 15Time (00 hrs)

pg/m

l

0

100

200

300

400

lux

46

Figure 4. Mean (- SEM) plasma melatonin concentrations of young boars. Pigs were allowed to adjust to the lighting (16 L : 8 D) for two weeks before the lighting was changed to the opposite (8 L : 16 D). A 48-hour sampling spanning the changeover with samples collected at two-hour intervals was performed at the time the lighting was changed. Black bar under axis illustrates the scotophase.

When the light-dark cycle was 12 L : 12 D, the photophase light intensity (40 lx, 200 lx or 10000

lx) did not affect the scotophase mean melatonin concentration (III). There was no difference in the

basal melatonin concentrations between the lighting regimens within or between the groups (Figure

5).

Figure 5. The mean (+SD) melatonin concentration during the photophase at all treatments (basal concentration) and the mean (+SD) melatonin concentrations during the scotophase at different treatments. The treatments are indicated under the x-axis.

Long day->Short day

0 2 4 6 8

10 12 14 16

12 16 20 24 4 8 12 16 20 24 4 8 12 Time

pg/m

l

Short day->Long day

0 2 4 6 8

10 12 14 16

12 16 20 24 4 8 12 16 20 24 4 8 12

Time

pg/m

l

0

2

4

6

8

10

12

14

Basal 40 lx 200 lx 10000 lx

pg/m

l

47

0

5

10

15

20

25

30

35

40

-1 3 6 10 13 17 20 24

Days

Ser

um p

roge

ster

one

(ng/

ml)

4.3. EARLY DISRUPTION OF PREGNANCY (IV, V)

The farrowing rate (V) in winter-spring was 72 % (n=58 inseminations) and in the summer-autumn

period it was 63 % (n=81 inseminations, Table 4). In winter – spring seven inseminated sows

returned to oestrus with regular return interval (20.9 ± 0.7) and four sows were classified as delayed

returns (25.5 ± 0.6). In summer – autumn seven sows returned to oestrus with regular return

intervals (22.3 ± 1.1) and three with delayed returns (26.7 ± 2.9). In addition to these, there was

one early disruption of pregnancy identified in the winter-spring, and nine such cases were found in

the summer – autumn (p < 0.05, Table 4). Compared with animals that remained pregnant, serum

progesterone concentrations tended to be lower on day 13 (5.6 ng/ml less) in animals eventually

loosing the pregnancy. On day 20, serum progesterone concentrations in sows eventually

interrupting their pregnancies (11.7 ng/ml) were lower than concentrations in animals remaining

pregnant (22.0 ng/ml) on day 20 (P < 0.05). However, there was no difference in serum

progesterone concentrations for animals remaining pregnant in summer – autumn compared with

pregnant animals in winter-spring when individual sows were monitored over a six-week period.

Table 4. Farrowing rates and observed EDP cases in different seasons. Numbers of regular and irregular returns are also shown

Figure 6. The serum progesterone concentrations (mean +/- SD) of sows in which early disruption of pregnancy (³) was detected (n=7) compared with nine control sows ( ) remaining pregnant in summer – autumn. Two sows that disrupted the pregnancy but remained pregnant beyond day 30 have been excluded from the figure.

Season N of inseminations

Farrowing rate %

N of regular returns

N of irregular returns

N of early disruptions

Feb-May 58 72 7 4 1

Aug-Nov 81 63 7 3 9

48

In the active GnRH-immunisation trial (IV), none of the 7 sows receiving the booster immunisation

10 days after mating were pregnant as detected by real time ultrasound 18 days post mating. There

was no visible evidence of abortion at any time. The effect of treatment on progesterone

concentrations was significant (Figure 7A; P=0.05). Plasma progesterone levels in immunised sows

started to decrease compared with pregnant controls after day 16, and by day 18 progesterone levels

were similar in immunised and non-pregnant sows. All sows that received the booster immunisation

on day 20 of pregnancy (n=4) aborted between 7 and 14 days later with a mean immunisation to

abortion interval of 10 ± 1.5 (SEM) days (Figure 7B). In these sows progesterone levels decreased

gradually until day 26 followed by a steep decrease between days 26 and 28. After progesterone had

reached a basal concentration, it took approximately two days before abortion occurred.

In the passive GnRH-immunisation trial, none of the gilts infused with antiserum (n=6) 12 days

post mating was pregnant 18 days after mating, whereas 5 out of 6 controls were pregnant on day

18 and remained pregnant until the end of the trial. The non-pregnant gilts on day 18 typically had a

small amount of fluid in the uterus with distribution of fluid distinguishably different from a typical

uterine oestrous fluid. The fluid was in a small isolated area rather than being distributed throughout

the uterus and between the endometrial folds. Despite daily close monitoring of animals there was

no visible evidence of abortion.

0

5

10

15

20

25

10 12 14 16 18 20 22 24 26 28

days after mating

plas

ma

prog

este

rone

ng/

ml

vaccinated preg contr non-preg contr

0

5

10

15

20

25

20 22 24 26 28 30 32 34 36 38

days after mating

plas

ma

prog

este

rone

ng/

ml

vaccinated preg contr

average abortion time

Figure 7. The mean (±SEM) plasma progesterone concentrations after active immunisation by a commercial GnRH-vaccine (Improvac®). The first immunisation was done on the day of farrowing and the second immunisation was done on day 10 (A) (n=7) or on day 20 (B) (n=4) post mating. The average abortion time after active immunisation on day 20 post mating is also shown (B).

A) B)

49

5. DISCUSSION

This study demonstrated that pigs secrete melatonin in rhythmic fashion similar to melatonin

secretion in other mammals studied. There is no difference in lighting – melatonin transduction

between wild boars and domestic pigs. Both the wild boar and the domestic pig express seasonal

variation in the circadian secretion pattern of melatonin provided that the artificial lighting in the

piggery is constant and the natural light is allowed through the windows. The seasonal variation in

melatonin secretion is directly related to the changes in the day length. It is justified to assume,

according to studies with other mammals, that also wild boars synchronize their breeding season

with a preferred season using the changing melatonin profile to anticipate the right time of the year.

It seems more than a coincidence that an increased number of pregnancy failures in the domestic

pig appears at the time when the European wild boar is in a sexual arrest (Peltoniemi et a., 1999). In

addition, when it is evident now that seasonal alterations in the melatonin secretion of these two

types of pigs are similar, it is obvious that changes in the day length contribute to the seasonal

infertility of the domestic pig.

These results provide further evidence that earlier confusion about the existence of a circadian

melatonin rhythm in the domestic pig related to inadequacies in the RIAs used, as demonstrated by

Klupiec et al. (1997) and confirmed later by Andersson et al (2000). The present study, together

with some earlier studies (Paterson et al., 1992), has conclusively demonstrated a circadian

melatonin rhythm in pigs under different lighting conditions, and further studies should be directed

towards the applicability of artificial lighting programs for commercial piggeries.

The circadian melatonin secretion can be entrained by artificial lighting. Pigs are able to respond

within relatively short acclimatization periods to abrupt and extreme changes in ambient lighting.

Indeed, pigs are able to respond immediately to an extension of the dark phase as demonstrated by

entrained melatonin profiles after the change from the eight-hour dark phase to the 16-hour dark

phase. It seemed to be more difficult for pigs to adjust to a change in the other direction (from short

day to long day). This was demonstrated by almost unchanged melatonin profiles during the first 24

hours after the changeover. However, in this group melatonin profiles analysed one week after the

changeover showed further entrainment to the new lighting regimen and the profiles were

practically adjusted within two weeks. The melatonin profiles were evident under all lighting

regimens and also during the first 24-hours after every changeover, even if they were not always

50

immediately adjusted to the ambient lighting. There was no evidence that extreme and abrupt

changes in lighting could abolish the circadian rhythm of melatonin secretion as proposed by

McConnell and Ellendorff (1987). It is evident that the response in circadian rhythms can be

achieved using simplified lighting regimens instead of step-wise changes in lighting (Paterson et al.,

1992). This has importance for application of artificial lighting programs to commercial piggeries.

Another important aspect of lighting regimens in commercial piggeries is the light intensity during

the photophase and its effects on melatonin secretion during the darkphase. Earlier studies implied

that pigs need relatively high light intensities during the photophase to entrain their melatonin

rhythms (Minton et al., 1989; Mack and Unshelm, 1997). It is important to use sufficient light

intensity during the phoptophase so that pigs can differentiate between the light- and darkphases. It

was demonstrated in the present study that the light intensity needed for differentiation between

photo- and scotophase is much lower in pigs than previously thought. As low intensity as 40 lx is

enough to entrain melatonin secretion so that high melatonin concentrations occur in the systemic

circulation during the darkphase. It appears that there is no difference in the darkphase melatonin

response when the lightphase light intensity varies between 40 and 10 000 lx. This suggests that

extremely high light intensities (10 000 lx) during the photophase do not provide any advantage

over a normal comfortable light intensity (200 lx) in attempts to manipulate reproduction by

artificial lighting regimens. It can be concluded that manipulation of circadian rhythms is possible

by simple artificial lighting programs and that most attention should be paid to the light-dark cycle

rather than the absolute light intensity, which is of little importance.

There are different manifestations of seasonal infertility; delayed puberty in gilts, prolonged

weaning to oestrus interval in sows and a reduction in the proportion of mated sows that farrow.

The last of these is of the greatest economic significance (Love et al. 1993). A major contribution to

the reduced farrowing rate is the increased number of irregular returns to oestrus 25 to 35 days after

mating (Love 1981). Early disruption of pregnancy, because of early embryonic deaths (Xue et al.

1994) or because of regression of corpora lutea (CL) (Wrathall et al. 1986), has been suggested to

cause these irregular returns in summer-autumn. The delayed return to oestrus, although commonly

reported in the literature and known as a manifestation of seasonal infertility, has never before been

shown to result from the disruption of an established pregnancy. This study demonstrated for the

first time a clear association between the seasonally lowered farrowing rate and early disruption of

pregnancies. Wrathall et al. (1986) have demonstrated a clear seasonal variation in the serum

progesterone levels that contradicts with the present study. However, they compared progesterone

51

levels at a later stage of pregnancy (between days 40 and 90), whereas in the present study the

comparison was at the early stage that is a more relevant time for studying this phenomenon. They

did compare progesterone also on days 25 – 30 in one single farm (250 sows), but the comparison

was made between single samples of animals rather than longer profiles as was done in the present

study. Interestingly, in the present study progesterone concentrations in animals eventually loosing

the pregnancy were somewhat lower as early as day 13 after mating. This implies that

endocrinological events leading to disruption of pregnancy in summer-autumn are demonstrable

only in “problem sows” and not in the general population. The endocrinological mechanisms of

early disruption of pregnancy are yet to be confirmed, but this study implies that something happens

to the normal development of embryos first and after that the CL regress.

A number of earlier studies have shown that pituitary LH-secretion is subject to seasonal variation

in the pig, even if there is some controversy about seasonal effects on LH-pulsatility. However,

most studies have found that LH-pulses are more regular and of higher amplitude in the winter (or

under short day lighting) when compared with summer (or long day lighting) (Peacock, 1991;

Paterson, 1992; Peltoniemi, 1997b). The LH-support for CL is lower during the time LH-pulses are

irregular and of lower amplitude, i.e. during the summer – autumn. It has been proposed that this

lowered pituitary support might cause regression of CL, when they are no longer autonomous after

day 12 of pregnancy, and this leads to early disruption of pregnancy typical for seasonal infertility

(Wrathall et al., 1986; Love et al., 1993). However, the studies by Anderson et al. (1967) and

Peltoniemi et al. (1995) have demonstrated that a prolonged complete blockade of LH-pulses is

needed to initiate regression of CL. Those studies together with the present work demonstrate that

from four to nine days’ complete LH-deprivation is needed to cause a destruction of CL. It is

unlikely that there could be such a complete seasonal LH-blockade.

The study also showed two different outcomes of active GnRH-immunization depending on the

timing of the second immunization. None of the sows immunized for the second time on day 10

post mating was detectably pregnant 18 days after mating, and there were no outer signs of

abortion. The passive immunization of gilts on day 12 post mating had a similar effect. The second

immunization on pregnancy day 20 resulted in abortion with normal fetuses expelled (Knight et al.,

1977). The mean interval between the second immunization and abortion was 10 ± 1.5 days. In both

actively immunized groups the progesterone decreased to the basal level eight days after the second

immunization. This means that in the 10-day immunization group the recognition and establishment

of pregnancy had already failed before the regression of CL, whereas in the 20-day vaccination

52

group the regression of CL caused the abortion. This is similar to the observation in the field trial,

where the normal development of embryos was first disturbed and the regression of CL was

initiated after that. This study implies that early disruption of pregnancy is caused by a mechanism

other than by a direct regression of CL because of inadequate pituitary LH-support.

The progesterone concentrations after pregnancy day 12 depend on pituitary support whereas before

day 12 the number of formed CL and a metabolic clearance of progesterone mainly affect the

progesterone concentrations. It is well known that progesterone controls a number of events

affecting embryos’ development rate during early pregnancy that is critical for signaling

commenced pregnancy and preventing the luteolysis (Geisert and Yelich, 1997). This study implies

that decreased progesterone concentration between days 14 and 18 post mating can disturb

development of embryos and prevent establishment of pregnancy before regression of CL occurs

(because of deprivation of LH-pulses). The recognition of pregnancy takes place in two phases

between days 12 and 18 as demonstrated by Pusateri et al. (1996a). This means that the first

oestrogen signal from the embryos occurs during the period of inadequate pituitary LH-support

(summer) because the pituitary LH does not affect progesterone and subsequently the development

rate of embryos before day 12. However, the seasonally decreased pituitary LH-support can affect

progesterone levels after day 12 and could interfere with the development of embryos, which in turn

could result in an inadequate or missing oestrogen signal on pregnancy day 18. If that was the case,

as suggested in the present study, it would lead to reorientation of PGF-2α secretion back to an

endocrine direction, luteolysis of CL, early disruption of pregnancy and a delayed return to oestrus

with interval of 25-30 days. A schematic figure of the proposed mechanism of early disruption of

pregnancy is presented in Figure 6.

It is concluded that the basis of lighting - melatonin transduction in the pig is similar to other

mammals. There is no difference with the melatonin secretion in the European wild boar and the

domestic pig. The circadian pattern of melatonin secretion can be modified by simple lighting

regimens. A light-dark cycle itself appears to be more important in producing a scotophase

melatonin response than the light intensity during the photophase. The lowered farrowing rate

typical for seasonal infertility is associated with early disruptions of pregnancy. It is possible that

the seasonal early disruption of pregnancy is mediated by decreased progesterone concentrations

leading to underdeveloped embryos at the time of recognition of pregnancy. This seasonal decrease

in progesterone appears to be under the influence of seasonally changing melatonin secretion

53

through the GnRH- and LH-secretion. Further studies are required to ensure the mechanism of

seasonal early disruption of pregnancy.

54

LONG DAY SCN SHORT DAY PINEAL GLAND

Nocturnal melatonin secretion

HYPOTHALAMUS

Effects on GnRH-neurons (sensitivity to oestradiol feed-back) Direct effects of neurotransmitters

GnRH-secretion

PITUITARY GLAND

LH-pulses CORPORA LUTEA ESTABLISHED

Progesterone secretion PREGNANCY beyond pregnancy day 12

UTERUS Secretion of histotrophes

UNDER-DEVELOPED OR DEAD NORMAL EMBRYOS EMBRYOS ON PREGNANCY DAY 18 ON PREGNANCY DAY 18

Figure 6. The schematic figure of the proposed mechanism of seasonal early disruption of pregnancy.

shor

t dur

atio

n

long

dur

atio

n

cate

chol

amin

e en

dog.

opi

oids

glut

amat

ergi

c ne

uron

s

decr

eas

incr

eas

low amplitude irregular

high amplitude regular

decr

ease

d

nor

mal

Oestrogen signal on day 18

PGF-2α secretion back to endocrine direction beyond day 18

Continuation of PGF-2α secretion to exocrine direction beyond day 18

decreased normal

EARLY DISRUPTION OF PREGNANCY

55

6. CONCLUSIONS

1. The basis of lighting – melatonin transduction in the pig is similar to other mammals.

2. There is no attenuation of melatonin secretion in the domestic pig due to altered genotype or

reduced light environment.

3. Simplified artificial lighting programs can modify the circadian pattern of melatonin

secretion. Adjustment of the secretion pattern takes place within two weeks even after

abrupt and extreme changes in lighting.

4. The photophase light intensity has no effect on the scotophase melatonin response, once the

threshold light intensity to differentiate the dark and light phases is exceeded. This threshold

is less than 40 lx.

5. The seasonally lowered farrowing rate is associated with an increased number of early

disruptions of pregnancies. There is/are factor(s) interfering first with the normal

development of embryos and the regression of CL initiates after that.

6. This retarded development of embryos might be caused by seasonally altered LH-secretion

decreasing the progesterone secretion of CL affecting the secretion of histotrophes of

endometrium.

56

ACKNOWLEDGEMENTS

The project has been financially supported by Finnish Ministry of Agriculture, Finnish

slaughterhouses and University of Helsinki, which are gratefully acknowledged.

I had no plans for PhD project when I came to ambulatory clinic of Saari to work as a clinic

veterinarian in 1997. However, Olli Peltoniemi, PhD, had a great vision of importance of

seasonality studies for Finnish pig production. Olli talked me into the project and ever since he has

been a principal supervisor of my PhD project and the closest fellow worker. His enthusiasm and

support has been irreplaceable, for which I will always be grateful.

I have got a lot of support and friendly guidance also from Professor Terttu Katila, the leader of the

PhD project, for which I will be indebted. Professor Robert Love was my supervisor when I spent

an unforgettable research year in Australia at University of Sydney. Ever since, Professor Love has

had a great contribution to my PhD project, of which I want to express my warmest thanks. I would

like to extend my gratitude to Professor Gareth Evans and Professor David Kennaway for their help

and cooperation. Professor Hannu Saloniemi and Professor Magnus Andersson have been my

supervisors giving me an incredible amount of encouragement during the course of years. They

both are gratefully acknowledged.

Håkan Andersson, PhD, my co-author, has always been ready for cooperation and stimulant

scientific discussions, for which I am grateful. My co-author, Outi Hälli, DVM was a great help

especially in the most demanding wild boar studies, for which I want to thank her. During my

project, Professor Satu Pyörälä, Erkki Pyörälä, DVM and Juhani Taponen, DVM, have all shown

their support to my PhD project in a number of ways for which I will always be indebted. I have

had a privilege to work with swine specialists like Mari Heinonen, PhD, Juha Virolainen, DVM,

Jonna Oravainen, DVM, Veikko Tuovinen, PhD, Taneli Tirkkonen, DVM, Tapio Laurila, DVM

and a number of others. They all are acknowledged with gratitude. I want to extend my gratitude to

all the staff of Saari Unit. I am particularly grateful to Reijo Jokivuori for his kindness.

I am also indebted to the pre-examiners of the dissertation, Professor Stig Einarsson and Professor

Erik Ropstad for their constructive criticism at the final stage of the present work. I also would like

to acknowledge the opponent, Professor Gerald Lincoln.

57

Finally, I would like to thank my beloved wife Kaija for her support and love. I am grateful for our

two sons, Julius and our beautiful baby. I also want to thank my parents and my brothers, who have

always supported me with love.

Anssi Tast

58

REFERENCES

Aherne, F.X. and Kirkwood, R. Nutrition and sow prolificacy. J. Reprod. Fertil. Suppl. 33, 1985:

169-183.

Almond, G.W., Friendship, R.M. and Bosu, W.T.K. Autumn abortions in sows. Can.Vet. J. 26,

1985: 162-163.

Andersson, H., Lillpers, K., Rydhmer, L. and Forsberg, M. Influence of light environment and

photoperiod on plasma melatonin and cortisol profiles in young domestic boars, comparing two

commercial assays. Domest. Anim. Enocrinol. 19, 2000: 261-274.

Anderson, L.L., Dyck, G.W., Mori, H., Henricks, D.M. and Melampy, R.M. Ovarian function in

pigs following hypophysial stalk transection or hypohysectomy. Am. J. Physiol. 1967: 1188-1194.

Anderson, L.L. Growth, protein content and distribution of early pig embryos. Anat. Rec. 190,

1978: 143-153.

Aoki, H., Yamada, N., Ozeki, Y., Yamane, H. and Kato, N. Minimum light intensity required to

suppress nocturnal melatonin concentration in human saliva. Neurosci. Lett. 252, 1998: 91-94.

Arendt, J., Symons, A.M. and Laud, C. Pineal function and photoperiod in the ewe. In:

Photoperiodism and Reproduction in Vertebrates. Colloq INRA 6, 1981: 219-229.

Bazer, F.W. and Thatcher, W.W. Theory of maternal recognition of pregnancy in swine based on

estrogen controlled endocrine versus exocrine secretion of prostaglandin F2alpha by the uterine

endometrium. Prostaglandins 14, 1977: 397-400.

Bazer, F.W., Geisert, R.D., Thatcher, W.W. and Roberts, R.M. The establishment and maintenance

of pregnancy. In: D.J.A. Cole and C.R. Foxcroft (eds.) Control of Pig Reproduction. Butterworths,

London, 1982: 227-252.

Boivin, D.B., Duffy, J.F., Kronauer, R.E. and Czeisler, C.A. Dose-response relationships for

resetting of human cicardian clock by light. Nature 379, 1996: 540-544.

59

Bollinger, A.L., Wilson, M.E., Pusateri, A.E., Green, M.L., Martin, T.G. and Diekman, M.A. Lack

of a nocturnal rise in serum concentrations of melatonin as gilts attain puberty. J. Anim. Sci. 75,

1997: 1885-1892.

Bonnefond, C. Walker, A.P., Stutz, J.A., Maywood, E., Juss, T.S., Herbert, J. and Hastings, M.H.

The hypothalamus and photoperiodic control of FSH secretion by melatonin in the male Syrian

hamster. J. Endocrinol. 122, 1989: 247-254.

Booth, P.J. Metabolic influences on hypothalamic-pituitary-ovarian function in the pig. J. Reprod.

Fertil. Suppl. 40, 1990: 89-100.

Brinkley, H.J., Norton, H.W. and Nalbandov, A.V. Role of a hypophysial luteotrophic substance in

the function of porcine corpora lutea. Endocrinology 74, 1964: 9-13.

Carson, A.F., Mc Caughey, W.J., Mackie, D.P., Higgins, H.C. and Steen, R.W.J. Production and

biological activity of Murine monoclonal antibodies against GnRH. Theriogenology 48, 1997:193-

207.

Carter, D.S. and Goldman, B.D. Antigonadal effects of timed melatonin ifusion in pinealectomized

male Djungarian hamsters (Phodopus sungorus sungorus): Duration is the critical parameters.

Endocrinology, 1983.

Cross, J.S. and Roberts, R.M. Porcine conceptuses secrete an interferon during the preattachment

period of early pregnancy. Biol. Reprod. 40, 1989: 1109-1118.

Dehoff, M.H., Bazer, F.W. and Collier, R.J. Ontogeny of prolactin receptors in porcine uterine

endometrium during pregnancy. In Proceedings: 4th International prolactin congress,

Charlottesville. 1984: 95 (abstr.).

Delcroix, I., Mauget, R. and Signoret, J.P. Existence of synchronization of reproduction at the level

of the social group of the European wild boar (Sus scrofa). J. Reprod. Fert. 89, 1990: 613-617.

Diekman, M.A., Clapper, J.A., Green, M.L. and Stouffer, D.K. Reduction in age of puberty in gilts

consuming melatonin during decreasing or increasing daylength. J. Anim. Sci. 69, 1991: 2524-

2531.

Diekman, M.A., Brandt, K.E., Green, M.L., Clapper, J.A. and Malayer, J.R. Lack of a nocturnal rise

of serum melatonin in prepubertal gilts. Domest. Anim. Endocrinol. 2, 1992: 161-167.

60

Diekman, M.A. and Green, M.L. Serum concentrations of melatonin in prepubertal or postpubertal

gilts exposed to artificial lighting or sunlight. Theriogenology, 47, 1997: 923-928.

Dziuk, P.J. Effect of migration, distribution and spacing of pig embryos on pregnancy and fetal

survival. J. Reprod. Fert. Suppl. 33, 1985: 57-63.

Ebling, F.J., Lincoln, G.A., Martin, G.B. and Taylor, P.L. LHRH and beta-endorphin in the

hypothalamus of the ram in relation to photoperiod and reproductive activity. Dom. Anim.

Endocrinol. 4, 1987: 149-156.

Ebling, F.J. and Hastings, M.H. The neural basis of seasonal reproduction. Ann Zootech. 41, 1992:

239-246.

Enne, G., Beccaro, P.V. and Tarocco, C. A note on the effect of climate on fertility in pigs in the

Padana Valley of Italy. Animal Production. 28, 1979: 115-117.

Fischer, H.E., Bazer, F.W. and Fields, M.J. Wsteroid metabolism by endometrial and conceptus

tissues during early pregnancy and pseudopregnancy in gilts. J. Reprod. Fert. 75, 1985: 69-78.

Foxcroft, G.R. Mechanisms mediating nutritional effects on embryonic survival in pigs. J. Reprod.

Fert. Suppl. 52, 1997: 47-61.

Frank, M., Bazer, F.W., Thatcher, W.W. and Wilcox, C.J. A study of prostaglandin F2alpha as the

luteolysin in swine: III effects of estradiol valerate on prostagalndin F, progestins, estrone and

estradiol concentrations in the utero-ovarian vein of nonpregnant gilts. Prostaglandins 14, 1977:

1183-96.

Gadsby, J.E., Heap, R.B. and Burton, R.D. Oestrogen production by blastocyst and early embryonic

tissue of various species. J. Reprod. Fert. 60, 1980: 409-17.

Gadsby, J.E., Fitz, T.A, Balapure, A.K. and Britt, J.H. PGF-2α receptors on pig luteal cells during

the oestrus cycle. Biol. Reprod. 38 Suppl.1, 1988: 181 (abstr.).

Garverick, H.A., Polge, C. and Flint, A.P. Oestradiol administration raises luteal LH receptor levels

in intact and hysterectomized pigs. J. Reprod. Fert. 66, 1982: 371-7.

61

Geisert, R.D., Pratt, T.N., Bazer, F.W., Mayes, J.S. and Watson, G.H. Immunocytochemical

localization and changes in endometrial progestin receptor protein during the porcine oestrus cycle

and early pregnancy. Reprod. Fertil. Dev. 6, 1994: 749-760.

Geisert, R.D. and Yelich, J.V. Regulation of conceptus development and attachment in pigs. J.

Reprod. Fert. Suppl. 52, 1997: 133-149.

Green, M.L., Clapper, J.A., Andres, C.J. and Diekman, M.A. Serum concentrations of melatonin in

prepubertal gilts exposed to either constant or stepwise weekly alteration in scotophase. Domest.

Anim. Endocrinol. 13, 1996.

Greenwood, F.C., Hunter, W.M. and Glover, J.S. The preparation of 131-I labeled human growth

hormone and high specific radioactivity. Biochemical Journal 89, 1963: 114-123.

Griffith, M.K. and Minton, J.E. Effect of light intensity on circadian profiles of melatonin, ACTH,

and cortisol in pigs. J. Anim. Sci. 70, 1992: 492-498.

Gross, T.S., Mirando, M.A., Young, K.H., Beers, S., Bazer, F.W. and Thatcher, W.W.

Reorientation of prostaglandin F secretion by calcium ionophore, estradiol and prolactin in

perifused porcine endometrium. Endocrinology 127, 1990: 637-42.

Guthrie, H.D. and Rexroad, C.E.Jr. Endometrial prostaglandin F release in vitro and plasma 13, 14-

dihydro-15-keto-prostaglandin F2alpha in pigs with luteolysis blocked by pregnancy, oestradiol

benzoate or human chorionic gonadotropin. J. Anim. Sci. 52, 1981: 330-9.

Hallford, D.M., Wettemann, R.P., Turman, E.J. and Omtvedt, I.T. Luteal function in gilts after

prostaglandin F2alpha. J. Anim. Sci. 41, 1975: 1706-10.

Hancock, R.D. Clinical observations on seasonal infertility in sows in Cornwall. Vet. Rec. 123,

1988: 413-416.

Hashimoto, S., Nakamura, K., Honma, S., Tokura, H. and Honma, K. Melatonin rhythm is not

shifted by lights that suppress nocturnal melatonin in humans under entrainment. Am. J. Physiol.

270, 1996: 1073-1077.

Heap, R.P. and Perry, J.S. The maternal recognition of pregnancy. Br. J. Hosp. Med. 12, 1974: 8-

14.

62

Henderson, K.M. and McNatty, K.P. A biochemical hypothesis to explain the mechanism of luteal

regression. Prostaglandins 9, 1975: 779-97.

Hughes, P.E. Factors affecting the natural attainment of puberty in the gilt. In: D.J.A. Cole and C.R.

Foxcroft (eds.) Control of Pig Reproduction. Butterworths, London. 1982: 117-138.

Hurtgen, J.P. and Leman, A.D. Seasonal influence on the fertility of sows and gilts. J. Am Vet.

Med. Ass. 177, 1980: 631-635.

Karch, F.J., Bittman, E.L., Foster, D.L., Goodman, R.L., Legan, S.J. and Robinson, J.E.

Neuroendocrine Basis of seasonal Reproduction. Recent Prog. Horm. 40, 1984: 185-232.

Karsch, F.J., Malpaux, B., Wayne, N.L. And Robinson, J.E. Characteristics of the melatonin signal

that provide the photoperiodic code for timing seasonal reproduction in the ewe. Repr. Nutr. Dev.

28,1988: 459-472.

Kennaway, D.J., Gilmore, T.A. and Seamark, R.F. Effect of melatonin feeding on serum prolactin

and gonadotropin levels and the onset of seasonal oestrus cyclicity in the sheep. Endocrinology 110,

1982: 1766-1772.

Klupiec, C., Evans, G., Love, R.J. and Kennaway, D.J. Clarifying plasma melatonin profiles in

domestic pigs: A critical and comparative evaluation of two radioimmunoassay systems. J. Pineal

Res. 22, 1997: 65-74.

Knight, J.W., Bazer, F.W., Thatcher, W.W., Franke, D.E. and Wallace, H.D. Conceptus

development in intact and unilaterally hysterectomized – ovariectomized gilts: interrelations among

hormonal status, placental development, fetal fluids and fetal growth. J. Anim. Sci., 44, 1977: 620-

637.

Krzymowski, T., Kotwicka, J., Okrasa, S. and Doboszynska, T. The function and regression of

corpora lutea during the sow’s oestrus cycle after 10 hours of prostaglandin F-2α infusion into the

anterior uterine vein. In Proceedings: VIII Int. Congr. Anim. Reprod. A. I., Krakow. 1976: 143.

La Bonnardière, C. Nature and possible functions of interferons secreted by the preimplantation pig

blastocyst. J. Reprod. Fert. Suppl. 48, 1993:157-170.

63

Lehman, M.N., Robinson, J.E., Karch, F.J. and Silverman, A.J. Immunocytochemical localization

of luteinizing hormone-releasing hormone (LHRH) pathways in the sheep brain during anestrus and

the mid-luteal phase of the estrous cycle. J. Comp. Neurol. 244(1), 1986: 19-35.

Lewy, A.J., Wehr, T.A. and Goodwin, F.K. Light suppresses melatonin secretion in humans.

Science. 210, 1980:1267-1269.

Lincoln, G.A. Photoperiod-pineal-hypothalamic relay in sheep. Anim. Reprod. Sci. 28, 1992: 203-

217.

Love, R.J. Definition of a seasonal infertility problem in pigs. Vet. Rec.104, 1978: 238-240.

Love, R.J. Seasonal infertility in pigs. Vet.Rec. 109,1981: 407-409.

Love, R.J., Evans, G. and Klupiec, C. Seasonal effects on fertility in gilts and sows. J. Reprod.

Fertil. Suppl. 48, 1993: 191-206.

Love, R.J., Klupiec, C., Thornton, E.J. and Evans, G. An interaction between feeding rate and

season affects fertility of sows. Anim. Reprod. Sci. 39, 1995: 275-284.

Lucia, T., Dial, G.D., Marsh, W.E. and P.R. Davies. Long-term breeding patterns in commercial

swine farms. In Proceedings: 13th International Pig Veterinary Society Congress, Bangkok,

Thailand. 1994.

Mack, A.C. and Unshelm, J. Cicardian rhythm of the light-dependant hormone melatonin as an

indicator for the photicrequirement of domestic pigs. In Proceedings: 9th International Congress in

Animal Hygiene, Helsinki, Finland. 1997: 216-220.

Mattioli, M., Prandi, A., Camporesi, A., Simoni, A. and Seren, E. Investigations on swine summer

infertility in Italy. In: E. Seren and M. Mattioli (eds.) Definition of the Summer Infertility Problem.

Commission of the European Communities, Brussels. 1987: 33-43.

Mauget, R. Reproductive biology of the European wild boar. In: E. Seren and M. Mattioli (eds.)

Definition of the Summer Infertility Problem. Commission of the European Communities, Brussels.

1987: 107-119.

McConnell, S.J. and Ellendorff, F. Absence of nocturnal plasma melatonin surge under long and

short artificial photoperiods in the domestic sow. J. Pineal. Res. 4, 1987: 201-210.

64

Melampy, R.M. and Anderson, L.L. Role of the uterus in corpus luteum function. J. Anim. Sci.

Suppl. 27, 1968: 77-96.

Meulen, J., Helmond, F.A., Oudenaarden, C.P.J. and Van der Meulen, J. Effect of flushing of

blastocysts on days 10-13 on the life-span of the corpora lutea in the pig. J. Reprod. Fert. 84, 1988:

157-162.

Minton, J.E., Davis, D.L. and Stevenson, J.S. Contribution of the photoperiod to cicardian

variations in serum cortisol and melatonin in boars. Domest. Anim.Endocrinol. 6, 1989: 177-181.

Moeljono, M.P., Thatcher, W.W., Bazer, F.W., Frank, M., Owens, L.J. and Wilcox, C.J. A study of

prostaglandin F2alpha as the luteolysin in swine: Characterization and comparison of prostaglandin

F, oestrogens and progestin concentrations in utero-ovarian vein plasma of nonpregnant and

pregnant gilts. Prostaglandins 14, 1977: 543-555.

Niswender, G.D., Reichert, L.E. and Zimmerman, D.R. Radioimmunoassay of serum levels of

luteinising hormone throughout the estrus cycle in pigs. Endocrinology 87, 1970: 576-580.

Paterson, A.M. and Pearce, G.P. Attainment of puberty in domestic gilts reared under long-day or

short-day artificial light regimens. Anim. Reprod. Sci. 23, 1990: 135-144.

Paterson , A.M., Pearce, G.P. and Antuono, M.F.D. Seasonal variation in attainment of puberty in

isolated and boar-exposed domestic gilts. Anim. Reprod. Sci. 24, 1991: 323-333.

Paterson, A.M., Martin, G.B., Foldes, A., Maxwell, C.A. and Pearce, G.P. Concentrations of plasma

melatonin and luteinizing hormone in domestic gilts reared under artificial long or short days. J.

Reprod. Fertil. 94, 1992: 85-95.

Peacock, A.J. Environmental and social factors affecting seasonal infertility of pigs. PhD-thesis.

University of Sydney, 1991.

Pearce, G.P. and Pearce A.N. Contact with sow in oestrus or a mature boar stimulates the onset of

oestrus in weaned sows. Vet. Rec. 130, 1992: 5-9.

Peltoniemi, O.A.T. Seasonal and Nutritional Effects on Gonadotrophins in the Early Pregnant Gilts.

MVet.Sc-thesis. University of Sydney, 1994.

65

Peltoniemi, O.A.T., Easton, B.G., Love, R.J., Klupiec, C. and Evans, G. Effect of chronic treatment

with a GnRH agonist (Goserelin) on LH secretion and early pregnancy in gilts. Anim. Reprod. Sci.

40, 1995: 121-133.

Peltoniemi, O.A.T., Heinonen, M., Tuovinen, V., Saloniemi, H. and Love, R.J. Characteristics of

seasonal infertility of the sow in Finland. In Proceedings: 9th International Congress in Animal

Hygiene, Helsinki, Finland. 1997: 208-211.(a)

Peltoniemi, O.A.T., Love, R.J., Klupiec, C. and Evans, G. Effect of feed restriction and season on

LH and prolactin secretion, adrenal response, insulin and FFA in group housed pregnant gilts.

Anim. Reprod. Sci. 49, 1997: 179-190.(b)

Peltoniemi, O.A.T., Love, R.J., Klupiec, C., Revell, D.K. and Evans, G. Altered secretion of LH

does not explain seasonal effects on early pregnancy in gilts. Anim. Reprod. Sci. 49, 1997: 215-

224.(c)

Peltoniemi, O.A.T., Love, R.J., Heinonen, M., Tuovinen, V. and Saloniemi, H. Seasonal and

management effects on fertility of the sow: a descriptive study. Anim. Reprod. Sci. 55, 1999: 47-61.

(a)

Peltoniemi, O.A.T., Heinonen, M., Leppävuori, A. and Love, R.J. Seasonal effects on reproduction

in the domestic sow in Finland – A herd record study. Acta Vet. Scand. 40, 1999: 133-144. (b)

Perry, J.S., Heap, R.B. and Amoroso, E.C. Steroid hormone production by pig blastocysts. Nature

245, 1973: 45-47.

Prunier, A., Quesnel, H., Messias de Braganca, M., Kermabon, A.Y. Environmental and seasonal

influences on the return-to-estrus after weaning in primiparous sows: a review. Liv. Prod. Sci. 45,

1996: 103-110.

Pusateri, A.E., Smith, J.M., Smith, J.W., Thomford, P.J. and Diekman, M.A. Maternal recognition

of pregnancy in swine. I. Minimal requirement for exogenous estradiol-17β to induce either short or

long pseudopregnancy in cycling gilts. Biol. Reprod. 55, 1996: 582-589.(a)

Pusateri, A.E., Wilson, M.E. and Diekman, M.A. Maternal recognition of pregnancy in swine. II.

Plasma concentrations of progesterone and 13,14-dihydro-15-keto-prostaglandin F-2α during the

66

estrous cycle and during short and long pseudopregnancy in gilts. Biol. Reprod. 55, 1996: 590-

597.(b)

Reiter, R.J. Melatonin: the chemical expression of darkness. Molecular and cellular endocrinology.

79, 1991: C153-C158.

Robinson, J.E., Radford, H.M. and Karsch, F.J. Seasonal changes in pulsatile luteinizing hormone

(LH) secretion in the ewe: Relationship of frequency of LH pulses to day length and response to

oestradiol negative feedback. Biol. Reprod. 33, 1985: 324-334.

Roche, J.F., Karch, F.J., Foster, D.L., Takagi, S. and Dzuik, P.J. Effect of pinealectomy on estrus,

ovulation and luteizing hormone in ewes. Biol. Reprod. 2, 1970: 251-256.

Sammelwitz, P.H., Aldred, J.P. and Nalbandov, A.V. Mechanism of maintenance of corpora lutea

in pigs and rats. J. Reprod. Fert. 2, 1961: 387-393.

Shille, V.M., Karlbom, I., Einarsson, S., Larsson, K., Kindahl, H. and Edqvist, L.E. Concentrations

of progesterone and 15-keto-13,14-dihydroprostaglandin F2alpha in peripheral plasma during the

estrous cycle and early pregnancy in gilts. Zentralbatt Für Veterinärmedizin - Reihe A. 26A, 1979:

169-81.

Stone, B.A. and Seamark, R.F. Steroid hormones in uterine washings and in plasma of gilts between

days 9 and 15 after oestrus and between days 9 and 15 after coitus. J. Reprod. Fert. 75, 1985: 209-

21.

Stork, M.G. Seasonal reproductive inefficiency in large pig breeding units in Britain. Vet. Rec. 104,

1979: 49-52.

Szafranska, B. and Tilton, J.E. Prolactin as a luteotrophin during late pregnancy in pigs. J. Reprod.

Fert. 98, 1993: 643-8.

Tast, A., Hälli, O., Peltoniemi, O.A.T. A Practical method for non-surgically inserting intra-arterial

catheters in European wild boars (Sus scrofa). Vet. Res. Comm. 24, 2000: 25-30.

Trout, W.E., Hall, J.A., Stallings-Mann, M.L., Galvin, J.M., Anthony, R.V. and Roberts, R.M.

Steroid regulation of the synthesis and secretion of retinol-binding protein by the uterus of the pig.

Endocrinology 130, 1992: 2557-2564.

67

Turek, F.W. and Campbell, C.S. Photoperiodic regulation of neuroendocrine-gonadal activity. Biol.

Reprod. 20, 1979: 32-50.

Vallet, J.L., Christenson, R.K., Trout, W.E. and Klemcke, H.G. Conceptus, progesterone, and breed

effects on uterine protein secretion in swine. J. Anim. Sci. 76, 1998: 2657-2670.

Van der Meulen, J., Helmond, F.A and Oudenaarden, C.P.J. Effect of flushing of blastocysts on

days 10-13 on life-span of the corpora lutea in the pig. J. Reprod. Fert. 84, 1988: 157-162.

Wayne, N.L., Malbaux, B. and Karsch, F.J. Social cues can play a role in timing onset of the

breeding season of the ewe. J. Reprod. Fertil. 87, 1989: 707-7 Frank et al., 197713.

Wettemann, R.P. and Bazer, F.W. Influence of environmental temperature on prolificacy of pigs. J.

Reprod. Fertil. Suppl. 33, 1985: 199-208.

Wilson, M.R. and Love, R.J. The effect of the presence of older females on the maintenance of

pregnancy in gilts in the seasonal infertile period. In Proceedings: The International Pig Veterinary

Society, Scientific Committee of 11th International Pig Sciences Society Fritz Weibel, Thun 1990:

442.

Wrathall, A.E., Wells, D.E., Jones, P.C. and Foulkes, J.A. Seasonal variations in serum

progesterone levels in pregnant sows. Vet. Rec. 118, 1986: 685-687.

Zavy, M.T., Bazer, F.W., Thatcher, W.W. and Wilcox, C.J. A study of prostaglandin F2alpha as the

luteolysin in swine: V. Comparison of prostaglandin F, progestins, estrone and estradiol in uterine

flushings from pregnant and nonpregnant gilts. Prostaglandins 20, 1980: 837-51.

Xue, J.L., Dial, G.D., Marsh, W.E. and Davies, P.R. Multiple manifestations of season on

reproductive performance of commercial swine. J. Am. Vet. Med. Ass. 204, 1994: 1486-1489.