Pre-ovulatory Graafian follicles are cooler than neighbouring stroma in pig ovaries

R.H.F. Hunter1,4, I.B. Bøgh1, N. Einer-Jensen2, S. Müller3 and T. Greve1

1 Department of Clinical Studies—Reproduction, Royal Veterinary and Agricultural University, Copenhagen, 2 Department of Physiology and Pharmacology University of Southern Denmark, Odense and 3 Department of Molecular Animal Breeding and Genetics, Ludwig Maximilian University, Munich, Germany


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Using an infra-red camera, domestic pig ovaries were thermo-imaged almost instantaneously at laparotomy or within a closed abdomen by endoscopy. Rectal and jugular vein temperatures were recorded using thermo-probes. Graafian follicles (7–10 mm diameter) were cooler than ovarian stroma in all experimental models examined, and both compartments were cooler than rectal and jugular temperatures. The mean difference between follicles and stroma in 73 observations was 1.3 ± 0.1°C. When thermo-imaged under the fimbriated extremity of the Fallopian tube, follicles and stroma could still be distinguished. Follicles cooled slightly more rapidly than adjoining stroma during the first 10 s of a 60 s recording interval, after which curves for the two tissues remained parallel. Arresting ovarian blood supply for 5 min had a negligible influence on the temperature differentials. Endoscopy in three models recorded mean differentials between follicles and stroma of 0.6 ± 0.1°C to 1.1 ± 0.1°C. It is concluded that temperature gradients do exist in the ovarian tissues of mature animals, and that these are generated at least in part as a consequence of endothermic reactions within Graafian follicles.

Key words: counter-current heat exchange/endothermic reactions/follicular fluid/infra-red endoscopy/ovary/pig


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Gonadal function in mammals is highly specialized. Whereas the ancestral gonad was principally a liberator of gametes under conditions of external (aquatic) fertilization, the mammalian gonad has developed a wide-ranging and sophisticated endocrine function. Associated with such evolutionary specialization in a majority of male eutherians so far examined has been descent of the testes through the inguinal canals into a scrotal sac. The overwhelming purpose of this migration away from the abdomen is thought to be for cooling, although whether such migration was initially prompted by the requirements of the epididymis for maturing and storing spermatozoa below abdominal temperature (Glover and Nicander, 1971Go; Glover, 1973Go; Bedford, 1978Go) or of the gonad for protecting the labile processes of spermatogenesis and spermiogenesis from temperature-induced errors (Ehrenberg et al., 1957Go; Setchell, 1978Go; Waites and Setchell, 1990Go) remains uncertain. Even so, as demonstrated by surgical restoration of a testis to the abdomen (Griffiths, 1893Go; Glover, 1960Go) or by nature's own experiment of cryptorchidism, the gonad itself is critically sensitive to ambient temperature.

Turning to the female, there has clearly been no pressing requirement for the ovary to assume an extra-abdominal location, although a minor migration from its initial embryonic site close to the kidneys and adrenal glands is usually demonstrable (Patten, 1948Go; Byskov and Høyer, 1988Go). However, if one or more stages of gametogenesis are vulnerable to deep body temperature in the male, with an accompanying increase in the risk of germ cell mutation (Ehrenberg et al., 1957Go; Cowles, 1965Go), then questions arise concerning species with intra-abdominal testes (Millar and Glover, 1973Go; Harrison, 1975Go; Glover et al., 1990Go) and, in the present context, concerning ovarian physiology. In brief, even though situated deep within the abdomen, do ovaries or particular compartments thereof always function at deep abdominal temperature? Infra-red scanning in both rabbits and man revealed the existence of gradients in ovarian tissue temperatures. Rabbit pre-ovulatory follicles were 1.4 ± 0.2°C cooler than adjoining stroma as measured by the infra-red approach or using micro-electrodes introduced during mid-ventral laparotomy (Grinsted et al., 1980Go). Human pre-ovulatory follicles could be as much as 2.3°C cooler than ovarian stroma (Grinsted et al., 1985Go).

In domestic farm animals, ovarian temperatures were also examined with thermistor probes sited in the follicular antrum or adjacent stroma, an invasive approach open to criticism but which nonetheless suggested that pig Graafian follicles of 8–10 mm (pre-ovulatory) diameter could be >=1.0°C cooler than the stroma (Hunter et al., 1995Go). In a subsequent approach using infra-red thermo-imaging of pig ovarian tissues, gradients were again found between pre-ovulatory follicular temperatures and those of ovarian stroma, the mean follicular temperature being 1.7 ± 0.4°C cooler (Hunter et al., 1997Go). Comparable studies in cattle revealed a follicle–stroma difference of 1.5°C (Greve et al., 1996Go). This locally reduced temperature was considered to be significant during resumption of meiosis in pre-ovulatory oocytes, and was suggested to be of immediate relevance to procedures of in-vitro oocyte maturation and in-vitro fertilization (IVF) (Shi et al., 1998Go). Inappropriate temperatures might lead to subtle derangements in oocyte maturation processes, not least in the spectrum of nuclear and cytoplasmic proteins expressed during preimplantation development. Inappropriate culture temperatures could, therefore, offer one explanation for the seemingly low overall success of in-vitro maturation and IVF procedures in generating viable fetuses after transplantation of embryos to suitable recipients. Accordingly, it appeared important to continue with measurements of ovarian temperature close to the time of ovulation, with a specific focus on mature Graafian follicles.

In the experiments described below, we have developed and extended previous studies based on infra-red sensing of ovarian temperatures in domestic pigs. The body of evidence derived from these studies strongly supports the existence of temperature gradients in reproductive tissues deep within the abdomen of mature animals.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Experimental animals
A total of 40 animals was used in these studies; their distribution among experiments is summarized in Table IGo. The animals were crossbred Large WhitexLandrace, purchased from Rudolf Larsen Farms (Roskilde, Denmark) close to the onset of puberty, and housed as groups of four in open-fronted indoor pens in the Department of Clinical Studies, Royal Veterinary and Agricultural University, Copenhagen. They were bedded on straw, fed a standard commercial diet of pelletted concentrates twice daily, and given free access to water. During the winter months, natural daylight was supplemented with overhead lighting from 07.00 to 18.00.


View this table:
[in this window]
[in a new window]
 
Table I. Sequence and distribution of experiments in which temperature differentials between Graafian follicles and neighbouring ovarian stroma were examined
 
Oestrous cycles were monitored twice daily in the presence of a mature boar who was housed in an adjoining pen. In addition to behavioural cues, particular attention was paid to vulval colour and swelling and to the nature of escaping secretions. More frequent checks were made with the boar throughout pro-oestrus until the onset of a standing reflex. Body weight at the time of surgical intervention after two or three cycles ranged from 110 to 135 kg. None of the animals had previously been subjected to any form of experimentation. The study was approved by the Danish Animal Ethics Committee.

Pre-operative procedures, anaesthesia and surgery
All animals were operated on during pro-oestrus or oestrus in a pre-heated and sealed theatre (28–30°C). In addition to background heating, the theatre and contents had been warmed for at least the previous 18 h by a space heater. This was switched off only at the commencement of the following procedures. Theatre temperatures had fallen to 26–28°C by the onset of surgery.

Animals were starved overnight, injected intravenously with Mebumal (pentobarbital sodium; 50 mg ml–1) to induce anaesthesia and raised hydraulically onto an insulated operating table. Using a Rowson laryngoscope (Penlon Ltd, Andover, Hants, UK) and cuffed McGill tube (Warne Surgical Products, Andover, Hants, UK), intubation was achieved with the animal on its side after which it was positioned on its back under semi-closed circuit administration of isofluran (Forene; Abbott Labs Ltd, Abbott Scandinavia, Kista, Sweden) and oxygen (1.0–1.5 l/min). Once full surgical relaxation was achieved, anaesthesia was routinely maintained at 2.0–2.5% isofluran in oxygen for the duration of the experiment (~1.0 h).

The abdominal field of operation was cleaned, shaved, sterilized and fully-draped. Strict aseptic procedures then followed throughout. A mid-ventral laparotomy was performed with temperature recordings (see below) on the body wall and sub-peritoneal fat. Using an incision no longer than 12 cm, the peritoneum was opened, the reproductive tract located swiftly, and an ovary brought into the incision site with minimum traction. Ovaries were visualized primarily by reflecting the body wall and peritoneum laterally rather than by lifting individual gonads to the level of the body surface. After almost instant infra-red sensing of temperatures with the fimbria in situ, this covering was displaced and the ovarian surface thermo-imaged whilst most of the reproductive tissues remained within the abdomen. The first ovary—picked at random—was then replaced and the same procedures followed for the contralateral ovary. A fine silk marker thread placed in the mesovarium was used to distinguish the first ovary. The incision was closed temporarily using haemostats applied to the skin and the reproductive organs left for a period of 5min for temperature equilibration before further recording. For the purposes of this report, a `sensing' was an individual temperature measurement; an `experiment' involved one or more sensings during the same surgical intervention.

Extensive measurements were made of the rate of ovarian cooling upon surgical exposure. Such measurements followed the initial sensing of ovarian temperatures in a total of 17 animals, and the period of equilibration. Cooling was observed over a period of ~20 s or, in a further group of six animals, during precisely 30 s with measurements at 10 s intervals. In a final group of three animals, the rate of cooling was assessed during a period of 60s (Table IGo). In these experiments, the ovary under observation rested within the incision site with the minimum of traction applied to its ligaments.

In two further sets of observations involving a total of 16 animals, the influence on ovarian tissue temperatures of arresting blood flow was examined by placing an atraumatic tubular rubber ligature around the ovarian pedicle after the initial readings (Figure 1Go). The ovary was then equilibrated in a closed abdomen for 5 min before infra-red sensing. It was treated similarly after removal of the rubber ligature. In each animal, only the ovary with the larger number of prominent follicles was examined in this manner.



View larger version (162K):
[in this window]
[in a new window]
 
Figure 1. (a) Porcine ovary exposed during mid-ventral laparotomy with the fimbriated infundibulum displaced to demonstrate the pre-ovulatory cohort of follicles of 8–10 mm diameter. An atraumatic rubber ligature has been positioned to arrest ovarian bloodflow. (b) Porcine ovary exposed in another animal during mid-ventral laparotomy to show mature Graafian follicles and corpora albicantia of the previous oestrous cycle. Blood flow has now been arrested by the fully tightened rubber ligature. (c) A further preparation with mature follicles and regressed corpora revealing the progressive influence on these tissues of arresting blood flow. Ovaries were restored to the abdomen and the incision closed by means of wound clamps placed on the skin for a full 5 min before a subsequent temperature recording (the marking so caused was temporary and had disappeared by completion of the operation). (d) Ovary with fimbriated infundibulum displaced to demonstrate the influence of arresting ovarian blood flow by means of an atraumatic rubber ligature for a full 5 min. Upon removal of the rubber ligature, the ovary rapidly resumed a normal healthy appearance and colour.

 
In three experiments (Table IGo), unilateral ovariectomy was performed after the initial temperature recordings. The supporting tissues were clamped close to the ovary and anchored ligatures of braided silk (Mersilk; Ethicon No. 3, Edinburgh, Scotland) placed around the mesovarium. The ovary was detached and then equilibrated in a closed abdomen for at least 5 min before infra-red sensing it whilst sitting loosely in the incision site.

In a further three experiments, the liberated ovary was suspended on a braided silk thread, plunged into liquid nitrogen at –196°C to deep freeze, thawed in warm phosphate-buffered saline (37°C), and re-equilibrated in a closed abdomen for at least 5 min before withdrawal into the incision site for sensing.

Post-operative procedures
After completion of all infra-red recording, and close inspection and photography of individual ovaries, the abdominal incision was closed in three layers. The peritoneum was joined with a continuous suture whereas interrupted sutures were placed in the body wall and skin. Animals remained in the post-operative pens of the Clinical Department. All recovered without incident. None was operated on a second time.

Infra-red temperature sensing
Temperatures were recorded using an infra-red camera for thermo-imaging. The camera was mounted on a fully adjustable tripod and positioned ~50 cm directly above the abdomen. Prior focusing enabled almost instantaneous (<2 s) recording of temperatures upon exposure and orientation of ovarian tissues.

The infra-red camera (THV550, without filter) was manufactured by Agema Thermovision (Stockholm, Sweden) and made available, together with associated computing equipment, by Praecisions Teknik in Copenhagen (DK-2610 Rødovre, Denmark). The computerized recording system enabled post-operative retrieval of individual pictures and analysis of all experimental data. The temperature scale was precise to within ±0.1°C, and the camera and recording system were periodically verified for such precision by Agema in Sweden. The temperature scale could be visualized in colour on a TV monitor screen during the surgical procedures, and the results could subsequently be recalled in colour for analysis of tissue differences (Figure 2Go). Precise pin-pointing of Graafian follicles or neighbouring stroma was thereby possible on screen, with both tissues visible in the same picture. All infra-red measurements and subsequent downloading of information were made by the same individual (I.B.B.). At least three follicular and three stromal observations were made per ovary for calculation of mean temperature differences.



View larger version (111K):
[in this window]
[in a new window]
 
Figure 2. Thermo-imaging of ovary to illustrate cooler Graafian follicles when compared with ovarian stroma (see colour scale to the right of picture). Infra-red sensing of tissues during mid-ventral laparotomy was almost instantaneous upon exposure.

 
Endoscopy
Because conventional endoscopes with quartz fibres do not transmit infra-red irradiation, a special endoscope with appropriate lenses was designed and constructed by Agema in Sweden for use in conjunction with the thermo-camera. In brief, it was of stainless steel, 36 cm longx3.25 cm external diameter (2.5 cm lens diameter) and attached directly to the camera. The final lens was recessed some 2.0 cm back from the external rim of the scope to protect it from body fluids. The endoscope was autoclaved on each occasion and pre-warmed (38.5°C) immediately before use. Application of anti-condensation liquid (Synoptik; Scanbech, Copenhagen, Denmark) on the final lens also diminished fogging. This liquid had been tested prior to the experiment to confirm that absolute temperatures were not influenced. The endoscope was introduced into the abdomen in the mid-line through an incision made marginally smaller than its own dimensions, thereby providing a gas tight seal. Final positioning relative to the ovaries was assisted by conventional quartz-fibre endoscopy. The ovaries themselves were not touched directly during these procedures although the reproductive tract had to be displaced forwards with the animal tilted head downwards. In this position, endoscopic forceps on the mesovarium enabled both orientation of an ovary and displacement of the fimbriated extremity of a Fallopian tube.

Jugular and deep rectal temperatures
Body temperatures were recorded during surgery by means of two Cu/CuNi thermo-probes (Ellab, DK-2610 Rødovre, Denmark). One (50 cm long, 0.8 mm diameter) was inserted its full length into the jugular vein through a short intravenous catheter (Venflon, 17 gauge, 45 mm) in a prominent ear vein; the other, a standard rectal probe, was positioned at least 10–15 cm into the rectum. The probes were connected to a four channel data acquisition and analysis system (Ellab, Rødovre, Denmark; CMC 96 + IBM Think Pad). Temperatures were measured, collected and stored every 2 s with an accuracy of within ±0.1°C.

Insertion into the jugular vein via an ear vein had to be abandoned in four cases because of failure to pass the intravenous probe around the contour of the vessel at the base of the ear. Successful recordings were obtained from 14 animals.

Experimental design
The distribution of animals used for the various individual measurements is summarized in Table IGo. A sequence of measurements was usually made in each animal, so addition of the numbers given in Table IGo does not give the total number of animals in the study.

Statistical analysis
Results were analysed by means of Microsoft Excel version 8.0 for Apple computers. The progamme was used to generate standard deviations, standard error of the mean, t-tests and in the plotting of graphs and tables.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Instant temperature measurements at laparotomy
The summarized results of infra-red sensing of ovarian temperatures at mid-ventral laparotomy are presented in Figure 3Go. Each result used for calculation was based on a mean of at least three observations on follicles and three of stromal tissues in the first field of view. The results were based on a total of 73 sensings in 21 animals and there were no exceptions to the finding that mature Graafian follicles (7–10 mm diameter) were always cooler than adjoining stromal tissues. Indeed this cardinal point remained true for all the experimental models examined in this study.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 3. Temperature differential for ovarian stroma plotted against large Graafian follicles. The solid line indicates temperature equality of the two tissues. In no instance did follicular temperature exceed that of stroma (n = 73).

 
The mean temperature difference (± SEM) between the follicles and stroma was 1.4 ± 0.1°C (n = 33) for the first ovary to be exposed and 1.2 ± 0.1°C (n = 40) for the second ovary, the overall mean difference being recorded as 1.3 ± 0.1°C (n = 73). The range in absolute temperatures is also shown in Figure 3Go and requires immediate comment. Animals received no premedication, had to be walked to the operating theatre (frequently involving manual assistance), and had to be restrained until the intravenous anaesthetic had taken effect. The excitement generated in some animals during these procedures in an already heated operating theatre would be reflected in the recorded temperatures. In fact, these were in line with the jugular vein and deep rectal temperatures.

In three animals in which recent ovulations rather than mature follicles were observed, within an estimated 4–6 h of ovulation, the collapsed haemorrhagic follicles were also cooler than adjoining stromal tissues. As determined in both ovaries, recent ovulation sites were 1.5 ± 0.2°C (range 0.9–2.2°C) cooler than ovarian stroma.

The influence of sensing the ovaries enveloped in the fimbriated extremity of the Fallopian tube in five animals compared with actual ovarian surface temperatures upon displacing the fimbria is shown in Figure 4Go. Although follicles were always cooler than stroma (38.5 ± 0.1°C versus 39.4 ± 0.1°C with the fimbria in situ), the apparent stromal temperature drop upon displacing the fimbria from the first exposed ovary (designated ovary 1) was 0.1°C greater than the follicular drop (data not shown). However, this difference was not apparent when the results from both ovaries were summarized.



View larger version (63K):
[in this window]
[in a new window]
 
Figure 4. Ovarian tissue temperatures (± SEM) recorded on both ovaries in oestrous animals (n = 5), indicating the difference between follicle and stromal compartments when the ovary was covered with the fimbriated infundibulum (WF) or when this was displaced (WOF).

 
Cooling rates of intact ovaries
An initial set of continuous registrations in nine animals assessed the extent of cooling over ~20 s (the duration differs slightly due to variation in access time to reach the free space available on the hard disk in the thermo-camera). As observed in 15 ovaries with suitable follicles in these nine animals, mature follicles were 1.4 ± 0.1°C cooler than stroma at the start of a mean 21.6 s cooling period (range 17.3–27.0 s) and 2.3 ± 0.2°C cooler at the end of this period. Follicles cooled by a mean of 1.9°C during this period and stroma by a mean of 1.0°C.

Further sets of cooling rate measurements in a new group of five animals were made at precisely 10 s intervals for a period of 30 s. These results are shown in Figure 5Go. The initial rate of cooling of mature follicles slightly exceeded that of stroma although the overall trend in cooling was essentially parallel.



View larger version (11K):
[in this window]
[in a new window]
 
Figure 5. Mean cooling rates (± SEM) for ovarian stroma and Graafian follicles in a group of five animals when exposed at mid-ventral laparotomy and thermo-imaged for a period of 30 s.

 
Follicles in pig nos. 1562 and 1592 and recent ovulations in pig no. 1557 all showed an initial rate of cooling which exceeded that of stroma, after which the curves became parallel. Figure 6Go illustrates the rate of cooling in one of three animals measured over a period of 60 s. Figure 7Go plots a cooling rate in one animal for ovaries resting on the reproductive tract deep within the incision site, and as visualized through the infra-red transmitting endoscope. This approach clearly acted to reduce the rate of cooling, which again appeared more or less parallel for the two tissues.



View larger version (12K):
[in this window]
[in a new window]
 
Figure 6. Mean cooling rates (± SEM) for ovarian stroma and Graafian follicles in pig no. 1562 when exposed at mid-ventral laparotomy and thermo-imaged for a period of 60 s.

 


View larger version (11K):
[in this window]
[in a new window]
 
Figure 7. Mean cooling rates (± SEM) of three distinct areas of ovarian stroma and in at least three large Graafian follicles in one animal, as determined with an endoscope introduced into the abdomen via a mid-ventral incision.

 
Extinguishing the surgical spotlights and background lighting in the operating theatre had no detectable influence on ovarian temperature gradients.

Influence of ovarian ligation
In the study involving placement of tubular rubber ligatures around the ovarian pedicle to arrest the blood supply (Figure 1Go), initial sensing of the ovaries recorded a mean temperature difference between follicles and stroma of 1.32 ± 0.1°C (n = 10). After precisely 5 min of ligation, during which the follicles had turned a deep blue in colour, the temperature difference was reduced to 1.27 ± 0.2°C. However, when blood flow resumed for a further 5 min during abdominal incubation — which restored the normal healthy appearance of the follicles — the differential was measured as 1.30 ± 0.1°C (Figure 8Go). Accordingly, ligation for 5 min had a rather minor influence on the ovarian tissue temperature differentials.



View larger version (42K):
[in this window]
[in a new window]
 
Figure 8. Histograms to illustrate the respective temperatures (± SEM, values shown above bars) of Graafian follicles and ovarian stroma before and during ligation of the ovarian blood supply for 5 min, and after removal of the ligature followed by 5 min of equilibration in the abdomen (n = 10).

 
Observations following the same procedures in a further group of six animals are presented in Figure 9Go. Although the follicle–stromal differentials were immediately apparent, the slight modifications in mean tissue temperatures during ligation and after restoration of the blood supply and abdominal incubation were of special interest: in one animal, we observed a prominent `rebound' in tissue temperatures after blood supply restoration. In this animal, the mean follicular rebound was 1.0°C compared with a mean stromal rebound of 0.3°C. The physiological significance of this `rebound' is uncertain, but it may have been associated with restoration of oxygen and metabolic substrates to the temporarily starved tissues.



View larger version (30K):
[in this window]
[in a new window]
 
Figure 9. Histograms to illustrate the respective temperatures (± SEM, values shown above bars) of Graafian follicles and ovarian stroma before and during ligation of the ovarian blood supply for 5 min, and after removal of the ligature followed by 5 min equilibration in the abdomen (n = 6).

 
Influence of ovariectomy
Initial observations on ovariectomy involved the previously reported approach (Hunter et al., 1997Go). The ovary with the largest number of pre-ovulatory follicles was removed, plunged into liquid nitrogen (–196°C) to kill the tissues, thawed in sterile phosphate-buffered saline at 37°C, and restored to the abdomen. Detailed observations were made in three animals. After 5 min of incubation in the closed abdomen, the detached ovary was withdrawn into the incision site and sensed immediately by the infra-red camera. There was no detectable temperature difference between the mature follicles and adjoining stromal tissue. Such a dead, frozen–thawed ovary was indistinguishable macroscopically from healthy, living tissue after ovariectomy.

In an extension of this approach, tissues were not killed. Rather, the detached ovary was left in the incision site and temperature-sensed for a period of 30 s (two animals) or 60 s (one animal). After such recording, it was suspended in the abdomen and the incision closed for 5 min. Then, once again, the ovary was withdrawn and temperatures recorded for 30 or 60 s. In 18 sets of observations (composed of 11 sets of observations in two animals during 30 s and seven sets of observations in one animal during 60 s), the ovarian temperature differentials were maintained or increased between mature follicles and stroma, although of course the tissues were cooling (Figure 10Go). There was a close parallel between these experimental observations on ovarian cooling rates in living tissues after ovariectomy and those in ligated ovaries as reported above.



View larger version (20K):
[in this window]
[in a new window]
 
Figure 10. The respective cooling rates of ovarian stroma and Graafian follicles before and after ovariectomy (OVX), the latter being determined with the detached ovary (OV) resting in the incision site. A total of 18 sets of observations was made in three animals: 11 sets were made in two animals during 30 s; seven sets were made in one animal during 60 s.

 
Preliminary endoscopy
In a preliminary trial (June 1998) to test the image building capability of a prototype endoscope, readings were made after initial laparotomy studies, having first closed the abdomen and permitted temperature equilibration for at least 5 min. In this approach, the ovary was not exteriorized. Instead, the endoscope was introduced via the previous incision and used to scan the ovary pre-positioned and resting on the reproductive tract below the level of the body wall. For three animals examined in this manner, the temperature differentials between large follicles and stroma are shown in Table IIGo.


View this table:
[in this window]
[in a new window]
 
Table II. Follicle and stromal temperatures measured by means of endoscopy or direct thermo-imaging with the infra-red camera. In the June and December 1998 studies, the ovaries were sensed by the endoscope deep within a mid-ventral incision whereas in the January 1999 series, the endoscope was introduced into the abdomen through a small gas-tight incision
 
The mean follicle–stroma difference for the five endoscopic readings was 1.1 ± 0.1°C, the same as the initial mean value obtained by direct sensing at laparotomy with the fimbria displaced. However, it should be stressed that this initial approach with the endoscope was certainly not gas-tight, for the ovary was simply viewed deep within the incision site.

In a further series of preliminary measurements (December 1998) to test a lens coating (anti-fogging) fluid (Synoptik) on the endoscope, thereby avoiding any problem of condensation, six sets of measurements were made on the two ovaries in one animal (no. 1558) with excellent pre-ovulatory follicles via a mid-ventral incision before exposing the ovary for conventional infra-red sensing. A 5 min period of temperature equilibration with the abdomen closed followed each pair of readings. The individual results are shown in Table IIGo. The mean follicle–stroma difference in this animal was 0.7 ± 0.1°C compared with a mean difference by direct scanning of the ovary in the incision site of 1.6 ± 0.1°C. This scanning was made after completion of the endoscopy readings and after a further 5 min of equilibration.

Due to scheduling problems of bringing the operating team from Odense, Munich and Edinburgh together in Copenhagen in January 1999, and because the six animals reserved for the experiment had not remained on precisely 21 day oestrous cycles from their last recorded heats in December 1998, a final set of preliminary endoscopy experiments had to be performed on Graafian follicles not exceeding 7 mm in diameter or on very recently ovulated follicles. The pre-heated infra-red endoscope was introduced through a gas-tight incision and, with a minimum of insufflation (air, not CO2), the ovaries were located with a quartz-fibre endoscope but not touched with the endoscopic forceps (Figure 2Go). Equilibration of the instruments occurred during the initial location and orientation of the gonad whilst the animal was angled head-down on an operating table tilted at ~24° below the horizontal. The mean temperature differential recorded between the 5–7 mm diameter follicles or recent ovulations and adjoining stroma in the six animals was 0.6 ± 0.1°C (Table IIGo). A larger differential would have been anticipated with Graafian follicles of a pre-ovulatory 9–10 mm diameter.

Urinary bladder temperatures
The moist surface of the distended urinary bladder was fully visible and sensed immediately upon opening the peritoneum during laparotomy in three animals. In the first animal, the bladder temperature was recorded as 39.1°C compared with a mean for the stroma of 39.2°C and for the follicles of 38.0° and 38.1°C on the two ovaries. In a second animal, the bladder temperature was recorded as 39.4°C, closely similar to the ovarian stroma (39.2°C), whereas the follicular mean was 38.0°C. In a third animal, the bladder temperature was again recorded as 39.4°C, the mean stromal temperature also 39.4°C, whereas the mean follicular temperature was 38.0°C. Hence, the surface temperatures of an exposed bladder did not match those of exposed follicles.

Jugular vein and deep rectal temperatures
Although no difference was recorded in three of 14 animals studied, 11 animals had a higher deep rectal than jugular vein temperature (0.25–1.0°C higher, usually 0.5°C) when these sites were measured simultaneously. The jugular temperature tended to remain reasonably constant whereas the rectal temperature either increased or decreased during surgery, suggestive of a thermo-regulatory mechanism. These findings are the subject of a separate report (Einer-Jensen et al., 1999Go). Ovarian follicular and stromal temperatures were invariably cooler than deep rectal temperatures at the start of recording (Figure 11Go).



View larger version (25K):
[in this window]
[in a new window]
 
Figure 11. The deep rectal, jugular and overall ovarian temperatures in one pig measured by indwelling thermo-probes. The recording period extended for ~1h 10min.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Taken together, the above body of evidence strongly supports the existence of temperature gradients in pig ovarian tissues deep within the abdomen. As already noted, there was not a single occasion when mature Graafian follicles were not cooler than neighbouring ovarian stroma. Such follicles were cooler than stroma in all the experimental models employed in the present study, and ovarian tissues were always initially cooler than deep rectal and jugular vein temperatures. These overwhelmingly consistent findings would seem to be of fundamental importance. The topic could be approached from various angles, but an original one would be to ask the question: `Why do pre-ovulatory follicles grow so large, in the present instance to a diameter of 8–10 mm?' Responses could conceivably concern the endocrine functions of the follicle or preparations for eventual discharge of the oocyte at ovulation. A more inspired response to the question might be to facilitate the generation of intra-ovarian temperature gradients, that is to enable expression of cooling reactions within appropriate follicles. Growth of a Graafian follicle will alter the ratio between the volume of fluid in the antrum and the volume of cells lining the structure. Whereas cells generate heat, a mature follicle is very largely a fluid reservoir in a relatively static condition. Accordingly, changing the ratio of the two components could be related to the development of temperature gradients.

Turning to the actual experiments, and in a context of possible artefactual cooling of ovarian tissues exposed in the operating theatre, the fact that temperature gradients existed when the ovaries were tightly enveloped by the undisturbed fimbriated extremities of the Fallopian tube comes as valuable supportive evidence. In this condition, loss of heat by evaporation from moist protruding follicles would be minimized by the intimate embrace of these thin but well-vascularized membranes. In other words, the fimbriae should act to retard or prevent surface cooling. The fact that follicle–stroma differences could be sensed at all with the fimbria in situ must be because the underlying ovarian structures were in immediate contact with portions of the membrane and were thereby able to transmit locally the influence of their own temperatures. This is even more impressive when the prominent capillary bed of such engorged peri-ovulatory fimbriae is taken into account.

Measurements of cooling rates for exposed ovaries are also instructive. If 7–10 mm diameter Graafian follicles protruding from the ovarian surface require ~20 s to cool by 1.5°C (Figure 5Go), then it is highly improbable that an almost instantaneous recording of a mean 1.3°C difference between pre-ovulatory follicles and neighbouring stroma could be accounted for simply by artefact. Although follicular tissue cooled initially more rapidly than stroma, perhaps in part because of the topography of follicles compared with the more compact adjoining tissues, Figures 5 and 6GoGo nonetheless reveal a significant temperature differential between the respective tissues during the first few seconds of recording. Furthermore, the subsequent rate of cooling of these tissues between 10 and 60 s is almost parallel, indicating maintenance of a temperature differential between large follicles and stroma. If the reasonable assumption is made that mild traction on the ligaments and mesenteries to exteriorise an ovary did not perturb follicular blood supply to a markedly greater extent than the stromal supply, then the conclusion must be that reactions are occurring in large Graafian follicles that actively generate such temperature differentials by removing heat. Establishing the precise nature of such putative endothermic processes is an important research objective.

Further lines of evidence for the existence of intra-ovarian temperature gradients come from the ligation experiments and those involving ovariectomy. Arresting blood flow for a period of 5 min while keeping the ovaries in a closed abdomen did not alter significantly the temperature differential between Graafian follicles and stroma. Indeed, the mean differential after 5 min ligation was almost identical with that found after removal of the ligature and a further 5 min of abdominal equilibration. The favoured interpretation here is that arrest of blood flow during a period of 5 min was insufficient to damage irreversibly or arrest the putative endothermic mechanisms that generate such gradients in the physiological situation; hence the consistent histograms in Figure 8Go. This interpretation for living tissues receives support from the ovariectomy data in which the differentials between large follicles and stroma were essentially maintained during a 60 s period of cooling of live tissues (Figure 10Go). By contrast, when ovarian tissues were killed upon plunging into liquid nitrogen and then thawed and sensed by the infra-red camera, the follicular-stromal temperature gradients were completely removed; these results thus confirm previous ones (Hunter et al., 1997Go). This evidence, based on prompt examination of ovariectomized tissues that had previously been sensed in situ, is also supportive of endothermic reactions in living follicles, for which there is now physico-chemical evidence. An abstract reporting preliminary studies (Luck and Griffiths, 1998Go) recorded a decrease in temperature in bovine follicular fluid subjected to saline dilution within an appropriate incubator; the decrease in temperature (0.14–0.2°C) was sustained for 7–13 min. It was concluded that such follicular fluid derived from undated slaughterhouse ovaries showed a net uptake of heat energy when diluted with saline, seemingly indicative of an endothermic reaction resulting from hydration of a large molecular weight component.

A current interpretation is that diverse components within mature Graafian follicles contribute to the intra-ovarian temperature gradients rather than the activities of a `single' molecule. This hypothesis may be incorrect but it would seem that the mature follicle is in such a dynamic condition— actively growing, remodelling, synthesizing and secreting, and doubtless influenced in a key way by its antral oocyte— that there is scope for more than one source of endothermic reaction. The present studies offer no guidance as to the precise nature of such reactions nor do current thoughts extend beyond our previous suggestions (Hunter et al., 1997Go). These included the synthesis and secretion of steroid hormones, prostaglandins and diverse peptides (including growth factors) and proteins (including heat shock proteins); extensive remodelling of follicular tissues including angiogenic proliferation of the thecal capillary bed; increased coagulability of the follicular fluid; and expansion and mucification of the cumulus oophorus. The last point remains of specific interest in relation to the hydration of proteoglycans — the principal intercellular cement substance — during pre-ovulatory loosening of follicular tissues and, latterly, during preparation for release of a secondary oocyte through a degraded follicular wall aperture. Putative endothermic reactions in Graafian follicles may still be active shortly after ovulation since very recently ovulated follicles also showed reduced temperatures compared with stroma in this study, in contrast to previous findings (Hunter et al., 1997Go).

If the assumption is made that endothermic reactions are indeed involved in the establishment of intra-ovarian temperature gradients, as previously considered (Grinsted et al., 1980Go), then a means must also be put forward for their maintenance deep within the abdomen. Two physiological systems may need to be functional to avoid local cooling of follicles being overridden by an influence of the systemic circulation. First, a counter-current exchange of heat might be required in the ovarian pedicle so that returning venous blood could cool the incoming arterial blood. Second, there would certainly need to be a counter-current exchange of heat in the blood supply to individual Graafian follicles (Figure 12Go). A counter-current exchange of heat in the ovarian plexus can be supported by the highly coiled and intimate arrangement of the ovarian artery on the ovarian vein. This shows close parallels with the male gonadal blood supply in which a counter-current heat exchange has been clearly demonstrated in species with scrotal testes (Setchell, 1978Go; Waites and Setchell, 1990Go; Glad Sørensen et al., 1991Go). A separate line of argument would be that the ovarian plexus facilitates a counter-current exchange of hormones between vein and artery, especially of relatively small molecules such as prostaglandins and steroids, as well as inert gases (McCracken, 1971Go; McCracken et al., 1971Go; Einer-Jensen, 1988Go, 1992Go), so there would seem to be no problem in principle for other forms of transfer across the walls of the ovarian artery and vein. As for the blood supply to individual follicles, although experimental demonstrations of hormone or gaseous transfer have yet to be made at such a level, follicular drawings (Andersen, 1926Go) in pigs and more recent scanning electron micrographs of ovarian vascular corrosion casts (Macchiarelli et al., 1997Go) in rabbits demonstrate the intimate network of blood vessels and lymphatics and thus the potential for an exchange of heat — and thereby the local maintenance of temperature gradients.



View larger version (72K):
[in this window]
[in a new window]
 
Figure 12. Diagrammatic representation of an ovary to indicate potential regions of heat exchange (i) in the blood supply to individual Graafian follicles and (ii) perhaps also in the ovarian pedicle where a counter-current transfer of hormones is widely accepted to occur. Arteries = solid fill; veins = stippled.

 
A practical context in which the results of the present experiments could be significant concerns in vitro technology. Information bearing on the precise nature of the putative endothermic reactions in Graafian follicles could be of direct relevance to systems of in-vitro oocyte maturation and fertilization that aim to achieve a more physiological maturation of follicular oocytes for subsequent use in procedures of IVF. Work on in-vitro maturation and IVF received a major stimulus in domestic species when it was found that incubation of maturing oocytes together with a sperm suspension at 39°C promoted a far higher incidence of fertilization than at 37.5°C or 38.0°C (Cheng et al., 1986Go). Indeed, attempts to obtain IVF in the large farm animal species at the lower temperatures achieved poor success prior to this modification. Whilst incubation at 39°C clearly facilitated sperm hyperactivation and penetration of the oocytes, this temperature may not have been beneficial for the oocyte and for the young embryo before transplantation into a recipient. The overall success from these techniques in yielding viable fetuses is only 15–20%, notably in humans, for whom there is by far the most extensive organization of IVF clinics. One interpretation of the low success rates is that although 39.0°C is beneficial for sperm penetration, incubation of secondary oocytes at this temperature perturbs the vitelline content of macromolecules, especially the proteins involved in subsequent development (Hunter, 1990Go; Hunter et al., 1997Go). Although fertilization and the first steps of cleavage may occur on a seemingly normal morphological basis, the major organizational steps required following hatching of a blastocyst would be compromised, leading to greater mortality. Accordingly, the observation that the incidence of bovine blastocyst formation, for example, is altered in only a minor way at temperatures of 37.0–38.5°C does not pose a contradiction (Shi et al., 1998Go). Transplantation into a recipient would have been required to reveal developmental deficiencies presumed to be expressed during the major steps of organogenesis.

In conclusion, what is clear from the body of evidence accumulated in these studies is that mature Graafian follicles are cooler than adjoining ovarian stroma, and that both compartments are cooler than jugular vein and deep rectal temperatures. What is not clear is the true extent of the temperature difference between follicles and stroma: is it the overall mean of 1.3°C revealed at laparotomy in a pre-heated operating theatre or the initial endoscopic mean of 1.1°C made following mid-ventral studies, or the final mean value of 0.6°C obtained after endoscopy without prior intervention (Table IIIGo)? The last value may be closer to the physiological differential, although it should be emphasized that this value comes from observations on small follicles (<7 mm diameter) representing only days 17–18, not day 21 of a normal oestrous cycle. A realistic suggestion might be that the value for pre-ovulatory porcine follicles of 9–10 mm diameter lies somewhere between 0.6 and 1.1°C. Even so, the caveat must be added that all of the methods employed in these studies were unphysiological. All were invasive to a greater or lesser degree, and all involved anaesthesia with the animal inverted on the operating table — an approach that could have influenced both ovarian bloodflow and the partition of such bloodflow. Although we plan to continue with the endoscopic approach in future studies using a modified instrument, the most satisfactory means of overcoming the above limitations would be some form of non-invasive whole body scanning of a fully conscious animal. Whilst this remains an important objective, current techniques do not offer the necessary precision and sensitivity for monitoring the temperature of individual Graafian follicles.


View this table:
[in this window]
[in a new window]
 
Table III. An overall summary of the differences in temperature recorded between Graafian follicles and neighbouring ovarian stroma when examined by different experimental approaches
 


    Acknowledgments
 
We wish to thank the following colleagues at the Royal Veterinary and Agricultural University: Dr Mette Schmidt for arranging the supply of animals; and Mrs Inger Heinze, Mrs Anni Petersen, Mrs Bente Synnestvedt, Mr Niels Raunkjaer and Mr Anders Andersen for technical assistance. We also thank Mr Thomas Dresler of Praecisions Teknik, Rødovre, Copenhagen, for making the infra-red apparatus available. Professor Hilary Dobson, Dr Martin Luck and Dr George Mann commented helpfully on a draft of the manuscript, and Mrs Frances Anderson kindly prepared the typescript. R.H.F.H. gratefully acknowledges the award of a Guest Professorship by the Carlsberg Foundation in Copenhagen. The Danish Agricultural and Veterinary Research Council sponsored the infra-red camera and construction of the endoscope, and the Danish Biotechnology Programme covered the cost of consumables.


    Notes
 
4 To whom correspondence should be addressed at:32 Gilmour Road, Edinburgh EH16 5NT, UK

Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Andersen, D.H. (1926) Lymphatics and blood vessels of the ovary of the sow. Contrib. Embryol. Carneg. Instn Wash., 17, 107–123.

Bedford, J.M. (1978) Anatomical evidence for the epididymis as the prime mover in the evolution of the scrotum. Am. J. Anat., 152, 483–508.[ISI][Medline]

Byskov, A.G. and Høyer, P.E. (1988) Embryology of mammalian gonads and ducts. In Knobil, E., Neill, J. et al. (eds), The Physiology of Reproduction. Raven Press, New York. Ch. 8, pp. 265–302.

Cheng, W.T.K., Polge, C. and Moor, R.M. (1986) In vitro fertilisation of pig and sheep oocytes. [Abstr.] Theriogenology, 25, 146.[ISI]

Cowles, R.B. (1965) Hyperthermia, aspermia, mutation rates and evolution. Quart. Rev. Biol. 40, 341–367.[ISI]

Ehrenberg, L., von Ehrenstein, G. and Hedgram, A. (1957) Gonadal temperature and spontaneous mutation rate in man. Nature, 180, 1433.[ISI][Medline]

Einer-Jensen, N. (1988) Counter-current transfer in the ovarian pedicle and its physiological implications. Oxford Rev. Reprod. Biol., 10, 348–381.[ISI][Medline]

Einer-Jensen, N. (1992) Counter-current transfer between blood vessels in the ovarian adnex. Acta Obstet. Gynecol. Scand., 71, 566–567.[ISI]

Einer-Jensen, N., Hunter, R.H.F., Bøgh, I.B. and Greve, T. (1999) Temperature gradients between the head and body in pigs. J. Anim. Physiol. Nutrition. In press.

Glad Sørensen, H., Lambrechtsen, J. and Einer-Jensen, N. (1991) Efficiency of the counter-current transfer of heat and 133Xenon between the pampiniform plexus and testicular artery of the bull under in vitro conditions. Int. J. Androl., 14, 232–240.[ISI][Medline]

Glover, T.D. (1960) Spermatozoa from the isolated cauda epididymidis of rabbits and some effects of artificial cryptorchidism. J. Reprod. Fertil., 1, 121–129.[ISI]

Glover, T.D. (1973) Aspects of sperm production in some East African mammals. J. Reprod. Fertil., 35, 45–53.[Medline]

Glover, T.D. and Nicander, L. (1971) Some aspects of structure and function in the mammalian epididymis. J. Reprod. Fertil. (Suppl. 13), 39–50.

Glover, T.D., D'Occhio, M.J. and Millar, R.P. (1990) Male life cycle and seasonality. In Lamming, G.E. (ed.), Marshall's Physiology of Reproduction, 4th edn. Reproduction in the Male. Churchill Livingstone, Edinburgh, Vol. 2, pp. 213–378.

Greve, T., Grøndahl, C., Schmidt, M. et al. (1996) Bovine preovulatory follicular temperature: implications for in vitro production of embryos. Arch. Tierzuch., Dummerstorf, 39 (Special Issue), 7–14.

Griffiths, J. (1893) The structural changes in the testicle of the dog when it is replaced within the abdominal cavity. J. Anat., 27, 483–500.

Grinsted, J., Blendstrup, K., Andreasen, M.P. and Byskov, A.G. (1980) Temperature measurements of rabbit antral follicles. J. Reprod. Fertil., 60, 149–155.[Abstract]

Grinsted, J., Kjer, J.J., Blendstrup, K. and Pedersen, J.F. (1985) Is low temperature of the follicular fluid prior to ovulation necessary for normal oocyte development? Fertil. Steril., 43, 34–39.[ISI][Medline]

Harrison, R.G. (1975) Effect of temperature on the mammalian testis. In Hamilton, D.W. and Greep, R.O. (eds), Handbook of Physiology. American Physiological Society, Washington DC, Vol. 5, pp. 219–223.

Hunter, R.H.F. (1990) Fertilisation of pig eggs in vivo and in vitro. J. Reprod. Fertil., 40 (Suppl.), 211–226.

Hunter, R.H.F., Grøndahl, C., Schmidt, M. and Greve, T. (1995) Graafian follicles are cooler than ovarian stroma. J. Reprod. Fertil., 16 (Abstr. Ser.), 8–9.

Hunter, R.H.F., Grøndahl, C., Greve, T. and Schmidt, M. (1997) Graafian follicles are cooler than neighbouring ovarian tissues and deep rectal temperatures. Hum. Reprod., 12, 95–100.[Abstract]

Luck, M.R. and Griffiths, S.D. (1998) Endothermic response of follicular fluid to aqueous dilution. J. Reprod. Fertil. (Abstr. Ser.), No. 22, 12–13.

McCracken, J.A. (1971) Prostaglandin F2{alpha} and corpus luteum regression. Annal. NY Acad. Sci., 180, 456–472.[ISI][Medline]

McCracken, J.A., Baird, D.T. and Goding, J.R. (1971) Factors affecting the secretion of steroids from the transplanted ovary of the sheep. Recent Progr. Horm. Res., 27, 537–582.[Medline]

Macchiarelli, G., Nottola, S.A. and Motta, P.M. (1997) On the presence of extensive arteriovenous contacts in the rabbit ovarian hilum and medulla. In Motta, P.M. (ed.), Recent Advances in Microscopy of Cells, Tissues and Organs. Antonio Delfino Editore, Rome, pp. 519–525.

Millar, R.P. and Glover, T.D. (1973) Regulation of seasonal sexual activity in an ascrotal mammal, the rock hyrax, Procavia capensis. J. Reprod. Fertil., 19 (Suppl.), 203–220.

Patten, B.M. (1948) Embryology of the Pig. McGraw-Hill, New York.

Setchell, B.P. (1978) The Mammalian Testis. Paul Elek, London.

Shi, D.S., Avery, B. and Greve, T. (1998) Effects of temperature gradients on in vitro maturation of bovine oocytes. Theriogenology, 50, 667–674.[ISI][Medline]

Waites, G.M.H. and Setchell, B.P. (1990) Physiology of the mammalian testis. In Lamming, G.E. (ed.), Marshall's Physiology of Reproduction, 4th edn, Vol. 2, Reproduction in the Male. Churchill Livingstone, Edinburgh, Ch. 1, pp. 1–105.

Submitted on June 21, 1999; accepted on October 12, 1999.