Localized increases in ovarian vascular permeability and leucocyte accumulation after induced ovulation in rabbits U.
Gerdes1, M. G\l=a%o\fvels2,
A.
Bergh1,3 and S. Cajander1
Departments of ' Pathology, 2Physiology and 3Anatomy, University ofUmeà, Sweden Summary. Colloidal carbon was injected i.v. in mature virgin rabbits at different times after induction of ovulation by human chorionic gonadotrophin (hCG, 100 iu) or mating. Before induction of ovulation, slight carbon leakage was observed in the inner vascular ring of the theca interna of antral follicles, but blood vessels in the other ovarian compartments
were
unstained. Between 4 and 10\m=.\5h after hCG-treatment
or
mating, vascular leakage was most marked in the blood vessels of the interstitial gland and in the theca interna of antral follicles. Just before ovulation, carbon particles were observed between granulosa cells and some carbon was seeping into the follicular fluid of preruptured follicles. Vascular leakage was also observed over the follicle dome before rupture as well as at the dorsomedial junction between the mesovarium and the ovary. The blood vessels stained with carbon were 7\p=n-\70\g=m\m diameter, representing capillaries and postcapillary venules. About 6 h after hCG injection, an increased number of polymorphonuclear leucocytes migrated from the vessels of these ovarian compartments into the surrounding interstitial tissue. The number of leucocytes seen in the follicular wall and ovarian medulla increased markedly towards ovulation. During early corpus luteum formation, the number of leucocytes decreased markedly. The localized vascular changes seen after mating and hCG stimulation were similar to an inflammatory reaction and could form the basis for the formation of peritoneal exudate after ovulation in rabbits and periovulatory ascitic accumulation seen in the peritoneal cavity of women during the menstrual cycle. Keywords: ovary; hCG; mating; vascular changes; leucocyte; ovulation;
rabbit
Introduction Ovulation in mammals is preceded by rapid follicle expansion under constant hydrostatic pressure (Blandau & Rumery, 1963; Espey & Lipner, 1963). This can occur because of increases in proteolytic enzyme activities in ovulatory follicles (Beers, 1975; Hsueh et ai, 1988; Cajander et ai, 1989) and an increase in vascular permeability (Bjersing & Cajander, 1974a, b, c; Okuda et ai, 1980, 1983; Damber et ai, 1987; Cajander, 1989). Hyperaemia, oedema and accumulation of leucocytes are noted around ovulatory follicles in rabbit ovaries (Burr & Davies, 1951; Morris & Sass, 1966; Zachariae et ai, 1958; Bjersing & Cajander, 1974a, b, c; Cavender & Murdoch, 1988) when ovulation is induced by mating or by treatment with human chorionic gonadotrophin (hCG). Treatment with luteinizing hormone (LH) and hCG increases release of ovarian histamine and concentration of eicosanoid, leukotriene, platelet-activating factor and bradykinin (Murdoch et al., 1986; Munalulu et ai, 1987; Espey et ai, 1989a, b,c; Krishna et ai, 1989). It has, thus, been suggested that these inflammatory mediators are involved in increases in ovarian blood flow and vascular permeability induced by LH and hCG (Espey, 1980; Cavender & Murdoch, 1988; Wallach "Present address: University of Pennsylvania, Department of Obstetrics and Gynecology, Division of Reproductive Biology, 7th Floor East-Clinical Research Building, 422 Curie Boulevard, Philadelphia, PA 19104-6147, USA.
et
ai, 1989). These observations have further substantiated the hypothesis that ovulation is related
inflammatory reaction in the ovulatory follicle (Espey, 1980), particularly as several studies reported that anti-inflammatory drugs, such as antihistamine and inhibitors of prostaglandin synthesis, lipoxygenase and platelet activators, blocked ovulation (Espey, 1980, 1986; Reich et ai, 1983; Abisogun et ai, 1988, 1989; Lipner, 1988). Other morphological studies have demonstrated the appearance of fenestrations and interendothelial cell gaps in perifollicular capillaries after hCG treatment (Bjersing & Cajander, 1974c; Okuda et ai, 1983; Otsuki et ai, 1986), but vascular changes in other regions of such ovaries have not been well characterized. The volume and protein composition of peritoneal fluid sampled during the menstrual cycle varies in relation to cycle day. to an
have
Studies indicate that this fluid could emerge from the ovaries
as a
result of transudation from
leaking blood vessels (Bouckaert et ai, 1986). The present study characterized the localization of leaking blood vessels and their relation to leucocyte accumulation in the ovary. The colloidal car¬ bon labelling technique (Majno et ai, 1961) was used for morphological investigation of changes in vascular permeability in rabbit ovaries after ovulation induced by mating or hCG treatment. Materials and Methods Animals White Swedish landrace rabbits, 6 months old and weighing 3-4 kg, were purchased from a local breeder. They kept in separate cages and fed a standard rabbit chow and water ad libitum for 2-3 weeks before the experiments. Rabbits were used because they ovulate 10-12 h after ovulation induced by hCG or mating (Harper, 1961; Cajander, 1976). The total number of rabbits in the permeability studies was 26. The animals were divided into three groups: one received 100 iu hCG (Gonadex: Leo Co., Helsingborg, Sweden or Pregnyl: Organon, Netherlands) i.v. in the lateral ear vein; the second group was mated; and the third group received no treatment and served as controls. In the experiments where leucocyte accumulation was tested, only hCG-treated rabbits (n 16) were used. Protocols for these studies were reviewed and approved by the Regional Ethical Committee for Animal experiments in Umeá. were
=
Assessment of vascular
permeability
Colloidal carbon (Pelikan Drawing Ink: Pelikan Werke Hannover, Germany) in a dose of 1-4 ml kg1 body weight was given i.v. 1 h before the rabbits were killed. Intravascularly injected carbon particles (average diameter 20-30 nm) are too large to penetrate an intact endothelium, but they may escape from the blood via open inter¬ endothelial cell junctions. Their further passage is, however, partly restricted by the basement membrane and thus sites of vascular leakage (open gaps or gaps that have been open and later closed) are thus labelled black and can later be localized in tissue sections. This technique has been used extensively to localize vascular leakage in various models of experimental inflammation (Cotran & Majno, 1964; O'Donnell et al, 1987) and in rat testis after hCG treatment (Bergh et al, 1987). The animals were anaesthetized with sodium pentobarbital (30mg kg"' i.v. of Mebumal vet.: ACO, Sweden) 1-5,4, 6, 8, 9-5 or 10-5 h after induction of ovulation and the ovaries and parts of the mesovarium were quickly removed and fixed for at least 24 h in 4% (v/v) formaldehyde, 3% (w/v) glutaraldehyde and 005% (w/v) picric acid in 01 mol Na-cacodylate buffer l"1. Animals were killed according to procedures described by Adams (1987), by overdose of barbiturates. Ovarian slices about 1 mm thick were dehydrated and embedded in glucolmethacrylate plastic (Histo-Resin, LKB, Sweden) and sections 1-2 pm thick were stained with haematoxylin and eosin. Other parts of the ovaries were fixed in 1 % (w/v) Os04, dehydrated and embedded in Epon. Thin sections were counterstained with lead citrate and uranyl acetate before observation in a Zeiss EM9 electron microscope. For morphological assessment of the distribution of carbon, the ovaries were categorized into cortex, interstitial gland and mesovarium. The follicles were divided into preantral, secondary antral and tertiary or ovulatory. Atresia was defined according to the criteria of Peters & McNatty (1980). Vessels were classified as capillaries when the wall was composed of an endothelial cell surrounded by occasional pericytes and the diameter was < 10 pm (Ham, 1971). Vessels with a larger diameter, but of the same appearance, were classified as large sinusoidal capillaries. Vascular spaces without erythrocytes were classified as lymphatics. The amount of carbon visible was classified from + to ++++ (+ indicated carbon particles just visible with the light microscope at 1000 magnification). Light micrographs were taken in a Zeiss microscope.
Assessment of leucocyte accumulation All rabbits in this part of the study were injected intravenously with lOOiu hCG At 0, 1,6, 10, 14-5 or 26 h after injection, the rabbits were oophorectomized. Three animals were operated at each time except at 1 h after hCG injection (n 1). The ovaries and about 1 cm of mesovarian tissue, including the ovarian vessels, were immersed in —
neutral 10% (v/v) formaldehyde. After 1-2 days of fixation, the ovaries were cut into thick. These specimens were dehydrated and embedded in glycol-metacrylate (Histo-Resin: LKB, Stockholm, Sweden). Sections 1-2 pm thick were cut on a Sorval, Porter-Blum microtome, JB-4, and stained with haematoxylin and eosin. Quantitative analysis of number of leucocytes in different ovarian regions was performed at a magnification of 24 pm. 120 or 120 500 with the aid of frames inserted in the eyepiece. The frames enclosed areas measuring 120 The square frame was used for morphometry of leucocytes in mesovarian tissue and ovarian medulla, while the rectangular one was used for tunica albugínea and follicle wall. Measurements on the Ovarian medulla' compartment was restricted to the loose connective tissue and the interstitial gland, i.e. immature follicles and oocytes, atretic follicles and large vessels were deliberately avoided. The analyses on the 'follicle wall' compartment were limited to the theca externa layers in healthy ovulatory follicles, i.e. the granulosa layer was not included. The apical, lateral and basal parts of the follicular wall were also compared. The tunica albugínea was analysed above ovulatory follicles and in areas not related to the follicles. After ovulation, leucocytes were counted in the ovarian medulla and mesovarian tissue. From each animal, three ovarian slices, each comprising at least one appropriate follicle or corpus luteum, were selected for morphometry. In the semithin plastic sections, interstitially located leucocytes in ten randomly chosen areas were counted, i.e. 30 separate areas/compartments per rabbit. In all compartments studied, effort was made to exclude counting of intravascular leucocytes. The average number of leucocytes per unit area was counted to give a semiquantitative estimation of number of leucocytes, but, since leucocyte size and section thickness were identical at each time, the method was adequate for monitoring changes in leucocyte abundance.
sodium-phosphate-buffered, slices about 1
Statistical
mm
analysis
Analyses were performed with the Mann-Whitney U test (Siegel, 1956). A significant.
value of 005
or
less
was
considered
Results
permeability during the ovulatory period In control animals, carbon staining of the vasculature was generally not observed, with the exception of a slight carbon leakage in the theca interna of some tertiary follicles (Table 1). A Vascular
similar situation
was
observed 1-5 h after hCG treatment.
Table 1. Intensity of carbon staining (arbitrary units) in compartments of rabbit ovaries at various times after treatment with human chorionic gonadotrophin (hCG) or mating Treatment
Control hCG hCG Mated hCG hCG hCG hCG Mated
(«)
Hours after stimulation
6 2 2 2 2 2 2 4 4
1-5 40 4-0 60 80 9-5 10-5 10-5
Preantral follicles
Secondary
Ovulatory
+ + + + + +
+ + +++ +++ +++ +++ +++ ++++ +++
follicles
-
—
+ + + + + +
-
follicles
Interstitial
gland -
At 4 h after hCG treatment, marked carbon staining was observed in blood vessels in the interstitial gland and a slight leakage was observed in the theca externa and interna of secondary ovulatory follicles (Table 1). Electron microscopy revealed that leaking carbon particles were trapped between the endothelial cells and the basement membrane (not shown). The leaking vessels were generally thin-walled without a muscular layer and 7-70 µ in diameter. Hence, the leakage was observed in capillaries and in large sinusoidal capillaries. The types of blood vessel labelled were similar in the various ovarian compartments. The tertiary follicles, now in the early phase of
demonstrated a considerably more intense carbon leakage in the theca interna (Table 1). Carbon leakage was also observed in the hilus region of the ovaries and in the mesovarian fat tissue close to the hilus. A similar leakage pattern was observed at 6, 8 and 9-5 h after hCG treatment.
rapid ovulatory development,
Fig. 1. Light micrographs of a rabbit ovulatory follicle 10-5 h after treatment with human chorionic gonadotrophin. (a) Apical (A) and interfollicular (I) regions of the ovarian surface; in the theca interna (ti) layer, carbon accumulation is evident; a marked accumulation of carbon particles is also seen just below the surface epithelium of the follicle dome, but no subepithelial carbon deposition has occurred in the interfollicular area, 115. (b) Closer view of the basal follicle wall showing dissociated granulosa cells and accumulation of carbon aggregates just visible between granulosa cells (arrowheads), 290.
Fig. 2. A freshly ruptured rabbit follicle at 10-5 h after treatment with human chorionic gonadotrophin. (a) Several areas demonstrate carbon aggregates dispersed between the granu¬ losa cells (arrows), 140. (b) Detail of part of Fig. 2a. Abundant amounts of carbon are demonstrated in the theca layer as well as aggregates that have seeped into the granulosa layer. x310.
At 10-5 h after hCG treatment, the ovary contained several ovulatory and a few ruptured follicles. Around ovulatory follicles, marked carbon leakage was observed in the prominent, apparently dilated, blood vessels in the theca interna (Fig. la) and there was no obvious difference in amount of leakage between the apical, basal or lateral parts of these follicles. Carbon particles were also observed between granulosa cells of ovulatory follicles (Fig. lb), indicating a flow of fluid into expanding follicles. This was particularly evident in the follicles where rupture had just
Fig.
3. Electron
micrograph
accumulation of carbon
apical part of an ovulatory rabbit follicle 9-5 h after gonadotrophin. Just below the surface epithelium (SE), clear
from the
treatment with human chorionic
particles is evident,
8500.
occurred (Fig. 2). Heavy carbon deposits were also observed just below the surface epithelium at the apex of ovulatory and ruptured follicles (Fig. la). Accumulation of carbon in the cortex area or under the ovarian surface epithelium was not observed outside the dome area of ovulatory follicles (Fig. la). The accumulated carbon was found mainly under the basement membrane of the epithelium as a sign of transudate-fluid passage into the abdominal cavity. As judged by electron microscopy of this area, carbon particles of the apical region of preovulatory follicles were demonstrated strictly below the epithelial basement membrane (Fig. 3). Moreover, at the meso¬ varian junction, similarly localized carbon accumulation was found regularly under the peritoneal mesothelium (Figs 4 and 5). This carbon deposition was always found at the mediodorsal part of the mesovarian folding (Figs 4c and 5) where the large ovarian blood and lymphatic vessels pass through (Fig. 4a, c). At this time after hCG, substantial leakage was observed in the interstitial gland and hilus region, with at least the same intensity as at 4 h after hCG treatment (Table 1). Only a slight carbon leakage was observed around secondary follicles at this time (Table 1). In animals treated with hCG, slight carbon leakage was occasionally observed around atretic follicles 4—10-5 h after treatment, but there was no increase in staining. In mated animals with endogenous LH release, a similar, but slightly weaker, carbon labelling pattern was observed at different times (Table 1). As an internal control, blood vessels of the
Fig. 4. Rabbit ovary at 9-5 h after treatment with human chorionic gonadotrophin, demon¬ strating carbon leakage at the mesovarian-ovarian junction, (a) Survey of the medial (med) and lateral (lat) aspect of the ovarian-peritoneal junction; a, artery; v, vein; 1, lymphatic vessels. 90. (b) Lateral part of the ovarian-peritoneal junction with ovarian tissue to the right and peritoneal surface to the left. No apparent carbon deposition, 220. (c) Medial part of the ovarian-peritoneal junction with prominent subepithelial carbon deposition, 220. psoas muscle before or after hormone stimulation observed (not shown).
were
studied, but
no
carbon
leakage
was
Leucocyte accumulation during induced ovulation Polymorphonuclear leucocytes were observed at an increased frequency in the interstitial gland as well as in the theca interna of ovulatory follicles and some were apparently migrating into the interstitial tissue. In ovaries before
hCG-treatment, there
were
four to six Graafian follicles. The
Fig. 5. Electron micrograph from the medial part of the rabbit ovarian-peritoneal 9-5 h after treatment with human chorionic gonadotrophin. Large deposits of carbon can be seen under the subepithelial basement membrane, 6600.
junction particles
pseudostratified granulosa layer contained closely arranged granulosa cells, but no leucocytes. In the theca interna-externa region, leucocytes were occasionally found (follicle wall 002 ± 004 leucocytes per unit area, Fig. 6). The medulla was dominated by solid clusters or cords of large epithelioid interstitial gland cells (Fig. 7) and some leucocytes were found (0-15 ± 011 leucocytes per unit area). In the tunica albugínea, no leucocytes were found. The mesovarian region showed loose connective tissue, large vessels, ordinary adipose tissue (Fig. 7) and occasional leuco¬ cytes (007 + 011 leucocytes per unit area). There was no interstitial oedema in any region at this avascular
stage.
Fig. 6. Histogram showing the relative frequency of leucocytes in the follicle wall (D), ovarian medulla ( ) and mesovarian tissue (E3) in rabbits at different times in relation to treatment with human chorionic gonadotrophin (hCG). sd is indicated by vertical bars.
Fig. 7. Photomicrographs of rabbit ovary before treatment, (a) Survey view showing part of the mesovarian tissue with adipose tissue, vessels and connective tissue. To the left is a part of the 76. (b) Detailed view ovary with solid clusters of epithelioid cells of the interstitial gland, showing interstitial gland tissue with large epithelioid cells with vacuolated cytoplasm. There is no interstitial oedema, 300. At 1 h after hCG treatment, one rabbit was investigated. There was no difference in number of leucocytes in any of the regions between treated and untreated animals and there were no obvious structural changes. At 6 h after induction of ovulation a marked and significant increase in number of leucocytes (follicle wall 0-45 + 0-50, medulla 1-5 + 0-26 and mesovarian tissue 1-2 + 0-38 leucocytes per unit < 005 versus 0 h) was registered in all regions, except the tunica albugínea where no leuco¬ area, cytes were found. At this stage, ovulatory follicles were easily identified by light microscopy. They were somewhat distended with dispersed granulosa cells and in the apical theca interna there was a slight interstitial oedema. No difference in number of leucocytes was registered among apical, lateral or basal regions of these ovulatory follicles. In the medulla and mesovarian tissue regions, there was also a moderate interstitial oedema. At 10 h after hCG treatment, no significant change in frequency of leucocytes was found compared with 6 h after hCG (Fig. 6), but the interstitial oedema had increased in the follicle wall, ovarian medulla and mesovarian tissue. Leucocytes were seen moving through the vascular endothelium of venules to accumulate in the interstitial compartment throughout the ovary (Figs 8a, b). In ovulatory follicles close to rupture, some leucocytes infiltrated the granulosa layer, but there was no leucocyte accumulation in the tunica albugínea above these follicles. Three follicles had ruptured at this time and considerable accumulation of leucocytes was observed close to the edges of the rupture hole. A few hours after follicular rupture, at 14-5 h after hCG injection, no significant change in leucocyte accumulation was registered (Fig. 6; medulla 0-9 + 006 and mesovarian tissue 1-3 + 0-84 leucocytes per unit area). In the newly ruptured follicles, leucocytes were regularly found between the granulosa cells and in the partly emptied antral cavity. At 26 h after hCG injection, the frequency of leucocytes in the ovarian medulla and the mesovarian tissue had decreased to the same range as before hCG stimulation (Fig. 6), and the ovarian oedema had disappeared.
Discussion Our results show not only an increase in perifollicular vascular permeability with flow of transudate into the antrum of ovulatory follicles and young corpora lutea, but also marked leakage from vessels and accumulation of leucocytes in the interstitial gland. Two new principal locations
Fig. 8. Photomicrographs of a rabbit ovary 10 h after treatment with human chorionic gonado¬ trophin. (a) Mesovarian tissue with many leucocytes in the interstitial oedematous fluid, 240. (b) Interstitial gland area in the ovarian medulla with many extravascular leucocytes in the interstitial fluid,
375.
for passage of transudate fluid from the ovary into the abdominal cavity were revealed: the dorsomedial mesovarian junction and the apical follicular dome. This transovarian fluid passage appearing after exposure to LH may be the main source of peritoneal fluid accumulating during the menstrual cycle in women. Leucocyte accumulation in different ovarian compartments and in mesovarian tissue was seen after ovulatory induction. Acute inflammation with vascular engorgement and exudation of fluid and leucocytes around ovulatory follicles have been reported in earlier studies (see Espey, 1980 for review). In the present study, hCG treatment and mating in rabbits induced a similar increase in carbon labelling of ovarian blood vessels, although the increase in permeability was slightly more accentuated after treatment with hCG. These observations indicate a marked and localized inflammation-like increase in the permeability of perifollicular blood vessels, supporting the hypothesis put forward by Espey (1980). In the present study, carbon leakage was demonstrated as early as 4 h after hCG injection. Vascular leakage occurring after hCG treatment or mating was mainly noted in larger vessels than previously recognized, i.e. in large sinusoidal capillaries, which probably represent postcapillary venules, and vascular leakage is not only confined to the ovulatory follicle but also noted in the interstitial gland. The physiological role of the increased permeability seen in the interstitial gland is unknown. As shown earlier, postcapillary venules are the type of blood vessel affected by changes in permeability in response to locally secreted inflammation mediators (Movat, 1985; Williams, 1985). The observations in the present study confirm earlier reports describing the appearance of large interendothelial cell gaps in perifollicular capillaries immediately before ovulation (Bjersing & Cajander, 1974c; Okuda et ai, 1983; Otsuki et ai, 1986). The role of increased permeability in blood vessels of the ovulatory ovary is unknown, but recent mechanistic studies have shown that macromolecules penetrate into the interstitial compartment of the extra¬ cellular space via more-or-less open interendothelial cell junctions in postcapillary venules and not via endothelial fenestrations. The size of these interendothelial gaps may be controlled by hormonal actions (Crone, 1987). As shown in the present study and previous observations in the testis (Bergh et ai, 1987, 1990), macromolecular permeability in gonadal blood vessels may also be controlled by hormonal mechanisms triggered by a sudden increase in exposure to LH, which may modulate the size of interendothelial cell junctions. However, slight carbon leakage was also observed in the theca interna of large unstimulated follicles in ovaries before hCG injection. These follicles in unstimulated rabbits develop up to the stage of tertiary follicles without prior hCG injection (Adams, 1987). It thus appears that growing follicles that have reached a certain develop¬ mental stage may influence their own supply of macromolecules; the mechanism regulating this is
unknown.
Leucocytes may play a role in macromolecular supply to gonadal cells. Both release of
endogenous LH and injection of LH and hCG induce an accumulation of leucocytes as seen in the ovary (this study) and testis (Bergh et ai, 1990). Thus, LH-stimulated cells may secrete leucotactic factors that attract leucocytes, which may be partly responsible for the appearance of the inflammation-like state in the ovary (Espey, 1980), similar to the mechanism described in other tissues (Movat, 1985; Williams, 1985). Our findings in the present study do not exclude the possibility that leucocytes in the ovulatory follicle wall contribute to dissolution of the follicle wall and subsequent rupture. However, the leucocytes were accumulating in equal amounts in the apical, lateral and basal follicle wall and there was no accumulation of leucocytes in the collagenrich tunica albugínea above follicles close to ovulation. The ovulation rate was doubled when polymorphonuclear leucocytes were added to media of rat ovaries perfused in vitro (Hellberg et ai, 1991). In the light of the results of the present study, we suggest that ovarian polymorphonuclear leucocytes may (as shown in the testis, Widmark et ai, 1987) increase vascular permeability in the ovary, possibly to secure an optimal ovulation rate. Why leucocytes accumulate in the ovary after mating or hCG treatment and the relation of this phenomenon to the detected increase in vascular permeability are unknown. However, it has been shown that ovarian cells secrete leucotactic factors such as platelet-activating factor, leukotriene B4 (LTB4), and factors of low molecular mass (Espey et al., 1986; Abisogun et ai, 1989; Espey et ai, 1989a, b, c; Murdoch & McCormick, 1989). Yoshimura et al. (1991) showed an increase in LTB4 secretion in response to hCG in rabbit ovaries perfused in vitro, with a maximum 6 h after the start of hCG infusion. Secretion of leukotriene B4 may explain the leucocytic accumulation seen in the present study as well as, by a synergistic effect to prostaglandin E2, the increase in vascular permeability recorded (Wedmore & Williams, 1981; Munalulu et ai, 1987). Histamine and prostaglandins are well-known candidates that have been ascribed a role in the vascular changes seen during the ovulatory period (Espey, 1980, 1986). Their vascular effect is probably not by attraction of leucocytes, since neither of them are leucotactic (Williams, 1985). Histamine release may explain the hyperaemia and increase in blood flow seen after the injection of LH and hCG (Szego & Gitin, 1964; Piacsek & Huth, 1971; Norjavaara et ai, 1987). In rabbits and hamsters, ovarian mast cells have been shown to degranulate during the ovulatory period, suggesting a functional role in ovulation (Krishna et ai, 1989). In conclusion, LH and hCG result in increased permeability in ovarian blood vessels, possibly increased by the secretion of leucotactic inflammatory mediators. The functional significance of this is unknown, but it may be involved in the rupture of follicles, uptake of macromolecules to steroid-producing cells and as a source of peritoneal fluid. The present study was supported by the Swedish Medical Research Council (Project Nos 9059, 8428 and 5935). Grants were also provided by the Foundation of Magnus Bergvall, The Swedish Society of Medicine and the Medical Faculty, Umeâ University. Skilful technical assistance was provided by I. Fransson and U. Westman. References
Abisogun, A.O., Daphna-Iken, I).. Reich, R., Kranzfelder, D. & Tsafriri, A. (1988) Modulatory role of eicosanoids in vascular changes during periovulatory period in the rat. Biology of Reproduction 38, 756-762.
Abisogun, A.O., Braquet, P. & Tsafriri, A. (1989) The involvement of platelet activating factor in ovulation.
Science 243,381-383. Adams, CE. (1987) The laboratory rabbit. In UFAW Handbook on the Care and Management ofLaboratory Animals, 6th edn, pp. 415—435. Longman Scientific & Technical, Harlow.
Beers, W.H. (1975) Follicular plasminogen and plas-
and the effect of plasmin on ovarian follicle wall. Cell 6, 379-386. Bergh, ., Rooth, P., Widmark, A. & Damber, J.-E. (1987) Human chorionic gonadotrophin treatment induces inflammation-like changes in testicular microcirculation. Journal of Reproduction and Fertility 73, 135-143. Bergh, ., Damber, J.-E. & Widmark, A. (1990) A physiological increase in LH may influence vascular permeability in the rat testis. Journal of Reproduction and Fertility 89, 23-31.
minogen activator
L. & Cajander, S. (1974a) Ovulation and the mechanism of follicle rupture. I. Light microscopic changes in rabbit ovarian follicles prior to induced ovulation. Cell and Tissue Research 149, 287-300. Bjersing, L. & Cajander, S. (1974b) Ovulation and the mechanism of follicle rupture. VI. Ultrastructure of theca interna and the inner vascular network surrounding rabbit Graafian follicles prior to induced ovulation. Cell and Tissue Research 153, 31^t4. Bjersing, L. & Cajander, S. (1974c) Ovulation and the mechanism of follicle rupture. IV. Ultrastructure of membrana granulosa of rabbit Graafian follicles prior to induced ovulation. Cell and Tissue Research 153, 1-14. Blandau, R.J. & Rumery, R.E. (1963) Measurements of intrafollicular pressure in ovulatory and pre¬ ovulatory follicles in the rat. Fertility and Sterility 14, 330-341.
Bjersing,
L.L. & Lipner, H. (1963) Measurements of intrafollicular pressures in the rabbit ovary. American Journal of Physiology 205, 1067-1072. Espey, L.L., Miller, D.H. & Margolius, H.S. (1986) Ovarian increase in kinin-generating capacity in PMSG/hCG-primed immature rat. American Journal of Physiology 251, E362-E365.
Espey,
(1986) Patterns of changes in proteins in the perito¬ neal fluid of women during the periovulatory phase of the menstrual cycle. Journal of Reproduction and Fertility 11, 329-336. Burr, J.R. & Davies, J.I. (1951) The vascular system of the rabbit ovary and its relationship to ovulation. Anatomical Record 111, 273-297. Cajander, S.B. (1976) On the mechanism of ovulation. Umeá University Medical Dissertations, New Series
Espey, L.L., Tanaka, N., Woodard, D.S., Harper, M.J.K. & Okamura, H. (1989a) Decrease in ovarian platelet-activating factor during ovulation in the gonadotropin-primed immature rat. Biology of Reproduction 40, 104-110. Espey, L.L., Tanaka, N. & Okamura, H. (1989b) Increase in ovarian leukotrienes during hormonally induced ovulation in the rat. American Journal of Physiology 256, E753-E759. Espey, L.L., Tanaka, N., Winn, V. & Okamura, H. (1989c) Increase in ovarian kallikrein activity during ovulation in the gonadotrophin-primed immature rat. Journal of Reproduction and Fertility 87, 503-508. Ham, A.W. (1971) The circulatory system. In Histology, 7th edn, pp. 559-592. J. B. Lippincott Company, Philadelphia. Harper, M.J.K. (1961) The time of ovulation in the rabbit following the injection of luteinizing hormone. Journal of Endocrinology 22, 147-152. Hellberg, P., Thomsen, P., Janson, P.O. & Brännström, M. (1991) Leukocyte supplementation increases the luteinizing hormone-induced ovulation rate in the in vi7ro-perfused rat ovary. Biology of Reproduction 44,
(1989) Periovulatory changes in the
Hsueh, A.J.W., Liu, Y.-X., Cajander, S., Peng, X.-R., Dahl, ., Kristensen, P. & Ny, T. (1988)
Bouckaert, P.X.J.M., Evers, J.L.H., Doesburg, W.H., Schellekens, L.A., Brombacher, P.H. & Rolland, R.
No. 23.
Cajander,
S.B.
Morphology and expression of tissue-type plasminogen activator. In Developments in Ultrastructure of Reproduction, pp. 91-101. Alan R. Liss, ovary.
Inc., New York.
Cajander, S.B., Hugin, M.P., Kristensen, P. & Hsueh, A.J.W. (1989) Immunohistochemical localization of tissue-type plasminogen activator in ovaries before and after induced and spontaneous ovulation in the rat. Cell and Tissue Research 257, 1-8. Cavender, J.L. & Murdoch, W.J. (1988) Morphological studies of the microcirculatory system of periovula¬ tory ovine follicles. Biology of Reproduction 39, 989-997. Cotran, R.S. & Majno, G. (1964) The delayed and prolonged vascular leakage in inflammation. I. Topography of the leaking vessels after thermal injury. American Journal of Pathology 45, 261-281. Crone, C. (1987) The Malpighi lecture. From porositates carnis to cellular microcirculation. International Journal of Microcirculation: Clinical and Expérimen¬ tale, 101-122. Damber, J.-E., Cajander, S., Gâfvels, M. & Selstam, G. (1987) Blood flow and vascular appearance in preovulatory follicles and young corpora lutea in immature pregnant mare serum gonadotropintreated rats. Biology of Reproduction 37, 651-658. Espey, L.L. (1980) Ovulation as an inflammatory reaction a hypothesis. Biology of Reproduction 22, 73-106. Espey, L.L. ( 1986) Effect of time and dose ofindomethacin on follicular prostaglandins and ovulation in the rabbit. Endocrinology 119, 746-754. -
791 797.
Gonadotropin-releasing hormone induces ovulation hypophysectomized rats: studies on ovarian tissue-type plasminogen activator activity, messenger
in
ribonucleic acid content, and cellular localization. Endocrinology 122, 1486-1495. Krishna, ., Beesley, K. & Terranova, P.F. (1989) Histamine, mast cells and ovarian function. Journal
of Endocrinology 120, 363-371. Lipner, H. ( 1988) Mechanism ofmammalian ovulation. In The Physiology of Reproduction, Vol. 1, pp. 447^188. Eds E. Knobil & J. D. Neil. Raven Press, New York.
Majno, G., Palade, G.E. & Schoefl, G.I. (1961) Studies on
inflammation. II. The site of action of histamine and serotonin along the vascular tree: a topographic study. Journal of Biophysics Biochemistry and
Cytology 11,607-626. Morris, B. & Sass, M.B. (1966) The formation of lymph in the ovary.
Proceedings of the Royal Society of Biology 164,577-591. Movat, H.Z. (1985) The vascular phenomena of acute inflammation. In The Inflammatory Reaction, pp. 43-75. Elsevier, Amsterdam. Munalulu, B.M., HiUier, K. & Peddie, M.J. (1987) Effect of human chorionic gonadotrophin and indomethacin on ovulation, steroidogenesis and prostaglandin synthesis in preovulatory follicles of PMSG-primed immature rats. Journal of Reproduction and Fertility 80, 229-234. Murdoch, W.J. & McCormick, R.J. (1989) Production of low molecular weight chemoattractants for leuco¬ cytes by periovulatory ovine follicles. Biology of Reproduction 40, 86-90.
Murdoch, W.J., Peterson, T.A., van Kirk, E.A., Vincent, D.L. & Inskeep, E.K. (1986) Interactive roles of
progesterone, prostaglandins, and collagenase in the ovulatory mechanism of the ewe. Biology of
Reproduction 35, 1187-1194. Norjavaara, E., Olofsson, J., Gâfvels, M. & Selstam, G. (1987) Redistribution of ovarian blood flow after injection of human chorionic gonadotropin and luteinizing hormone in the adult pseudopregnant rat.
Endocrinology 120, 107-114. O'Donnell, S.R., Braun, R.A., Reid, J.J. & Wholohan, T. (1987) A histological method for studying the effect
of drugs on mediation-induced airway microvascular leakage in rodents. Journal ofPharmacology Methods
17,205-217. Okuda, V., Okamura, II.. Kanzaki, II.. Takenaka, ., Morimoto, K. & Nishimura, T. (1980) An ultrastruc¬ tural
study of capillary permeability of perifollicular capillaries of rabbit ovary during ovulation using car¬ bon particles as tracer. Acta Obstétrica et Gynaecologica Japónica 32, 859-870.
Okuda, Y., Okamura, H., Kanzaki, H. & Takenaka, A.
of rabbit ovarian ovulation. Journal of Anatomy 137,
(1983) Capillary permeability
follicles prior 263-269.
to
Otsuki, Y., Magari, S. & Sugimoto, O. (1986) Lymphatic
in rabbit ovaries during ovulation: an ultrastructural study. Lymphology 19, 55-64. Peters, H. & McNatty, K. (1980) The Ovary. A corre¬ lation of structure and function in materials, pp. 31-34 and 28. Canada Publishing, London. Piacsek, B.E. & Huth, J.F. (1971) Changes in ovarian venous blood flow following cannulation: effect of luteinizing hormone (LH) and antihistamine. Proceedings of Society for Experimental Biology and Medicine 138, 1022-1024.
capillaries
Reich, R., Kohen, F., Naor, . & Tsafriri, . (1983)
Possible involvement of lipoxygenase products of arachidonic acid pathway in ovulation. Prostaglandins 6, 1011-1020. Siegel, S. (1956) Non-parametric Statistics for the Behavioral Sciences. McGraw-Hill. Kogakusha,
Tokyo. Szego, CM. & Git in, E.S. (1964) Ovarian histamine depletion during acute hyperaemic response to luteinizing hormone. Nature 201, 682-684. Wallach, E.E., Atlas, S., Dharmarajan. A.M., Oski, J.A. & Santulli, R. (1989) The periovulatory interval: Physiologic and endocrinologie implications. In Development of Preimplantation Embryos and their Environment, pp. 87-100. Alan R. Liss Inc., New
York.
Wedmore, CV. & Williams, T.J. (1981) Control of vas¬
cular permeability by polymorphonuclear leucocytes in inflammation. Nature 289, 646-650. Widmark, ., Bergh, ., Damber, J.-E. & Smedegárd, G. (1987) Leukocytes mediate the hCG induced increase in testicular venular permeability. Molecular and Cellular Endocrinology 53, 25-31. Williams, I.J. (1985) Vascular responses and their suppression; vasodilation and edema. In The Pharmacology of Inflammation, pp. 49-59. Eds I. L. Bontà, M. A. Bray & M. J. Parnham. Elsevier,
Amsterdam.
Yoshimura, Y., Nakamura, Y., Shiraki, M., Hirota, Y., Yamada, 11., Ando, M., Ubukata, Y. & Suzuki, M. Involvement of leukotriene B¿ in ovulation in the rabbit. Endocrinology 129, 193-199. Zachariae, F., AsIkk-I lausen. G. & Boseila, A.W.A. (1958) Studies on the mechanism of ovulation. Acta Endocrinologica 28, 547-552. Received 12 April 1991
(1991)
Gerdes1, M. G\l=a%o\fvels2,
A.
Bergh1,3 and S. Cajander1
Departments of ' Pathology, 2Physiology and 3Anatomy, University ofUmeà, Sweden Summary. Colloidal carbon was injected i.v. in mature virgin rabbits at different times after induction of ovulation by human chorionic gonadotrophin (hCG, 100 iu) or mating. Before induction of ovulation, slight carbon leakage was observed in the inner vascular ring of the theca interna of antral follicles, but blood vessels in the other ovarian compartments
were
unstained. Between 4 and 10\m=.\5h after hCG-treatment
or
mating, vascular leakage was most marked in the blood vessels of the interstitial gland and in the theca interna of antral follicles. Just before ovulation, carbon particles were observed between granulosa cells and some carbon was seeping into the follicular fluid of preruptured follicles. Vascular leakage was also observed over the follicle dome before rupture as well as at the dorsomedial junction between the mesovarium and the ovary. The blood vessels stained with carbon were 7\p=n-\70\g=m\m diameter, representing capillaries and postcapillary venules. About 6 h after hCG injection, an increased number of polymorphonuclear leucocytes migrated from the vessels of these ovarian compartments into the surrounding interstitial tissue. The number of leucocytes seen in the follicular wall and ovarian medulla increased markedly towards ovulation. During early corpus luteum formation, the number of leucocytes decreased markedly. The localized vascular changes seen after mating and hCG stimulation were similar to an inflammatory reaction and could form the basis for the formation of peritoneal exudate after ovulation in rabbits and periovulatory ascitic accumulation seen in the peritoneal cavity of women during the menstrual cycle. Keywords: ovary; hCG; mating; vascular changes; leucocyte; ovulation;
rabbit
Introduction Ovulation in mammals is preceded by rapid follicle expansion under constant hydrostatic pressure (Blandau & Rumery, 1963; Espey & Lipner, 1963). This can occur because of increases in proteolytic enzyme activities in ovulatory follicles (Beers, 1975; Hsueh et ai, 1988; Cajander et ai, 1989) and an increase in vascular permeability (Bjersing & Cajander, 1974a, b, c; Okuda et ai, 1980, 1983; Damber et ai, 1987; Cajander, 1989). Hyperaemia, oedema and accumulation of leucocytes are noted around ovulatory follicles in rabbit ovaries (Burr & Davies, 1951; Morris & Sass, 1966; Zachariae et ai, 1958; Bjersing & Cajander, 1974a, b, c; Cavender & Murdoch, 1988) when ovulation is induced by mating or by treatment with human chorionic gonadotrophin (hCG). Treatment with luteinizing hormone (LH) and hCG increases release of ovarian histamine and concentration of eicosanoid, leukotriene, platelet-activating factor and bradykinin (Murdoch et al., 1986; Munalulu et ai, 1987; Espey et ai, 1989a, b,c; Krishna et ai, 1989). It has, thus, been suggested that these inflammatory mediators are involved in increases in ovarian blood flow and vascular permeability induced by LH and hCG (Espey, 1980; Cavender & Murdoch, 1988; Wallach "Present address: University of Pennsylvania, Department of Obstetrics and Gynecology, Division of Reproductive Biology, 7th Floor East-Clinical Research Building, 422 Curie Boulevard, Philadelphia, PA 19104-6147, USA.
et
ai, 1989). These observations have further substantiated the hypothesis that ovulation is related
inflammatory reaction in the ovulatory follicle (Espey, 1980), particularly as several studies reported that anti-inflammatory drugs, such as antihistamine and inhibitors of prostaglandin synthesis, lipoxygenase and platelet activators, blocked ovulation (Espey, 1980, 1986; Reich et ai, 1983; Abisogun et ai, 1988, 1989; Lipner, 1988). Other morphological studies have demonstrated the appearance of fenestrations and interendothelial cell gaps in perifollicular capillaries after hCG treatment (Bjersing & Cajander, 1974c; Okuda et ai, 1983; Otsuki et ai, 1986), but vascular changes in other regions of such ovaries have not been well characterized. The volume and protein composition of peritoneal fluid sampled during the menstrual cycle varies in relation to cycle day. to an
have
Studies indicate that this fluid could emerge from the ovaries
as a
result of transudation from
leaking blood vessels (Bouckaert et ai, 1986). The present study characterized the localization of leaking blood vessels and their relation to leucocyte accumulation in the ovary. The colloidal car¬ bon labelling technique (Majno et ai, 1961) was used for morphological investigation of changes in vascular permeability in rabbit ovaries after ovulation induced by mating or hCG treatment. Materials and Methods Animals White Swedish landrace rabbits, 6 months old and weighing 3-4 kg, were purchased from a local breeder. They kept in separate cages and fed a standard rabbit chow and water ad libitum for 2-3 weeks before the experiments. Rabbits were used because they ovulate 10-12 h after ovulation induced by hCG or mating (Harper, 1961; Cajander, 1976). The total number of rabbits in the permeability studies was 26. The animals were divided into three groups: one received 100 iu hCG (Gonadex: Leo Co., Helsingborg, Sweden or Pregnyl: Organon, Netherlands) i.v. in the lateral ear vein; the second group was mated; and the third group received no treatment and served as controls. In the experiments where leucocyte accumulation was tested, only hCG-treated rabbits (n 16) were used. Protocols for these studies were reviewed and approved by the Regional Ethical Committee for Animal experiments in Umeá. were
=
Assessment of vascular
permeability
Colloidal carbon (Pelikan Drawing Ink: Pelikan Werke Hannover, Germany) in a dose of 1-4 ml kg1 body weight was given i.v. 1 h before the rabbits were killed. Intravascularly injected carbon particles (average diameter 20-30 nm) are too large to penetrate an intact endothelium, but they may escape from the blood via open inter¬ endothelial cell junctions. Their further passage is, however, partly restricted by the basement membrane and thus sites of vascular leakage (open gaps or gaps that have been open and later closed) are thus labelled black and can later be localized in tissue sections. This technique has been used extensively to localize vascular leakage in various models of experimental inflammation (Cotran & Majno, 1964; O'Donnell et al, 1987) and in rat testis after hCG treatment (Bergh et al, 1987). The animals were anaesthetized with sodium pentobarbital (30mg kg"' i.v. of Mebumal vet.: ACO, Sweden) 1-5,4, 6, 8, 9-5 or 10-5 h after induction of ovulation and the ovaries and parts of the mesovarium were quickly removed and fixed for at least 24 h in 4% (v/v) formaldehyde, 3% (w/v) glutaraldehyde and 005% (w/v) picric acid in 01 mol Na-cacodylate buffer l"1. Animals were killed according to procedures described by Adams (1987), by overdose of barbiturates. Ovarian slices about 1 mm thick were dehydrated and embedded in glucolmethacrylate plastic (Histo-Resin, LKB, Sweden) and sections 1-2 pm thick were stained with haematoxylin and eosin. Other parts of the ovaries were fixed in 1 % (w/v) Os04, dehydrated and embedded in Epon. Thin sections were counterstained with lead citrate and uranyl acetate before observation in a Zeiss EM9 electron microscope. For morphological assessment of the distribution of carbon, the ovaries were categorized into cortex, interstitial gland and mesovarium. The follicles were divided into preantral, secondary antral and tertiary or ovulatory. Atresia was defined according to the criteria of Peters & McNatty (1980). Vessels were classified as capillaries when the wall was composed of an endothelial cell surrounded by occasional pericytes and the diameter was < 10 pm (Ham, 1971). Vessels with a larger diameter, but of the same appearance, were classified as large sinusoidal capillaries. Vascular spaces without erythrocytes were classified as lymphatics. The amount of carbon visible was classified from + to ++++ (+ indicated carbon particles just visible with the light microscope at 1000 magnification). Light micrographs were taken in a Zeiss microscope.
Assessment of leucocyte accumulation All rabbits in this part of the study were injected intravenously with lOOiu hCG At 0, 1,6, 10, 14-5 or 26 h after injection, the rabbits were oophorectomized. Three animals were operated at each time except at 1 h after hCG injection (n 1). The ovaries and about 1 cm of mesovarian tissue, including the ovarian vessels, were immersed in —
neutral 10% (v/v) formaldehyde. After 1-2 days of fixation, the ovaries were cut into thick. These specimens were dehydrated and embedded in glycol-metacrylate (Histo-Resin: LKB, Stockholm, Sweden). Sections 1-2 pm thick were cut on a Sorval, Porter-Blum microtome, JB-4, and stained with haematoxylin and eosin. Quantitative analysis of number of leucocytes in different ovarian regions was performed at a magnification of 24 pm. 120 or 120 500 with the aid of frames inserted in the eyepiece. The frames enclosed areas measuring 120 The square frame was used for morphometry of leucocytes in mesovarian tissue and ovarian medulla, while the rectangular one was used for tunica albugínea and follicle wall. Measurements on the Ovarian medulla' compartment was restricted to the loose connective tissue and the interstitial gland, i.e. immature follicles and oocytes, atretic follicles and large vessels were deliberately avoided. The analyses on the 'follicle wall' compartment were limited to the theca externa layers in healthy ovulatory follicles, i.e. the granulosa layer was not included. The apical, lateral and basal parts of the follicular wall were also compared. The tunica albugínea was analysed above ovulatory follicles and in areas not related to the follicles. After ovulation, leucocytes were counted in the ovarian medulla and mesovarian tissue. From each animal, three ovarian slices, each comprising at least one appropriate follicle or corpus luteum, were selected for morphometry. In the semithin plastic sections, interstitially located leucocytes in ten randomly chosen areas were counted, i.e. 30 separate areas/compartments per rabbit. In all compartments studied, effort was made to exclude counting of intravascular leucocytes. The average number of leucocytes per unit area was counted to give a semiquantitative estimation of number of leucocytes, but, since leucocyte size and section thickness were identical at each time, the method was adequate for monitoring changes in leucocyte abundance.
sodium-phosphate-buffered, slices about 1
Statistical
mm
analysis
Analyses were performed with the Mann-Whitney U test (Siegel, 1956). A significant.
value of 005
or
less
was
considered
Results
permeability during the ovulatory period In control animals, carbon staining of the vasculature was generally not observed, with the exception of a slight carbon leakage in the theca interna of some tertiary follicles (Table 1). A Vascular
similar situation
was
observed 1-5 h after hCG treatment.
Table 1. Intensity of carbon staining (arbitrary units) in compartments of rabbit ovaries at various times after treatment with human chorionic gonadotrophin (hCG) or mating Treatment
Control hCG hCG Mated hCG hCG hCG hCG Mated
(«)
Hours after stimulation
6 2 2 2 2 2 2 4 4
1-5 40 4-0 60 80 9-5 10-5 10-5
Preantral follicles
Secondary
Ovulatory
+ + + + + +
+ + +++ +++ +++ +++ +++ ++++ +++
follicles
-
—
+ + + + + +
-
follicles
Interstitial
gland -
At 4 h after hCG treatment, marked carbon staining was observed in blood vessels in the interstitial gland and a slight leakage was observed in the theca externa and interna of secondary ovulatory follicles (Table 1). Electron microscopy revealed that leaking carbon particles were trapped between the endothelial cells and the basement membrane (not shown). The leaking vessels were generally thin-walled without a muscular layer and 7-70 µ in diameter. Hence, the leakage was observed in capillaries and in large sinusoidal capillaries. The types of blood vessel labelled were similar in the various ovarian compartments. The tertiary follicles, now in the early phase of
demonstrated a considerably more intense carbon leakage in the theca interna (Table 1). Carbon leakage was also observed in the hilus region of the ovaries and in the mesovarian fat tissue close to the hilus. A similar leakage pattern was observed at 6, 8 and 9-5 h after hCG treatment.
rapid ovulatory development,
Fig. 1. Light micrographs of a rabbit ovulatory follicle 10-5 h after treatment with human chorionic gonadotrophin. (a) Apical (A) and interfollicular (I) regions of the ovarian surface; in the theca interna (ti) layer, carbon accumulation is evident; a marked accumulation of carbon particles is also seen just below the surface epithelium of the follicle dome, but no subepithelial carbon deposition has occurred in the interfollicular area, 115. (b) Closer view of the basal follicle wall showing dissociated granulosa cells and accumulation of carbon aggregates just visible between granulosa cells (arrowheads), 290.
Fig. 2. A freshly ruptured rabbit follicle at 10-5 h after treatment with human chorionic gonadotrophin. (a) Several areas demonstrate carbon aggregates dispersed between the granu¬ losa cells (arrows), 140. (b) Detail of part of Fig. 2a. Abundant amounts of carbon are demonstrated in the theca layer as well as aggregates that have seeped into the granulosa layer. x310.
At 10-5 h after hCG treatment, the ovary contained several ovulatory and a few ruptured follicles. Around ovulatory follicles, marked carbon leakage was observed in the prominent, apparently dilated, blood vessels in the theca interna (Fig. la) and there was no obvious difference in amount of leakage between the apical, basal or lateral parts of these follicles. Carbon particles were also observed between granulosa cells of ovulatory follicles (Fig. lb), indicating a flow of fluid into expanding follicles. This was particularly evident in the follicles where rupture had just
Fig.
3. Electron
micrograph
accumulation of carbon
apical part of an ovulatory rabbit follicle 9-5 h after gonadotrophin. Just below the surface epithelium (SE), clear
from the
treatment with human chorionic
particles is evident,
8500.
occurred (Fig. 2). Heavy carbon deposits were also observed just below the surface epithelium at the apex of ovulatory and ruptured follicles (Fig. la). Accumulation of carbon in the cortex area or under the ovarian surface epithelium was not observed outside the dome area of ovulatory follicles (Fig. la). The accumulated carbon was found mainly under the basement membrane of the epithelium as a sign of transudate-fluid passage into the abdominal cavity. As judged by electron microscopy of this area, carbon particles of the apical region of preovulatory follicles were demonstrated strictly below the epithelial basement membrane (Fig. 3). Moreover, at the meso¬ varian junction, similarly localized carbon accumulation was found regularly under the peritoneal mesothelium (Figs 4 and 5). This carbon deposition was always found at the mediodorsal part of the mesovarian folding (Figs 4c and 5) where the large ovarian blood and lymphatic vessels pass through (Fig. 4a, c). At this time after hCG, substantial leakage was observed in the interstitial gland and hilus region, with at least the same intensity as at 4 h after hCG treatment (Table 1). Only a slight carbon leakage was observed around secondary follicles at this time (Table 1). In animals treated with hCG, slight carbon leakage was occasionally observed around atretic follicles 4—10-5 h after treatment, but there was no increase in staining. In mated animals with endogenous LH release, a similar, but slightly weaker, carbon labelling pattern was observed at different times (Table 1). As an internal control, blood vessels of the
Fig. 4. Rabbit ovary at 9-5 h after treatment with human chorionic gonadotrophin, demon¬ strating carbon leakage at the mesovarian-ovarian junction, (a) Survey of the medial (med) and lateral (lat) aspect of the ovarian-peritoneal junction; a, artery; v, vein; 1, lymphatic vessels. 90. (b) Lateral part of the ovarian-peritoneal junction with ovarian tissue to the right and peritoneal surface to the left. No apparent carbon deposition, 220. (c) Medial part of the ovarian-peritoneal junction with prominent subepithelial carbon deposition, 220. psoas muscle before or after hormone stimulation observed (not shown).
were
studied, but
no
carbon
leakage
was
Leucocyte accumulation during induced ovulation Polymorphonuclear leucocytes were observed at an increased frequency in the interstitial gland as well as in the theca interna of ovulatory follicles and some were apparently migrating into the interstitial tissue. In ovaries before
hCG-treatment, there
were
four to six Graafian follicles. The
Fig. 5. Electron micrograph from the medial part of the rabbit ovarian-peritoneal 9-5 h after treatment with human chorionic gonadotrophin. Large deposits of carbon can be seen under the subepithelial basement membrane, 6600.
junction particles
pseudostratified granulosa layer contained closely arranged granulosa cells, but no leucocytes. In the theca interna-externa region, leucocytes were occasionally found (follicle wall 002 ± 004 leucocytes per unit area, Fig. 6). The medulla was dominated by solid clusters or cords of large epithelioid interstitial gland cells (Fig. 7) and some leucocytes were found (0-15 ± 011 leucocytes per unit area). In the tunica albugínea, no leucocytes were found. The mesovarian region showed loose connective tissue, large vessels, ordinary adipose tissue (Fig. 7) and occasional leuco¬ cytes (007 + 011 leucocytes per unit area). There was no interstitial oedema in any region at this avascular
stage.
Fig. 6. Histogram showing the relative frequency of leucocytes in the follicle wall (D), ovarian medulla ( ) and mesovarian tissue (E3) in rabbits at different times in relation to treatment with human chorionic gonadotrophin (hCG). sd is indicated by vertical bars.
Fig. 7. Photomicrographs of rabbit ovary before treatment, (a) Survey view showing part of the mesovarian tissue with adipose tissue, vessels and connective tissue. To the left is a part of the 76. (b) Detailed view ovary with solid clusters of epithelioid cells of the interstitial gland, showing interstitial gland tissue with large epithelioid cells with vacuolated cytoplasm. There is no interstitial oedema, 300. At 1 h after hCG treatment, one rabbit was investigated. There was no difference in number of leucocytes in any of the regions between treated and untreated animals and there were no obvious structural changes. At 6 h after induction of ovulation a marked and significant increase in number of leucocytes (follicle wall 0-45 + 0-50, medulla 1-5 + 0-26 and mesovarian tissue 1-2 + 0-38 leucocytes per unit < 005 versus 0 h) was registered in all regions, except the tunica albugínea where no leuco¬ area, cytes were found. At this stage, ovulatory follicles were easily identified by light microscopy. They were somewhat distended with dispersed granulosa cells and in the apical theca interna there was a slight interstitial oedema. No difference in number of leucocytes was registered among apical, lateral or basal regions of these ovulatory follicles. In the medulla and mesovarian tissue regions, there was also a moderate interstitial oedema. At 10 h after hCG treatment, no significant change in frequency of leucocytes was found compared with 6 h after hCG (Fig. 6), but the interstitial oedema had increased in the follicle wall, ovarian medulla and mesovarian tissue. Leucocytes were seen moving through the vascular endothelium of venules to accumulate in the interstitial compartment throughout the ovary (Figs 8a, b). In ovulatory follicles close to rupture, some leucocytes infiltrated the granulosa layer, but there was no leucocyte accumulation in the tunica albugínea above these follicles. Three follicles had ruptured at this time and considerable accumulation of leucocytes was observed close to the edges of the rupture hole. A few hours after follicular rupture, at 14-5 h after hCG injection, no significant change in leucocyte accumulation was registered (Fig. 6; medulla 0-9 + 006 and mesovarian tissue 1-3 + 0-84 leucocytes per unit area). In the newly ruptured follicles, leucocytes were regularly found between the granulosa cells and in the partly emptied antral cavity. At 26 h after hCG injection, the frequency of leucocytes in the ovarian medulla and the mesovarian tissue had decreased to the same range as before hCG stimulation (Fig. 6), and the ovarian oedema had disappeared.
Discussion Our results show not only an increase in perifollicular vascular permeability with flow of transudate into the antrum of ovulatory follicles and young corpora lutea, but also marked leakage from vessels and accumulation of leucocytes in the interstitial gland. Two new principal locations
Fig. 8. Photomicrographs of a rabbit ovary 10 h after treatment with human chorionic gonado¬ trophin. (a) Mesovarian tissue with many leucocytes in the interstitial oedematous fluid, 240. (b) Interstitial gland area in the ovarian medulla with many extravascular leucocytes in the interstitial fluid,
375.
for passage of transudate fluid from the ovary into the abdominal cavity were revealed: the dorsomedial mesovarian junction and the apical follicular dome. This transovarian fluid passage appearing after exposure to LH may be the main source of peritoneal fluid accumulating during the menstrual cycle in women. Leucocyte accumulation in different ovarian compartments and in mesovarian tissue was seen after ovulatory induction. Acute inflammation with vascular engorgement and exudation of fluid and leucocytes around ovulatory follicles have been reported in earlier studies (see Espey, 1980 for review). In the present study, hCG treatment and mating in rabbits induced a similar increase in carbon labelling of ovarian blood vessels, although the increase in permeability was slightly more accentuated after treatment with hCG. These observations indicate a marked and localized inflammation-like increase in the permeability of perifollicular blood vessels, supporting the hypothesis put forward by Espey (1980). In the present study, carbon leakage was demonstrated as early as 4 h after hCG injection. Vascular leakage occurring after hCG treatment or mating was mainly noted in larger vessels than previously recognized, i.e. in large sinusoidal capillaries, which probably represent postcapillary venules, and vascular leakage is not only confined to the ovulatory follicle but also noted in the interstitial gland. The physiological role of the increased permeability seen in the interstitial gland is unknown. As shown earlier, postcapillary venules are the type of blood vessel affected by changes in permeability in response to locally secreted inflammation mediators (Movat, 1985; Williams, 1985). The observations in the present study confirm earlier reports describing the appearance of large interendothelial cell gaps in perifollicular capillaries immediately before ovulation (Bjersing & Cajander, 1974c; Okuda et ai, 1983; Otsuki et ai, 1986). The role of increased permeability in blood vessels of the ovulatory ovary is unknown, but recent mechanistic studies have shown that macromolecules penetrate into the interstitial compartment of the extra¬ cellular space via more-or-less open interendothelial cell junctions in postcapillary venules and not via endothelial fenestrations. The size of these interendothelial gaps may be controlled by hormonal actions (Crone, 1987). As shown in the present study and previous observations in the testis (Bergh et ai, 1987, 1990), macromolecular permeability in gonadal blood vessels may also be controlled by hormonal mechanisms triggered by a sudden increase in exposure to LH, which may modulate the size of interendothelial cell junctions. However, slight carbon leakage was also observed in the theca interna of large unstimulated follicles in ovaries before hCG injection. These follicles in unstimulated rabbits develop up to the stage of tertiary follicles without prior hCG injection (Adams, 1987). It thus appears that growing follicles that have reached a certain develop¬ mental stage may influence their own supply of macromolecules; the mechanism regulating this is
unknown.
Leucocytes may play a role in macromolecular supply to gonadal cells. Both release of
endogenous LH and injection of LH and hCG induce an accumulation of leucocytes as seen in the ovary (this study) and testis (Bergh et ai, 1990). Thus, LH-stimulated cells may secrete leucotactic factors that attract leucocytes, which may be partly responsible for the appearance of the inflammation-like state in the ovary (Espey, 1980), similar to the mechanism described in other tissues (Movat, 1985; Williams, 1985). Our findings in the present study do not exclude the possibility that leucocytes in the ovulatory follicle wall contribute to dissolution of the follicle wall and subsequent rupture. However, the leucocytes were accumulating in equal amounts in the apical, lateral and basal follicle wall and there was no accumulation of leucocytes in the collagenrich tunica albugínea above follicles close to ovulation. The ovulation rate was doubled when polymorphonuclear leucocytes were added to media of rat ovaries perfused in vitro (Hellberg et ai, 1991). In the light of the results of the present study, we suggest that ovarian polymorphonuclear leucocytes may (as shown in the testis, Widmark et ai, 1987) increase vascular permeability in the ovary, possibly to secure an optimal ovulation rate. Why leucocytes accumulate in the ovary after mating or hCG treatment and the relation of this phenomenon to the detected increase in vascular permeability are unknown. However, it has been shown that ovarian cells secrete leucotactic factors such as platelet-activating factor, leukotriene B4 (LTB4), and factors of low molecular mass (Espey et al., 1986; Abisogun et ai, 1989; Espey et ai, 1989a, b, c; Murdoch & McCormick, 1989). Yoshimura et al. (1991) showed an increase in LTB4 secretion in response to hCG in rabbit ovaries perfused in vitro, with a maximum 6 h after the start of hCG infusion. Secretion of leukotriene B4 may explain the leucocytic accumulation seen in the present study as well as, by a synergistic effect to prostaglandin E2, the increase in vascular permeability recorded (Wedmore & Williams, 1981; Munalulu et ai, 1987). Histamine and prostaglandins are well-known candidates that have been ascribed a role in the vascular changes seen during the ovulatory period (Espey, 1980, 1986). Their vascular effect is probably not by attraction of leucocytes, since neither of them are leucotactic (Williams, 1985). Histamine release may explain the hyperaemia and increase in blood flow seen after the injection of LH and hCG (Szego & Gitin, 1964; Piacsek & Huth, 1971; Norjavaara et ai, 1987). In rabbits and hamsters, ovarian mast cells have been shown to degranulate during the ovulatory period, suggesting a functional role in ovulation (Krishna et ai, 1989). In conclusion, LH and hCG result in increased permeability in ovarian blood vessels, possibly increased by the secretion of leucotactic inflammatory mediators. The functional significance of this is unknown, but it may be involved in the rupture of follicles, uptake of macromolecules to steroid-producing cells and as a source of peritoneal fluid. The present study was supported by the Swedish Medical Research Council (Project Nos 9059, 8428 and 5935). Grants were also provided by the Foundation of Magnus Bergvall, The Swedish Society of Medicine and the Medical Faculty, Umeâ University. Skilful technical assistance was provided by I. Fransson and U. Westman. References
Abisogun, A.O., Daphna-Iken, I).. Reich, R., Kranzfelder, D. & Tsafriri, A. (1988) Modulatory role of eicosanoids in vascular changes during periovulatory period in the rat. Biology of Reproduction 38, 756-762.
Abisogun, A.O., Braquet, P. & Tsafriri, A. (1989) The involvement of platelet activating factor in ovulation.
Science 243,381-383. Adams, CE. (1987) The laboratory rabbit. In UFAW Handbook on the Care and Management ofLaboratory Animals, 6th edn, pp. 415—435. Longman Scientific & Technical, Harlow.
Beers, W.H. (1975) Follicular plasminogen and plas-
and the effect of plasmin on ovarian follicle wall. Cell 6, 379-386. Bergh, ., Rooth, P., Widmark, A. & Damber, J.-E. (1987) Human chorionic gonadotrophin treatment induces inflammation-like changes in testicular microcirculation. Journal of Reproduction and Fertility 73, 135-143. Bergh, ., Damber, J.-E. & Widmark, A. (1990) A physiological increase in LH may influence vascular permeability in the rat testis. Journal of Reproduction and Fertility 89, 23-31.
minogen activator
L. & Cajander, S. (1974a) Ovulation and the mechanism of follicle rupture. I. Light microscopic changes in rabbit ovarian follicles prior to induced ovulation. Cell and Tissue Research 149, 287-300. Bjersing, L. & Cajander, S. (1974b) Ovulation and the mechanism of follicle rupture. VI. Ultrastructure of theca interna and the inner vascular network surrounding rabbit Graafian follicles prior to induced ovulation. Cell and Tissue Research 153, 31^t4. Bjersing, L. & Cajander, S. (1974c) Ovulation and the mechanism of follicle rupture. IV. Ultrastructure of membrana granulosa of rabbit Graafian follicles prior to induced ovulation. Cell and Tissue Research 153, 1-14. Blandau, R.J. & Rumery, R.E. (1963) Measurements of intrafollicular pressure in ovulatory and pre¬ ovulatory follicles in the rat. Fertility and Sterility 14, 330-341.
Bjersing,
L.L. & Lipner, H. (1963) Measurements of intrafollicular pressures in the rabbit ovary. American Journal of Physiology 205, 1067-1072. Espey, L.L., Miller, D.H. & Margolius, H.S. (1986) Ovarian increase in kinin-generating capacity in PMSG/hCG-primed immature rat. American Journal of Physiology 251, E362-E365.
Espey,
(1986) Patterns of changes in proteins in the perito¬ neal fluid of women during the periovulatory phase of the menstrual cycle. Journal of Reproduction and Fertility 11, 329-336. Burr, J.R. & Davies, J.I. (1951) The vascular system of the rabbit ovary and its relationship to ovulation. Anatomical Record 111, 273-297. Cajander, S.B. (1976) On the mechanism of ovulation. Umeá University Medical Dissertations, New Series
Espey, L.L., Tanaka, N., Woodard, D.S., Harper, M.J.K. & Okamura, H. (1989a) Decrease in ovarian platelet-activating factor during ovulation in the gonadotropin-primed immature rat. Biology of Reproduction 40, 104-110. Espey, L.L., Tanaka, N. & Okamura, H. (1989b) Increase in ovarian leukotrienes during hormonally induced ovulation in the rat. American Journal of Physiology 256, E753-E759. Espey, L.L., Tanaka, N., Winn, V. & Okamura, H. (1989c) Increase in ovarian kallikrein activity during ovulation in the gonadotrophin-primed immature rat. Journal of Reproduction and Fertility 87, 503-508. Ham, A.W. (1971) The circulatory system. In Histology, 7th edn, pp. 559-592. J. B. Lippincott Company, Philadelphia. Harper, M.J.K. (1961) The time of ovulation in the rabbit following the injection of luteinizing hormone. Journal of Endocrinology 22, 147-152. Hellberg, P., Thomsen, P., Janson, P.O. & Brännström, M. (1991) Leukocyte supplementation increases the luteinizing hormone-induced ovulation rate in the in vi7ro-perfused rat ovary. Biology of Reproduction 44,
(1989) Periovulatory changes in the
Hsueh, A.J.W., Liu, Y.-X., Cajander, S., Peng, X.-R., Dahl, ., Kristensen, P. & Ny, T. (1988)
Bouckaert, P.X.J.M., Evers, J.L.H., Doesburg, W.H., Schellekens, L.A., Brombacher, P.H. & Rolland, R.
No. 23.
Cajander,
S.B.
Morphology and expression of tissue-type plasminogen activator. In Developments in Ultrastructure of Reproduction, pp. 91-101. Alan R. Liss, ovary.
Inc., New York.
Cajander, S.B., Hugin, M.P., Kristensen, P. & Hsueh, A.J.W. (1989) Immunohistochemical localization of tissue-type plasminogen activator in ovaries before and after induced and spontaneous ovulation in the rat. Cell and Tissue Research 257, 1-8. Cavender, J.L. & Murdoch, W.J. (1988) Morphological studies of the microcirculatory system of periovula¬ tory ovine follicles. Biology of Reproduction 39, 989-997. Cotran, R.S. & Majno, G. (1964) The delayed and prolonged vascular leakage in inflammation. I. Topography of the leaking vessels after thermal injury. American Journal of Pathology 45, 261-281. Crone, C. (1987) The Malpighi lecture. From porositates carnis to cellular microcirculation. International Journal of Microcirculation: Clinical and Expérimen¬ tale, 101-122. Damber, J.-E., Cajander, S., Gâfvels, M. & Selstam, G. (1987) Blood flow and vascular appearance in preovulatory follicles and young corpora lutea in immature pregnant mare serum gonadotropintreated rats. Biology of Reproduction 37, 651-658. Espey, L.L. (1980) Ovulation as an inflammatory reaction a hypothesis. Biology of Reproduction 22, 73-106. Espey, L.L. ( 1986) Effect of time and dose ofindomethacin on follicular prostaglandins and ovulation in the rabbit. Endocrinology 119, 746-754. -
791 797.
Gonadotropin-releasing hormone induces ovulation hypophysectomized rats: studies on ovarian tissue-type plasminogen activator activity, messenger
in
ribonucleic acid content, and cellular localization. Endocrinology 122, 1486-1495. Krishna, ., Beesley, K. & Terranova, P.F. (1989) Histamine, mast cells and ovarian function. Journal
of Endocrinology 120, 363-371. Lipner, H. ( 1988) Mechanism ofmammalian ovulation. In The Physiology of Reproduction, Vol. 1, pp. 447^188. Eds E. Knobil & J. D. Neil. Raven Press, New York.
Majno, G., Palade, G.E. & Schoefl, G.I. (1961) Studies on
inflammation. II. The site of action of histamine and serotonin along the vascular tree: a topographic study. Journal of Biophysics Biochemistry and
Cytology 11,607-626. Morris, B. & Sass, M.B. (1966) The formation of lymph in the ovary.
Proceedings of the Royal Society of Biology 164,577-591. Movat, H.Z. (1985) The vascular phenomena of acute inflammation. In The Inflammatory Reaction, pp. 43-75. Elsevier, Amsterdam. Munalulu, B.M., HiUier, K. & Peddie, M.J. (1987) Effect of human chorionic gonadotrophin and indomethacin on ovulation, steroidogenesis and prostaglandin synthesis in preovulatory follicles of PMSG-primed immature rats. Journal of Reproduction and Fertility 80, 229-234. Murdoch, W.J. & McCormick, R.J. (1989) Production of low molecular weight chemoattractants for leuco¬ cytes by periovulatory ovine follicles. Biology of Reproduction 40, 86-90.
Murdoch, W.J., Peterson, T.A., van Kirk, E.A., Vincent, D.L. & Inskeep, E.K. (1986) Interactive roles of
progesterone, prostaglandins, and collagenase in the ovulatory mechanism of the ewe. Biology of
Reproduction 35, 1187-1194. Norjavaara, E., Olofsson, J., Gâfvels, M. & Selstam, G. (1987) Redistribution of ovarian blood flow after injection of human chorionic gonadotropin and luteinizing hormone in the adult pseudopregnant rat.
Endocrinology 120, 107-114. O'Donnell, S.R., Braun, R.A., Reid, J.J. & Wholohan, T. (1987) A histological method for studying the effect
of drugs on mediation-induced airway microvascular leakage in rodents. Journal ofPharmacology Methods
17,205-217. Okuda, V., Okamura, II.. Kanzaki, II.. Takenaka, ., Morimoto, K. & Nishimura, T. (1980) An ultrastruc¬ tural
study of capillary permeability of perifollicular capillaries of rabbit ovary during ovulation using car¬ bon particles as tracer. Acta Obstétrica et Gynaecologica Japónica 32, 859-870.
Okuda, Y., Okamura, H., Kanzaki, H. & Takenaka, A.
of rabbit ovarian ovulation. Journal of Anatomy 137,
(1983) Capillary permeability
follicles prior 263-269.
to
Otsuki, Y., Magari, S. & Sugimoto, O. (1986) Lymphatic
in rabbit ovaries during ovulation: an ultrastructural study. Lymphology 19, 55-64. Peters, H. & McNatty, K. (1980) The Ovary. A corre¬ lation of structure and function in materials, pp. 31-34 and 28. Canada Publishing, London. Piacsek, B.E. & Huth, J.F. (1971) Changes in ovarian venous blood flow following cannulation: effect of luteinizing hormone (LH) and antihistamine. Proceedings of Society for Experimental Biology and Medicine 138, 1022-1024.
capillaries
Reich, R., Kohen, F., Naor, . & Tsafriri, . (1983)
Possible involvement of lipoxygenase products of arachidonic acid pathway in ovulation. Prostaglandins 6, 1011-1020. Siegel, S. (1956) Non-parametric Statistics for the Behavioral Sciences. McGraw-Hill. Kogakusha,
Tokyo. Szego, CM. & Git in, E.S. (1964) Ovarian histamine depletion during acute hyperaemic response to luteinizing hormone. Nature 201, 682-684. Wallach, E.E., Atlas, S., Dharmarajan. A.M., Oski, J.A. & Santulli, R. (1989) The periovulatory interval: Physiologic and endocrinologie implications. In Development of Preimplantation Embryos and their Environment, pp. 87-100. Alan R. Liss Inc., New
York.
Wedmore, CV. & Williams, T.J. (1981) Control of vas¬
cular permeability by polymorphonuclear leucocytes in inflammation. Nature 289, 646-650. Widmark, ., Bergh, ., Damber, J.-E. & Smedegárd, G. (1987) Leukocytes mediate the hCG induced increase in testicular venular permeability. Molecular and Cellular Endocrinology 53, 25-31. Williams, I.J. (1985) Vascular responses and their suppression; vasodilation and edema. In The Pharmacology of Inflammation, pp. 49-59. Eds I. L. Bontà, M. A. Bray & M. J. Parnham. Elsevier,
Amsterdam.
Yoshimura, Y., Nakamura, Y., Shiraki, M., Hirota, Y., Yamada, 11., Ando, M., Ubukata, Y. & Suzuki, M. Involvement of leukotriene B¿ in ovulation in the rabbit. Endocrinology 129, 193-199. Zachariae, F., AsIkk-I lausen. G. & Boseila, A.W.A. (1958) Studies on the mechanism of ovulation. Acta Endocrinologica 28, 547-552. Received 12 April 1991
(1991)
