Expression of prostaglandin synthesizing enzymes (cyclooxygenase 1 and cyclooxygenase 2) in the ovary of the ostrich (Struthio camelus)

Acta Histochemica 117 (2015) 69–75

Contents lists available at ScienceDirect

Acta Histochemica journal homepage: www.elsevier.de/acthis

Expression of prostaglandin synthesizing enzymes (cyclooxygenase 1 and cyclooxygenase 2) in the ovary of the ostrich (Struthio camelus) Daniela Rodler, Fred Sinowatz ∗ Institute of Anatomy, Histology and Embryology, Department of Veterinary Sciences, Ludwig-Maximilians University Munich, Veterinaerstrasse 13, D-80539 Munich, Germany

a r t i c l e

i n f o

Article history: Received 28 August 2014 Received in revised form 5 November 2014 Accepted 9 November 2014 Keywords: Ostrich Ovary Cyclooxygenase Immunocytochemistry In situ hybridization

a b s t r a c t Cyclooxygenase is the rate limiting enzyme in the production of prostaglandins. In birds two isoforms are present: cyclooxygenase 1 (COX-1) and cyclooxygenase 2 (COX-2). Despite evidence implicating that cyclooxygenases and PGs are critical factors in female reproduction in birds, little is known about COX expression in the avian ovary. In birds, cyclooxygenases have been studied in very few species only. In this study we report on the expression of COX-1 and COX-2 in the ovary of the ostrich (Struthio camelus) using immunohistochemistry and non-radioactive in situ hybridization techniques. Our results demonstrate that COX-1 is strongly expressed in the cytoplasm of oocytes of previtellogenic follicles, whereas COX-2 shows the strongest immunostaining in the granulosa cells of previtellogenic follicles. The signals of both isoenzymes fade significantly with increasing diameter and finally nearly vanish in the vitellogenic follicles with a size >1.8 cm. This expression pattern in the ostrich (S. camelus) is, therefore, completely different from the localization of COX-1 and COX-2 in the hen (Gallus gallus), a finding which also suggests different functions of the cyclooxygenases in the ostrich species. Non-radioactive in situ hybridization confirmed that COX-1 is synthesized in the ooplasm and COX-2 in the granulosa layers of early previtellogenic follicles. According to the results of this study it appears unlikely that COX-1 or COX-2 play a major role in ovulation and oviposition in the ostrich. © 2014 Elsevier GmbH. All rights reserved.

Introduction The ovulated oocyte of the ostrich (Struthio camelus) is regarded as the largest cell in the animal kingdom. Little is still known about the folliculogenesis and ovulation in this bird species but studies in mammalian ovaries suggest that prostaglandins produced by cyclooxygenases (COX) may be involved in these processes (Sirois et al., 2004). Cyclooxygenase is the rate limiting enzyme in the production of prostaglandins. It exists in two isoforms: the constitutively expressed cyclooxygenase 1 (COX-1) and the usually induced form, cyclooxygenase 2 (COX-2). Possessing two separate, but linked active sites, COX catalyses the bis-dioxygenation and subsequent reduction of arachidonic acid (AA) to an intermediate prostaglandin, PGH2 . Downstream enzymes, such as prostaglandin E synthase (PGE synthase), convert PGH2 to a family of prostaglandins (PGs), each member of which exerts a range of physiological effects through G-protein coupled receptors. COX-1

∗ Corresponding author. E-mail addresses: [email protected] (D. Rodler), [email protected] (F. Sinowatz). http://dx.doi.org/10.1016/j.acthis.2014.11.005 0065-1281/© 2014 Elsevier GmbH. All rights reserved.

is involved in homeostatic functions, while COX-2 has been implicated in pathological processes such as inflammation and cancer (Smith et al., 2000). Both isoforms of cyclooxygenase have been shown to be involved in normal and pathological conditions of the female reproductive tract, like ovulation, menstruation, implantation and parturition (Jabbour et al., 2009). All of these events are associated with upregulation in the expression of an array of inflammatory mediators, which include cytokines, growth factors and lipid mediators, influencing growth and functions of the immune and vascular compartment. Ovulation, the release of the female germ cell, the ovum, from the ovary is a key event in mammalian and avian reproduction. Since many years, this event has captured the attention of investigators, who have approached the phenomenon in a number of different ways. In recent times, ovulation has been linked to an inflammatory response with expression of mediators of inflammation, like the prostaglandin synthesizing enzymes COX-1 and COX-2. An important question is the identity and location of the principal mediators of the ovulatory process within the ovary. To clarify this issue, one must consider several functional domains of the follicle (the cumulus cells, the granulosa cells, and the thecal cells) but also the stromal cells of the ovary. In the mammalian

70

D. Rodler, F. Sinowatz / Acta Histochemica 117 (2015) 69–75

ovary, COX-1 is constitutively expressed and confined to the interstitial cells and cells of the corpus luteum (Sirois and Richards, 1993). COX-2 expression is induced by the LH-wave preceding ovulation in mammals and has been localized to granulosa cells. Besides the COX-2 gene (Sirois et al., 1992), LH-induced genes include that for progesterone receptor (PR) (Park and Mayo, 1991), CAATenhancer binding protein beta (C/EBP) (Sirois and Richards, 1993), and early growth regulatory factor-1 (Egr-1) (Espey et al., 2000; Espey, 2006). Genes involved in luteinization are also induced rapidly by LH. As there are no corpora lutea formed in birds, it has been hypothesized that post ovulatory follicles (POFs) serve as endocrine tissue and may also express COX-1 and COX-2. Previous immunohistochemical studies in the ovary of hen have shown that COX-1 is expressed in the surface epithelium of the female gonad and in cortical stromal cells adjacent to the growing follicles. In the few studies performed in avian species it was observed that in the hen ovary COX-2 is also expressed in granulosa cells and interstitial tissue of the ovary. Despite evidence implicating that cyclooxygenases and the resulting PGs are critical factors in female reproduction, little is known about COX expression in the vertebrate ovary only. In birds, cyclooxygenases have been studied only in very few species. In the domestic hen (Gallus domesticus), Hales et al. (2008) was able to localize COX-1 in the granulosa cell layer and cortical stroma, ovarian surface epithelium and postovulatory follicles. COX-2 was highly expressed in the interstitial tissue of the normal ovary and to a lesser degree in the granulosa cells of the follicles (Hales et al., 2008). The present study was undertaken to evaluate the hypothesis that cyclooxygenase may also play a role in ovulation in the ostrich. Materials and methods Animals and tissue preparation A total of 25 female (Zimbabwe blue/African black hybrid) ostriches (S. camelus) aged between 13 and 15 months and weighing 95–110 kg were used in the present study. Eighteen of the birds had active ovaries which contained 10–25 small, predominantly yellow-yolk follicles, the diameters of the largest follicles ranging from 10 to 19 cm. The immature ovaries showed an average size of 10–12 cm. The ostriches were slaughtered at a commercial ostrich abattoir at the German ostrich farm Donaumoos (Engelhardt family), Leipheim, employing a standard slaughter protocol. Ovarian tissue samples were obtained from the birds immediately after slaughter and immediately fixed in Bouin’s solution (1500 ml of picric acid, 500 ml of glacial acetic acid, and 100 ml of 37% formalin) for 12 h. After fixation, tissues were immersed for 3 days in 70% ethanol. Then, appropriate tissue samples of different follicle sizes were cut and dehydrated in a graded series of ethanol before being embedded in paraffin wax using an automatic tissue processor (Shandon Duplex Processor, Frankfurt, Germany) and a Histostat Tissue Embedding Center (Reichert-Jung, Vienna, Austria). Paraplast sections of 5 ␮m were cut with a rotary microtome (Leitz Microm HM 340E, Wetzlar, Germany), and the sections were collected on “Superfrost” glass slides (Carl Roth, Karlsruhe, Germany). Immunohistochemistry Paraffin sections were dewaxed and then washed 3 times for 5 min with PBS buffer at pH 7.4. Sections for COX-1 were pretreated with 10 mM sodium citrate buffer at pH 6.0 (9 ml 0.1 M citric acid and 41 ml 0.1 M sodium citrate (Merck, Germany) diluted in 500 ml distilled water) in a microwave oven with a power of

600 W for 3 × 10 min. Endogenous peroxidase activity was blocked with 7.5% H2 O2 (diluted in distilled water) at room temperature for 10 min. Non-specific antibody binding was blocked with Dako Protein Block Serum Free (Dako Deutschland GmbH, Hamburg, Germany) for 10 min. The sections were incubated with polyclonal primary antibodies against COX-1 (diluted 1:250, host rabbit, ab53766 (256); Abcam, Cambridge, UK) and COX-2 (diluted 1:400, host goat, ab23672 (266); Abcam, Cambridge, UK) at 6 ◦ C overnight (immunogens: synthetic peptides derived from human COX-1/COX-2). Localization of the antigen was achieved using the avidin–biotin complex (ABC) technique (Hsu et al., 1981; Romeis, 2004). The appropriate biotinylated secondary antibodies against COX-1 (diluted 1:300, anti-rabbit from pig; Dako, Hamburg, Germany) and COX-2 (diluted 1:400, anti-goat from rabbit; Dako, Hamburg, Germany) were incubated with the sections for 16 h at room temperature. Subsequently, treatment with StreptABComplex/HRP (Dako Deutschland, Hamburg, Germany) was performed for 30 min at room temperature, and treatment with 1 mg/ml 3,3 diaminobenzidine tetrahydrochloride (DAB Tablets, 10 mg; Biotrend Chemikalien, Köln, Germany) was performed for 5 min. All of the incubations were performed in a humidified chamber. Sections were either left unstained or were counterstained with hematoxylin (20 s), dehydrated, and mounted with Eukitt quickhardening mounting medium for microscopy (Fluka Analytical© , Sigma–Aldrich Laborchemikalien Seelze, Germany). Negative controls were performed by either replacing the primary antibody with buffer or non-immune serum or by incubating with the 3,3 -diaminobenzidine reagent alone to exclude the possibility of detecting non-suppressed endogenous peroxidase activity. A lack of detectable staining in the negative controls demonstrated that the reactions were specific. The images were captured with a Leica Labo-Lux® microscope equipped with a Zeiss Axiocam® camera (Zeiss, Munich, Germany). As positive controls, ovarian tissue from several mammalian species (bull, dog) of proven immunoreactivity was used. Non-radioactive in situ hybridization (NISH) In situ hybridization was performed using a standard protocol used in our laboratory. Small pieces of ovarian tissue (side length: 5 mm) were fixed in 3.7% formalin for 12 h and embedded in paraffin wax. Serial sections (5 ␮m) were mounted on amino propylene ethoxysilane-coated slides and dried at 50 ◦ C. All steps during in situ hybridization were conducted under strict RNAse-free conditions, and all solutions for RNA in situ hybridization were prepared using diethylpyrocarbonate (DEPC)-treated water. Sections were deparaffinized with xylene (3 × 10 min), immersed in absolute ethanol (2 × 5 min), and air-dried. Immediately after drying, the sections were dipped in 2% saline sodium citrate (SSC) pre-warmed in a water bath (80 ◦ C) for 10 min, followed by cooling off for 20 min at room temperature. Slides were then sequentially washed in distilled water (2 × 5 min) and Tris buffer (pH 7.4; 2 × 5 min) and incubated for 20 min with 0.05% proteinase E (VWR, Ismaning, Germany) in Tris buffer at room temperature. Sections were subsequently washed in Tris buffer (2 × 5 min) and distilled water (2 × 5 min) and postfixed for 10 min in freshly prepared 4% paraformaldehyde/PBS (pH 7.4). After further washing in PBS (2 × 5 min) and distilled water, slides were dehydrated in an ascending graded series of ethanol and air-dried. Oligonucleotide probes labeled with biotin were diluted with RNA hybridization buffer (DAKO, Munich, Germany) to a final concentration of 5 pmol/ml. Hybridization was carried out by overlaying the dry sections with 40 ␮l of the hybridization mixture. Incubation was done under a cover slip in a humidified chamber at 38 ◦ C overnight. Then, slides were washed in SSC

D. Rodler, F. Sinowatz / Acta Histochemica 117 (2015) 69–75

71

Table 1 Antisense and sense biotinylated oligonucleotides. Oligo nucleotides COX-1 COX-2

Sense/Antisense

Alias

5 tca ggt ggt tct ggg aca tca 3 5 tgt agc cgt act ggg agt tga a 3 5 ctg ctc cct ccc atg tca ga 3 5 cac gtg aag aat tcc ggt gtt 3

Sense Antisense Sense Antisense

(2 × 15 min) prewarmed to 38 ◦ C, distilled water (2 × 5 min), and Tris Buffer (2 × 5 min). Detection of transcripts was performed using a streptavidin–biotin-peroxidase complex kit and diaminobenzidine (DAKO, Hamburg, Germany) according to the manufacturer’s instructions. For negative controls, parallel sections were hybridized either with the sense oligonucleotide probe or with buffer alone. Tissue from bovine uterus was used as a positive control. Used antisense and sense biotinylated oligonucleotides for COX-1 and COX-2 are shown in Table 1.

Results Overview and organization of the ostrich ovary The ovary of the ostrich was composed of the cortex, medulla, and ovarian stalk. The tongue-shaped ovary of immature ostriches had a size of about 10 × 5 cm, showed rather compact design and contained small follicles up to 0.5 cm (Fig. 1a). The ovary of mature ostriches was filling the whole abdominal cavity and had a size up to 70 cm and contained follicles up to 19 cm in diameter. The prominent, protruding healthy follicles of the mature bird lent the characteristic grape-shaped appearance to the avian ovary. Various sizes of follicles could be seen in the cortex whereas the medulla and ovarian stalk contained large nerve bundles and blood vessels. Deep surface crypts partially divided the ovarian cortical region into compartments. White primordial and previtellogenic follicles, as well as vitellogenic follicles with yellow yolk content could be differentiated (Fig. 1b). White-colored primordial follicles were distributed in the peripheral area of the cortex, directly underneath the ovarian surface. Based on an earlier avian follicle size scheme (Ohtsuki et al., 2004) the size of primordial follicles varied between 90 and 100 ␮m in diameter. They consisted of a primary oocyte enclosed by flat or cuboidal follicle cells, the granulosa cells. In the outer periphery, beyond the granulosa cells, a layer of fibroblasts with interspersed melanocytes surrounded the primordial follicles. The ooplasm of the oocyte showed a characteristic avian feature, the Balbiani body,

Label

Gene bank

Material

5 -Biotin

XM 425326

Chicken specific

5 -Biotin

XM 422297

Chicken specific

which consisted of an aggregation of organelles, such as mitochondria and endoplasmic reticulum and which is localized close to the nucleus. Early previtellogenic follicles, which were white and measured 100–150 ␮m in diameter, were also situated near the ovarian surface epithelium and often protruded from the surface of the organ. Late previtellogenic follicles (150–400 ␮m) were located in the deeper regions of the cortex. Vitellogenic and preovulatory follicles (>400 ␮m in diameter and up to 19 cm), which had incorporated a large amount of yolk during follicular maturation and, therefore, appeared yellow, were surrounded by a monolayer of flat to cuboidal granulosa cells that sat upon a distinct basement membrane. Inner and outer thecal layers that are composed of fibroblasts and smooth muscle cells surrounded these follicles.

Immunohistochemical observations On the basis of visual examination by two independent observers, the relative intensities of COX-1 and COX-2 immunostaining were rated as negative (−), weak (+), moderate (++) or strong (+++). The immunostaining intensities are summarized in Table 2. The expression of COX-1 in the ovary of the ostrich showed a characteristic pattern during folliculogenesis. A weak immunostaining for COX-1 was observed in the cytoplasm of primordial follicles, which often occur singly, but also in groups of four to six follicles in the peripheral areas of the ovary. The oocyte was surrounded by a layer of flat granulosa cells, which did not express COX-1. During their growth to early previtellogenic follicles, a crescent-shaped Balbiani’s vitelline body was noted in the ooplasm in contact with the nuclear membrane, which stained distinctly stronger for COX-1 than the rest of the cytoplasm (Fig. 2a). Their simple cuboidal or columnar granulosa cell layer, surrounding the oocyte, showed no immunopositivity for COX-1. Concordantly, the undifferentiated thecal layer, which can be discerned at this stage of folliculogenesis, was negative for COX-1. The strongest COX-1 expression was found in the cytoplasm of oocytes of early (Fig. 2b)

Fig. 1. Ostrich ovary. (a) Immature gonad; (b) mature gonad. Scale bars = 8 cm.

72

D. Rodler, F. Sinowatz / Acta Histochemica 117 (2015) 69–75

Table 2 Immunostaining intensity of different ovarian structures. Antibody

COX-1 COX-2

Previtellogenic follicle

Vitellogenic follicle

GC

Primordial follicle OOP

BB

TH

GC

OOP

TE

TI

GC

OOP

ST

Other ovarian tissue V

Nv

OSE

TA

− +/++

+/++ −

+++ −

− −/+

− +++

+++ −

− ++

− +

− +

+/− −

− +

− ++

− −

− ++

− +

− = negative; + = weak; ++ = moderate; +++ = strong. BB – Balbiani body; GC – granulosa cells; Nv – nerve; OOP – ooplasm; OSE – ovarian surface epithelium; ST – ovarian stroma cells; TH – theca around previtellogenic follicles; TE – theca externa; TI – theca interna; TA – tunica albuginea; and V – vessel.

and late previtellogenic follicles (Fig. 2c). The oocytes of vitellogenic follicles contained a variable amount of yolk depending on the stage of development. With transition to the vitellogenic follicle stage and increasing size of the follicles, the immunostaining for COX-1 gradually declined in the cytoplasm of the oocyte. In smaller vitellogenic follicles, the peripheral area of the oocyte, although still displaying stronger reaction than the inner ooplasm, was yet to show decreasing staining intensity with follicular growth (Fig. 2d), which eventually was lost in vitellogenic follicles with a size of 1.8 cm or larger. The simple cuboidal to flat granulosa cell layer, which surrounds the oocyte of vitellogenic follicles, was completely negative for COX-1. Similarly, the now fully differentiated theca layers were immunonegative for COX-1. No immunostaining for COX-1 was found in the surface epithelium of the ostrich ovary, in the endothelial cells of the blood vessels and in the smooth muscles cells located in the ovarian stroma. A complex and rather different immunostaining pattern could be demonstrated for COX-2. Whereas the granulosa cell layer of primordial follicles showed weak to moderate immunostaining for COX-2, the pseudostratified columnar granulosa cell layer in

early (Fig. 3a) and late previtellogenic follicles (Fig. 3b and c) was distinctly to strongly immunopositive for this enzyme. The simple cuboidal granulosa cell layer, which surrounds the large vitelline follicle oocyte, and which contains a variable amount of yolk depending on the stage of development, showed only a weak immunostaining for COX-2. During the follicular development, the growing oocytes themselves were always negative for COX-2 in all classes of follicles. The basal areas of the ovarian surface epithelium and the endothelium of the blood vessels displayed a distinct immunostaining for COX-2, whereas the smooth muscle cells stained weakly to distinctly positive for this enzyme (Fig. 3d). Results of non-radioactive in situ hybridization In situ hybridization with the antisense probe against COX-1 and COX2 gave a positive signal in the ooplasm (COX-1, Fig. 4a and b) and the granulosa layers of early previtellogenic follicles (COX-2, Fig. 4c and d), respectively. The incubations with the sense probes showed only very weak reaction. This confirms the

Fig. 2. COX-1 staining. (a) Primordial follicles, showing weak staining for COX-1 in the ooplasm (Oop), and stronger intensity in the Balbiani body (BB). N = nucleus. (b) Early previtellogenic follicle. A distinct expression for COX-1 is observed in the ooplasm (Oop). Although the staining of the granulosa layer at all follicular stages is negative, this image nevertheless shows a non-specific reaction, which highlights some granulosa cells (GC). (c) Late previtellogenic follicle. The reaction with COX-1 in the ooplasm (Oop) of the left follicle becomes slightly weaker than at the early previtellogenic stage. Right side: vitellogenic follicle. The granulosa cells (GC) of all follicular stages remain negative. (d) Small vitellogenic follicle. The ooplasm (Oop) of this follicle, that is incorporating yolk to become a vitellogenic follicle, shows decreasing staining reaction with the COX-1 antibody. With follicular maturation, the staining intensity further diminishes. GC = granulosa cells; N = nucleus. Scale bars: (a–c) = 50 ␮m; (d) = 12 ␮m.

D. Rodler, F. Sinowatz / Acta Histochemica 117 (2015) 69–75

73

Fig. 3. COX-2 staining, (a) Early previtellogenic follicle showing distinct immunopositivity for COX-2 in the granulosa cells (GC) of this follicle, while the ooplasm (Oop) remains negative at all follicular stages. The ovarian surface epithelium cells (OSE) display some staining at their basal area. (b) Late previtellogenic follicles. The granulosa cells (GC) of previtellogenic follicles display the strongest reactions with the COX-2 antibody. Oop = ooplasm; OSE = ovarian surface epithelium; N = nucleus. (c) Healthy and atretic previtellogenic follicles. The strong positive staining for COX-2 in the granulosa cells (GC) can still be observed in atretic follicles (small arrows), even though with weaker signals. Oop = ooplasm. (d) Late previtellogenic follicle. During maturation of the follicle the number of COX-2 positive cells in the granulosa cell layer (GC) decreases. They can be barely seen in vitellogenic stages. The endothelium of vessels (V) always displays positive reactions with COX-2 antibodies. Oop = ooplasm; Th = Theca layer. Scale bars: a = 50 ␮m; b and d = 25 ␮m; c = 100 ␮m.

immunocytochemical results that the cyclooxygenases are synthesized in these locations. Discussion Prostaglandins (PGs) were first isolated from seminal fluid from man and several mammals (Von Euler, 1936) and erroneously thought to be derived from the prostate. Subsequent studies proved that the prostaglandins from the seminal fluid primarily originate from the seminal vesicles and that they play an important role in male and female reproduction (Eliasson, 1959). Recent data have shown that in birds as in mammals, PGs are associated with ovulation, ovum transport through the oviduct and oviposition (Hales et al., 2008). It has been also shown that the ability of PGs to stimulate oviductal contractions varies regionally along the oviduct. Cyclooxygenase (COX) is regarded as the major enzyme that catalyses the conversion of arachidonic acids to PG and other eicosanoids (Kirschenbaum et al., 2000). In birds, two isoforms of COX have been identified so far: COX-1, which has been reported to be constitutively expressed in several tissues, is necessary for various homeostatic functions (Smith et al., 2000), and COX-2, which is regarded to be inducible by various factors, including steroid hormones, cytokines and growth factors (Hales et al., 2008). COX-2 has therefore been shown to be the induced form of the enzyme in mammals and birds (Sirois et al., 2004; Hales et al., 2008). The results obtained in the present study do not support this generally accepted dichotomy of cyclooxygenase isoenzymes in a constitutively expressed (COX-1) and an induced isoform (COX-2). Both enzymes showed a characteristic, but different expression pattern in the ovary of the ostrich,

which was obviously dependent on the stage of follicular development. Ovulation in birds involves a complex series of biochemical and biophysical processes that ultimately leads to the rupture of the preovulatory follicle and release of the megalecithal oocyte. The process of ovulation shows all the signs of an acute, self-controlled inflammatory reaction, including hyperemia, leukocyte extravasation, edema, and induction of proteolytic activities (Espey, 1980; Espey and Lipner, 1994; Song et al., 1998). Prostaglandins, which are known to play an important role in inflammation, have been recognized as key mediators of ovulation in mammals for more than 30 years. COX-2 production is induced by the preovulatory LH surge in the granulosa cell layer of the follicle prior to ovulation (Smith et al., 2000). Non-steroid anti-inflammatory drugs (NSAID) such as aspirin and indomethacin inhibit ovulation in many mammalian species including rats, rabbits, pigs, sheep, cows and humans (Sirois et al., 2004). In these species, the LH surge causes a marked increase in the concentrations of PGE2 and PGF2 in ovarian follicles just prior to ovulation, with granulosa cells as the primary sites of prostaglandin synthesis. Different mechanisms have been suggested as potential mechanisms responsible for the subsequent increase in follicular prostaglandins due to the LH surge (Sirois et al., 2004), including an increase in arachidonic acid release, induction of COX-2 expression in the granulosa cell layer of follicles and/or rise in prostaglandin synthase activity (Clark et al., 1978). As in mammalian species, COX-2 is responsible for the generation of PGE2 in the preovulatory follicles that are destined to ovulate in the hen (Hales et al., 2008; Johnson, 2012). However, a notable difference in the expression of COX-2 between the hen

74

D. Rodler, F. Sinowatz / Acta Histochemica 117 (2015) 69–75

Fig. 4. In situ hybridization. (a) COX-1, antisense oligonucleotide: the expression of COX-1 can be reproduced by the positive staining of the ooplasm (Oop) of especially early previtellogenic follicles. (b) COX-1, sense oligonucleotide: the ooplasm of the same follicle as in Fig. 4a remains almost negative. (c) COX-2, antisense oligonucleotide: the expression of COX-2 is corroborated by the positive staining of the granulosa cells (GC) of early previtellogenic follicles. (d) COX-2, sense oligonucleotide: the granulosa cell layers (GC) of the same follicles as in Fig. 4c remain largely negative. Scale bars = 50 ␮m.

and the ostrich appears in the time course of COX-2 expression. In the hen, COX-1 is localized to the granulosa cell layer and cortical interstitium, ovarian surface epithelium and the postovulatory follicle (Hales et al., 2008). Changes in prostaglandin levels in the ovary of the hen, contractions of the ovary and occurrence of oviposition suggest that prostaglandins produced by the follicle from which the ovulation will occur within 1 hour are responsible for stimulating uterine contractions and hence expulsion of the egg from the uterus. In the hen, there is a significantly higher expression of COX-1 in the first preovulatory follicle (POF-1), compared to POF-2 and POF-3. POF-1 is, therefore, regarded as an important site of prostaglandin production in the hen. Hales et al. (2008) demonstrated a strong immunostaining for COX-1 in POF-1 and therefore suggested that COX-1 is important for the production of prostaglandins in POF-1 and for oviposition. In the ostrich, COX-1 had its strongest expression in the previtellogenic follicles and was very weak or nearly negative in vitellogenic follicles with a size >1.8 cm. No immunostaining for COX-1 was found in the POFs. Contrary to the hen, COX-1 appears neither to be involved in ovulation nor in oviposition in the ostrich. Also, the attenuation of COX-2 protein expression in the follicular wall of vitellogenic follicles does not support the idea that COX-1 and COX-2 are involved in ovulation in this species. Recent studies in cynomolgus monkeys showed that COX-2 inhibitors significantly reduced the rates of nuclear maturation of oocytes (Duffy, 2011). The authors concluded that COX-2 expression is necessary for certain stages of oocyte nuclear maturation. A similar scenario with a temporarily high expression of COX-2 may influence nuclear maturation processes during folliculogenesis in the ostrich, which appears interesting, but certainly additional studies are necessary to prove this hypothesis also in the avian system.

Acknowledgements The authors gratefully acknowledge the technical assistance of Mrs Monica Settles, Mrs Angela Servatius, Mrs Andrea Kügel, and Mr Yilmaz Göck. The authors are also grateful to the Engelhardt family of the German ostrich farm Donaumoos in Leipheim for providing ovaries from their ostriches. References Clark MR, Marsh JM, LeMaire WJ. Mechanisms of luteinizing hormone regulation of prostaglandin synthesis in rat granulosa cells. J Biol Chem 1978;253:7757–61. Duffy DM. Maturation and fertilization of non-human primate oocytes are compromised by oral administration of a COX2 inhibitor. Fertil Steril 2011;95:1256–60. Eliasson R. Studies on the prostaglandin occurrence, formation and biological actions. Acta Physiol Scand 1959;158(Suppl.):1–73. Espey LL. Ovulation as an inflammatory response – a hypothesis. Biol Reprod 1980;22:73–106. Espey LL. Comprehensive analysis of ovarian gene expression during ovulation using differential display. Methods Mol Biol 2006;317:219–41. Espey LL, Lipner H, Ovulation. Knobil E, Neil JD, editors. Physiology of reproduction. New York: Raven Press; 1994. p. 725–81. Espey LL, Ujoka T, Russell DL, Skelsey M, Vladu B. Induction of earlygrowth response protein-1 (Egr-1) gene-expression in the rat ovary in response to an ovulatory dose of hCG. Endocrinology 2000;141:2385–91. Hales DB, Zunge Y, Lagman JA, Ansenberger K, Mahon C, Barua A, et al. Cyclooxygenase expression and distribution in the

D. Rodler, F. Sinowatz / Acta Histochemica 117 (2015) 69–75

normal and their role in ovarian cancer in the domestic hen (Gallus domesticus). Endocrinology 2008;33:235–44. Hsu SM, Raine I, Fanger H. Use of avidin–biotin-peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabelled antibody (PAP) procedure. J Histochem Cytochem 1981;29:577–80. Jabbour HN, Sale KJ, Catalano RD, Norman JD. Inflammatory pathways in female reproductive health and disease. Reproduction 2009;138:903–19. Johnson PA. Follicle selection in the avian ovary. Reprod Domest Anim 2012;47(Suppl. 4):283–7. Kirschenbaum A, Liotta DR, Yao S, Liu X, Klausner AP, Unger P, et al. Immunohistochemical localization of cyclooxygenase1 and cyclooxygenase-2 in the human fetal and adult male reproductive tracts. J Clin Endocrinol Metabol 2000;85: 3436–41. Ohtsuki M, Hanafy AM, Mori M, Sasanami T. Involvement of interaction of ZP1 and ZPC in the formation of quail perivitelline membrane. Cell Tissue Res 2004;318:565–70. Park OK, Mayo K. Transient expression-of progesterone receptor messenger-RNA in ovarian granulosa cells after-the preovulatory luteinizing hormone-surge. Mol Endocrinol 1991;5:967–78.

75

Romeis B. Färbemethoden. In: Mulisch M, Welsch U, editors. Romeis – Mikroskopische Technik. Berlin: Springer Spectrum; 2004. p. 198–281. Sirois J, Richards JS. Transcriptional regulation of the rat prostaglandin endoperoxide synthase 2 gene in granulosa-cells. J Biol Chem 1993;268:21931–8. Sirois J, Sayasith K, Brown KA, Stock AE, Bouchard N, Dore M. Cyclooxygenase-2 and its role in ovulation: a 2004 account. Hum Reprod Update 2004;10:375–85. Sirois J, Simmons DL, Richards JS. Hormonal regulation of messenger-ribonucleic acid encoding a novel isoform of prostaglandin endoperoxide H-synthase in rat preovulatory follicles. J Biol Chem 1992;267:11586–92. Smith WL, Witt DL, Garavito RM. Cyclooxygenase: structural, cellular and molecular biology. Annu Rev Biochem 2000;69:145–82. Song JH, Sirois J, Houde A, Murphy BD. Cloning, developmental expression and immunohistochemistry of cyclooxygenase 2 in the endometrium during embryo implantation and gestation in mink (Musculus vision). Endocrinology 1998;140:2685–95. Von Euler US. On the specific vaso-dilatating and plain muscle stimulating substances from accessory genital glands in man and certain animals (prostaglandin and vesiglandin). J Physiol 1936;88:213–34.