Oxytocin Gene Expression and Oxytocin Immunoactivity in the Ovary ...

BIOLOGY OF REPRODUCTION 50, 1216-1222 (1994)

Oxytocin Gene Expression and Oxytocin Immunoactivity in the Ovary of the Common Marmoset Monkey (Callithrix jacchus) A. EINSPANIER,'

2

R. IVELL,3 G. RUNE, 4 and J.K. HODGES 2

Department of Reproductive Biology, 2 German Primate Centre, Go5ttingen, Germany Institutefor Hormone and Fertility Research,3 University of Hamburg, Hamburg, Germany Institute of Anatomy, 4 University of Berlin, Berlin, Germany ABSTRACT Oxytocin was identified in ovaries recovered on Day 5 (± 1) of the luteal phase from three female marmoset monkeys (Callitbrixjacchus). With use of a reverse transcription-polymerase chain reaction assay, expression of mRNA for oxytocin and oxytocin receptor was detected in both luteal tissue and in the ovarian remnant. Evidence for ovarian synthesis of oxytocin was provided by immunohistochemistry, which showed positive staining for oxytocin and neurophysin in the cytoplasm of the luteal cells. Some luteal cells had a more intensely stained perinuclear region than others for oxytocin immunoreactivity, whereas the staining for neurophysin was evenly distributed. Granulosa and theca cells of antral follicles also showed positive staining for oxytocin immunoreactivity; no reactivity was found in fibroblast or endothelial cells. Oxytocin immunoreactivity was also detected in the luteal tissue of all animals by immunoassay, with values ranging from 2.8 to 12.1 ng/g wet weight. The oxytocin concentration for the ovarian remnant was either very low (0.55-0.75 ng/g wet weight) or nondetectable (< 0.5 ng/g wet weight). Local production of oxytocin within the ovary was suggested by the measurement of higher oxytocin concentrations in the blood from ovaries containing corpora lutea compared with peripheral blood. Collectively, these results provide evidence for ovarian biosynthesis of oxytocin and suggest the possibility of a paracrine role in the regulation of primate ovarian function.

INTRODUCTION

The presence of oxytocin (OT) in the ruminant ovary and its localization primarily within the corpus luteum (CL) is well established [1, 2]. The available data describe an involvement of ovarian OT in determining the functional life span of the CL [3-5] and in regulating ovarian steroidogenesis [6, 7]. OT has also been shown to be present in the ovaries of women [8, 9] and nonhuman primates, such as baboons [10, 11] and cynomolgus monkeys [12], although the reported levels are generally much lower than those in ruminants. The major source of OT within the primate ovary is still unclear. Several studies have reported that immunoreactive OT (OT-ir) as well as OT mRNA could be detected only in the CL [8-13] whereas a more recent study reported the presence of higher levels of immunoassayable OT in the non-CL-bearing ovary than in the luteal tissue [14]. Thus, although there is good evidence for the local production of OT in the primate ovary, the low levels measured and the uncertainty over the relative importance of a luteal or nonluteal source has made it difficult to ascribe a physiological role for the peptide at the ovarian level. A luteotrophic function has been suggested by several authors on the basis of the stimulation of progesterone by luteal tissue in vitro [8, 15]. On the other hand, a luteolytic effect of OT has been demonstrated for luteal cells both in vivo and in vitro [11, 16-19], and a possible dual action of OT has also been proposed [8]. Accepted January 14, 1994. Received October 28, 1993. 'correspondence: Dr. A. Einspanier, Department of Reproductive Biology, German Primate Centre, Kellnerweg 4, 37077 G6ttingen, Germany. FAX: +49/5513851228.

As a part of a longer study to investigate the function of OT in the primate ovary using the common marmoset monkey, Callithrixcjacchus, we report here the identification and characterization of ovarian OT in this species. Luteal and nonluteal tissue from ovaries collected during the early luteal phase was examined with the following specific aims: 1) to positively identify the expression of OT and OT receptor (OTR) genes by mRNA detection, 2) to determine the presence of OT-ir by means of immunohistochemical staining techniques, and 3) to provide evidence for OT-ir biosynthesis by use of a specific assay combined with high pressure liquid chromatography (HPLC). MATERIALS AND METHODS

Animals Three adult female common marmoset monkeys (Callitbrixjacchus) with normal ovarian cycles were used in this study. The animals were housed in pairs under controlled conditions (24°C, 30% humidity, 13 h light photoperiod) in the German Primate Centre, Gottingen. Commercially prepared pellets (Sniff, Soest, Germany) were supplemented daily with fresh fruits and vegetables; water was available ad libitum. Blood samples (0.2 ml) were collected twice weekly by puncture of the Vena saphena, and plasma progesterone content was determined in order to monitor ovarian cyclicity. The day of ovulation was defined as that preceding a rise in plasma progesterone concentration above 10 ng/ ml [20]. Ovarian cycles were controlled by the application of a luteolytic dose of prostaglandin (PG) F2a (0.8 pLg Es-

1216

OVARIAN OXYTOCIN IN THE MARMOSET MONKEY trumate; Pitman-Moore, GmbH, Burgwedel, Germany) on Day 12 of the luteal phase [21]. Ovulation in the marmoset monkey occurs on average 10.7 days after PGF2,-induced luteolysis [21]. In order to ensure that the females did not conceive during the study, they were separated from their male partners 5 days after PGF2,, application and were returned to them 4 days after ovulation. Sample Collection Sample collection was carried out at the same stage of the ovarian cycle (early luteal phase) in all three animals. On Day 5 of the luteal phase, animals were sedated with ketanest/rompun mixture (0.2 ml/animal, i.m.) and narcotized with halothane-nitrous oxide-oxygen mixture. Blood samples (0.5 ml) were taken from the right and left V. ovarica, and the ovaries were subsequently removed. The CL (n = 7) were carefully removed under a dissecting microscope, and ovarian remnants (ovary minus CL, n = 6) were also saved. The pituitary gland was also removed from two of the animals after death. All tissue samples were immediately placed in liquid nitrogen and stored at -70°C until analysis. Peripheral (V. saphena) and heart blood samples were also taken. All blood samples were collected into chilled heparinized syringes and centrifuged at 1500 x g for 20 min; plasma was stored at -20°C for OT determination. Reverse Transcription-PolymeraseChain Reaction Assay (RT-PCR) Identification of mRNA for OT and OTR RNA was prepared from small amounts of CL or ovarian remnants by means of the microprocedure of Chomczynski and Sacchi [22]. Five micrograms of this total RNA was then used to program the synthesis of single-strand cDNA in a 100-pxl reaction mixture with use of an oligo(dT) primer and the enzyme AMV reverse transcriptase as described in detail by Ivell et al. [13]. One microliter of this cDNA was then used as template for the RT-PCR reaction, which was carried out in a 50-xl volume containing 100 ng of each primer oligonucleotide specific for either human OT or human OTR gene transcripts. The RT-PCR reaction proceeded for 30 cycles of 1 min at 95°C, 1 min at the annealing temperature, and 1 min at the elongation temperature of 72 0C. A hot-start touchdown protocol was used, wherein the annealing temperature was shifted progressively from 61°C to 51°C. The RT-PCR products were subsequently subjected to agarose gel electrophoresis followed by ethidium bromide staining. Confirmation of the product identity was obtained first by hybridization of a Southern blot from the agarose gel, using as probe a radiolabeled oligonucleotide encoding a sequence from within the specific gene transcripts (see below), followed by autoradiography. Further confirmation was obtained by eluting RT-PCR products from the agarose gel, subcloning these into the pCRII T-tailed plasmid vector (Invitrogen, Heidelberg, Germany), and sub-

1217

jecting them to double-stranded DNA sequencing. As positive controls, RNA from human or baboon CL and marmoset pituitary were used (not shown). As negative control, water was used instead of cDNA. No signals were detected for mRNA extracted from other tissues (liver, spleen, and muscle; not shown). The following oligonucleotides were used as probes. For OT mRNA [23,13]: 5'primer - ACCTCCGCCTGCTACATCCAGAACT (exon 1) 3'primer - CTCCGCGTCGCAGGCAGGGTCGGCGTG (exon 3) internal probe - 60 bp Bal I-Ava I fragment of the rat OT gene (exon 2) For OTR mRNA [24]: 5'primer - ATGGAGGGCGCGCTCGCAGCCAACT 3'primer - AAGAAGAAAGGCGTCCAGCACACGA internal probe - AGGCGGCACAGCAGGTCGGGCCCGTAGAAGCGGAAGGTGATGTCCCA Immunohistochemistry Frozen cryostat sections (8 Kxm) of CL, ovarian remnant, and pituitary were mounted on gelatine-coated slides. The sections were preincubated for 1 h with 10% normal human serum (Serotec, Berlin, Germany), rinsed three times with PBS (pH 7.2), and incubated overnight at 40C in a moist chamber with anti-OT (1:100) [25] or anti-neurophysin (1:100) [26, 27] antibodies. The immunoenzymatic labeling was carried out with use of immune complexes of alkaline phosphatase and monoclonal anti-alkaline phosphatase (APAAP-complexes; Dianova, Hamburg, Germany) [28]. To test specificity of the staining, control sections were set up without the hormone-specific antibody, and also with antiserum that had been preabsorbed with excess of OT. A minimum of two sections from luteal tissue as well as from the ovarian remnant were examined from each animal. Sections were mounted in glycerine jelly and examined through a Zeiss photomicroscope. Preparationof Samples for OT Assay The preparation of CL, ovarian remnant, and pituitary for OT-ir determination was similar to that described by Amico and Zeleznik [14]. The tissues were placed in hydrochloric acid (10 mg/100 1l 0.1 N HCI tissue) at 40C, homogenized, and centrifuged at 6000 x g for 30 min. The supernatants were stored at -20 0C in aliquots (10, 100 .l1). Plasma and tissue extracts were then applied to Sep-Pak columns (Waters Associates, Milford, MA) prewashed with 5 ml isopropanol and 10 ml distilled water. After sample application (100 ul1), columns were washed with 30 ml 0.01 M trifluoroacetic acid (TFA) and eluted with 3 ml each of 25%, 50%, and 75% acetonitrile. The eluates were air-dried and reconstituted in HPLC solvent (250 pl) overnight. All analyses of tissue extracts were performed after HPLC. Preliminary comparison of plasma samples before and after

1218

EINSPANIER ET AL. known amount of OT standard was also injected onto the column on each day that HPLC of sample extracts was performed. Between the different HPLC runs of extracted tissue and standard, solvents A and B alone were applied to the HPLC column, and fractions were collected and taken to assay in order to control for cross-contamination with the synthetic nonapeptide. Recovery values for both tracer and standard were greater than 86%.

FIG. 1. Autoradiograms of radioactively probed PCR products derived from various marmoset monkey tissues representing mRNA of OTR and OT. Samples were derived from individual CL (lanes 1, 2, and 6), ovarian remnants (lanes 3 and 5), and pituitary (lane 4). The correctly sized PCR products are indicated by the size in basepairs given at the side of each panel. K, control using water instead of DNA; M, DNA size marker (Haellldigested PhiX174DNA).

HPLC purification showed no difference in the values of OT measured. Therefore, results presented here for plasma are without HPLC. Extraction efficiency for tissue samples was monitored by adding tracer amounts of 125I-OT at the beginning of the extraction procedure. The recovery value was 88 ± 2%. All results were expressed without correction for recovery. For plasma, the extraction efficiency was 86 ± 3%. HPLC Chromatography was carried out on a C18u Bondapak column (3.9 mm x 30 cm, Waters Associates) with use of two solvents (solvent A: 0.01 M distilled TFA; solvent B: 0.01 M TFA containing acetonitrile [1:1]) as previously described [14]. The samples were chromatographed on a linear gradient system of 20-50% pump B for 15 min at 2 ml/min. The fractions (2 ml) were collected into glass tubes, airdried, and reconstituted in OT buffer (500 ,al) for detection in the OT assay. OT tracer ( 3H-OT, DuPont, Bad Homburg, Germany) and synthetic OT (Syntocinon, Sandoz; Basel, Switzerland) were used as markers to monitor column efficiency and indicate the elution position of authentic OT. Initially, 3 H-OT was injected together with extracted sample to monitor for co-elution of the tracer and measured immunoreactivity. However, because of 3H tracer interference with the 12 5I-OT assay, co-injection was not performed when HPLC was used for the quantitative assessment of tissue OT content. Instead 3H tracer was applied separately to the column immediately before and after each sample run, and a

Immunoassayfor OT OT was determined by RIA as described by Schams et al. [25]. The antiserum, generously provided by Dr. Schams (Freising, Germany) was used at a dilution of 1:25000 with '25 I-OT, 5000 cpm/tube) as tracer. Synthetic OT (Sandoz) was used as standard over the range 0.1-32.0 pg/tube. Assay sensitivity determined at 90% B/Bo was between 0.3 and 0.5 pg/tube. The antiserum cross-reacted less than 0.01% with arginine-vasopressin, lysine-vasopressin, arginine-vasotocin, neurophysin, angiotensin, LH, or FSH. The intraassay CV varied from 7.1% (low quality control [QC]) to 12.5% (high QC) and the interassay CV from 10.4 to 16.3%. Serial dilutions of the extracted tissue and plasma samples were parallel to the OT standard. The results for both plasma and tissue concentrations of OT-ir were not corrected for procedural losses. Values in plasma and tissue are expressed as pg/ml and ng/g wet weight, respectively. OT-ir was not detectable in OT buffer. RESULTS RT-PCR Identification of mRNA for OT and OTR The marmoset monkey ovarian and pituitary extracts showed positive signals for mRNA of OT and OTR by autoradiogram (Fig. 1). Both OT mRNA and OTR mRNA could be detected in luteal tissue (lanes 1, 2, and 6) and ovarian remnant (lanes 3 and 5) from the early luteal phase. Clear signals for the transcripts for both the hormone and its receptor were also detected in the pituitary (lane 4). Although staining intensities differ between the samples it should be stressed that PCR is not a quantitative method and conclusions regarding relative mRNA amounts in different tissues should not be made. The partially sequenced RT-PCR products all indicated gene structures closely homologous (> 95%) but not identical with the corresponding human or baboon transcripts. Localization of OT Immunoreactivity Immunoreactivity for OT and neurophysin was found in the cytoplasm of luteal cells as seen in Figure 2. In the vast majority of luteal cells, the dye was evenly distributed. A small proportion of luteal cells showed more intense staining for OT in the perinuclear zone. Regional differences in staining within CL from the same or different animals were not observed. No reactivity was seen in fibroblasts or endothelial cells.

1219

OVARIAN OXYTOCIN IN THE MARMOSET MONKEY

FIG. 2. A) Immunoreactivity of OT in marmoset monkey CL (Day 5 luteal phase) localized in luteal cells. Most cells exhibited an even cytoplasmic staining, although a small proportion showed a more intensely stained perinuclear region (arrowheads). B) Control sections showed no reactivity when first antibody was omitted. x350.

OT-ir was also found in the ovarian remnant. Granulosa and theca cells from antral follicles showed positive staining. Stromal cells were devoid of staining activity. The pituitary was also analyzed for OT-ir and, as expected, showed strong positive staining for OT. Staining intensity and distribution were identical whether the fixation protocol used paraformaldehyde (not shown) or cryopreservation. However, antigenicity was best preserved in native cryostat sections. No staining of tissue sections was observed after control incubation. Characterizationof OT Immunoreactivity OT-ir in the pituitary, CL, and plasma eluted from HPLC as a single peak at the same position as the synthetic OTstandard and 3H-OT (Fig. 3). In all HPLC runs performed, the peak of OT occurred in fraction 15, with occasional smaller amounts also present in fraction 16. OT-ir was detectable only in fractions 15 and 16. The levels of OT in individual CL determined after HPLC are shown in Table 1. Immunoreactivity was detectable in all extracted luteal tissues from each animal; concentrations ranged from 2.8 to 12.1 ng/g wet weight. In three out of five individual ovarian remnants examined, OT-ir was detected between 0.55 and 0.75 ng/g wet weight, whereas the other ovarian remnants had no detectable OT-ir (detection limit 0.5 ng/g wet weight). Levels of OT-ir were therefore much lower than those in the corresponding luteal tissue. Levels of OT-ir in heart blood and in peripheral and ovarian venous blood from each animal are shown in Table

2. OT levels in blood draining ovaries with two CL were higher than those in blood from ovaries with 1 CL or no CL, and also higher than values measured in heart and peripheral blood. Irrespective of the number of CL, samples from blood draining CL-containing ovaries contained much higher concentrations of OT than peripheral blood samples. DISCUSSION The present report describes a combined approach incorporating mRNA analysis, RIA, and immunohistochemical detection to provide information on the presence and distribution of OT within the luteal phase ovary of the common marmoset monkey, Callithrixjacchus. Evidence for mRNA expression for both OT and its cognate receptor in luteal tissue has been shown in this study. The sequenced products are closely homologous but not identical with human and baboon transcripts obtained by the same method [13]. Specificity for the transcripts was based on the sequence of the 5' primer and 3' primer, the size of the amplification product, and hybridization to an internal specific sequence, and was confirmed by partial sequence analysis of the cloned RT-PCR products. The presence of mRNA for OT in the pituitary, though unexpected, can probably be explained by the axonal transport of hypothalamic mRNA, as recently demonstrated in the rat [29]. Detection of OT receptor mRNA has also been described in the human ovary [24], although details of its localization in luteal and nonluteal tissues were not reported. Further-

1220

EINSPANIER ET AL. 1,000

3

H Tracer

TABLE 1. OT concentration (ng/g wet weight) in tissue extracts of individual CL from 3 animals.

800

E

Q.

Number of CI

OT content (ng/g wet weight)

w15

3

4.0 10.7 12.1

w27

2

2.8 3.3

w46

2

4.4 9.7

Animal

600 400 200 0 12 10 8 6 4 2

Standard

0 0) .0

3

0. .G

2

I.

1

CI extract

Z

0

0

3

ovarian remnant

2 0 0

_

' tjt*tol

.

>t lt

1 3 5 7 9 11131517192123252729

fraction FIG. 3. Profile of OT immunoreactivity after HPLC separation of OT standard and extracts of luteal tissue and ovarian remnant. In each case, the peak of immunoreactivity was measured in fraction 15 with smaller amounts in fraction 16. The OT RIA detection limit was 0.3-0.5 pg/tube.

more, a more recent study by Khan-Dawood et al. [30], reported the presence of specific binding sites for OT in baboon CL from different stages of the luteal phase. These results for the human and baboon together with our present findings in the marmoset monkey thus indicate that the primate ovary not only expresses the OTR gene but that functional receptor proteins are also produced. Collectively, these data provide strong evidence for a local action of OT within the primate ovary by a receptor-mediated mechanism. The detection of OT mRNA was confirmed by our immunohistochemical results, which indicated that OT-ir is

mainly localized in the cytoplasm of the luteal cells. The majority of luteal cells exhibited homogenous staining, whereas in a smaller proportion of cells staining was localized in the perinuclear zone. Several studies have reported positive OT-ir staining in the cytoplasm of only some luteal cells, mainly in the large luteal cells of ruminants [31, 32] and in the large cell type of both human and nonhuman primates [9-11]. In the marmoset monkey, however, clearly distinguishable populations of large and small luteal cells could not be demonstrated [33] and this apparent monophasic distribution of cell size might explain why in the present study all cells were found to stain positively for OT-ir. The differential intensity of staining observed in individual cells may nevertheless indicate differences between cells with respect to intracellular storage of OT and/ or indicate that the luteal cells differ in their capacity for OT synthesis, but not in their size distribution or differentiation state. Neurophysin, a component of the same precursor as that for OT, was also identified within the same cells, strongly supporting the view that these luteal cells are sites of de novo OT biosynthesis. The presence of OT-ir within the luteal tissue was also indicated by RIA measurements. The similarity between the HPLC elution profiles for sample extracts and for OT tracer and standard provide good evidence that the values reported here reflect authentic OT. Nevertheless, absolute specificity of measurement is not guaranteed with HPLC, TABLE 2. Concentrations of OT immunoreactivity in heart blood and V. saphena and V. ovarica of the 3 animals examined. Ovarian blood' Heart blood

Peripheral blood

L*

R**

w15

8.5

9.2

8.2 (1)

16.2 (2)

w27

13.0

22.0