Catecholaminergic innervation of the chicken ovary

Cell Tissue Res (1983) 230:431-450 Cell and Tissue Research �9 Springer-Verlag 1983

Catecholaminergic innervation of the chicken ovary With special reference to the follicular wall*

K. Unsicker, F. Seidel, H.-D. Hofmann, T.H. Miiller, R. Schmidt, and A. Wilson

Department of Anatomy and Cell Biology, Philipps University, Marburg, Federal Republic of Germany

Summary. The innervation of the chicken ovary was investigated with special emphasis on adrenergic nerves in the follicular wall. Quantitative determinations of catecholamines (CA) by high-performance liquid chromatography and electrochemical detection (hplc-ed) revealed 15.4 _+ 3.3 ng/mg protein of norepinephrine (NE) and 3.14 ng/mg protein of epinephrine (E), with even larger amounts in the cranial part of the ovary close to the adrenal gland. Serial sections that had been processed for the visualisation of aminergic nerves (Falck-Hillarp- or glyoxylic acid technique) showed CA localized in nerve-fibre bundles; cell bodies of chromaffin and sympathetic neurons were only found at the ovarian- adrenal junction suggesting that ovarian nerves stored considerable quantities of E. An antiserum against bovine phenylethanolamine N- methyltransferase (PNMT, the E-synthesizing enzyme) produced no im- munostaining in chicken ovary or adrenal gland, due to a lack of cross- reactivity between the antiserum and chicken PNMT.

Serial sections processed alternately for the visualisation of aminergic nerves and myosin (from chicken gizzard) immunoreactivity revealed a scarce nerve supply of contractile cells in the theca externa compared to an extraordinarily dense innervation of the endocrine interstitial tissue of the theca interna. This distribution pattern of nerve fbres in the follicular wall was confirmed by electron microscopy in ovarian tissue that had been pretreated with 5- or 6-hydroxydopamine (OHDA). More than 90% of the terminal axons were specifically labeled by these false adrenergic transmitters. Many of these terminals were seen in close contact (20 nm) with steroidogenic cells suggesting a neuromodulatory function of CA in hormone synthesis and/or release. It is yet unclear

Send offprint requests to : Professor K. Unsicker, Department of Anatomy and Cell Biology, Philipps-University Marburg, Robert-Koch-StraBe 6, Phone 06421/284030, D-3550 Marburg, Federal Republic of Germany

* Dedicated to Professor H. Leonhardt, Kiel, in honour of his 65th birthday

432 K. Uns icker et al.

whether E and NE are stored in separate or identical axon moieties and within the same organelles. Choline acetyltransferase activity, which was taken as a measure for a cholinergic nerve component in the ovary, amounted to only 7% of its adrenal activity. It is suggested that the chicken ovary may serve as an excellent model to investigate the modula- tory role of nerves in the endocrine function of the ovary.

Key words: Chicken ovary - Adrenergic innervation Cholinergic innervation - Steroidogenic interstitial cells - Myosin immunoreactivity - P N M T Choline acetyltransferase

Control of reproductive and endocrine ovarian functions by the hypotha- lamic pituitary-gonadal axis is a well-established fact. Although the occurrence of intraovarian nerves has been extensively documented (for references, see Bahr et al. 1974; Burden 1978a; Stefenson et al. 1981), the extent to which they may modulate ovarian functions is not entirely clear. Detailed analyses employing the Falck-Hillarp fluorescence technique for the visualisation of aminergic nerves (Bj6rklund et al. 1972; Owman et al. 1979), acetylcholinesterase staining for the demonstration of (presumed) cholinergic nerves (Bulmer 1965; Jacobowitz and Wallach 1967), and electron microscopy have suggested that contractile cells in the wall of the follicle (Burden 1972; Owman et al. 1975; Walles et al. 1975, 1978), endocrine cells in the theca interna and interstitial tissue (Unsicker 1970; Svensson et al. 1975; see Unsicker 1982, for review), and blood vessels may represent targets for neuronal control mechanisms. The concept of an important neural component in the mediation of follicular contractility has been essentially sub- stantiated by Owman and his group (Owman et al. 1975; Walles et al. 1974, 1976, 1977a, b), who have demonstrated autonomic receptors in follicle strips and obtained contractions of the follicular smooth musculature in response to electrical stimulation of the sympathetic nerves. Evidence for an involvement of nerves in ovarian hormone synthesis and release is far less complete. It has been shown that norepinephrine (NE) decreases proges- terone secretion in slices of the ovarian interstitial gland from pregnant rats (Burden 1978a). In hypophysectomized rats stimulation of the ovarian plexus causes regressed interstitial cells to assume the fine structural features of active steroidogenic cells (Capps et al. 1978).

Strategies designed for elucidating the role of intraovarian nerves in ovulation, maturation of follicles and modulation of endocrine celi activities would have to employ ovaries from species that are particularly well innervated. The relative amount and distribution of sympathetic and parasympa- thetic nerves vary considerably in the ovary of different mammalian species (cf. Stefenson et al. 1981). For example, the concentration of NE, which reflects the density of catecholaminergic nerves, is tenfold higher in the cat (4.93-1-1.30 ~tg/g tissue; Rosengren and Sj6berg 1967) than in the rat (0.50_+0.03 lag/g; Stefenson et al. 1981).

We have recently reported (Unsicker et al. 1980) that the chicken ovary contains large amounts of catecholamines (CA), which greatly exceed those

Catecholaminergic innervation of the chicken ovary 433

found in the mammalian ovary and include both NE and epinephrine (E). Although the chicken ovary reportedly receives a rich supply of catecholaminergic nerves (Bennett and Malmfors 1970; Burden 1978b), it was not clear whether these amines are exclusively stored in axons or might occur in ovarian nerve cells (Gilbert 1965, 1969) or chromaffin tissue.

In the present study we determined the CA content of the chicken ovary by the highly sensitive and accurate HPLC and electrochemical detection method; we also determined cholineacetyltransferase (Chat) activity as a measure for a possible cholinergic neuronal component. Moreover, we addressed the question of the distribution of catecholaminergic nerves at var- ious postnatal ages, their relation to myosin-immunoreactive contractile cells and steroidogenic cells, and the problem of axonal versus cellular storage of E.

Materials and methods

Animals. White HNL and brown Warren SSL chickens aged 6 weeks to 1 year were purchased from breeding stations near Kiel and Marburg. Numbers of animals used in the different experiments are listed in Table 1. The ovaries were rapidly dissected out under Nembutal | (40 mg/kg i.v. or i.p.) or ether anesthesia or after decapitation.

Histochemistry of catecholaminergic nerves

1. Falck-Hillarp method. Organs were frozen in liquid propane cooled by liquid nitrogen. The specimens were freeze-dried for 4 days and treated with formaldehyde vapor at 80~ for 1 h. The paraformaldehyde used had previously been equilibrated in air at 70% relative humidity. Controls remained untreated. After vacuum embedding in paraffin, specimens were sectioned (8-15 gm) and the sections mounted in liquid paraffin. For further technical details, see Bj6rklund et al. (1972).

2. Glyoxylic acid method (Lindvall and Bj6rklund 1974, modified according to De La Torre and Surgeon 1976). Native frozen sections (14~40 gm) were cut on a Dittes-Duspiva Cryostat, mounted on slides, dried in a stream of cold air for 15 min and immersed for 3 sec in an ice-cold 2% glyoxylic acid solution (in 0.1 M phosphate buffer, pH 7.4). Sections were subse- quently dried, incubated at 80 ~ C for 5 min and mounted in liquid paraffin.

Table 1. Animals used in the present study

Technique/Experiments

Age of Histo- Myosin Light 5-OHDA 6-OHDA Quanti- Chat P N M T animals chem- immuno- and tative activity immuno- used istry reac- electron deter- reac-

of adren- tivity micro- mina- tivity ergic scopy tions nerves of CA

6 weeks 2 2 2 - 2 - - --

14-20 weeks 6 2 4 1 2 5 6 2

>30weeks 2 2 4 -- 2 - - --

434 K. Unsicker et al.

Phenylethanolamine N-methyltransferase (PNMT) immunoreactivity

PNMT-antibodies. P N M T was purified by the method of Connett and Kirshner (1970) from bovine adrenal medulla. The enzyme preparation contained minor impurities, which were partially removed by an additional gel filtration chromatography on Sephadex G 75 SF. For the production of antibodies we used either the native enzyme after the gel filtration step or the denatured protein that was eluted from a preparative sodium dodecylsulphate polyacryl- amide gel. The antibodies were raised in rabbits by one injection of 200 I-tg protein followed by three injections of 100 ~tg protein in 2-3 week intervals.

PNMT-immunoreactivity. The ovaries and adrenal glands were removed from two chickens and fixed overnight by immersion in 4% paraformaldehyde in 0.1 M phosphate buffer, pH 7.4, at 4 ~ C. Tissues were rinsed in 0.1 M phosphate buffer for 1 day and washed overnight in the same buffer containing 30% sucrose. After freezing the tissues at - 2 5 ~ in Tissue-Tek embedding medium (Miles Laboratories), 104tm cryostat sections were cut, melted onto gelatin- coated glass slides and allowed to dry in air. Sections were incubated for 1, 2 or 3 nights at room temperature in ant i -PNMT antisera diluted 1:40 in phosphate-buffered saline (PBS). After washing in PBS, sections were incubated for 1 h at room temperature in FITC-labelled goat anti-rabbit IgG (Miles) diluted i :15 in PBS, again washed in PBS, and mounted in PBS/glycerol (1 : 3, pH 7.4).

Normal rabbit serum was used in place of immune serum in the above procedure as a control for staining specificity. Bovine and rat adrenal glands were prepared in the same way as the chicken tissues, and sections were stained in parallel with chicken tissues in order to check the cross-reactivity of the antiserum.

Myosin immunoreactivity. 5-1am thick cryostat-cut sections, frequently adjacent to those taken for the localisation of adrenergic nerves, were air-dried for 1 h and incubated for 30 min at room temperature with specific, y-globulin-enriched rabbit antibodies and their corresponding controls as follows:

(1) antiserum (1 mg/ml) raised against purified myosin from chicken gizzard smooth muscle;

(2) anti-smooth muscle myosin antibody previously adsorbed to chicken gizzard myosin; or

(3) pre-immune rabbit 7-globulin. After washing in PBS the sections were incubated for 30 min with FITC-labelled goat

anti-rabbit IgG (1:15, Miles, Frankfurt , F.R.G.), washed in PBS, and mounted in a glycerol/ 0.1 M-glycine buffer (7:3 v/v), pH 8.6. For further biochemical and technical details of the purity of the antigen used and the specificity of the antibody, see Gr6schel-Stewart et al. (1976).

Fluorescence microscopy. CA histofluorescence and FITC immunofluorescence were viewed under a Zeiss Universal fluorescence microscope equipped with 4 mm BG 38, 2 x KP 420 and K 530 filters. Photographs were taken on Kodak-Tri-X-Pan film.

Electron microscopy. Ovaries of 6-week-old chickens were sliced and fixed by immersion in ice-cold 3.5% phosphate-buffered (0.1 M, pH 7.4) glutaraldehyde for 2 h. Ovaries of older animals were fixed by perfusion of the fixative (see above) via the descending aorta for 15-20 min. All specimens were rinsed in phosphate buffer for at least 2 h, postfixed with 2% aqueous OsO4 for 2 h, dehydrated in a graded series of ethanol and embedded in Araldite. Thin sections were cut on a Reichert-Sitte 0 m U 2 ultramicrotome, stained with uranyl acetate (saturated solution in 70% methanol) and lead citrate and viewed under Zeiss EM 9 A and Siemens 101 electron microscopes, l - to 24tin thick semithin sections served for orientation and were stained with azur II-methylene blue.

5-and 6-Hydroxydopamine (OHDA) treatment. One-mm thick slices of ovaries were incubated in an oxygenated (95% O2, 5% CO2) organ bath for 30, 60 or 120 min. The bathing solution consisted of Hanks ' Balanced Salt Solution (Flow Laboratories, Bonn, F.R.G.), ascorbic acid (0.1 g/l) and 6-OHDA (0.1 g/l; Labkemi, G6teborg, Sweden) or 5-OHDA (0.2 g/l; Labkemi). In vitro-application of 5- and 6-OHDA was given preference to in vivo-administration in


order to obtain maximal labelling and to save substances. Control organ slices were incubated without false adrenergic transmitters for the respective periods of time. After incubation all tissue samples were processed for electron microscopy (see above).

Quantitative determinations of cateeholamines (CA). Five ovaries from 18-week-old chickens were each divided into three sections (caudal, intermediate and cranial=adrenal segment), which were separately analyzed for their content of NE and E. Dopamine (DA) was added to each sample as an internal standard. The finely cut tissue was homogenized with 500 Ixl of Tris-HC1 buffer (pH 8.6) containing 0.2% (w/v) Triton X-100, 5 mM EDTA-Na2 and 15 mM NaHSO3 in an Elvium glass potter at 20-4 ~ C. 15 min after addition of 20 ~tl of 9 M perchloric acid the vigorously vortexed homogenate was centrifuged at 10,000 g for 10 min (2~ ~ C). After an A120 3 adsorption step, the CA were quantified in 50 Ix1 aliquots of the supernatant by high performance liquid chromatography with an electrochemical detector (hplc-ed) as described in detail by Miiller and Unsicker (1981).

Determinations of cholinacetyltransferase (Chat) activity

The acetylcholine-synthesizing enzyme choline acetyltransferase (acetyl-CoA-choline 0-acetyltransferase, EC 2.3.1.6) was determined by Fonnum's radiochemical microassay with minor modifications (Fonnum 1975). In brief, 200 mg (wet wt.) of tissue were homogenized in 2 ml of 10 mM EDTA containing 0.5% (v/v) Triton X-100 in a glass and Teflon homogenizer. Fifty microliters of the homogenate were incubated for 30 min at 37 ~ C in 200 lal of an incubation mixture containing the following constituents (final concentrations) : 50 ~tM acetylcoenzy- me A (mixed from radioactive and nonradioactive compounds to give a total of 100000 dpm per incubation tube), 300 mM NaC1, 50 mM sodium phosphate buffer (pH 7.4), 8 mM choline chloride, 20 mM EDTA (adjusted to pH 7.4) and 0.1 mM eserine sulfate. Incubations were performed in the caps of small plastic centrifuge tubes (Eppendorf, Hamburg, F.R.G.) and terminated by dropping the caps into 4 ml of 10 mM sodium phosphate buffer (pH 7.4) within a scintillation vial. Two milliliters of acetonitrile containing 10 mg sodium tetraphenylboron and 10 ml of scintillation cocktail (Beckman Ready-Solv R NA) were added quickly and shaken gently to extract the formed radioactive acetylcholine into the organic phase. Samples were counted in a Packard Tri-Carb 460C liquid scintillation counter equipped with automatic external standardization.

In order to correct for other acetylated compounds such as acetylcarnitine, all incubations were repeated by omitting the eserine sulfate from the incubation mixture and adding 10 units of acetyl cholinesterase. The radioactive products measured under these conditions were sub- tracted from the original readings.

Other biochemical methods

Protein was determined by the method of Lowry et al. (1951) using bovine serum albumin as a standard.

Chemicals used

1-14C acetyl-coenzyme A (specific activity, 54 mCi/mmol) was obtained from Amersham Buchler, Braunschweig, F.R.G. Nonradioactive acetyl coenzyme A (lithium salt), acetylcholinesterase (acetylcholine acetylhydrolase, EC 3.1.1.7, from electric eel), bovine serum albumin, choline chloride, eserine sulfate (physostigmine sulfate) and sodium tetraphenylboron (Kalig- nost) were all from Sigma, Munich. Acetonitrile, ethylenediamintetraacetic acid (EDTA sodium salt), and Folin and Ciocalteu's reagent were purchased from Merck, Darmstadt, F.R.G.

Results

Histology of semithin sections

In 6-week-old animals the ovarian cortex was relatively well demarcated from the medulla (Fig. 1). The cortex contained numerous follicles and the


oocytes stromct

] vessels e bundles nective tissue)

t 1 grand

corpus vertebrae

Fig. 1. Semi-schematic drawing showing a horizontal section through an ovary of a 6-week-old chicken

interfollicular stroma exhibited islets of interstitial cells. A loose vascular connective tissue and bundles of nerve fibres were the major constituents of the medulla. The clear demarcation of cortex and medulla disappeared in older chickens, when follicles and interstitial cells were increasingly found within the medulla. General structural features of the follicle (diameter 2 mm) are presented in Fig. 2: there were two layers surrounding the oocyte with its follicular epithelium, the theca interna and the theca externa. The theca interna could be subdivided into an inner layer of collagen fibres and fibroblast-like cells abutting the basal lamina of the follicular epithelium. Endocrine lipid-loaden (=interstitial) cells forming the thecal gland and blood vessels were the main constituents of the outer theca interna. Smooth muscle-like cells were the dominating cell type in the theca externa.

Special attention was paid to the histological features of the follicular stalk of larger follicles (diameter > 5 mm), since Gilbert (1965) had reported the occurrence of ganglion cells at this site. The stalk contained a core of vascular connective tissue with very small follicles, interstitial cells, smooth muscle-like cells and thick, partly myelinated nerve bundles being interspersed. Serial sections did not reveal any nerve cell bodies, but showed cells with large, rounded nuclei within the nerve bundles, which, at the electron-microscopic level, could be identified as Schwann cells. However, nerve cells were frequently observed in ganglia associated with the ovarian plexus ("ovarian stalk ganglia", cf. Gilbert 1969) (Fig. 1).

Catecholaminergic innervation of the chicken ovary

oocyte nucleus stigma epithe[ium

fe tissue

437

basal ta

folticular epitheliu

inner fibrobtas layer of theca int

interstith

iat cells

intern a

muscle celLs ,ca externa

xterna

S terna

foLiicula

blood vessels, nerve fibre bundles

Fig. 2. Mature ovarian follicle and follicular stalk of an adult hen

Myosin imrnunofluorescence

The most brightly fluorescent cells were found in the wall of arterial vessels and in the perifollicular region. Six-week-old chickens displayed only immunoreactivity of vascular smooth muscle cells without any non-vascular fluorescence around follicles (Fig. 7). In older animals follicles with a diameter of at least 1.1-1.4 mm exhibited up to 30 layers of intensely fluorescent cells located in the theca externa (Fig. 8). These cells became most prominent in the large pendulating follicles (Fig. 6) of adult hens. The fibroblast-like cells of the theca interna immediately adjacent to the follicular basal lamina (see below, Electron microscopy) and endocrine thecal cells did not fluoresce at any developmental stage of follicle maturation (Fig. 8).

Incubation with control sera revealed only a very low background staining.

Catecholamine (CA) histofluorescence

Sections through ovaries from 6-week-old chickens displayed large numbers of yellow-green- to yellow-fluorescent fibre bundles (Fig. 3). These bundles were found both along blood vessels and running independently from them in the medulla and in the cortex, where they formed coarse meshworks around the evenly sized follicles (Figs. 3, 4). Fine varicose fibres were rarely seen. The density of catecholaminergic innervation was not clearly different from that seen in older animals. In the 10-15 times larger ovaries of 16-


Fig. 3. Catecholamine histochemistry, Falck-Hillarp technique. Bundles of intensely fluorescent axons surrounding follicles (,10 and running in the stroma (s); 6-week-old chicken ovary, x 190

Fig. 4. Tangential section through a follicle of a 6-week-old chicken showing a very dense network of catecholaminergic nerve fibres within the follicular wall. Glyo• acid technique. x 190

Fig. 5. Network of adrenergic axons (glyoxylic acid technique) surrounding interstitial cells (/) in the theca interna of a large follicle. Follicular epithelium (e). x 470


to 20-week-old chickens the amount of catecholaminergic nerves increased proportionally. Small follicles showed an innervation pattern identical to that observed in the 6-week-old animals. Medium-sized (1.1-1.4 ram) and large follicles, however, exhibited a dense network of smooth and varicose fluorescent fibres in the theca interna, while the theca externa was only sparsely innervated (Figs. 5, 10). The high density of catecholaminergic nerves in the theca interna was particularly obvious in tangential sections through the follicular wall (Fig. 4).

All CA-specific histofluorescence in the intermdeiate and caudal segments of the ovary was contained in nerve fibres; serial sections did not provide evidence to suggest a localisation of CA in intraovarian chromaffin or sympathetic nerve cells. However, intensely yellow-fluorescent chromaffin cells and sympathetic principal neurons were predominant at the ovarian- adrenal junction.

Correlation of myosin-immunoreactive perifollicular cells and catecholaminergic nerves

The catecholaminergic innervation of myosin-immunoreactive perifollicular cells was studied using consecutive sections that were incubated with glyoxylic acid or processed for immunocytochemistry. Figures 8 and 9 represent two adjacent sections through a follicle wall and clearly demonstrate that the density of catecholaminergic fibres is by far more pronounced in the theca interna than in the region of the cells with strong myosin immunoreactivity ( = theca externa).

Phenylethanolamine N-methyltransferase ( PNMT) immunofluorescence

PNMT-like immunoreactivity was not visible in either the chicken adrenal glands or ovaries. Parallel staining of bovine and rat adrenal glands with the same antiserum produced PNMT-like immunoreactivity in the cytoplasm and nuclei of a majority of cells in the adrenal medulla of both species.

Electron microscopy

1. Ultrastructure of perifollicular cells

Fibroblasts andfibroblast-like cells formed the innermost layer of the follicular wall immediately adjacent to the basal lamina surrounding the follicular epithelium (Fig. 11). Up to ten layers of these cells were encountered in the largest follicles. The fine structure of these cells varied depending on whether they were located next to the follicular basal lamina or in the immediate vicinity of endocrine thecal cells (Fig. 11). In the latter location they were stout, rich in cytoplasm with dilated cisternae of rough endoplasmic reticulum (ER), some smooth ER and occasional lipid droplets, suggesting a possible relationship to the steroid-producing cells of the thecal gland in terms of a precursor function. Fibroblasts of the innermost layers were


slender, with scant dark cytoplasm and scarce ER. The cells were embedded in a matrix of collagen fibres and were not connected by specific junctions.

The fine structure of the interstitial tissue in the follicular (thecal gland) was essentially as described by Dahl (1970) and compatible with the idea of its steroid-hormone secretory function. Follicles larger than 1.1-1.4 mm in diameter displayed a typical theca externa that consisted of numerous (up to 30) layers of myoid cells. Both the thickness of the theca externa and the degree of differentiation of the cells depended on the size of the follicle. In the largest follicles, most of the cells displayed ultrastructural features typical of metabolically active smooth muscle cells (Fig. 13), with few caveolae aligned along the cell membrane. Gap junctions were not observed. The putative precursors of these highly differentiated contractile cells contained fewer microfilaments and were rich in rough ER and free ribosomes (Fig. 12).

2. Ultrastructure of perifoUicular nerve fibres

The relative distribution of nerve fibres as seen with the histofluorescence technique was confirmed at the ultrastructural level indicating that non- catecholaminergic nerves did not make a major contribution to the innervation of those perifollicular layers that had a sparse supply of catecholaminergic nerves. Thus, very few nerve-fibre bundles were detected in the innermost layer of the theca interna formed by fibroblast-like cells (Fig. 14) and in the theca externa, whilst the endocrine interstitial cells of the theca interna were densely innervated. Single axons and axons in small nerve bundles were rarely encountered and, when present, occurred mostly in the theca interna. Terminal axons (" varicosities ") that were devoid of Schwann cell investment usually contained small electron-lucent vesicles with a diameter approximately 50 nm and larger dense-cored vesicles (80-120 nm in diameter). Such varicosities approached interstitial cells, penetrated their basal

Fig. 6. Myosin immunofluorescence in smooth muscle cells of the theca externa (th.e.) of a large follicle (2 cm in diameter, mature ovary). Theca interna (th.i.) with blood vessels. x 190

Fig. 7. Myosin immunofluorescence in the ovary of a 6-week-old chicken. Strong immunoreactivity is only seen in the walls of blood vessels (v). A typical theca externa containing cells with strong myosin immunoreactivity has not yet been formed. Groups of interstitial cells (i). Follicular epithelium (e). • 310

Figs. 8, 9. Myosin immunofluorescence (Fig. 8) and adrenergic innervation (Fig. 9) in the different layers of a follicular wall (approximate follicle diameter 2 mm) as shown in two adjacent sections. Follicular epithelium (e), theea interna (th.i.), theca externa (th.e.), interstitial cells (i.). The bulk of fluorescent nerve fibres is concentrated within the theca interna, few fibres are found in the theca externa. Ovary of a 16-week-old chicken, x 350

Fig. 10. Distribution of adrenergic nerves within the wall of a follicle (4 mm in diameter) of a 23-week-old chicken. Note the high density of nerve fibres in the theca interna (th.i.) compared to the theea externa (th.e.). x 190


lamina and were often found abutting their target cells with an intercellular gap of 20 nm (Fig. 15). Synaptic vesicles were seen frequently in close prox- imity to the "presynaptic" membrane, which, in a few cases, showed accu- mulations of electron-dense material at its inner surface. Postsynaptic membrane specializations were not developed. In contrast, the shortest distance observed between terminal axons and other putative target cells (smooth muscle, fibroblast-like cells) was never less than 200 nm.

Studies using 5- or 6-OHDA for specifically labelling catecholaminergic axons clearly showed that non-aminergic profiles constituted only a minor proportion of perifollicular nerves. Over 90% of the perifollicular axons exhibited uptake of the drugs, with an even higher proportion of catecholaminergic terminals contacting interstitial cells (Fig. 17). Application of 6-OHDA induced the well-known degenerative changes (Fig. 16), whilst the effect of 5-OHDA uptake was clearly shown by an increase in electron density of the large granular vesicles and the appearance of dense cores in the small (50 nm) vesicles (Fig. 17). Controls that were incubated without the specific drugs showed good to excellent tissue preservation without the respective structural changes within axons.

Quantitative determinations of CA

Table 2 gives the results of the determinations for NE and E in chicken ovaries obtained by hplc-ed. The values are surprisingly high, not only when compared to the mammalian ovary (cf. Introduction), but also with

Table 2. Norepinephrine (NE) and epinephrine (E) content of different sections (caudal, middle, cranial) of the chicken ovary. Values represent the means of five independent samples with their standard errors. Determinations were made in triplicate

NE/ng .mg NE/pmol-mg E/ng.mg E/NE + E E/pmol.mg E/NE + E Protein Protein Protein g% Protein mol%

15.4 +3.3 91.0 3.14+1.5 17 17.1 15.8 Caudal and middle section

Cranial section

19.57_+3.3 115.7 9.9 _+1.3 33.6 54 31.8

Fig. 11. Theca interna of a small follicle. Flattened fibroblast-like cells 09 form the innermost layer, whilst rounded and polyhedral cells containing lipid droplets (/), tubular mitochondria and patches of smooth endoplasmic reticulum are more peripherally located. Bundles of axons (a) are rare. Follicular epithelium (e). x 4500

Figs. 12, 13. Cells in the theca externa of an immature (1.2 mm in diameter; Fig. 12) and a ruptured follicle (Fig. 13). The cells of Fig. 12 resemble fibroblasts, contain dilated cisternae of rough endoplasmic reticulum (ER) and bundles of microfilaments (mf). Cells at this developmental stage of the theca externa do not exhibit clearly positive myosin-immunofluorescence (cf. Fig. 7). In contrast, the theca externa of fully mature follicles (Fig. 13) contains typical smooth muscle cells (of. Fig, 6) Fig. 12: x 9000; Fig. 13: x 4500


Fig. 14. Bundle of axons invested by Schwann cell (sc) cytoplasm in the theca interna. Axons contain small granular (sg), small agranular (sa) and large dense-cored (ld) vesicles. The axon bundle is surrounded by a basal lamina (b/) and fibroblast processes, x 15000

Fig. 15. Two terminal axon profiles (al, a2) in close contact with an endocrine interstitial cell that can be identified by the presence of tubular mitochondria (tin), lipid droplets (/) and smooth endoplasmic reticulum (sER). Profile al contains small agranular and large dense- cored vesicles and exhibits a "presynapt ic" membrane specialisation (arrow), whereas no small vesicles are discernible in a 2. x 28 000

Fig. 16. Group of terminal and preterminal axons (al-a7) in the ovarian cortex in close proximi- ty to an interstitial cell (0- The organ had previously been incubated for 1 h with 6-OHDA. Profiles al through a6 exhibit the degenerative changes observed after 6-OHDA treatment, such as axonal swelling, formation of dense bodies and uptake of the false transmitter into vesicles; a 7 containing small agranular inclusions and a large dense-cored vesicle has remained unaffected suggesting that it utilizes non-adrenergic transmitter, x ! 5 000

Fig. 17. Two terminal axons that have been filled with 5-OHDA as indicated by the highly electron-dense cores in small (arrows) and large (arrowheads) granular vesicles in close apposi- tion to a steroid-producing interstitial cell (i). x 36000

446

Table 3. Choline acetyltransferase activity in tissue samples

K. Unsicker et al.

nmol acetylcholine formed per min and g wet wt

pmol acetylcholine formed per min and mg protein

Adrenal medulla

Ovary, close to the adrenal medulla

Ovary, far distant from the adrenal medulla

Pectoralis muscle

0.800 ___+__ 0.268 7.24 + 1,30

0.126+__0,044 1.60+0.42

0.056 + 0.022 0.50__+ 0.26

0.054 + 0.020 0.84 _____ 0.60

All values are the means of six independent determinations with their standard errors. From each ovary two samples were taken: one was dissected out far distant from the adrenal medulla, the other one relatively close to the adrenal medulla in order to estimate the possible contribution from contamination with adrenal tissue

regard to CA levels in other sympathetical ly innervated organs o f the chick, e.g. the hear t (6.7 ng E and 2.0 ng N E / m g protein; Ignarro and Shideman 1968). Moreover , it is evident that there is a considerable p ropor t ion o f E, with 15.8 m o l % of the total CA content in the lower and 31.8 m o l% in the upper ( = adrenal) segment. The higher p ropor t ion o f E in the upper segment may presumably be explained by its close relation to the adrenal gland.

Quantitative determinations of Chat

In order to estimate the cont r ibut ion o f cholinergic nerves to the ovar ian innervat ion, the amoun t of the transmitter-synthesizing enzyme Chat was determined in the ovary. F r o m each chicken one tissue sample o f the ovary far distant f rom the adrenal medulla and one close to the adrenal were dissected out and compared with samples o f the adrenal medulla and the pectoralis muscle. Measurements o f Chat activities are summarized in Ta- ble 3. Cha t activities in the ovary were low, corresponding to 7% of the values determined for the adrenal medulla. Tissue samples taken f rom the ovary close to the adrenal showed higher enzyme activities, presumably due to con tamina t ion with tissue f rom the adrenal medulla.

Discussion

The present results conf i rm and extend earlier observations on the innerva- t ion of the chicken ovary. An impor tan t issue addressed in our study was to conf i rm the large amounts o f ovar ian CA and to investigate its cellular versus axonal storage sites. It is evident f rom this study that cellular storage o f CA in cell bodies o f chromaff in cells or sympathet ic neurons only occurs in the region close to the adrenal gland, where these cells are ar ranged in a ganglion-like fashion (of. Gilbert 1969).

Large amounts o f CA could be measured in ovar ian segments distant f rom the adrenals, where serial sections did not reveal any chromaff in or nerve cells, indicating that CA are exclusively stored in nerve fibres. The


extremely high concentrations of CA are reflected by an extraordinarily dense catecholaminergic nerve supply, which is in accordance with observations by Bennett and Malmfors (1970) but differs from the report of Gilbert (1965), who considered the catecholaminergic component to be only small. E contributes substantially to the total CA content of the chick ovary. Although it is proportionately greater in the cranial than in the middle and caudal ovarian segments, suggesting a contamination by ovarian plexus or adrenal chromaffin tissue, it still makes up 17% of total CA in the caudal segment. E has generally been reported to occur in higher relative concentrations in avian than in corresponding mammalian tissues (cf. Holz- bauer and Sharman 1972). Since chromaffin tissue is absent from the caudal part of the chicken ovary, storage of E in nerve fibres is very likely. However, PNMT-like immunoreactivity could not be detected in the chicken ovary, either in axons or in other components. This finding appears to be due to a lack of cross-reactivity between chicken P N M T and antiserum to bovine PNMT, since the same antiserum produced PNMT-like immunoreactivity in rat and bovine adrenal medulla, but not in chicken adrenal gland. In addition, high P N M T activity has been measured biochemically in the chicken ovary (L6ffelholz, personal communication). It is not yet clear whether E and P N M T are contained in a separate moiety of ovarian nerves or co-stored with NE. At the ultrastructural level all axons observed exhibited a uniform dual population of small agranular or granular and large dense-cored vesicles, which did not permit a distinction to be made between NE- and E-storing nerve fibres. If E-containing axons originate from chromaffin or small intensely fluorescent (SIF) cells, as may be expected since there is no evidence to date for storage of E in sympathetic principal neurons, these E-producing cells either lack "chromaff in" storage vesicles larger than 120 nm in diameter or do not transport them into their processes.

The occurrence of chromaffin (-like) cells with granular vesicles in the size range of those found within axons of the chicken ovary is well documented (see Taxi 1979). However, there is also evidence that chromaffin cells may transport small storage vesicles into their neurites and, simulta- neously, contain larger ones in their cell bodies (Unsicker et al. 1981). Co- storage of NE and E may occur in embryonic and early postnatal chromaffin cells of mammals (Coupland 1972), but has not been documented to take place in neurites.

The main site of origin of NE- and E-containing nerve fibres in the chicken ovary seems to be outside the organ, since Gilbert and Wood-Gush (1969) have found very few nerves in ovarian transplants. A crucial question with regard to the intraaxonal storage of NE and E is, whether both amines may serve as transmitters in the chicken ovary and, if so, are released simul- taneously and have identical or different targets. Lindmar and DeSantis (1974) have shown that sympathetic nerve stimulation and tyramine cause release of NE and E in the chicken heart with a ratio that is similar to that found in the unstimulated heart. Release from chromaffin cells was unlikely, since CA output could not be induced by applying acetylcholine.

The large quantities of CA determined by hplc-ed, together with the small number of non-catecholaminergic nerve profiles encountered after

448 K. Uns icker et al.

application of 5- or 6-ODHA suggest that CA act as the main neurotransmitters in the chicken ovary. However, the demonstration of Chat activity, although amounting to 7% of its adrenal activity only, and nerve profiles with small agranular vesicles that did not show uptake of 5- and 6-OHDA, support the idea of a small cholinergic component in the innervation of the chicken ovary.

Another main topic of this study was the morphological identification of the putative target cells of the vast catecholaminergic innervation in the follicular wall of the chicken ovary. The principal message arising from this examination is that, in contrast to the mammalian ovary, endocrine (interstitial) cells in the theca interna appear to receive the major part of catecholaminergic and non-catecholaminergic terminal axons rather than the contractile cells in the theca externa.

In the mammalian ovarian follicle autonomic nerves can be frequently found in close contact with smooth muscle or smooth muscle-like cells located in the theca externa (Burden 1972; Walles et al. 1974, 1975, 1976). Moreover, good morphological correlation exists with regard to the distribution of actin- and myosin-immunoreactive cells in the theca externa and histochemically visualised catecholaminergic nerves (Walles et al. 1978). These nerves clearly serve a functional role, since contractions of the follicular wall elicited by sympathetic nerve stimulation have been shown to be mediated by adrenergic a-receptors (Walles et al. 1975, 1977). The paucity of catecholaminergic nerves in the cell layers of the chicken theca externa, which display strong myosin-specific immunofluorescence, certainly does not rule out a nerve-mediated contractile function of these layers, since diffusion of the neurotransmitters may occur over great distances. Soliman and Walker (1976) have provided evidence for in vitro-motility of avian ovarian follicles, which could be induced by the addition of acetylcholine, NE, and E.

Quantitatively more impressive is the amount of nerve fibres in the chicken theca interna and the large number of axon terminals in close contact with steroidogenic cells (cf. Dahl 1970), which by far outnumber those found on interstitial cells in the mammalian ovary (Unsicker 1970; Svensson et al. 1975). With few exceptions these terminals are catecholaminergic, suggesting participation of NE and E in steroid hormone synthesis and/or release.

Physiological data supporting such a concept are not yet available. The dense innervation of the ovarian interstitial cells of the chicken make this tissue a favorable model for future studies on neuronal modulation of steroi- dogenesis in general and on the ovarian endocrine function in particular. Experiments to be suggested might include determinations of steroid hormone synthesis and output in response to nerve stimulation and ablation of nerves, in transplants and in culture, where the effect of the catecholaminergic transmitters can be monitored after direct application or in co-culture with sympathetic ganglion explants.

It is noteworthy that the interstitial tissue of the avian testis is also densely innervated (Baumgarten and Holstein 1968, 1971; Haase 1973; for review, see Unsicker 1982). This might be indicative of a general principle


in the regulation of endocrine cell activities in avian reproductive organs, differing from that found in mammals.

Acknowledgements. This work was supported by a grant from the Kempkes-Foundation (Marburg, F.R.G.).

The authors gratefully acknowledge a generous gift of anti-smooth muscle myosin antibodies from Professor U. Gr6schel-Stewart, TH Darmstadt, F.R.G. Thanks are also due to M. Johannsen, R. Korff, Ch. Fiebiger, H. Schneider, W. B6rner, and I. Ganski for technical assistance and secretarial help.

References Bahr J, Kao L, Nalbandov AV (1974) The role of catecholamines and nerves in ovulation.

Biol Reprod 10: 273-290 Baumgarten HG, Holstein AF (1968) Adrenerge Innervation im Hoden und Nebenhoden

vom Schwan. Z Zellforsch 91:402-410 Baumgarten HG, Holstein AF (1971) Noradrenerge Nervenfasern im Hoden von Mammaliern

and anderen Vertebraten. J Neuro-Visc Rel, Suppl 10:563-572 Bennett T, Malmfors T (1970) The adrenergic nervous system of the domestic fowl. Z Zellforsch

mikrosk Anat 106:22-50 Bj6rklund A, Falck B, Owman Ch (1972) Fluorescence microscopic and microspectrofluoro-

metric techniques for the cellular localization and characterization of biogenic amines. In: Berson SA (ed) Methods of investigative and diagnostic endocrinology, Vol 1. Rall JE, Kopin IJ (eds) The thyroid and biogenic amines. North Holland Publ Co, Amsterdam, pp 318-368

Bulmer D (1965) A histochemical study of ovarian cholinesterases. Acta Anat 62:254-265 Burden HW (1972) Adrenergic innervation in ovaries of the rat and guinea-pig. Am J Anat

133:455-462 Burden HW (1978a) Neural modulation of ovarian function. TINS 1:85-86 Burden HW (1978b) Ovarian Innervation. In: Jones RE (ed) The vertebrate ovary. Compara-

tive Biology and Evolution. Plenum Press, New York and London, pp 615-638 Capps ML, Lawrence IE Jr, Burden HW (1978) Ultrastructure of the cells of the ovarian

interstitial gland in hypophysectomized rats. Cell Tissue Res 193:433-442 Connett RJ, Kirshner N (1970) Purification and properties of bovine phenylethanolamine

N-methyltransferase. J Biol Chem 245:329-334 Coupland RW (1972) The Chromaffin System. In: Blaschko H, Muscholl E (eds) Catechol-

amines. In: Handb Exp Pharm XXXIII, Springer, Berlin-Heidelberg-New York, pp 16~5 Dahl E (1970) Studies of the fine structure of ovarian interstitial tissue. 3. The innervation

of the thecal gland of the domsetic fowl. Z Zellforsch 109:212-226 De La Torre JC, Surgeon JW (1976) Histochemical fluorescence of tissue and brain mono-

amines: results in 18 min using the sucrose-phosphate-glyoxylic acid (SPG) method. Neu- roscience 1:451-453

Fonnum F (1975) A rapid radiochemical method for the determination of choline acetyltransferase. J Neurochem 24:407 409

Gilbert AB (1965) Innervation of the ovarian follicle of the domestic hen. Q J Exp Physiol 50:437-445

Gilbert AB (1969) Innervation of the ovary of the domestic hen. Q J Exp Physiol 54:40~415 Gilbert AB, Wood Gush DGM (1969) Innervation of ovarian transplants in the domestic

hen. J Reprod Fertil 18:550-551 Gr6schel-Stewart U, Schreiber J, Mahlmeister C (1976) Production of specific antibodies to

contractile proteins and their use in immunofluorescence microscopy. I. Antibodies to smooth and striated chicken muscle myosins. Histochemistry 46 : 229-236

Haase E (1973) Histochemische und elektronenmikroskopische Untersuchungen fiber die Inner- vation des Hodens vom Bergfink (Fringilla montifringilla). Verh Dt Zool Ges 66:106-110

Holzbauer M, Sharman DF (1972) The distribution of catecholamines in vertebrates. In: Blaschko H, Muscholl E (eds) Catecholamines. In: Handb Exp Pharm XXXIII, Springer, Berlin-Heidelberg-New York, pp 110-185

Ignarro LJ, Shideman FE (1968) Appearance and concentrations of catecholamines and their biosynthesis in the embryonic and developing chick. J Pharmacol Exp Ther 159 : 38-48


Jacobowitz D, Wallach EE (1967) Histochemical and chemical studies of the autonomic innervation of the ovary. Endocrinology 81 : 113~1139

Lindmar R, DeSantis VP (1974) The significance of noradrenaline and adrenaline as adrenergic transmitters in the chicken. Naunyn-Schmiedeberg's Arch Pharmacol 282:R 58

Lindvall O, Bj6rklund A (1974) The glyoxylic acid fluorescence histochemical method: A detailed account of the methodology for the visualization of central catecholamine neurons. Histochemistry 39 : 97-127

Lowry OH, Rosebrough NJ, Farr AL, Randall RJ (1951) Protein measurement with the Folin phenol reagent. J Biol Chem 193:265-275

Miiller TH, Unsicker K (1981) High performance liquid chromatography with electrochemical detection as a highly efficient tool for studying catecholaminergic systems. I. Quantification of noradrenaline, adrenaline and dopamine in cultured adrenal medullary cells. J Neurosci Meth 4:39-52

Owman Ch, Sj6berg N-O, Svensson KG, Walles B (1975) Autonomic nerves mediating contractility in the human Graafian follicle. J Reprod Fertil 45:553-556

Owman Ch, Sj6berg N-O, Wallach EE, Walles B, Wright KH (1979) Neuromuscular mechanisms of ovulation. In: Hafez ESE (ed) Human ovulation: Mechanisms, prediction, detection and regulation. Elsevier/North Holland, Amsterdam, pp 57-100

Rosengren E, Sj6berg N-O (1967) The adrenergic nerve supply to the female reproductive tract of the cat. Am J Anat 121:271-284

Soliman KFA, Walker ChA (1976) In vitro motility of the avian ovarian follicles. Gen Comp Endocrinol 30:372-377

Stefenson A, Owman Ch, Sj6berg NO, Sporrong B, Walles B (1981) Comparative study of the autonomic innervation of the mammalian ovary, with particular regard to the follicular system. Cell Tissue Res 215 : 47-62

Svensson KG, Owman Ch, Sj6berg NO, Sporrong B, Walles B (1975) Ultrastructural evidence for adrenergic innervation of the interstitial gland in the guinea pig ovary. Neuroendocrinol 17: 4 0 4 7

Taxi J (1979) The chromaffin and chromaffin-like cells in the autonomic nervous system. Int Rev Cytol 57:283-343

Unsicker K (1970) Zur Innervation der interstitiellen Drfise im Ovar der Maus (Mus musculus L.). Z Zellforsch 109:46-54

Unsicker K (1982 in press) Innervation of endocrine tissues. In: Motta PM (ed) Electron microscopy in biology and medicine. Nijhoff Publ Hingham MA USA

Unsicker K, Seidel F, Gr6schel-Stewart U, Lindmar R, L6ffelholz K, Wolf U (1980) Zur adrenergen Innervation des Hfihnerovars. Verh Anat Ges 74:443445

Unsicker K, Tschechne B, Tschechne D (1981) Differentiation and transdifferentiation of adrenal chromaffin cells of the guinea-pig. Cell Tissue Res 215:341-367

Walles B, Edvinsson L, Nybell G, Owman Ch, Sj6berg NO (1974) Amine-induced influence on spontaneous ovarian contractility in the guinea-pig and the cat. Fertil Steril 25:602-611

Walles B, Edvinsson L, Falck B, Owman Ch, Sj6berg NO, Svensson KG (1975) Evidence for a neuromuscular mechanism involved in the contractility of the ovarian follicular wall : Fluorescence and electron microscopy and effects of tyramine on follicle strips. Biol Reprod 12 : 239-248

Walles B, Edvinsson L, Owman Ch, Sj6berg NO, Sporrong B (1976) Cholinergic nerves and receptors mediating contraction of the Graafian follicle. Biol Reprod 15:565-572

Walles B, Falck B, Owman Ch, Sj6berg NO (1977a) Characterization of autonomic receptors in the smooth musculature of human Graafian follicle. Biol Reprod 17:4234-431

Walles B, Edvinsson L, Owman Ch, Sj6berg NO, Sporrong B, Stefenson A (1977b) Influence of sympathetic nerves, amine receptors and anti-adrenergic drugs on follicular contractility and ovulation. Acta Physiol Scand Suppl 452:113-120

Walles B, Gr6schel-Stewart U, Owman Ch, Sj6berg NO, Unsicker K (1978) Fluorescence histochemical demonstration of a relationship between adrenergic nerves and cells containing actin and myosin in the rat ovary, with special reference to the follicle wall. J Reprod Fertil 52:175-178

Accepted December 15, 1982

Documents

Catecholaminergic innervation of the chicken ovary