HAL Id: hal-00899955https://hal.archives-ouvertes.fr/hal-00899955

Submitted on 1 Jan 1997

HAL is a multi-disciplinary open accessarchive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come fromteaching and research institutions in France orabroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, estdestinée au dépôt et à la diffusion de documentsscientifiques de niveau recherche, publiés ou non,émanant des établissements d’enseignement et derecherche français ou étrangers, des laboratoirespublics ou privés.

Comparison between oogenesis and related ovarianstructures in a reptile, Pseudemys scripta elegans

(turtle) and in a bird Coturnix coturnix japonica (quail)M Callebaut, L van Nassauw, F Harrisson

To cite this version:M Callebaut, L van Nassauw, F Harrisson. Comparison between oogenesis and related ovarian struc-tures in a reptile, Pseudemys scripta elegans (turtle) and in a bird Coturnix coturnix japonica (quail).Reproduction Nutrition Development, EDP Sciences, 1997, 37 (3), pp.233-252. �hal-00899955�

Original article

Comparison between oogenesis and relatedovarian structures in a reptile, Pseudemys

scripta elegans (turtle) and in a birdCoturnix coturnix japonica (quail)

M Callebaut L Van Nassauw, F Harrisson

RUCA, UA, Laboratory of Human Anatomy and Embryology, 171 Groenenborgerlaan,B-2020 Antwerp, Belgium

(Received 24 October 1996; accepted 27 January 1997)

Summary ― The aspect of the oogonia during their premitotic DNA synthesis and of the premeio-cytes during their premeiotic DNA synthesis was studied in turtles by autoradiography, after injec-tion of 3H-thymidine. As in the adult laying quail, the intrafollicular oocytes of the adult turtle gothrough three successive stages: prelampbrush, lampbrush and postlampbrush. During the prelamp-brush and lampbrush stage two kinds of nucleoli exist: peripheral and central. In contrast to avian yolk,during its final rapid growth, no polyhedric protein yolk units were found in turtle yolk. As in the yel-low yolk of quail, highly osmiophilic alcohol insoluble satellite yolk (egg oil) accumulates betweenthe protein globular yolk of Pseudemys. Turtle yolk globules increase in volume by fusion. The pen-etration of peripherally assembled yolk in the turtle germinal disc is analogous to what we havedescribed in the quail. Also in postlampbrush germinal discs subcortical ooplasmic organelles are pre-sent. Below the turtle germinal disc no structure comparable to the avian nucleus of Pander could beobserved. No pyriform cells (as in squamate reptiles) and no pyriform-like cells (as in birds: Calle-baut, 1991b) were found in the chelonian ovarian granulosa layer. We could not demonstrate func-tional lacunoperitoneal communications via openings in the hilus ovarii of the turtle as is the case inbirds.

ovary / turtle / oogenesis / avian oocyte

Résumé ― Comparaison entre l’ovogenèse et structures de l’ovaire du reptile Pseudemysscripta elegans (tortue) et d’un oiseau (caille japonaise). L’aspect des ovogonies pendant leursynthèse d’ADN prémitotique et des prémeiocytes pendant leur synthèse d’ADN prémeiotique aété étudié chez Pseuderrcys par autoradiographie après injection de thymidine tritiée. Chez la tortueadulte, les ovocytes intrafolliculaires passent par trois stades successifs : avant le stade en écou-

* Correspondence and reprintsTel: (32) 03 218 03 90; fax: (32) 03 218 03 98; e-mail: macal@nets.ruca.va.ac.be

villon (prelampbrush), stade en écouvillon (lampbrush), stade après écouvillon (postlampbrush).Pendant la période de croissance rapide de l’ovocyte de tortue, on ne trouve pas de masses polyh-édriques de vitellus comme chez l’oiseau. Pendant la période de croissance finale, on trouve desgouttes de lipides hautement osmiophiles, insolubles dans l’alcool entre les globules vitellins de tor-tue. Les globules vitellins de tortue grandissent par fusion. L’assemblage du vitellus qui participe àla formation du disque germinatif est analogue à ce que nous avons décrit chez la caille. Des organitesooplasmiques sous-corticaux sont également présents. Nous n’avons pas trouvé chez Pseudemys destructure analogue au noyau de Pander tel qu’il existe chez les oiseaux. Nous n’avons pas pu démon-trer l’existence de communications entre les lacunes de l’ovaire de tortue et la cavité péritonéale.

ovaire / tortue / ovogenèse / ovocytes d’oiseau

INTRODUCTION

Only a small number of studies have beenperformed on oogenesis sensu latu andrelated ovarian structures in Chelonia: Mun-son (1904), Loyez (1906), Thing (1918),Bhattacharya (1925), Altland (1951), Guraya(1959) and Rahil and Narbaitz (1973). In aprevious study (Callebaut and Van Nas-sauw, 1987) some typical aspects of theovary of the turtle Pseudemys scripta ele-gans were described. We have demonstratedthe existence of a well developed smoothmuscle-like layer (desmin immunoreactive)in the ovarian tunica albuginea, giving aparticular aspect to the follicular stigmaregion. We have found that all the turtleoocytes with a diameter of 200 pm or moreare surrounded by voluminous lacunae andare enclosed in a follicle with a superficialcrater, bulging at the surface of the ovary.Also in the turtle ovary, smooth muscle-likecells were found in a suspensory apparatusformed by chordae, the tunica albuginea,and the theca extema of the ovarian folli-cles (Van Nassauw et al, 1991). An excellentcomparative review of oogenesis and fol-liculogenesis in birds and reptiles has beenpublished by Guraya (1989). In the presentwork, we have studied some aspects of oog-enesis sensu latu and related ovarian struc-tures in the turtle Pseudemys scripta ele-gans. This was performed at different ages,using histochemical and autoradiographicmethods. Whenever possible a comparison

was made with oogenesis and ovarian struc-tures in birds.

MATERIALS AND METHODS

Female red-eared turtles (Pseudemys scripta ele-gans) of different ages, reared in our laboratory,were used in this study. The turtles were kept intap water at 25-28 °C in 50 L containers undercontinuous infrared illumination. A stone emerg-ing from the water in the centre of the containerpermitted the turtles to take a sunbath. The babyturtles were fed with small earth worms. Aftera few months they received quail embryos. Tur-tles older than 1 year received newly hatchedquails. This procedure permitted us to study ooge-nesis without disturbing the turtles. Adult andprepuberal female turtles could be distinguishedfrom the males by the presence of a shorter tailand the absence of long claws on their forelegs.Seven baby turtles (weight: 16-20 g) receivedan im injection of 100 pci methyl 3H-thymidine(Amersham; 82 Ci/mmol) in water, five 1-year-old turtles (weight: 130-150 received an iminjection of 500 pci methyl H-thymidine (82Ci/mmol). Six young adult female turtles (weight:850-1 000 g) received im injections of 5 mCiL-[3,5-3H] Tyrosine (Amersham, 51 Ci/mmol)during February or the beginning of March (justbefore or at the moment of egg laying under ourexperimental conditions). The turtles were killed1 h to several days after the injection(s). Babyturtles or 1-year-old turtles still with a soft cara-pace were decapitated by a transverse sectionthrough the middle of their carapace. Older tur-tles with a hard carapace were, first anesthesized

by an im injection of 2-6 mL Nembutal(60 mg/mL) solution in saline, before decapita-

tion. After removal of the ventral part of theircarapace and the opening of the abdomen, theovaries were fixed in situ (for baby turtles) orremoved by sectioning through the hilus ovariiand then fixed. The volume of the ovaries varies

enormously and depends on the age and physio-logical or hormonal state of the turtle. In the adultturtle both ovaries are functional in contrast tobirds, and during the egg laying period they con-tain follicles of all sizes with a maximum dia-meter of approximately 20 mm. Several largefollicles of the same diameter are seen. In contrastto birds both ovaries of turtles (and reptiles ingeneral) contain permanent groups of extrafol-licular germ cells (oogonia and oocytes) whichare called germinal centres or beds. Using char-coal marks, the germinal disc of the largest oo-cytes was labelled. The oocytes of approximatelythe same diameter were fixed in different fixati-ves for comparison. Fixation was performedovernight at room temperature in Heidenhain’sSusa without sublimate (Romeis, 1948), in cal-cium formalin for 1 night at room temperatureor in a mixture of 2% paraformaldehyde and0.1 % glutaraldehyde in 1.0 mM phosphatebuffered saline for 4 h at 4 °C. Other oocytes orpart of the ovaries were fixed in Bouin for I nightat room temperature followed by a postfixation in70% alcohol (Callebaut et al, 1991) or rinsed di-rectly in tap water to study the presence of lipidsin the oocytes during vitellogenesis. Some of theoocytes, fixed in Bouin and postfixed in 70%alcohol or not, were rinsed in distilled water and

placed in a 1 % solution of osmium tetroxide inwater for I h, followed by rinsing in tap water.The germinal disc of the largest oocytes wasexcised and pieces of ovary were dehydrated inthe alcohol series, cleared in xylene and embed-ded in paraffin. After sectioning at a thicknessof 8-10 0 pm, most of the slides were preparedfor autoradiography.

Lipids were studied on the sections withoutdeparaffination and without staining. Indeed dur-ing a study of lipids in avian follicles we observedthat a second passage through xylene (for depa-raffination) removes all lipids (Callebaut,1988b). Some of the sections were Feulgen stai-ned, according to Demalsy and Callebaut (1967).For the study of the labelling of cytogical de-tails, photographs were take of some of the secti-ons, before and after the autoradiographic proces-sing (Callebaut and Demalsy, 1967). Comparablesections of similar oocytes of non-injectedfemales were used as controls both before andafter the autoradiographic procedure.

The radioactively labelled and some controlsections (PAS or Feulgen stained or not) weredipped in nuclear emulsion L4 (Ilford, UK). After5 weeks of exposure in the dark, the autoradio-graphs were developed according to Caro andVan Tubergen (1962). The unstained autoradio-graphs from calcium formalin fixed tissues werecoloured with Unna. The tissue sections obtainedafter other fixations were stained with iron hema-

toxylin and eosin. Since Callebaut ( 1979a,b;1988a) demonstrated the existence in birds oflacunoperitoneal communications, functional viathe hilus ovarii we decided to investigate if thiswas also the case in Pseudemys scripta elegans.To do this, finely divided yolk of five unincu-bated quail eggs were suspended in 1.5 mLRinger solution containing 1 % trypan blue (VanNassauw and Callebaut, 1991). From this sus-pension 0.3-0.5 mL were injected intraperi-toneally ventrally from the hindleg via the open-ing in the carapace in four baby turtles, or 10mL were injected in the same way into threeadult turtles. Five to 16 hours later the turtleswere killed and their ovaries fixed in Susa

(Romeis, 1948). After fixation for 2 days at roomtemperature in this fixative, the ovaries wereplaced directly in 96% alcohol for 3 days. Afterfurther dehydration in absolute alcohol and clea-ring in xylene they were embedded in paraffin.After sectioning, the sections were deparaffinizedand counterstained with Kernechtrot.

RESULTS

The general aspect of the baby or 1-year-old turtle ovary is shown in a transverse sec-tion (fig 1): larger oocytes bulge at the sur-face. At least on each side of themesovarium a germinal bed is visible. How-ever, by the growth of intra follicularoocytes in the neighbourhood, these germi-nal beds seem to become divided. A long,narrow mesovarium fixed to a narrow linearhilus ovarii is seen. Openings are not visiblein the mesovarium, or in the hilus ovarii.After an intraperitoneal injection of trypanblue-labelled yolk, we found no yolk gran-ules in the lacunae of the ovary, indicatingthat there is no functional communication .

between the peritoneal cavity and the ovar-ian lacunae.

After im injection of 3H-thymidinein the baby or 1-year-old turtle

Shortly after injection (l h), groups of oogo-nia localized in germinal centres directlybelow the ovarian surface epithelium (orsometimes on the surface) but superficiallyto the tunica albuginea, are seen to belabelled mainly on the peripheral part oftheir nuclei (containing a fine reticular Feul-gen positive network with positive chro-matin granules) surrounding a non-labelledcentral nucleolar area (compare fig 2A,before autoradiography and fig 2B, afterautoradiography). This labelling of oogo-nia is due to premitotic DNA synthesis dur-ing their multiplication period. But also wepresume that premeiotic DNA synthesistakes place in some larger germ cells thatare already partially surrounded and separa-ted from each other by prefollicular cells(fig 3). In these germ cells, which form atransition stage from oogonia to primaryoocytes, relatively voluminous central nucle-oli are seen. During the days following aninjection of 3H-thymidine, the number of

labelled germ cells of the latter typeincreases in the germinal beds, because theydifferentiate after divisions from pre-exist-ing labelled oogonia into primary oocytes(fig 4). The intensity of the labelling per nu-cleus, however, decreases but finallyremains constant, since these divisions endwith a final premeiotic division. The aspectof the nucleoli can most clearly be seen insections through germinal beds of 1-year-old turtles after staining of the sections withiron hematoxylin and eosin before autoradio-graphy (fig 5). We observed a typicalchromosome-free clearer area of nucleo-

plasm surrounding each of the nucleoli. Withthis staining procedure two kinds of nucle-oli can be discerned: (1) intensely red(eosinophilic) stained nucleoli, usually local-ized centrally in a chromosome-poor cleararea, (2) dark stained nucleoli. This aspect ofthe nucleoli is even more pronounced whenthe oocyte grows: the central red stained(eosinophilic) nucleoli become surroundedby a large fine granular nucleoplasm, freefrom chromosomes, whilst the vacuolizeddark stained nucleoli (only surrounded by

a clear halo) are localized at the inner side ofthe germinal vesicle membrane (fig 6).

Cytology of the oocytesin the adult turtle ovary

As in the oocyte of adult laying quail wecan distinguish three successive develop-mental stages of intrafollicular oocytes inadult turtles before ovulation.

(1) The prelampbrush stage with chromo-somes in diplotene without clearly definedlampbrush lateral loops and with individualnucleoli surrounded by a clearer chromo-some-free halo (fig 7). In the ooplasm a volu-minous paranuclear Balbiani complex closeto an eccentric germinal vesicle could beseen (figs 7-9). At some distance from thesurface a lipid granular layer can be obser-ved, whilst at the surface of the Balbianicomplex a lipid-containing layer is also seen.Between both lipid layers the intensely PASstaining ooplasm is visible (fig 9).

(2) The lampbrush stage with chro-mosomes presenting the lampbrush chro-mosome configuration with lateral loops asalready described by Loyez (1906) in Cis-tudo europaea and with numerous volumi-nous pyroninophilic nucleoli against theinner side of the germinal vesicle wall(fig 10), often bulging at its surface. Thegerminal vesicle has a central or paracen-tral localization and increases strongly involume. This stage begins when the oocytehas a diameter of 600-700 pm. During thelampbrush stage, two kinds of nucleoli canbe distinguished (particularly obvious afterBouin fixation, alcohol postfixation, osmiumtetroxide treatment and Unna staining)(fig 10): (a) the large peripheral py-roninophilic nucleoli, already visible at thebeginning of the lampbrush stage; (b) themuch smaller grey stained nucleoli close toor on the chromosomes, localized in thelarge central part of the germinal vesicle.

(3) The postlampbrush stage starts whenthe oocyte has a diameter of 3-4 mm and

the germinal vesicle penetrates the deeperpart of the cortex where the beginning ofthe assemblage of small yolk spheres takesplace. At the beginning of the postlamp-brush stage, the nucleoli are no longer loca-lized at the inner side of the germinal vesi-cle wall but deeper at some distance on asphere surrounding the shortened chromo-somes in the central area (fig 11 When thegerminal vesicle penetrates the cortex, a ger-minal disc forms. In this germinal disc wesee the development of subcortical oo-plasmic organelles (fig 12). These organellesare most clearly visible after Bouin fixation,osmium tetroxide treatment and Unna stain-

ing and are pressed between the yolk gra-nules, as is also the case in the quail. In the10-mm-diameter oocyte, the germinal vesi-cle has penetrated through the cortex (fig13). Only a narrow cap of small yolk gran-ules separates it from the vitelline mem-brane. The short chromosomes are hardlyvisible among the voluminous vacuolizednucleoli, aggregated in the centre of the ger-minal vesicle forming an oval caryosphere.At its maximum of development, the oocyte(at the end of the postlampbrush stage, justbefore maturation) reaches a diameter of19-20 mm and the germinal vesicle is foundflattened at the surface of the ooplasm (fig14). However, the yolk cap above the ger-minal vesicle is still present. The smallernucleoli are still visible as a flat group inthe centre of the germinal vesicle. Theprominent presence of nucleoli seems thusto be a characteristic feature of the chelo-nian oocyte during its whole intrafollicularperiod. This is not the case during the cor-responding period in avian oocytes. In theintrafollicular Pseudemys oocytes, afterFeulgen staining, the chromosomes are notor very faintly visible. Particularly duringthe postlampbrush stage no spherical con-tracted chromosomes are seen, as is the casein the quail (Callebaut, 1973a). Although atthe end of their final growth period (end ofthe postlampbrush stage), both the quail andturtle oocyte reach the same maximal diam-

eter (approximately 20 mm), the prelamp-brush stage in the turtle develops to a greaterdiameter (600 pm), than in the quail (only150 Lun diameter: Callebaut, 1975). Alsothe lampbrush stage in the turtle becomesmore voluminous (3 mm diameter) contra1.5 mm in the quail. During the final growthperiod in the oocyte of Pseudemys scriptaelegans alcohol insoluble satellite yolk oregg oil (as described in birds: Callebaut et al,1991) localized between the protein yolkglobules (fig 15), could be demonstrated byfixation in Bouin followed by postfixation inalcohol 70% and osmium tetroxide treat-ment. As in birds, lipid drops were alsofound in the theca intema. An accumulationof large clumps of intensely staining osmiop-hilic material was also locally observed,

pressed against the outer side of the granu-losa basement membrane. A paranuclearlipid drop could be seen in the granulosacells and sometimes between the granulosacells. After Bouin fixation and rinsing inwater the satellite yolk is no longer visible.It disappears also from the sections after de-paraffination by xylene (Callebaut, 1988b).This yellow stained satellite yolk seems tocontribute to the yellow colour of the largeroocytes in Pseudemys. After ip injection oftrypan blue-labelled yolk in adult turtles wefound no yolk granules in the lacunae of theovary, which indicates that also in the adultfemale no functional communications existbetween the peritoneal cavity and the ovar-ian lacunae.

Adult turtle ovary,after an im injection of 3H-tyrosine,38 h before fixation

The labelling on the autoradiographs wasmost clearly seen after Bouin fixation andFeulgen staining before dipping. Theradioactive labelling was, however, nolonger visible in Feulgen stained sectionsafter tissue fixation in Susa without subli-

mate, calcium formalin or paraformalde-hyde and glutaraldehyde. After calcium for-malin fixation, staining with Unna after theautoradiographic processing also gives agood contrast with the autoradiographiclabelling.

Labelling of the nuclei of the extrafol-licular germ cells (visible as clearer areasin the germinal beds) is seen (fig 16). Thelabelling of the germinal vesicle of thesmallest intrafollicular oocytes (prelamp-brush stage) is very intense and visible evenat low magnification (fig 17). Also a lessdense labelling is seen on the whole ooplasmat higher magnification (fig 16). In some-what larger oocytes (lampbrush stage, eg,700 pm diameter) the central ooplasmicmass presents a labelling whilst the periph-eral cortex is not or poorly labelled (fig 18):this indicates a central, perinuclear metabolicactivity (protein synthesis). In still largeroocytes (lampbrush chromosome stage),small yolk globules begin to accumulatebelow the cortex, and the labelling in thelatter increases. From this moment on, thereare clearly two kinds of labelling from dif-ferent origins: (1) the central extravitellinelabelling (most pronounced around the ger-minal vesicle), and (2) the peripherallabelling localized on the yolk globulesforming a subcortical layer. Some largermore centrally localized yolk globules arenot labelled and are obviously alreadyformed before the injection. At this momentthe germinal vesicle strongly increases involume and the nucleoplasm and the largevacuolized peripheral nucleoli are intenselylabelled (fig 19). In early postlampbrush

oocytes (approximately 4 mm in diameter),a broad labelled cortex containing smalllabelled yolk granules is seen (fig 20).Below this cortex, larger unlabelled yolkspheres are seen. Small labelled granulesadhere to the surface of this unlabelled yolkspheres, and seem to fuse with them afterpenetration through the cortex. In the neigh-bourhood of the germinal vesicle, which haspenetrated the subcortical yolk layer, thefirst Anlage of the germinal disc forms(fig 21 ).

Laterally from and below the germinalvesicle, labelled yolk granules are visible(comparable with the first Anlage of the so-called polar granules in birds: Bartelmez,1912, Callebaut, 1974) suggesting a slidingmovement. In 8-10-mm-diameter postlamp-brush oocytes (exterior to the germinal discregion) we see that the intensely labelledlarge yolk globules are localized below theless or non-labelled smaller yolk globules(fig 22). This indicates that the 3H-tyrosinepool was already exhausted before fixation.In the germinal disc region of 8-10-mm-diameter postlampbrush oocytes, the pene-tration of the peripherally assembled labelledyolk granules seems to be impaired by thepresence of weakly stained subcorticalooplasmic organelles (fig 23).

Sometimes the labelled penetrating yolkgranules form a cap over these organelles, aswas also seen in the quail germinal disc(Callebaut, 1974). In the germinal discregion of postlampbrush oocytes of approx-imately 15 mm in diameter, we see that thelabelled yolk layer just reaches the anglesof the germinal vesicle (fig 24) without slid-ing laterally from it. Above the germinalvesicle a narrow yolk cap is seen composedof two parts: the deepest part, close to thegerminal vesicle is unlabelled (and thusformed first) whilst the superficial part islabelled and formed after the injection of3H-tyrosine. The penetration of yolk intothis region is thus strongly reduced. In thelargest postlampbrush oocytes, exterior to

the germinal disc, we can see that, at theboundary of labelled and unlabelled yolk,large yolk globules often have a partiallypolarized labelling pattern, ie, that they pre-sent a radioactively labelled cap directedtowards the surface of the oocyte (fig 25).Sometimes large yolk spheres also containone or more round central unlabelled parts,whilst the surrounding periphery of the yolksphere is strongly labelled (fig 26). Even inthe largest oocytes of Pseudemys scriptaelegans, only round yolk units (yolk spheres)are assembled and no polyhedric yolk units,

as has been described during the final rapidgrowth period in the oocytes of birds (Perryand Gilbert, 1985; Callebaut et al, 1991).

Therefore, we prefer not to speak of yolkplatelets because the protein chelonian yolkis present in a globular form. This seemsalso to be the case in oocytes of other rep-tiles, eg, in the lizard Sceloporus torquatustorquatus (Uribe et al, 1995). In the largestpostlampbrush oocytes the yolk cap abovethe germinal vesicle no longer containslabelled yolk, which indicates that here thepenetration of yolk precursors is negligible

or absent (fig 27). The narrow labelled yolklayer in the germinal disc ends at the anglesof the very flattened germinal vesicle. Thegranulosa cells above the germinal vesicleare very flat whilst more laterally in theimmediate neighbourhood, they form a twonuclei thick layer (figs 27 and 28). This isexceptional since it is usually assumed thatthe turtle granulosa layer is only one cellthick. The labelling in the granulosa layeris extranuclear, both above the germinal discregion (fig 28) and exterior to this region(fig 25). This suggests that the transport ofprotein yolk precursors mainly occurs viathe intercellular spaces as is the case in the

quail (D’Herde and Vakaet, 1992) andeventually through the granulosa cytoplasmas also observed in the quail (Callebaut,

1991 a). In contrast to the intense labellingfound in the germinal vesicle of prelamp-brush and lampbrush turtle oocytes, duringthe postlampbrush stage, no or very slightlabelling is seen. During the latter periodthe germinal vesicle no longer increases involume and the metabolic activity of thechromosomes and nucleoli decreases con-

siderably.

Adult turtle ovaryafter two im injections of 3H-tyrosine,4 and 8 days before fixation

Labelling in the smaller turtle oocytes issimilar to the autoradiographic observationsafter a single injection. In larger oocytes

with the beginnings of peripheral proteinyolk assemblage, two labelled layers can berecognized below the relatively broad cortex(fig 29). In still larger oocytes (beginningpostlampbrush stage), the two labelled yolklayers become localized in a more superficialposition (fig 30). Finally in the largest post-lampbrush oocytes (where the cortex isabsent) the last-formed labelled yolk layer islocalized on the surface of the ooplasm(fig 31). Most obvious in the largest post-

lampbrush oocytes is the labelled yolk capover the unlabelled yolk, localized justsuperficially to the central part of the ger-minal vesicle (fig 32). This again indicatesthat the penetration of yolk above the ger-minal vesicle is very slow (more than 8days) and there is only one labelled layervisible above the germinal vesicle. The pro-gressive sliding of the yolk layers below therims of the germinal vesicle is also veryslow (8 days or more). This sliding of the

deep yolk layers below the germinal vesicleis the origin of the polar granules (fig 32)which are analogous to the polar granulesin birds (Bartelmez, 1912; Callebaut, 1974).The formation and the structure of the ger-minal disc of Pseudemys scripta elegansseems to be approximately analogous towhat has been described in the quail (Calle-baut, 1974, 1975). The sliding is, however,slower than in the quail oocyte since in thelatter, after 8 days, the labelled layer wasfound deep in the nucleus of Pander or inthe latebra neck.

The nucleus of Pander, first described inbirds by Pander (1817) as a large cushion-like mass of white yolk below the germinalvesicle, is however not clearly defined in

the turtle. A large column extends from thecentre of the oocyte to below the centre ofthe germinal vesicle (fig 33).

The alternating morphologically differentyolk layers end perpendicularly to and lat-erally from this column. After fixation withcalcium formalin and Unna staining of thesections after the autoradiographic process-ing, we could find at least 14-15 differentlayers alternatively composed of the twotypes of yolk and disposed in concentriclayers encircling at a large distance the cen-tre of the oocyte (figs 14 and 33). Thisressembles the banding pattern describedby Cuellar ( 1971 ) in the parthenogeneticlizard Cnemidophorus uniparens and byUribe et al ( 1996) in Ctenosaura pectinata.

Comparison of the thickness and number oflabelled yolk layers in the oocytes of dif-ferent diameter suggests that 2-3 such lay-ers are laid down in approximately 8 days.

Consequently, we may conclude that theperiod of relatively rapid growth in the stud-ied chelonian oocyte lasts longer than in thequail oocyte, although both finally reachapproximately the same volume.

DISCUSSION

By contrast to the presence of several ger-minal beds in the Pseudemys ovary, no ger-minal beds occur in the adult avian ovary.Indeed in the latter all the oogonia are trans-formed into primary oocytes at the end ofembryonic and the beginning of postnatallife. In birds there are no stem germ cellsfor repeating the seasonal multiplication ofoogonia by mitosis. The premitotic and pre-meiotic DNA synthesis in the chelonianoogonia and premeiocytes have somehomology with what we have described inchicken and quail, respectively (Callebaut,1967, 1973b). The labelling after 3H-thymi-dine administration is localized on the Feul-

gen positive chromatin threads surrounding gthe central nucleoli. A difference with

respect to the corresponding developmen-tal stage in birds is that the germinal bedsin the turtle are much less voluminous. Inthe turtle, oogonia or preoogonia are oftennot localized below the tunica albuginea ofthe ovary but still more superficially at thesurface of the ovary in direct contact with the

peritoneal cavity (fig 5). This very superfi-cial position constitutes perhaps a migra-tion path of undifferentiated female germcells on the ovarian surface. Indeed sincethe studies of Waldeyer (1870) we knowthat in most female reptiles (in contrast tobirds and mammals) oogonia persist untilafter sexual maturity and give rise to oocytesby mitotic divisions (Zuckerman, 1962). Afollicular hierarchy with a limited number offollicles as in the avian ovary seems not to

exist in the chelonian ovary. Our study alsoindicates that before the chelonian oocytematures, it develops through three successivestages: prelampbrush, lampbrush and post-lampbrush. As in the oocytes of birds (Cal-lebaut, 1973a) there exists a correlationbetween the aspect and localization of thegerminal vesicle and the ooplasm in theimmediate neighbourhood. For instance inthe prelampbrush stage, a voluminousparanuclear Balbiani complex is present.The postlampbrush stage begins exactly atthe moment when the germinal vesiclemigrates into the cortex. In the lampbrushstage, the germinal vesicle increases stronglyin volume probably in relation to the lamp-brush chromosome activity and the appe-arance of giant nucleoli. The labelling inthe germinal vesicle after 3H-tyrosine injec-tion is most pronounced during the prelamp-brush and lampbrush stages, whilst thelabelling in the postlampbrush germinalvesicle is minimal or absent and its volumeremains constant. Our study indicates that inthe chelonian oocyte there also exist sub-cortical ooplasmic organelles distributed atthe animal pole around the germinal vesi-cle. These organelles present homology(similar aspect, relationship with penetratingyolk and development at a comparablestage) with the geometrically distributedsubcortical cytoplasmic organelles or ticosin the quail oocyte (Callebaut, 1972, 1983;D’Herde et al, 1995), and with the radiallydisposed mitochondrial clusters describedby Tourte et al (1984) in the Xenopus laevisoocyte. In the present study, at the begin-ning of peripheral yolk formation, wedemonstrated that after penetration throughthe oocytal cortex, granules adhere to thedeeper, larger and earlier formed unlabelledyolk globules. This observation visualizesthe fusion hypothesis of Ho (1987) andGuraya (1989). Moreover, the polarized orpartially superficial radioactive labellingafter 3H-tyrosine administration often seenin the large yolk globules of the large post-lampbrush oocytes also provides evidence

for this fusion hypothesis. In the largestPseudemys postlampbrush oocyte, we foundno particular yolk structure similar to thenucleus of Pander (1817), visible in birds.Even in the largest Pseudemys postlamp-brush oocyte, we could (in contrast to theavian oocytes) distinguish morphologicallythe successive yolk layers in the deeper yolkmass below the germinal disc by the use ofan appropriate fixative (calcium formalin)and Unna staining (fig 33). So it can be seenthat the successive yolk layers in Pseude-mys form only concentric spheres aroundthe centre of the oocyte. These layers do notcircumscribe a second centre (nucleus ofPander) localized below the germinal disc aswe have demonstrated in the quail oocyte(Callebaut, 1974, 1975) by autoradiographicstudies. In the parthenogenetic lizard, Cne-midophorus uniparens, Cuellar (1971)described that the germinal disc lies moreor less isolated from the rest of the yolk.This does not seem to be the case in the che-lonian oocyte where we observed in the pre-sent study that there is always continuitybetween the smaller radioactively labelledlayers in the germinal disc and the broadersuperficial layers exterior to the germinaldisc. Among Sauropsidia, a remarkable dif-ference in the structure of the ovarian gran-ulosa has long been recognized (Loyez,1906). In turtles, it consists, throughout thefollicle growth of a simple, homogenouscuboidal or columnar epithelium in which nopyriform cells were described (Munson,1904; Loyez, 1906; Thing, 1918; Callebautand Van Nassauw, 1987). In contrast, thegranulosa of lizards and snakes (Squamata)develops into a complex layer composed ofseveral cell types (Loyez, 1906; Betz, 1963;Varma, 1970; Uribe et al, 1995, 1996). Theheterogeneity of the granulosa cell popula-tion in squamate reptiles is unlike thatknown for any other vertebrate: most notableis the transitory presence of very large clearpyriform cells. In the quail granulosa layerwe have described pyriform-like cells(Callebaut, 1991b) with a structure and tem-

porospatial distribution, that has much anal-ogy with the pyriform cells described insquamate reptiles. We found no such pyri-form-like cells in the chelonian granulosalayer. The peripheral arrangement of manynucleoli in the previtellogenic oocytes offish, amphibia and turtles suggests somephylogenetic relationships among them(Guraya, 1989). By contrast the oocytes ofsquamates (Arronet, 1973) and of birds(Callebaut, 1975) have a few nucleoli thatare localized in the central karyoplasm. Thisis a second similar aspect of the follicularand oocytal structures in squamate reptilesand birds. Moreover, the obvious and poly-morphic development of the granulosa layerin squamates (with a.o. pyriform cells) andin birds (with pyriform-like cells: Callebaut,1991 b) seems to be inversely proportional tothe feeble nucleolar development in the ger-minal vesicle of their oocytes. By contrast inChelonians the granulosa remains monomor-phic and mainly single-layered, whilst thenucleoli are very large and numerous andpersist throughout the oocytal development.Perhaps only the central nucleoli observedduring the prelampbrush and the lampbrushstage in Pseudemys are homologous to thenucleoli seen in aves and in squamate rep-tiles.

In the present study we demonstrate, by ipinjection of trypan blue-labelled yolk, that nofunctional communications exist betweenthe lacunae of the chelonian ovary and the

peritoneal cavity. By contrast, in the quail ithas been shown by different labelling pro-cedures that lacuno peritoneal commu-nications are present (Callebaut, 1979a,b,1988a). The reason for the difference is notknown. Between the hilus ovarii of both

species there are indeed obvious differences.In birds the hilus ovarii is very irregular inform. Moreover, the avian ovary is fixed,without a mesovarium, directly to the dorsalwall of the peritoneal cavity in the imme-diate neighbourhood of the vena cava infe-rior. The eggs in birds are laid one by one,sometimes over long periods. The rapid

growth phase of the avian oocytes (end ofthe postlampbrush stage) is faster than inturtle oocytes. This requires perhaps theexistence of an expansion space and escaperoute for the increasing pressure in the ova-ry.

ACKNOWLEDGMENTS

We thank S Van Rompaey and V Van der Stockfor excellent technical assistance, F De Bruynfor artwork and V De Maere for typing themanuscript. This study was supported by grant3.0029.93 of the National Fund of ScientificResearch Belgium and a matching fund from theUniversity Centre of Antwerp (RUCA).

REFERENCES

Altland P ( 1951 ) Observations on the structure of thereproductive organs of the box turtle. J Morphol89,599-621 1

Arronet VN (1973) Morphological changes of nuclearstructures in the oogenesis of reptiles (Lacertidae,Agamidae). J Herpetol7, 163-193

Bartelmez G (1912) The bilaterality of the pigeon’s segg. A study in egg organization from the first

, growth period of the oocyte to the beginning ofcleavage. J Morphol 23, 269-329

Bhattacharya DR ( 1925) Les inclusions cytoplasmiquesdans 1’oogenese de certains reptiles. These. Vign6Ed, Paris, France

Betz TW (1963) The ovarian histology of the diamond-backed water snake, Natrix Rhombifera, during thereproductive cycle. J Morphol 113, 245-260

Callebaut M (1967) Premeiosis and premeiotic DNAsynthesis in the left ovary of the female chickembryo. J Embryol Morph 18, 299-304

Callebaut M (1972) Cytoplasmic uniaxial!radial sym-metry during the final growth period in the oocytesof the Japanese quail (Coturnix coturnix japonica).Experientia 28, 62-63

Callebaut M (1973a) Correlation between germinalvesicle and oocyte development in the adultJapanese quail (Coturnix coturnix japonica). Acytochemical and autoradiographic study. JEmbryol Exp Morph 29, 145-157

Callebaut M (1973b) La synth6se de 1’ADN pr6mito-tique et pr6meiotique dans les cellules germinalesde 1’embryon femelle de la caille japonaise(Coturnix cotumixjaponica). Experientia 29, 471-472

Callebaut M ( 1974) La formation de l’oocyte d’oiseau.Étude autoradiographique chez la caille japonaisepondeuse a I’aide de la leucine triti6e. Arch Biol(Bruxelles) 85, 201-233

Callebaut M ( 1975) Bijdrage tot de studie van de ooge-nesis van de vogels. PhD thesis. RUCA, Antwerp,Belgium

Callebaut M (1979a) Lacunoperitoneal communica-tions in the ovary of the baby quail. IRCS Med Sci7,384

Callebaut M ( 1979b) The avian ovary is an open organ;a study of the lacunar system. Anat Embryol 158,103-119

Callebaut M (1983) Electron microscope study of ticos(3H-thymidine incorporating cytoplasmicorganelles) in the germinal disc of the postlamp-brush oocytes of Japanese quail. IRCS Med Sci 11, I,491-492

Callebaut M ( 1988a) The ovarian chordolacunar systemin birds. Arch Biol (Brussels) 99, 1-IS

Callebaut M (1988b) Localization of voluminous highlyosmiophilic material in the largest ovarian folliclesof laying Japanese quails. Med Sci Res 16, 1307-1309

Callebaut M (1991a) Electron microscopic evidencefor the delivery of lipids by the granulosa cells to thesuperficial avian ooplasm. Eur Arch Biol (Brus-sels) 102, 47-53

Callebaut M (1991b) Pyriform-like and holding gran-ulosa cells in the avian ovarian follicle wall. EurArch Biol (Brussels) 102, 135-145

Callebaut M, Demalsy P (1967) Improved resolution offine caryological details by staining and examina-tion before and after autoradiography. Stain Tech-nol 42, 227-229

Callebaut M, Van Nassauw L (1987) Demonstration bymonoclonal anti-desmin of a myoid tissue coat inthe preovulatory ovarian tunica albuginea of theturtle Pseudemys scripta elegans. Med Sci Res 15,1129-1130

Callebaut M, D’Herde K, Hermans N, Van Nassauw L(1991) Localization and transport of lipids in theavian ovarian follicular layers and the structuralrelationship of theca and granulosa to the basementmembrane. J Morphol 209, 143-163

Caro L, Van Tubergen R (1962) High resolution autora-diography. J Cell Biol 15, 173-188

Cuellar 0 (1971) Reproduction and the mechanism ofmeiotic restitution in the parthenogenetic lizardCnemidophorus uniparens. J Morphol 133, 139-166

Demalsy P, Callebaut M (1967) Plain water as a rins-ing agent preferable to sulfurous acid after the Feul-gen nuclear reaction. Stain Technol 42,133-136.

D’Herde K, Vakaet L (1992) Study of yolk precursortransport in the avian ovary with the use of horseradish peroxidase.IntJ Dev Biol 36, 435-438

D’Herde K, Callebaut M, Roels F, De Prest B, VanNassauw L (1995) Homology between mitochon-driogenesis in the avian and amphibian oocyte.Reprod Nutr Dev 35, 305-311

Guraya SS (1959) Histochemical studies of lipids inoocytes VII. Lipids in the oogenesis of the fresh-water turtle Lissemys punctata. Res Bull PanjabUniv 10, 305-313

Guraya SS (1989) Ovarian Follicles in Reptiles andBirds. Springer Verlag, Berlin, Germany

Ho SM (1987) Endocrinology of vitellogenesis. In:Horntones and Reproduction in Fishes. Amphibiansand Reptiles (DO Norris and RE Jones, eds), Ple-num Press, New-York, USA, 145-170

Loyez M (1906) Recherches sur le d6veloppementovarien des oeufs meroblastiques a vitellus nutritifabondant. Arch Anat Microscop Morph Exp 8, 69-398

Munson J (1904) Researches on the oogenesis of thetortoise Clemmys marorata. Am J Anat 3, 311-346

Pander C (1817) Historiam metamorphoseas quamovum incubatum prioribus quinque diebus subit.FE Nitribitt Wirceburgi

Perry M, Gilbert AB (1985) The structure of yellowyolk in the domestic fowl (Gallus domesticus). JUltruct Res 90, 313-322

Rahil KS, Narbaitz R (1973) Ultrastructural studies onthe relationship between follicular cells and grow-ing oocytes in the turtle Pseudemys scripta ele-gans. JAnat 115, 175-186

Romeis W (1948) Mikroskopische Technik 15. Leib-nitz, Aufl Munchen, Germany

Thing A (1918) The formation and structure of thezona pellicida in the ovarian eggs of turtles. Am JAnat 23, 237-258

Tourte M, Mignotte F, Mounolou JC ( 1984) Hetero-geneous distribution and replication activity ofmitochondria in Xenopus laevis oocytes. EurJ CellBiol34, 171-178

Uribe M, Mendez Omana M, GonzAlez Quintero J,Guillette L Jr (1995) Seasonal variation in ovarianhistology of the viviparous lizard Sceloporustorquatus torquatus. JMorphol226, 103-119 9

Uribe M, Portales GL, Guillette L Jr (1996) Ovarian fol-liculogenesis in the oviparous Mexican lizard Cte-nosaura pectinata. J Morphol 230, 99-112 2

Van Nassauw L, Callebaut M (1991) Structural andimmunohistochemical aspects of the postovulatoryfollicle in Japanese quail. Anat Rec 229, 27-30

Van Nassauw L, Harrisson F, Callebaut M (1991)Localization of smooth-muscle markers in theovaries of some ectothermic vertebrates. Anat Rec229, 439-446

Varma S (1970) Morphology of ovarian changes inthe garden lizard Calotes versicolor. J Morphol131, 195-210

Waldeyer W (1870) Eierstock und Ei. Engelmann,Leipzig, Germany

Zuckerman S (1962) The Ovary. Academic Press, NewYork, USA