HAL Id: hal-00890787https://hal.archives-ouvertes.fr/hal-00890787

Submitted on 1 Jan 1989

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.

Membrane-bound iron-rich granules in fat cells andmidgut cells of the adult honeybee (Apis mellifera L.)

H. Raes, W. Bohyn, P.H. de Rycke, F. Jacobs

To cite this version:H. Raes, W. Bohyn, P.H. de Rycke, F. Jacobs. Membrane-bound iron-rich granules in fat cells andmidgut cells of the adult honeybee (Apis mellifera L.). Apidologie, Springer Verlag, 1989, 20 (4),pp.327-337. �hal-00890787�

https://hal.archives-ouvertes.fr/hal-00890787https://hal.archives-ouvertes.fr

Original article

Membrane-bound iron-rich granules in fat cells andmidgut cells of the adult honeybee (Apis mellifera L.)

H. Raes W. Bohyn P.H. De Rycke F. Jacobs

! Labo/’afo/y!o/’Zoop/tys/o/ot! State University of Gent, K.L. Ledeganckstraat 35, B-9000 Gent,Belgium;2 Laboratory for Electron Microscopy, State University of Gent, St. Pietersnieuwstraat 25, B-9000Gent, Belgium

(received 25 April 1988; accepted 18 May 1989)

Summary — Previous studies have found iron-containing granules in fat cells of adult workerhoneybees. By using different preparation techniques, we collected additional data on theiroccurrence, origin, formation, relative composition and possible function.

In addition to fat cells, we found the same type of iron-rich granules in columnar cells of themidgut; in both cell types they are formed within the cisternae of the rough endoplasmatic reticulum.In the midgut their occurrence is limited to the period of pollen nutrition; during that period thereticulum forming the granules is arranged in a fingerprint design.

Energy-dispersive X-ray analyses of air-dried granule preparations show that, apart from Fe, Pand some Ca and K which could be demonstrated on thin sections, Na, Mg, S, Cl, Mn and Zn arealways present. All our results suggest that their function is similar to that of other mineralizedgranules, i.e. storage of surplus ions and toxic metals; in support of this is the fact that perorallyadministered Pb is accumulated in the iron rich granules as well as in the spherocrystals.

Apis mellifera - biomineralization - iron-rich granule - fat cell - midgut cell

Introduction

Many invertebrates are capable of dealingwith surplus ions and toxic metals by aprocess of biomineralization. In severaltissues - mostly involved with digestion,storage or excretion - mineralizedgranules are formed within the Golgivesicles or the cisternae of the granularendoplasmatic reticulum (Jeantet et aL,1977; Simkiss, 1979, 1981; Brown, 1982).

The presence of calciferous granulesin midgut epithelial cells of the honeybeehas been known for some time (Koehler,1920); they clearly correspond to thespherocrystals described in other insects(Martoja et at, 1984).

Iron-rich granules were described forthe first time by Kuterbach et aL (1982) inthe fat body of adult honeybees; however,these fat cells were mistakenly calledoenocytes. This slip was adopted by Loper(1985), who described iron-rich granules

in the oenocytes of drones. Later,Kuterbach and Walcott (1986a,b)provided a more thorough description ofthe granules in the trophocytes of adultworkers. They examined the possible roleof these structures in the detection of theearth magnetic field by honeybees; thegranules were described as membrane-bound, but &dquo;without association with anycellular organelle such as mitochondria orendoplasmatic reticulum&dquo;.

Finally, similar granules were des-cribed in 1987 by Da Cunha in the fatcells of a stingless bee, Scaptotrigonaportica Latr.

Pending a study on the ultrastructureof the fat body and the midgut in the adulthoneybee, our attention was drawn by theoccurrence of electron-dense, apparentlymineralised structures both in the

trophocytes and in the midgut columnarcells. By using different preparationtechniques, we collected additional dateon their occurence, origin, formation,relative composition and possiblefunction, which adds considerably to theknowledge of these cellular inclusions.

Materials and Methods

Animals

Honeybees of.known age were used, both froman outside hive and from experimental cages.Adult bees less than 24-h old, which emergedin the incubator, were color-marked and joinedto their colony, or kept in the laboratory inrearing cages with 50 to a cage. The animalswere provided with a piece of comb foundation,drinking water, sugar syrup (1 : 1 ) and amixture of comb pollen and honey (2 : 1 ). Insome experiments where the effect of pollennutrition on the occurrence of iron-rich granuleswas studied, the mixture of comb pollen andhoney was omitted. The rearing cages werekept in a dark, climatized room at 30 °C.

HistologyDissection and preparation of the fat body wascarried out as described by Raes et al. (1985);additional care was taken to adjust theosmolarity of the fixation buffer at 440 mosm(Raes et aL, 1987). The midgut was removedfrom the abdomen by gently pulling the lastabdominal segment. The honey stomach, therectum and the Malpighian tubules werecarefully cleared away.

For fluorescence and phase contrastmicroscopy, unstained paraffin sections or freshtissue were used. In order to visualise the iron-rich granules at the light microscope level,paraffin sections were stained with the Prussianblue method according to Hutchison (1953).

The ultrastructural study was hampered bythe fact that honey bee tissue show only aslight affinity for uranium and lead, resulting invery poor membrane contrast. The method

adopted here is the result of muchexperimentation and could probably be furtherimproved :- Dissection in ice-cold fixative (2%

gluteraldehyde and 2% paraformaldehyde in0.1% cacodylate buffer, containing 0.12 M ofsucrose and 0.05% CaC’2, and adjusted to apH of 7.4).- Long fixation times (between 6 and 18 h)

to improve membrane contrast. Post-fixation in1 % osmium tetroxide in the same buffer.

- During dehydration 2 successive &dquo;enbloc&dquo; staining steps : first with 2% uranylacetate in 50% ethanol; secondly in saturatedlead acetate in ethanol-acetone (1 : 1)(Kushida, 1966); both steps take 1 h.- Embedding in a mixture of Epon and

Araldite (1 : 1), thus combining the goodpenetration of the former with the bettercontrast given by the latter (penetration in 3steps, within 48 h).- Silver and gray sections were post-

stained with Reynolds lead solution and 0.8%uranyl solution in 50% methanol; alternatively,the sections were stained with alkaline bismuth(Hayat, 1981 ).

The last method gives a much bettercontrast than the first, but the coarser grain ofthis stain interferes with magnifications of> 30 000.

The ultrastructural studies were performedon a Philips EM 420 analytical transmissionmicroscope.

Energy-dispersive X-ray analysis (EDX)EDX analysis was performed on 2 types ofsamples : 1 ) Thin gold sections made fromtissue prepared without post fixation, excludingany contact with heavy metals. 2) Air-dried,granule preparations processed as follows :- Both dissection and homogenising wereperformed in deionized water.- The fat body or the midgut of one honeybeewas homogenized in an Eppendorf tubecontaining 200 ,.11 of deionized water, with anepoxy-resin pestle. This system was especiallyconceived for small quantities and to minimizemetal contamination. A small droplet of thehomogenate was placed on a Formvar-supported single slot copper or nickel grid andair-dried. In these preparations the mineralizedgranules could easily be recognized andanalyzed under the transmission microscope.According to Morgan (1984) this and similarpreparation procedures are acceptable for theEDX analysis of relatively insoluble mineralizedgranules.

The analyses were performed with anEDAX-9100 system and the same microscope,at an accelerating voltage of 60 kV and a probesize of 200 11m. All spectra were recorded for atleast 60 live s.

Results

HistologyFat cells. Initially we noticed the granuleson paraffin sections of old winter bees,where they appeared as a blackishgranulation in the cytoplasm, but were toosmall to be discerned individually. Thegranules are most obvious when themicroscope is slightly out of focus. Theirpresence causes a strong PAS (Periodicacid - Schiff) coloration in the fat cellcytoplasm. In fresh preparations, underphase contrast, they can be seen as tinyblack dots, whereas under UV light, theyshow a bright yellow autofluorescence.These results suggest that they are atleast partly organic in nature.

After treatment of paraffin sections withPrussian blue, the positive reactionshowed that the granules were rich iniron.

On thin sections we found that the

granulation corresponded with theoccurrence of very electron-dense,membrane-bound granules, measuringbetween 400-800 nm. In fat cells of old

bees, they are the predominant feature inthe cytoplasm and seem to be distributedat random.

Their extreme hardness makes

sectioning of the granules difficult; evenwith a diamond knife, the granules tend tobreak or to be torn from the resin.

Their form is irregular and they appearcoarsely granular. From sections of tissuethat had not been stained with heavymetals, it can be concluded that thegranules are naturally electron-dense.However, in unstained sections, theirappearance is less homogenous, sug-gesting the existence of a matrix which isnot naturally electron-dense, but whichhas a strong affinity for lead and uranyl.

In order to determine the origin ofthese granules, we studied fat cells ofyoung honeybees. In 24-h-old bees, thegranules are relatively rare, but they canalready be found in every fat cell. At thistime they measure between 100—150nm. New granules are formed within thecisternae of the rough endoplasmaticreticulum (RER), close to both the plasmamembrane and the nucleus. Younggranules bud from the RER with theirmembrane still carrying some ribosomes(Fig. 1, inset). They increase by theaccumulation of coarsely granularmaterial which is most obvious onunstained sections; sometimes, fusion of2 small granules can be seen. The largestor mature granules in older animals havelost their ribosomes and are at this pointrandomly distributed in the cytoplasm.

In animals reared without pollen,granules are very rare and theirultrastructure differs from that describedin pollen-fed bees. They are considerablyless electron-dense and show a finelygranular material which is very similar toferritin (Figs. 2, 3). In addition, theelectron-dense material does not fill thevesicle formed by the RER as in thepollen-fed bees.

Midgut epithelial cells. In midgutepithelial cells we found the same type of

granules as in the fat cells. However, inthe former the granules only occur duringa limited period of the animal’s life. Bymeans of the Prussian blue stain, wefollowed the distribution, appearance anddisappearance of the iron-rich granules inthe midgut epithelium on longitudinalparaffin sections. In a first experiment theanimals were reared in cages as

described; they were studied after 1, 4, 8,10, 14, 18, 20 days. In a secondexperiment, pollen nutrition was stoppedon the 4th day; we studied the midgut

tissue on the 5th, 6th, 7th and the 8th dayrespectively.

In control bees fed pollen, the granulescannot be found on day 1; they are veryabundant on day 4. Their formation doesnot occur in the most apical and caudalparts of the midgut which are free of iron-rich granules. They can be seen in all thecolumnar cells but not in the crypt cells.From the 10th day onwards, theiroccurrence decreases gradually. Betweenthe 10th and the 20th day they no longeroccur in the young epithelial cells and thearea where they are found becomesrestricted to the most central part of themiddle region of the midgut. On the l8thday they are generally absent and whenthey still occur, it is only in the oldest cells,which are ready to be shed.

When pollen feeding is terminated onthe 4th day, the amount of granules on the5th day is the same as on the 4th day. Onthe 6th day the occurrence of granulesshifts gradually to the older epithelial cells.By the 8th day, all the granules havedisappeared from the epithelium.

At the EM level, the RER in the cellsclose to the regenerative crypts isarranged in lamellar stacks around thenucleus in 4-day-old bees (Fig. 4) : inthese cisternae the granules are formedthe same way as in the fat cells. However,this lamellar arrangement was not foundin the fat body cells. In 18-day-old beesthe granules have become extremelyrare. At that age the RER has lost its well-ordered appearance and the cisternaehave become swollen and vesiculated.

EDX analysisQualitative EDX analyses of granules inthin sections of unstained, unosmicatedmaterial show large peaks for Fe and P,and smaller peaks for Ca and K (Fig. 5)(the Cu peak is an artefact from thecopper grid; it disappears when nickelgrids are used; the Si peak is an artifactfrom the detector). Analyses of severalgranule-free areas in the cytoplasm showa small peak for P and only trace amountsof Ca and K. These results are notinfluenced by the location of the granulesor by the age of the animal.

On osmicated sections, the granulesshow only trace amounts of osmium; thusthey are not osmiophilous.

In air-dried granule preparations of fatbody tissue, the granules can very easilybe identified under the transmissionelectron microscope due to their electrondensity. In midgut tissue this identity mustbe confirmed by the typical high iron peakin the EDX spectrum because confusionwith the so-called &dquo;calciferous granules&dquo;is possible. Typical spectra of thegranules in air-dried preparations (Fig. 6)show several elements which cannot bedetected on thin sections. The granules,irrespective of their origin, are mainlycomposed of P, Fe, Ca and K. Smallerpeaks for Na, Mg, S, CI, Mn and Zn arealways present; again, the Cu- and Sipeaks are artefacts.

Discussion and Conclusion

Fat cells and midgut epithelial cells fromadult honeybees contain electron-densemineralized granules which are formedwithin the cisternae of the rough

endoplasmatic reticulum. After theirisolation from the reticulum they appearsurrounded by a ribosome-coveredmembrane and are dispersed throughoutthe cytoplasm. They increase in size,probably until they have lost theirribosomes.

Their place of origin, PAS positivereaction and autofluorescence indicatethat they are partly composed of organicmatter. This would agree with one of thecharacteristics of mineralized granules(Gouranton, 1968; Martoja et al., 1984).At this point the nature of this organicmatrix has not been elucidated.

EDX analyses of granules on thinsections yield the same results asdescribed by Kuterbach and Walcott(1986a) : 2 large peaks for P and Fe, anda smaller peak for Ca; in some cases wealso found traces of K and Na. However, itis known that the preparation proceduresnecessary for thin resin sections involveserious extraction of both organic andinorganic material (Brown, 1982; Masonand Simkiss, 1982). This is reconfirmedby the comparison of the analytical resultsof thin sections with those of air-dried

granule preparations. This extremelysimple preparation method provides a

considerably better conservation of themineral components of the granules.Apparently the granules are composed ofa wide variety of elements, which ischaracteristic of mineralized granules(Ballan-Dufrancais, 1970; Sohal et aL,1976). Their occurrence in trophocytes,high iron content, formation within theRER and the absence of concentric stratadifferentiate them from the well-known

spherocrystals.In the fat cells the number of granules

increases considerably as the animalgrows older; in accordance withKuterbach and Walcott (1986b), we foundthat this increase takes place primarily inthe young adult. However, we canot agreewith the authors’ conclusion that thenumber of granules would be limited by&dquo;a finite number of sites within the cellavailable for granule formation&dquo;, ratherthan by the amount of iron in the diet. Theunderlying experiments are unconvincingbecause the difference between the &dquo;iron-deficient&dquo; group and the &dquo;iron-enriched&dquo;

group was insignificant : both groupsreceived pollen, which is a much richeriron source than the iron supplied in thesugar syrup of the latter group.

The period in which the number ofgranules increases the most coincideswith the period in which the animals feedon pollen. In old winter bees, who feed onpollen much longer, the amount of iron-rich granules is markedly higher than inold summer bees. Moreover, when wekept bees in experimental cages withoutpollen, very few granules were formed.We therefore suggest that the number ofgranules is limited by the amount of pollenconsumed, and most of all by their ironcontent.

At the low iron diet (without pollen), wefound that the fine structure of the

granules was different. Normally, even ongray sections, it was impossible to discern

any ordered substructure in the granules;without pollen, their ultrastructure re-vealed the occurrence of finer granules,resembling ferritin. While it cannot beexcluded that in normal granules the finersubstructure may be masked by their highmineral content, it is also possible thatnormally the iron is bound to anamorphous matrix like haemosiderin orapoferritin. The latter would follow atheory developed by Locke and Leung(1984) while studying the occurrence offerritin in Calpodes larvae. These authorsreport for the first time the occurrence offerritin within the cisternae of the ER inseveral tissues of this insect. At a normallow iron intake, ferritin molecules areformed; in contrast, when the animals areloaded with unnaturally high ironconcentrations, ferritin is replaced by afluffy amorphous material. The authorsinterpret this material as apoferritin whichhas gained affinity for iron beforeassuming its shell structure. It is notimpossible that in an insect with a highdietary iron intake like the honeybee, thelatter system is the rule rather than theexception.

To our knowledge, iron-rich granulessimilar to those described here have not

yet been found in midgut cells of insects.In the columnar cells of adult honeybeesthey appear only during a limited period ofadult life, and their occurrence isrestricted to the middle half of the midgut.In 4-day-old bees, the RER in which theyare formed shows a whirl-like organi-sation, not unlike that described byStaubli et al. (1966) in midgut cells ofmosquitoes. This has been interpreted asan adaptation to the temporal need todigest a protein-rich food. This isparticularly interesting, because, as inblood-sucking insects, the midgut ofhoneybees must only temporarily digestprotein-rich food (pollen); afterwards, bothinsects are nectar feeders.

In contrast to the fat cells, the midgutepithelial cells are constantly renewed,and thus some days after the granulesare formed they are eliminated from thetissue. Two days after the prematureending of the pollen nutrition in oursecond experiment, granules can nolonger be seen in the younger columnarcells where normally they are mostactively synthetized. This indicates thatno new granules are formed past thisperiod and that their formation is closelylinked to pollen nutrition.

Neither at the light microscope levelnor at the ultrastructural level have weever seen the extrusion of individual

granules from the cells; their eliminationseems bound to the expulsion of old,spent cells which are filled withspherocrystals, iron-rich granules andsecondary lysosomes. Therefore, theresults of this experiment indicate that therenewal of the midgut epithelium in adulthoneybees takes about 4 days.We cannot exclude the possibility that

the iron-rich granules have a functionrelated to the detection of earth

magnetism as suggested by Loper (1985)and by Kuterbach and Walcott (1986a).However, we think it more probable that,like the typical spherocrystals which arealso numerous in the midgut epithelium,the iron-rich granules serve in eliminatingsurplus ions by a process of biomine-ralization. As the midgut spherocrystalscontain only traces of iron, their affinity forthis element is apparently too small todeal with the high concentrations presentin pollen nutrition. A further intensificationof the iron elimination system describedby Locke and Leung (1984) wouldtherefore form a useful adaptation.

Apart from iron, the granules containseveral other elements and it is

interesting to note that after peroraladministration of lead chloride, they store

this heavy metal just like thespherocrystals (Raes et al., 1988). In thiscontext we are now engaged in a study onthe detoxification of heavy metals byhoneybees, and their potential use asbiomonitors for this form of pollution.

Acknowledgments

We are grateful to Mrs. U. Rzeznik for skillfultechnical assistance. This research has beenpartly supported by grant No. 2.0015.88 fromthe FKFO (Belgian Fund for Collective Funda-mental Research).

Résumé — Granules riches en ferassociés à la membrane dans lescellules adipeuses et les cellules del’intestin moyen chez l’abeille adulte

(Apis mellifica L.). Kuterbach et al.(1982) et Kuterbach et Walcott (1986a, b)ont décrit chez l’abeille des granulesriches en fer dans les cellules adipeuses.Nous vous présentons ici des donnéescomplémentaires sur la présence,l’origine, la formation, la compositionrelative et la fonction possible de cesgranules.

Les abeilles, d’âge connu, proviennentd’une ruche de plein air et de cagettesexpérimentales. Pour la microscopie enfluorescence et en contraste de phase,nous avons utilisé des coupes paraffinéesnon colorées ou du tissu frais. Les

granules riches en fer ont été mis enévidence sur les coupes paraffinées par laméthode du bleu de Prusse. A cause dufaible contraste membranaire des tissusde l’abeille il a fallu mettre au point unetechnique spéciale de préparation de cestissus pour la microscopie électronique àtransmission. La composition relative des

granules a été étudiée par l’analyse pardiffraction aux rayons X (EDX) sur lescoupes fines réalisées dans du tissu

préparé sans fixation ou coloré avec desmétaux lourds et sur des préparations degranules séchés à l’air. Les analyses ontété faites avec un système EDAX-9100sur un microscope analytique àtransmission Philips EM 420.

Les concrétions minéralisées seforment dans les cellules adipeuses et lescellules épithéliales de l’intestin moyendes abeilles adultes; elles sont riches enfer, naturellement opaque aux électronset possèdent probablement une matriceorganique.

Un spectre EDX typique de granulesséchés à l’air a montré plusieurs élémentsqui n’ont pu être détectés sur les coupesfines. Outre les principaux picscorrespondant à P, Fe, Ca et K, des picsplus petites pour le Na, Mg, S, CI, Mn etZn étaient toujours présents.

Des granules riches en fer se formentà l’intérieur des cisternae du reticulum

endoplasmique rugueux. Apèrs avoirbourgeonné, à partir du reticulum, ilsrestent entourés d’une membranecouverte de ribosomes qui leur permet degrandir. Dans les cellules adipeuses, ilsse forment soit contre la membranecellulaire, soit contre le noyau; leurnombre s’accroît principalement durant lapériode où les abeilles se nourrissent depollen. Chez les abeilles plus vieilles, lesgranules ont perdu leurs ribosomes etsont alors distribués au hasard dans le

cytoplasme. Le même type de granulesse forme dans les cellules columnaires.Dans les cellules jeunes, le RER à partirduquel elles se forment se présente defaçon typique. Tandis que dans lescellules adipeuses les granules sontaccumulés durant toute la vie de l’insecte,dans l’intestin moyen, ils disparaissentavec les cellules épithéliales utilisées. Ils

sont très abondants le 4e jour mais leurnombre décroît à partir du 10e jour,lorsque les abeilles arrêtent leuralimentation pollinique. Lorsque l’onsupprime l’alimentation pollinique le 4ejour, les granules disparaissent del’épithélium de l’intestin moyen avant le8e jour fournissant ainsi une indication surla durée de vie des cellules columnaires.Chez les abeilles élevées sans pollen, lesgranules sont très rares et leurultrastructure différente : ils sont moins

opaques aux électrons et présentent defins granules semblables à de la ferritine.

Nous proposons l’hypothèse suivante :comme les sphérocristaux typiques,également nombreux dans l’intestinmoyen, les granules riches en fer serventà éliminer les ions en surplus, par unprocessus de biominéralisation. Leurformation semble être liée à la nutrition

pollinique. Il s’agit probablement d’uneadaptation particulière à la forte teneur enfer de cette source de protéine.

Apis mellifera - biominéralisation -

granule riche en fer - cellule adipeuse- cellule épithéliale de l’intestin

Zusammenfassung — An Membranengebundene eisenhaltige Granula in denZellen des Fettkörpers und des Mittel-darmes der erwachsegen Honigbiene(Apis mellifera L.). Eisenhaltige Granulain Fettzellen wurden bei Honigbienen vonKuterbach et al. (1982) und vonKuterbach und Walcott (1986a, b)beschrieben. In dieser Arbeit legen wirzusätzliche Daten über Vorkommen,Ursprung, Bildung, anteilmäßige Zusam-mensetzung und mögliche Funktiondieser Granula vor.

Für die Versuche wurden Honigbienenbekannten Alters aus freifliegenden

Völkem so wie aus Versuchskäfigenbenutzt. Für Untersuchungen mit demFluoreszenz- und Phasenkontrast-Mikroskop wurden ungefärbte Paraffin-schnitte oder frisches Gewebe benutzt.Die eisenhaltigen Granula wurden inParaffinschnitten mit der Preußisch-Blau-Methode sichtbar gemacht.

Wegen des schwachen Kontrastes derMembranen des Bienengewebes mußtefür TEM-Untersuchungen dieses Gewebeseine besondere Technik entwickeltwerden.

Für die Untersuchung der relativenZusammensetzung der Granula wurdenEDX-Analysen der Dünnschnitte vonGeweben benutzt, die ohne Fixierungoder Färbung mit Schwermetallenpräpariert waren, oder luftgetrockneteGranula-Präparate. Die Analysen wurdenmit einem EDAX-9100-System an einemPhilips EM 420 analytischen Trans-missions- Mikroskop durchgeführt.

Die mineralisierten Konkretionenwerden sowohl in Fettzellen wie in

Epithelzellen des Mitteldarms dererwachsenen Honigbiene gebildet. Siehaben einen hohen Eisengehalt, sindelektronendicht und besitzen wahr-scheinlich eine organische Matrix. Eintypisches EDX -Spektrum von luftgetrock-neten Granula zeigte mehrere Elemente,die an Dünnschnitten nicht gefundenwerden konnten. Außer den Hauptpeaksfür P, Fe, Ca und K waren immer auchkleinere Peaks für Na, Mg, S, CI, Mn undZn vorhanden.

Die eisenhaltigen Granula werdeninnerhalb der Zistemen des RER gebildet.Sobald sie sich aus dem Retikulum als

Knospen entwickelt haben, bleiben sievon einer Ribosomen-bedeckten Mem-bran eingeschlossen, wodurch eineGrößenzunahme ermöglicht wird. InFettzellen werden sie entweder an der

Zellmembran oder am Kern gebildet; ihreZahl steigt besonders während derPeriode, in der sich die Tiere von Pollenemähren. Bei älteren Bienen haben sieihre Ribosomen verloren und die Granulasind dann zufällig im Zytoplasma verteilt.

Derselbe Typ von Granula wird in densäulenförmigen Zellen in der Zentral-region des Mitteldarmes gebildet. Injungen Zellen ist das RER, aus dem siesich bilden, in einer sehr typischen Weiseausgebildet. Während die Granula in denFettzellen während des ganzen Lebensdes Tieres erhalten bleiben, werden sie imMitteldarm zusammen mit den Epithel-zellen abgestoßen; am 4.Tag sind sie sehrreichlich, aber nach dem 10.Tag - wenndie Biene die Pollenaufnahme beendet -

nimmt ihre Zahl ab. Wird die

Pollenemährung am 4.Tag unterbrochen,dann sind die Granula am 8.Tag aus denEpithelzellen des Mitteldarms versch-wunden, wodurch auch ein Hinweis aufdie Lebensdauer der säulenförmigenZellen gegeben wird. In pollenfreiaufgezogenen Bienen sind die Granulasehr selten und deren Struktur istverschieden : Sie sind wenigerelektronendicht and zeigen feine Ferritin-ähnliche Kömchen.

Wir vermuten, daß die eisenhaltigenGranula ähnlich wie die typischen,zahlreichen Sphärokristalle des Mittel-

’

darms dazu dienen, überschüssige Ionendurch einen Prozeß der Biomineralisation

auszuscheiden; ihre Bildung scheint andie Pollenemährung gekoppelt zu seinund stellt wahrscheinlich eine Anpassungan den hohen Eisengehalt dieserProteinquelle dar.

Apis melllfera - Blominerallsatlon -

eisenhaftige Granula - Fettzell -Epithelzelle des Mitteldarms

References

Ballan-Dufranqais C. (1970) Données cytophy-siologiques sur un organe excréteur particulierd’un insecte : Blatella germinica (L). Z.Zellforsch. Mikrosk. Anat. 109, 336-355

Brown B.E. (1982) The form and function ofmetal-containing &dquo;granules&dquo; in invertebratetissue. Biol. Rev. 57, 621-667

Da Cunha M.A.S. (1987) Iron containing cells inthe stingless bee Scaptotrigona posdca Latr.(Hym., Apidea). In : Chemistry and Biology ofSocial Insects (J. Eder & H. Rembold, eds),p. 91Gouranton J. (1968) Composition, structure etmode de formation des concrétions min6ralesdans fintestin moyen des Homopt6resCercopides. J. Cell Biol. 37, 316-328Hayat M.A. (1981) Principles and Techniquesof Electron Microscopy. Biological Applications(A.E. Kent, ed), p. 413 3Hutchison H.E. (1953) The significance ofstainable iron in sternal marrow sections.Blood 8, 236-248Jeantet A.Y., Batian-Dufrancais C. & Marjota R.(1977) Insect resistance to mineral pollution.Importance of spherocrystals in ionicregulation. Rev. Ecol. Sol. 14, 563-582Koehler A. (1981) Uber die Einschliisse derEpithelzellen des Bienendarmes und die damitin Beziehung stehenden Probleme derVerdauung. Z. Mgew. Entomol. 7 (1) 68-91Kushida H. (1966) Staining of thin sections withlead acetate. J. Electron. Microsc. 15, 93Kuterbach D.A. & Walcott B. (1986a) Iron-containing cells in the honeybee (Apismellifera). I. Adult morphology and physiology.J. Exp. Biol. 126. 375-387Kuterbach D.A. & Walcott B. (1986b) Iron-containing cells in the honeybee (Apismellifera). II. Accumulation during development.J. Exp. Biol. 126, 389-401Kuterbach D.A., Walcott B., Reeder R.J. &Frankel R.B. (1982) Iron-containing cells in thehoneybee (Apis mellifera). Science 218, 695-697

Locke M. & Leung H. (1984) The induction anddistribution of an insect ferritin - a new

function for the endoplasmatic reticulum.Tissue Cell 16, 739-766

Loper G.M. (1985) Influence of age on thefluctuation of iron in the oenocytes of honeybee(Apis mellifera) drones. Apidologie 16, 181-184

Martoja R. & Ballan-Dufrangais C. (1984) Theultrastructure of the digestive and excretoryorgans. In : Insect Ultrastructure 2 (R.C. King& Akai, eds), Plenum Press, pp. 199-268Mason A.Z. & Simkiss K. (1982) Sites ofmineral deposition in metaf-accumulafing cells.Exp. Cell Res. 139, 383-391

Morgan A.J. (1984) The localisation of heavymetals in the tissue of terrestrial invertebratesby electron micropobe X-ray analysis. In : .’Scanning Electron Microscopy IV (O’Hare, ed.),Chicago, pp. 1847-1865Raes H., Bohyn W. & Jacobs F. (1988) Etudede la d6toxication du plomb de fabeille (Apismellifera L.). Actes 4 Colloq. Insectes Sociaux,pp. 95-101

Raes H., Jacobs F. & Mastyn E. (1985) Apreliminary qualitative and quantitative study ofthe microscopic structure of the dorsal fat bodyin adult honeybees (Apis mellifera), including atechnique for the preparation of whole sections.Apidologie 16, 275-290Raes H., De Coster W. & Bohyn W. (1987)Light and electron microscopic study ofoenocytes in adult honeybees with particularemphasis on the fixative osmolarity. In : .’Chemistry and Biology of Social Insects (J.Eder & H. Rembold, eds.), p. 89

Simkiss K. (1979) Metal ions in cells.Endeavour NS 3, 2-6Simkiss K. (1981) Cellular discrimination pro-cesses in metal accumulating cells. J. Exp.Biol. 94, 317-327

Sohal R.S., Peters P.D. & Hall TA. (1976) Finestructure and X-ray microanalysis ofmineralised concretions in the Malpighiantubules of the house fly, Musca domestica.Tissue Cell 8, 447-458

Staubli W., Freyvogel TA. & Suter J. (1966)Structural modification of the endoplasmaticreticulum of midgut epithelial cells ofmosquitoes in relation to blood intake. J.Microsc. 5, 189-204