A

mprsta©

K

1

wbktthttItta

lfi

0d

Tissue and Cell 40 (2008) 103–112

Circulating hemocytes from larvae of the paperwasp Polistes dominulus (Hymenoptera, Vespidae)

Fabio Manfredini a,∗, Romano Dallai a, Enzo Ottaviani b

a Department of Evolutionary Biology, University of Siena, via A. Moro 2, Siena 53100, Italyb Department of Animal Biology, University of Modena and Reggio Emilia, via Campi 213/D, 41100 Modena, Italy

Received 29 August 2007; received in revised form 10 October 2007; accepted 12 October 2007Available online 11 December 2007

bstract

Circulating hemocytes from larval stages of the paper wasp Polistes dominulus were characterized by light and transmission electronicroscopy. Three types were identified: prohemocytes, plasmatocytes and granulocytes. The first two are agranular cells while the latter

resent typical cytoplasmic inclusions called granules. Plasmatocytes differ from prohemocytes being larger, showing lower nucleus/cytoplasmatio and they possess many phagolysosomes. The substantial uniformity of most subcellular features and the presence of “intermediate forms”upport the “single-cell theory” i.e., there is only one cell line that originates from the prohemocyte and leads to the granular cell passing

hrough the plasmatocyte. This hypothesis seems to be confirmed by functional tests. Indeed, most part of cells adheres to the glass and isble to phagocytize fluorescent microspheres.

2007 Elsevier Ltd. All rights reserved.

eywords: Hemocytes; Ultrastructure; Behavior; Hymenoptera

tf1oDOmotciaube

. Introduction

In an attempt to characterize the hemocytes of the paperasp Polistes dominulus larvae it might be suitable to remem-er that anyone venturing into the study of insect hemocytesnows only too well how discouraging it is in the beginningo have to contend with a bewildering array of terminologieshat exists in the literature (Gupta, 1985). More than 20 yearsave passed and the situation has even worsened because ofhe large amount of newly published data, mainly inherento the ultrastructural and functional analysis of these cells.nstead of simplifying and clarifying previous observations,his modern approach has uncovered new possible parameterso be evaluated and consequently it has further on complicatednomenclature that was itself controversial and misleading.

In this work we present a description of hemocytes fromarvae of P. dominulus (3rd and 4th instars), in the aim oflling in the gap present in the literature about this topic:

∗ Corresponding author. Tel.: +39 0577234492; fax: +39 0577234476.E-mail address: manfredini2@unisi.it (F. Manfredini).

oRdioc

040-8166/$ – see front matter © 2007 Elsevier Ltd. All rights reserved.oi:10.1016/j.tice.2007.10.003

here are scanty and outdated published data on hemocytesrom wasps or other Hymenoptera (Chauvin, 1968; Ahmad,988) and this is in great contrast with the large amountf information about other insect orders like Dictyoptera,iptera, Lepidoptera, Orthoptera (Giulianini et al., 2003).ur study is based on light microscopy, transmission electronicroscopy and foremost functional tests i.e., phagocytosis

f fluorescent beads and adhesion on glass slides. We knowoday that a proper methodological approach requires theombination of different analysis techniques when studyingnsect hemocytes. This may include analysis of thin sectionst an ultrastructural level, examination of semithin section,sing the light microscopy for hemocytes samples fixed atlood removal; observation of cells behavior in monolay-rs by means of phase contrast microscopy or SEM; employf antibody and genetic markers (Lavine and Strand, 2002;ibeiro and Brehelin, 2006). We concentrate on few specific

evelopmental stages in order to reduce the large variabil-ty due to the physiological state of both the cell and therganism which may condition the morphology of hemo-ytes (mainly at an ultrastructural level), which is another

1 ue and

apmitad

2

2

sw2fstw

bamrt

2

voaMo

2

dac3acF5paoetaaam

2

lVctTom

2

satspm

3

3

3

uTocmasptvnit

3

PppsbAab

04 F. Manfredini et al. / Tiss

spect often neglected by hematologists. For example, theattern of the cellular elements in the hemolymph of a larvaay be quite different from that present in an adult, mostly

n holometabolous insects (Meylaers et al., 2007). In addic-ion to this, we should not forget that experimental conditionslways represent a stress source for cells, therefore in vitroynamics can quite differ from in vivo behavior.

. Materials and methods

.1. Animal rearing and hemolymph collection

Polistes dominulus Christ came from Asciano, in theurroundings of Siena (Tuscany, Italy). Hibernating groupsere collected at the end of winter and arranged in0 cm × 20 cm × 20 cm Plexiglas cages in order to let themound new colonies. In the field, nests were picked up inummer and housed in a similar way. Both groups were main-ained at 15L/9D and 28 ± 2C◦; they were fed with sugar,ater and Sarcophaga sp. larvae ad libitum.Hemolymph was collected from larvae of P. dominulus

y operating a small incision on the cuticle of chilled larvaend collecting the oozing hemolymph drop with a precisionicroliter pipette. Samples were kept in ice and immediately

esuspended in Grace’s Insect Medium (Invitrogen–Gibco)o prevent hemolymph coagulation and cell impairment.

.2. Light microscopy

Hemolymph-solution samples derived from wasp lar-ae (50–100 �l for each individual) were cytocentrifugednto slides with a Shandon Instrument Cytospin II runningt 400 rpm for 2 min. Hemocytes were then stained withay Grunwald–Giemsa for morphological examination and

bserved with a Leitz Dialux 22 light microscope.

.3. Transmission electron microscopy (TEM)

For ultrastructural analysis hemolymph samples wererawn from 3rd to 4th instar larvae (20 �l × larva), poolednd directly transferred in 1.5 ml test tubes (5 larvae × tube)ontaining 0.1 M, pH 7.2 phosphate buffer (PB) to which% sucrose had been added. The material was centrifugedt 1100 rpm for 5 min in order to obtain a pellet of hemo-ytes and this operation was repeated after each passage.ixation occurred in 2.5% glutaraldehyde in PB at 4 ◦C forh and then the material was rinsed in PB several times,ost-fixed in 1% osmium tetroxide in PB at 4 ◦C for 1 hnd rinsed again in PB. After dehydration in a graded seriesf ethanol to 100% ethanol and a step in propylene oxide,mbedding in an Epon-Araldite mixture followed. Ultra-hin sections, obtained with a Reichert Ultracut II E and

LKB Ultratome III, were routinely stained (1% uranylcetate followed by 1% lead citrate) and observed with

TEM Philips CM 10 at 80 kV. All measurements areeans ± S.D.

voma

Cell 40 (2008) 103–112

.4. Adhesion test

Drops of hemolymph (∼20 �l each) from P. dominulusarvae were collected and placed on glass slides where aaseline ring was delimitated. The slides were set in a wethamber successively covered and hemocytes were made set-le for 20–30 min in presence of Grace’s Insect Medium.hereafter the slide was removed from the chamber andbserved in phase contrast with a Leitz Dialux 22 lighticroscope.

.5. Phagocytosis assay

For each P. dominulus larva 10 �l hemolymph wereampled and added to 100 �l Grace’s Insect Medium in

0.2 ml test tube previously covered. The material washen incubated with 0.1 �l of a fluorescent beads suspen-ion for 30 min in soft oscillation. After that, samples werelaced on glass slides and observed under the fluorescenceicroscope.

. Results

.1. Morphological observations

.1.1. Light microscopyThe examination of the stained hemocytes from P. domin-

lus larvae (Fig. 1A) indicates the presence of three cell types.he first (type a) (Fig. 1B) is a small cell, with the nucleusccupying most part of the cellular body and a thin basophileytoplasm which stains up blue after May Grunwald–Giemsaethod. The nuclear chromatin appears as deep red–violet

nd the cytoplasm keeps substantially round in shape. Theecond (type b) (Fig. 1B) is a larger cell, with abundant cyto-lasm softly stained in a range from pink to light grey. Alsohe nucleus appears paler than “type a” and the cell shapearies from round to irregular or elongated. In four cases weoticed a circulating hemocyte dividing by mitosis: generallyt was an infrequent process (Fig. 1C–F). A round cell withhe characteristics of a prohemocyte was also observed.

.1.2. Ultrastructural analysisThe investigation on the fine structure of hemocytes from

. dominulus larvae revealed a substantial uniformity in theresence and morphology of the principal subcellular com-onents. Apparently, this might contrast with their complextructure which shows many peculiar organelles as requiredy the numerous different tasks performed by these cells.ctually, we did not perceive striking ultrastructural diversity

mong hemocytes, even when samples came from specimenelonging to different developmental stages; at the most, a

ariability in size and shape was often observed for somerganelles, correlated with the activity of the cell at theoment of fixation. All these considerations seem to justifygeneralized description of these cells.

F. Manfredini et al. / Tissue and Cell 40 (2008) 103–112 105

F r light( s of the

3itcrtaqcg(er

3paTadait

ig. 1. (A) Circulating hemocytes from Polistes dominulus larvae stained foa and b). (C–F) Circulating hemocytes dividing by mitosis at different step

.1.2.1. The nucleus. Nuclei show round, ovoid, sometimesrregular profiles. The chromatin is generally uniformly dis-ributed and often few to many patches of electron-opaquehromatin are evident (Fig. 2A and B): the size of theseegions is extremely variable (from 0.06 to 0.7 �m diame-er) and they can be clumped in the middle of the nucleuss well as adherent to the nuclear envelope. Nucleoli are fre-uently visible (Fig. 2A and B). In some cases, perinuclearisternae are considerably distended and may show small

roups of vesicles in addition to finely dispersed materialFig. 2B). The nucleus/cytoplasm ratio is variable: in gen-ral, hemocytes may have a high value or a low value for thisatio.

3nd

microscopy observations. (B) The two cell types most frequently observedprocess.

.1.2.2. The cytoplasm. The abundance of functional cyto-lasmic organelles observed in P. dominulus hemocytes isn evidence for the intense activity performed by these cells.hey consist of standard organelles which are widely diffusedmong hemocytes and also in other cell lines because fun-amental for the cell survival; phagolysosomes, whose sizend quantity varies from cell to cell; granules i.e., peculiarnclusions which characterize some hemocytes that we callherefore “granular cells” or “granulocytes”.

.1.2.3. Standard organelles. Mitochondria are generallyumerous, variable in size (0.55 ± 0.35 �m) and with welleveloped cristae; they are moderately electron-dense but

106 F. Manfredini et al. / Tissue and Cell 40 (2008) 103–112

Fig. 2. General subcellular features of P. dominulus hemocytes. TEM. (A–B) Nuclear region. Note the nucleoli (N) and the chromatin forming electron-opaquep vesiclesd les (mtr ng-like

wralwoi

reu

atches (arrowheads). The perinuclear cisternae (pc) form small groups ofroplets. (C) Detail of a hemocyte showing microtubules organized in bundeticulum arranged in enlarged cisternae (RER) or rolled up in concentric ri

ell discernible at TEM (Fig. 2E). A large amount ofough endoplasmic reticulum is present, arranged in stacksnd characterized by enlarged cisternae filled with floccu-

ent material (Fig. 2E), mainly in hemocytes from youngasp larvae. In later stages, RER appears frequently devel-ped in long and narrow cisternae, occasionally rolled upn typical concentric ring-like figures (Fig. 2D and F). Free

oudc

(arrows) and finely dispersed material (asterisks). r: ribosomes; LD: lipid). (D–F) Cytoplasmic region. Note the well developed rough endoplasmicfigures (asterisks). m: mitochondria.

ibosomes are abundant and often set in clusters (Fig. 2A),specially in hemocytes poor in phagolysosomes or gran-les. Golgi apparatus, when evident, has the appearance

f an active organelle, with both its saccular and vesic-lar parts well discernible (Fig. 3A). Commonly, lipidroplets may be present (Fig. 4D), especially in olderells.

F. Manfredini et al. / Tissue and Cell 40 (2008) 103–112 107

F centriolo mocytes

mnaettiiaic

cgFsr(

3

ig. 3. The prohemocyte. TEM. (A–C) Details of prohemocytes showing af a prohemocyte. (E) Coated pits (cp) on the plasma membrane of a proheide of the forming vesicles.

Other frequently detected cytoplasmic elements areicrotubules (Fig. 2C), dispersed or arranged in bundles and

ormally located in close proximity to the plasma membranend then numerous coated pits (Fig. 3E), visible at differ-nt moment of the pinocytotic process and characterized byhe presence of the clatrin protein in the cytoplasmic side ofhe forming vesicles. In three cases we noticed a centriolen longitudinal section (Fig. 3A–C): these observations are

n line with the mitotic figures identified at light microscopynd confirm hemocytes ability to multiplicate while circulat-ng in the hemolymph. The three hemocytes provided withentrioles present recurrent features also observed in other

igow

e in longitudinal section (asterisks). G: Golgi apparatus. (D) Whole section: note the presence of the clatrin protein (arrowheads) on the cytoplasmic

ells and remind us of the prohemocyte, the stem cell whichives rise to the other types (Brehelin and Zachary, 1986;ranchini et al., 1996). These hemocytes present roundedhape, little size (diameter <5 �m), high nucleus/cytoplasmatio, absence of granules and few small phagolysosomesFig. 3D).

.1.2.4. Phagolysosomes. They are membrane-limited

nclusions highly irregular in shape and filled with hetero-eneous material: lamellated degraded cellular componentsr other residues (Fig. 4A, C). These vacuoles are tiedith the autophagic-digestive activity of the cell and they

108 F. Manfredini et al. / Tissue and Cell 40 (2008) 103–112

F lled wito of phag

adedltnai

3rmnptmgcsc

oR

wam(eosFnsata

ig. 4. The plasmatocyte. TEM. (A–C) Different kind of phagolysosomes fir other residues (C). (D and E) Whole agranular cells: note the abundance

ppear very numerous and of large size (up to 1.6 �miameter) in hemocytes from P. dominulus specimen at thend of the larval development. Phagolysosomes are welliffused and they are prevalent in those agranular cells i.e.,acking in granules (Fig. 4D and E) that are different fromhe prohemocytes for the size (diameter >5 �m) and theucleus/cytoplasm ratio (much lower): we refer to these cellss “plasmatocytes” for the analogy with what was observedn other insects.

.1.2.5. Granules. These membrane-limited inclusions,ather regular in shape, are quite variable in size (min 0.1 �max 0.9 �m) and electron density. Their function is still

ot clear but they seem involved in the defense reactionsrocesses of the organism against intruders: frequentlyhey were observed while releasing their content in the

edium through exocytosis (Fig. 5G). When abundant,

ranules identify a precise kind of hemocyte: the granularell (Fig. 5A and H). A primary distinction is betweentructured and unstructured or amorphous granules: theseategories were already employed in the past to characterize

eboa

h heterogeneous material: lamellated (A) degraded cellular components (B)olysosomes (asterisks) and lipid droplets (LD).

ther insect groups (Scharrer, 1972; Butt and Shields, 1996;ibeiro and Brehelin, 2006).

Structured granules present electron-opaque filamentshich are arranged in bundles stacked at different

ngles within the granule (Fig. 5C and E). The fila-ents resemble microtubular structures in cross-sections

0.0131 ± 0.0039 �m diameter) and their opacity isxtremely variable. These inclusions are prevalently circular,void or kidney-shaped and some images help to uncover theteps of their formation, as observed by Scharrer (1972) too.irst tubular elements appear loosely scattered or restrictedearby the bounding membrane (Fig. 5D); then they progres-ively fill the entire inclusion body by forming densely packedrray of bundles. In some cases the filaments do not invadehe whole granule and the central zone become occupied bygranular or filamentous matrix of variable electron density.

Within the second type of granules we include both some

lements analogous to what Moran (1971) called “opaqueodies” and other similar inclusion with different degreesf electron density. As a rule, these granules are withoutny discernable substructure and appear filled with granular-

F. Manfredini et al. / Tissue and Cell 40 (2008) 103–112 109

Fig. 5. The granulocyte. TEM. (A and H) Whole cells in different stages of growth: a young granular cell (A), with thin cytoplasm and few small granules (gr)a lusionse d the elT on withG is.

floifis

nd a mature granulocyte whose large cytoplasm shows many granular inclectron-transparent granules (B) filled with granular-flocculent material anhe structured granules: note the early steps in the formation of the inclusiranular inclusions releasing their content in the medium through exocytos

occulent material (electron-transparent granules; Fig. 5B)

r with a dense, homogeneous matrix (electron-opaque bod-es; Fig. 5F). Generally these inclusions are larger than therst type and present approximately a circular or ellipsoidhape.

bfov

(H). LD: lipid droplets. (B and F) The two types of “opaque bodies”: theectron-opaque bodies (H) filled with a dense, homogeneous matrix. (C–E)first tubular elements appearing nearby the bounding membrane (D). (G)

In appearance there are no specific patterns in the distri-

ution of the two types of granules: their frequency changesrom cell to cell and many hemocytes show both types whilethers can present only one type. Moreover there is a greatariability in morphology and degree of electron density and

110 F. Manfredini et al. / Tissue and Cell 40 (2008) 103–112

F ent thes yte, wh(

wau(par

3

3

tBipcadtsrn

da

wwucp

3

tmsw

4

ig. 6. Functional tests. (A and B) Adhesion on glass slides. Here are evidpherical and very refractive at phase contrast and the type “b” or plasmatocB, inset). (C–F) Phagocytosis of fluorescent beads.

e detected many intermediate forms between structurednd amorphous granules: occasionally, unstructured gran-les revealed “masked” microtubular elements, as Scharrer1972) had already observed. All these considerations sup-ort a strong relationship between the two types of inclusionsnd let us suppose that one type can convert into the otherapidly.

.2. Functional tests

.2.1. Adhesion on glass slidesIn monolayers, most of P. dominulus hemocytes are able

o adhere to a glass slide after 20 min incubation (Fig. 6A and). We have observed two modalities of spreading accord-

ng to as many cellular morphologies, recurrent in equivalentroportions. Some small hemocytes, very refractive in phaseontrast, keep spherical and do not spread extensively: theydhere to the substrate with thin pseudopodia and lamellipo-ia which sometimes encircle the cell. Other hemocytes, on

he contrary, larger and less refractive than the previous ones,pread rapidly and become thin cells characterized by a bidi-ectional fibroblastic extension (Fig. 6B, inset): they developumerous pseudopodia and chiefly long and wide lamellipo-

hos

two types of hemocytes yet described: the type “a” or granular cell, small,ich is larger, less refractive and shows a bidirectional fibroblastic extension

ia. Their cytoplasm varies from round to enlarged, elongatednd spindle-shaped.

Because of the size, the shape and the spreading modality,e can correlate the first variety of these adhering hemocytesith “type a” identified at light microscopy and with the gran-lar cell of TEM observations; the other hemocytes, on theontrary, seem to be analogous to optical “type b” and TEMlasmatocytes.

.2.2. Phagocytosis of fluorescent beadsMost of P. dominulus hemocytes were able to phagocy-

ize fluorescent microspheres after 30 min incubation in theedium (Fig. 6C–F). We observed both hemocytes with a

ingular bead within the cytoplasm (Fig. 6D–F) and otherith two or more (up to seven; Fig. 6C).

. Discussion

The first problem we dealt with while characterizing theemocytes of P. dominulus larvae was the difficulty to homol-gize our results with those of studies focused on otherpecies which mostly belonged to other insect orders; for-

ue and

tf(a1fw1Bddafsshteh

tbat1lhaptr(pfcscoolapa

taamrdolaamh

aoacig1ipZcaiwobe1egt

tkufielRsfctit(

scctwtpfofrgyig

F. Manfredini et al. / Tiss

unately we uncovered a certain similarity with hemocytesrom Lepidoptera which is a well studied group of insectsPrice and Ratcliffe, 1973; Brehelin et al., 1978; Brehelinnd Zachary, 1983; Butt and Shields, 1996; Ribeiro et al.,996; Ribeiro and Brehelin, 2006). We have followed a dif-erent approach with respect to the procedure adopted in lastorks on similar matter (for example, Kaaya and Ratcliffe,982; Hillyer and Christensen, 2002; Giulianini et al., 2003;rayner et al., 2005). These Authors normally identify 4–7ifferent cellular types (variables depending whether theyecide to include also prohemocytes and adipocytes or not)nd they accurately describe them one by one chasing theollowing order: general morphological observations, ultra-tructural analysis, functional assays. What’s more, thesetudies usually support the “multiple-cell theory” aboutematopoiesis and hemocyte differentiation, which suggesthe existence of separate immutable cell lines, each differ-ntiating from a single germinal stem, that give rise to theemocyte types (Akai and Sato, 1973; Gupta, 1985).

In the case of P. dominulus larvae, we recognize no morehan three types of hemocytes (we do not consider adipocytes)ut, due to the substantial uniformity in external morphologynd to the strong repetitiveness of most subcellular struc-ures, we feel much closer to a “single-cell” theory (Scharrer,972). This is a more unitarian interpretation of the cellu-ar elements of hemolymph which suppose that the variousemocyte types are merely stages, with separate functions, ofsingle cell line derived from a unique germinal stem i.e., therohemocyte. This is in line with the great functional versa-ility of these complex and highly specialized cells which isequired in order to achieve to so a disparate multitude of tasksRowley and Ratcliffe, 1981; Gotz and Boman, 1985): a widehysiological flexibility is necessary to undergo ready trans-ormation in response to environmental stimuli. The single-ell theory relies on the presence of transitional stages ofome hemocyte types and on the existence of only prohemo-ytes and plasmatocytes in tissue culture and hematopoieticrgans (Lavine and Strand, 2002). A more recent reviewn insect immunorecognition (Ottaviani, 2005) supports aimited number of hemocytes involved in defense reactionsgainst invading organisms suggesting that, apart from therohemocytes, probably two immunocytes (plasmatocytesnd granular cells) appear to play a role in immune functions.

In our opinion there is only one feature that marks dis-inctively a precise hemocyte variety: the granules, whichre a prerogative of granular cells. The other two types aregranular cells and they differ each other for the size (prohe-ocytes are evidently smaller) and for the nucleus/cytoplasm

atio which is much lower in plasmatocytes. In the case of P.ominulus no other feature seems to justify the identificationf further hemocyte types and this is supported by first cyto-ogical investigations on honey bees (Chauvin, 1968), which

re other hymenopteran insects presumably close to the waspss regards to the cellular components of the hemolymph. Forany reasons we consider the small, agranular cells as pro-

emocytes: the size first of all, the thin cytoplasm which is

dOfs

Cell 40 (2008) 103–112 111

proof for a cell that is beginning its growth, the absencef phagolysosomes or granules, whose increase in numbernd size is a consequence of the activity of the cell, theentrioles (three cases observed). Despite the difficulty tondividuate them, which is in line with the low mitotic indexenerally recorded in other insects (Rowley and Ratcliffe,981), these are extremely important organelles since theyndicate that the cell has mitotic capacity (this confirms therohemocyte, able of dividing as stem cell, see Brehelin andachary, 1986; Franchini et al., 1996) and because they are alear sign of hemocytes ability to separate in circle i.e., whenlready circulating in the hemolymph. One should not taket for granted because still not long time ago it was not clearhether maintenance of hemocyte numbers depended mainlyn hematopoiesis or on in circle division. Now we know thatoth processes probably operate, depending on the speciesxamined and the stage in the life cycle (Rowley and Ratcliffe,981; Nappi and Carton, 1986; Lavine and Strand, 2002)ven though it has been demonstrated that hemocytes canrow and multiply in vitro for an indefinite period, withouthe involvement of hematopoietic organs (Gupta, 1985).

There is a certain agreement about the origin and func-ion of the typical granular cells inclusions, although accuratenowledges about them are still scanty nowadays. Gran-les seem to be synthesized by the Golgi apparatus and thenal stage of their maturation is apparently the structureless,lectron-dense body containing acid mucopolysaccharide-ike substances (Rowley and Ratcliffe, 1981; Gupta, 1985;ibeiro and Brehelin, 2006). With reference to their function,

everal observations suggest these inclusions are the storageorm of the clotting protein, which reacts with the plasmaoagulogen (a lipophorin) when released into the hemolympho form the clot; moreover, it has been supposed that thesenclusions contains phenol oxidase, an enzyme implicated inhe process of melanization and in the tanning of the cuticleMoran, 1971; Scharrer, 1972; Butt and Shields, 1996).

The substantial uniformity in the principal subcellulartructures of P. dominulus hemocytes makes it difficult tolassify them as distinct cellular types: these forms show aommon history and probably it might be more correct to seehem as functional states of a unique cellular line, which beginith the prohemocyte and leads to the granular cell passing

hrough the plasmatocyte. As we know from literature, a mainroof to sustain the “single-cell” theory are the “intermediateorms” i.e., hemocytes whose morphology do not satisfy anyf these categories completely, because of an overlapping ofeatures. In our observations, we have evidence of many rep-esentatives of this cell group. Hemocytes with very few littleranules are no more plasmatocytes (= agranular cells) but notet typical granular cells: they are transient hemocytes, evolv-ng from one form to the other and we suppose they will pro-ressively enrich their cytoplasm with granular inclusions,

ependently to the activity of both the cell and the organism.n the other side, the controversial size and the presence of

ew small phagolysosomes make it difficult to assert whetherome particular cells are prohemocytes or plasmatocytes. In

1 ue and

tsatatcatodm

opictot1plfdtasiswpioacfiiconf

A

f

R

A

A

B

B

B

B

B

C

F

G

G

G

H

K

L

M

M

N

OP

R

R

Rdemic Press, London, pp. 421–488.

Scharrer, B., 1972. Cytophysiological features of hemocytes in cockroaches.Z. Zellforsch. 129, 301–319.

12 F. Manfredini et al. / Tiss

his case functional tests are helpful since they uncover thetrong ability of plasmatocytes of adhering on glass slidesnd phagocytizing fluorescent beads, which are typical func-ions of a macrophage (Ottaviani, 2005). These observationsre strongly supported by the analysis of Gupta (1985) onhe evolution of hemocyte types: in taxa were only prohemo-ytes, plasmatocytes and granular cells are present, first twore evanescent stages and are not discernible as types whilehe differentiation after granular cells (i.e., spherule cells,enocytoids, coagulocytes, etc.) is generally accompanied byistinct prohemocytes and plasmatocytes. In addition, in theore highly evolved taxa any of the types may be suppressed.Hematopoiesis and hemocytes differentiation are still

pen questions since if we suppose that hematopoietic organsroduce just prohemocytes and only prohemocytes can dividen circle, we should accept as a consequence that the hemo-yte load of the organism is maintained exclusively by thesewo processes. We must also consider that hemocytes turnver is quite rapid referring to the autoradiographic studies onhe lepidopteron Galleria mellonella (Rowley and Ratcliffe,981): these Authors counted 3 days for differentiation (fromrohemocyte to granular cell) and other 3 days to completeife cycle under in vitro conditions, which can quite differrom the situation in vivo. The only consideration that we mayo after our observation is that during P. dominulus life spanhere must be a change in the mechanism of hematopoiesist a certain point, probably at the end of larval development,ince we found a much lower hemocyte load in the adults thann 3rd/4th instar larvae: this trend is in line with the analy-is about the relationship between hemocyte load and age oforkers in honey bees (Chauvin, 1968). From an ecologicalerspective, we might explain this disparity thinking on themmune role of hemocytes and highlighting the importancef fitness maintenance in the whole organism. Adult waspsre provided with a toughened cuticle which is itself an effi-ient barrier against parasites while larvae are an easier targetor invading organisms and must rely on a more competentmmune system to survive: this is the reason why they mustnvest more energetic resources to maintain a greater hemo-yte load. Moreover, from the ‘ecological immunology’ pointf view (Siva-Jothy et al., 2005) contrary to larvae adultseed to find a balance between the application of resourcesor survival and reproduction (Meylaers et al., 2007).

cknowledgments

The authors are grateful to Professor Antonella Franchinior her technical assistance.

eferences

kai, H., Sato, S., 1973. Ultrastructure of the larval haemocytes of thesilkworm, Bombyx mori (L) (Lepidoptera, Bombycidae). Int. J. InsectMorphol. Embryol. 2, 207–231.

S

Cell 40 (2008) 103–112

hmad, A., 1988. Free haemocytes in adult Polistes hebroeus Fabr.(Hymenoptera: Vespidae). J. Ent. Res. 12, 28–35.

rayner, F.A., Araujo, H.R.C., Cavalcanti, M.G.S., Alves, L.C., Peixoto,C.A., 2005. Ultrastructural characterization of the hemocytes of Culexquinquefasciatus (Diptera: Culicidae). Micron 36, 359–367.

rehelin, M., Zachary, D., 1983. About insect plasmatocytes and granularcells. Dev. Comp. Immunol. 7, 683–686.

rehelin, M., Zachary, D., Hoffmann, J.A., 1978. A comparative ultrastruc-tural study of blood cells from nine insect orders. Cell Tissue Res. 195,45–57.

rehelin, M., Zachary, D., 1986. Insect Haemocytes: A New Classification toRule Out the Controversy. Immunity in Invertebrates. Springer-Verlag,Berlin, pp. 36–48.

utt, T.M., Shields, K.S., 1996. The structure and behavior of gypsy moth(Lymantria dispar) hemocytes. J. Invertebr. Pathol. 68, 1–14.

hauvin, R., 1968. La cytologie de l’hemolymphe (Lensky). Traite de zoolo-gie de l’abeille 1, 316–318.

ranchini, A., Miyan, J.A., Ottaviani, E., 1996. Induction of ACTH-and TNF-�-like molecules in the hemocytes of Calliphora vomitoria(Insecta, Diptera). Tissue Cell 28, 587–592.

iulianini, P.G., Bertolo, F., Battistella, S., Amirante, A.G., 2003. Ultrastruc-ture of the hemocytes of Chetonischema aeruginosa larvae (Coleoptera,Scarabaeidae): involvement of both granulocytes and oenocytoids in invivo phagocytosis. Tissue Cell 35, 243–251.

otz, P., Boman, H.G., 1985. Insect Immunity. Comprehensive Insect Phys-iology, Biochemistry and Pharmacology. Pergamon Press, Oxford, pp.453–485.

upta, A.P., 1985. Cellular Elements in the Hemolymph. ComprehensiveInsect Physiology, Biochemistry and Pharmacology. Pergamon Press,Oxford, pp. 400–451.

illyer, J.F., Christensen, B.M., 2002. Characterization of hemocytes fromthe yellow fever mosquito, Aedes aegypti. Histochem. Cell Biol. 117,431–440.

aaya, G.P., Ratcliffe, N.A., 1982. Comparative study of hemocytes andassociated cells of some medically important dipterans. J. Morphol. 173(3), 351–365.

avine, M.D., Strand, M.R., 2002. Insect hemocytes and their role in immu-nity. Insect Biochem. Mol. Biol. 32, 1295–1309.

eylaers, K., Freitak, D., Schoofs, L., 2007. Immunocompetence of Galle-ria mellonella: sex- and stage-specific differences and the physiologicalcost of mounting an immune response during metamorphosis. J. InsectPhysiol. 53, 146–156.

oran, D.T., 1971. The fine structure of cockroach blood cells. Tissue Cell3, 413–422.

appi, A.J., Carton, Y., 1986. Cellular Immune Responses and Their GeneticAspects in Drosophila. Immunity in Invertebrates. Springer-Verlag,Berlin, pp. 171–187.

ttaviani, E., 2005. Insect immunorecognition. Inv. Surv. J. 2, 142–151.rice, C.D., Ratcliffe, N.A., 1973. A reappraisal of insect haemocyte classifi-

cation by examination of blood from fifteen insect orders. Z. Zellforsch.147, 537–549.

ibeiro, C., Simoes, N., Brehelin, M., 1996. Insect Immunity: the haemo-cytes of the Armyworm Mythimna unipuncta (Lepidoptera: Noctuidae)and their role in defence reactions. In vivo and in vitro studies. J. InsectPhysiol. 42, 815–822.

ibeiro, C., Brehelin, M., 2006. Insect hemocytes: what type of cell is that?J. Insect Physiol. 52, 417–429.

owley, A.F., Ratcliffe, N.A., 1981. Insects. Invertebrate Blood Cells. Aca-

iva-Jothy, M.T., Moret, Y., Rolff, J., 2005. Insect immunity: An evolution-ary ecology perspective. Adv. Insect Physiol. 32, 1–48.