Hirudo medicinalis: a new model for testing activators and inhibitors of angiogenesis

Original article

Hirudo medicinalis: A new model for testing activators and inhibitors of angiogenesis

M. de Eguileor1, A. Grimaldi1, G. Tettamanti1, R. Ferrarese1, T. Congiu2, M. Protasoni2, G. Perletti1, R.Valvassori1 & G. Lanzavecchia11Department of Structural and Functional Biology and 2Department of Clinical and Biological Science, University ofInsubria, Varese, Italy

Received 26 January 2001; accepted in revised form 5 December 2001

Key words: angiogenesis, angiogenesis inhibitors, ECM, grafts, growth factors, Hirudo medicinalis

Abstract

An increasing body of evidence indicates that in the leech Hirudo medicinalis the angiogenic process is finelyregulated and coordinated by the botryoidal tissue. In this paper we provide evidence on the involvement ofbotryoidal tissue cells in angiogenesis induced in H. medicinalis by a variety of stimuli including surgical wounds orthe administration of modulators of neovascularization. Interestingly, we show that either human activators ofvascular cell growth, or anti-angiogenic peptides like angiostatin and endostatin, or the drug mitomycin, can inducea prompt biological response in H. medicinalis. We show as well that angiogenesis in this invertebrate shares asurprising degree of similarity with neovascularization in vertebrates, both at the biochemical and cellular levels,because it involves similar growth factors/growth factor receptors, and relies on analogous cell–cell or cell–matrixinteractions. For these reasons we suggest that H. medicinalis can be used as a reproducible model for testingactivators or inhibitors of angiogenesis, and for investigating the biochemical, ultrastructural and cellular processesinvolved in new vessel formation.

Introduction

The leech Hirudo medicinalis (Annelida, Hirudinea) is aninvertebrate characterized by a relative anatomicalsimplicity. The body of Hirudo consists of a musculo-cutaneous sac which contains several organs embeddedin a loose connective tissue [1, 2]. Hirudo medicinalis ischaracterized by the absence of a vascular system withinthe muscular body wall of the animal, and by thepresence of the botryoidal tissue, located within theparenchyma between the gut and the body wall [3].Besides being involved in angiogenesis, this coelothe-lium-derived tissue displays myelo/erythroid and storagefunctions [4, 5]. It is composed by clusters of roundish/oval multifunctional botryoidal cells [5], as well as bysmaller, flattened, endothelial-like cells [3, 6]. We haverecently demonstrated that the botryoidal tissue plays amajor role in angiogenesis in this animal model [6]. Thistissue can change its shape from a solid cord of cells to atubular, pre-vascular structure through a dehiscenceprocess. In this system, remodeling is characterized bymarked cellular changes, i.e. by flattening, lengthening

and stretching of both botryoidal and endothelial-likecells.

Besides presenting further evidence of the involvementof Hirudo botryoidal tissue cells in angiogenesis inducedby a variety of biochemical and surgical stimuli, weshow that either human activators of vascular cellgrowth or anti-angiogenic compounds can induce aprompt biological response in Hirudo. Due to themarked similarities with the biochemical and morpho-logical features of neovascularization in vertebrates, thework described in this paper suggests that adult Hirudomay be a useful model for studying at different levels theangiogenic and the vasculogenic processes.

Materials and methods

Animals and treatments

Leeches (Hirudo medicinalis, Annelida, Hirudinea, fromRicarimpex, Eysines, France) measuring about 10 �1:00 cm were kept in water at 22–23 �C in aerated tanks,and were fed weekly with calf liver. Before eachexperiment, leeches were starved for four weeks.

Sixty animals were randomized into four groups,which were treated as follows:

Group (1): untreated, control leeches.

Correspondence to: Magda de Eguileor, Department of Structural and

Functional Biology, University of Insubria, Via J.H. Dunant 3, 21100

Varese, Italy. Tel: +39-0332-421310; Fax: +39-0332-421300; E-mail:

[email protected]

Angiogenesis 4: 299–312, 2001. 299� 2002 Kluwer Academic Publishers. Printed in the Netherlands.

Group (2): leeches injected with three different solublevascular growth factors:

• group (2a), vascular endothelial growth factor(VEGF) (Pepro Tech, London). Ten microliters of a5 ng/ll solution in phosphate buffered saline (PBS,pH 7.4) were administered as a single dose by injec-tion into the body wall of the animal;

• group (2b), basic fibroblast growth factor (bFGF)(Pepro Tech, London). The total volume of 10 ll(5 ng/ll bFGF dissolved in PBS, pH 7.4) was injectedas a single dose;

• group (2c) granulocyte–macrophage-colony stimu-lating factor (GM-CSF) (Biogenesis, Kingston, NewHampshire). Ten microliters were injected as in theprevious experimental groups (concentration of thedrug: 5 ng/ll dissolved in PBS, pH 7.4). GM-CSFwas chosen because it was shown to be involved invessel growth during wound repair processes [7].

• A subgroup of control leeches was injected with 10 llof vehicle only (PBS, pH 7.4).

Group (3): leeches subjected to lesions consisting of atissue explant (about 2� 2� 2 mm) affecting the entirebody wall. The tissue explant was surgically removedwith microdissecting scissors. Within this group, threeleeches were used to monitor cell proliferation at theexplant site using Bromodeoxyuridine (BrdU) labelingtechniques [8].

Group (4): leeches treated with angiogenesis inhibitorsafter surgical induction of neovascularization. As ingroup (3), leeches were subjected to lesions consisting ofa tissue explant (about 2� 2� 2 mm) affecting theentire body wall. Immediately after surgery, angiostatin(group 4a), endostatin (group 4b) and mitomycin (group4c) were administered to groups of six lesioned animals.Angiostatin treatment (group 4a) was performed byintramuscular injection of 200 ng protein dissolved in10 ll PBS. Three injections were performed, allowing4 h spans between treatments. Recombinant yeast en-dostatin (group 5b) was administered i.m. (total vol-ume: 10 ll) at the concentration of 30 lg proteindissolved in 10 ll PBS. Endostatin solution was injectedthree times. All injections were performed in thesurgically-wounded area, allowing intervals of 4 h be-tween treatments.

Within the group 4a, three leeches were used tomonitor cell proliferation at the explant site by BrdUlabeling technique. A subgroup of six control leecheswas injected for three times (4 h between treatments)with 10 ll PBS (vehicle).

Mitomycin treatment (group 4c) (Kyowa HaccoKogyo, Tokyo, Japan) was performed by topicallyadministering 10 ll of a 0.02% solution of the drug inPBS. Each leech was treated twice, allowing a 4 h spanbetween treatments, according to routine ophtalmicsurgical protocols. During treatment, animals were keptin moist chambers, to avoid dispersion in tank water oftopically administered drugs.

A subgroup of three control, surgically-stimulatedleeches was topically treated with 10 ll PBS (vehicle).

Surgical lesions and growth factor injections wereperformed at the distal portion of the animal, about 2/3from the oral extremity (at about the 80th superficialmetamere).

Before surgical procedures, treatments and fixation,leeches from all groups were anesthetized with asaturated solution of mephenesin (3-o-toloxy-1,2-pro-panediol).

Light and electron microscopy

Leeches were dissected and fixed for 2 h in 0.1 Mcacodylate buffer, pH 7.2, containing 2% glutaralde-hyde. Specimens were then washed in the same bufferand postfixed for 2 h with 1% osmic acid in cacodylatebuffer, pH 7.2. After a standard step of serial ethanoldehydration, specimens were embedded in an Epon-Araldite 812 mixture. Sections were obtained with aReichert Ultracut S ultratome (Leica, Wien, Austria).Semithin sections were stained by conventional methods(crystal violet and basic fuchsin) according to Mooreet al. [9], and subsequently observed at a light micro-scope (Olympus, Tokyo, Japan). Thin sections werestained by uranyl acetate and lead citrate and observedwith a Jeol 1010 EX electron microscope (Jeol, Tokyo,Japan).

5-Bromodeoxyuridine (BrdU) labeling

Cell proliferation was assayed in treated leeches bymonitoring the incorporation of the substituted nucleo-tide 5-BrdU into the DNA of S-phase cells. Leecheswere immersed in 0.05% BrdU dissolved in tap waterfor 1, 2 or 3 h. After BrdU incubation, leeches werewashed in clean running tap water and then prepared forelectron microscopy using routine techniques [8]. Semi-thin sections were exposed for 2 min to a resin-removingmixture (2 g KOH in 5 ml propylene oxide and 10 mlmethanol), rinsed with methanol and placed in PBS, pH7.3. Sections were subsequently incubated overnight at4 �C with an anti-BrdU monoclonal antiserum diluted1:60 in PBS. After a 20-min preliminary exposure to0.3% H2O2 in PBS – performed to exclude the activityof endogenous peroxidase – sections were incubated for3 h with a peroxidase-conjugated anti-mouse IgG(1:100) at room temperature. After four washes inPBS, sections were incubated with a solution containing0.05% 3,3¢-diaminobenzidine and 0.03% H2O2 in PBS.Sections were then rinsed in distilled water. Controlsections were prepared by omitting the primary anti-body from the process.

Immunocytochemistry

Anaesthetised leeches from groups 1 and 2 weredissected in a cold Ringer solution [10] into smallblocks, which were immediately embedded in polyfreeze

300 M. de Eguileor et al.

cryostat embedding medium (OCT) (Polyscience Eu-rope, Eppelheim, Germany), and stored in liquid nitro-gen according to Geiger et al. [11]. Cryosections (10 lmthick) of unfixed leeches were obtained with a Reichert-Jung Frigocut 2800; slides were immediately used orstored at �20 �C. Sections were incubated for 15 minwith a 0.05 M solution of NH4Cl to inhibit autofluo-rescence, washed with PBS, and incubated for 30 minwith primary antibodies/antisera at the following dilu-tions: anti-human CD34 monoclonal antibody (cloneQBEnd/10, diluted 1:50, Novocastra Laboratories,Newcastle upon Tyne, UK), anti-human integrin aVb3monoclonal antibody (1:100 dilution, Chemicon Inter-national, Temecula, California), anti-human vascularendothelial-cadherin (VE-cadherin) polyclonal antibody(1:500 dilution, Santa Cruz Biothecnology, Santa Cruz,California), anti-human fibronectin polyclonal antibody(1:200 dilution, Sigma, Italy), anti-human VEGFR-1(flt-1) receptor for VEGF, polyclonal antibody (1:100dilution, Upstate Biotechnology, Lake Placid, NewYork), anti-human VEGFR-2 (flk/KDR) monoclonalantibody (1:40 dilution, Sigma, Italy), anti-humanbFGF receptor 1 antibody (1:20 dilution, Calbiochem,Cambridge, Massachusetts). After incubation with theprimary antibody, specimens were washed and incu-bated with appropriate secondary antibodies conjugatedwith tetramethylrhodamine or fluorescein isothiocya-nate (FITC) (1:50 dilution, Jackson, Immuno ResearchLaboratories, West Grove, Pennsylvania). Incubationswere performed for 1 h in a dark moist chamber. ThePBS buffer used for washing steps and antibodydilutions contained 2% bovine serum albumin (BSA).In control samples, primary antibodies were omitted,and sections were treated with BSA-containing PBS.

Coverslips were mounted in Vectashield mountingmedium for fluorescence (Vector Laboratories, Burlin-game, California); slides were examined with a confocallaser microscope (laser 568 nm for rhodamine, laser492 nm for fluoresceine; MRC 1024, Bio-Rad Labora-tories, Hemel Hempstead, UK) using �40 and �63objectives (NA 1.30, 1.25). Confocal images weresuperimposed using the Photoshop 5.0 program: fluo-rescent images were then overlayed onto transmissionimages showing the corresponding tissue sections.

Quantitative evaluation of angiogenesis

Angiogenesis was evaluated by calculating the vascularsurface density [12] in semithin sections of treated andcontrol leeches. An IMAGE-PRO plus (IPP) imageanalysis system (Media-cybernetics, Silver Spring,Maryland) was used for unbiased surface estimation ofthe vascular system, characterized by an evident longi-tudinal anisotropy, according to the stereological tech-niques described by Weibel [13], and by Cruz-Orive andWeibel [14]. A series of six semithin cross sections (1 lmthick) of the body wall were taken at increasingdistances [0.4, 0.8, 1.2, 1.6, 2.0 and 15 mm (as control)]

starting from the center of the surgical lesions, or fromthe exact site of vascular growth factor injection. Thedistances were measured along the longitudinal axis ofeach animal. For each section we have randomly ana-lyzed four sample body wall areas of 250 · 250 lm2

(62500 lm2) and for each point we have analyzed fiveseriate sections (in all 20 areas).

Results

Anatomical features of untreated H. medicinalis bodywall

Control, untreated animals [group (1)] displayed thenormal anatomical features of the body wall of adultH. medicinalis, which is mainly composed by tightly-packed bundles of muscle fibers embedded in a connec-tive tissue, ‘wrapped’ in a monolayer of epithelial cells(Figures 1 and 2). It is crucial to underline that bloodvessels are virtually absent in the body wall of normal,untreated adult animals (Figures 1 and 2), since theHirudo circulatory system is mainly composed by fewlongitudinal vessels located between the gut and thebody wall (Figure 1). Within this system, two contractilelateral longitudinal vessels assure unidirectional bloodflow, while dorsal and ventral longitudinal vessels areresponsible for distribution to different organs. Vesselscontain blood carrying out gaseous exchange via respi-ratory pigments, and transporting nutrients, excreta andcells with immune defence activity [2]. The cross-sectional drawing in Figure 1 summarizes the majoranatomical features of H. medicinalis.

The botryoidal tissue was found to be localized withinthe loose connective tissue between the thick musclelayer and the gut (Figures 1 and 2) and consisted ofcompact cords, containing clusters of large, oval gran-ular cells concealing smaller, flattened endothelial-likecells (Figure 3).

Angiogenic effect of surgical lesions in H. medicinalis

The botryoidal tissue of animals subjected to surgicalexplants was characterized by a standard responsepattern. After a short time (3–6 h) from surgery, thebotryoidal tissue underwent a marked structuralmodification characterized by the loss of its compactcord-like structure (Figure 4), accompanied by gradualdecrease of cell–cell contacts, and by tapering andthinning of granular botryoidal cells (Figures 4 and 5).The consequence of progressive cell scattering was theformation of the new vessel cavity, consisting of acentral lumen lined by botryoidal tissue cells (Figures 4and 5). Immunodetection of BrdU in the nuclei ofbotryoidal tissue cells indicated that the formation ofnew vessels involved extensive cellular proliferation(Figure 6).

These structural modifications included the emergencewithin the botryoidal tissue of a second cell type

Hirudo as a new in vivo model of angiogenesis 301

(referred to as: endothelial-like cells) characterized bysmaller size, flattened shape, and by an agranular,electrondense cytoplasm (Figure 7). Large, granularbotryoidal cells were tightly connected through desmo-some-like junctions to these flattened, electrondense cells(Figure 8). These junctions were also observed betweenadjacent endothelial-like cells (Figure 9), which posi-tively reacted to antibodies against the endothelialmarkers CD34 (Figures 10a–d) and VE-cadherin (Fig-ures 11a–d).

After 8–12 h from surgery the newly-forming vessel,consisting of a lumen surrounded by botryoidal andendothelial-like cells, was organized into finger-likestructures, ‘stretching’ toward the surgical lesion (Fig-ure 12). Immunocytochemistry revealed that, when

compared to quiescent control botryoidal tissue cells(group (1)), endothelial-like cells (undergoing activevasculogenesis/angiogenesis in a surgically stimulatedregion of Hirudo) upregulated three different kinds ofreceptors, which reacted with anti-human VEGFR-1/flt-1 (Figures 13a–c), with anti-human VEGFR-2/KDR(Figures 14a, b), and with anti-human bFGF receptor 1(Figures 15a, b) antibodies. Similar staining patternswere observed at later times (24 h).

In subsequent stages of new vessel formation, endo-thelial-like cells, lying at opposite sides in the vascularlumen, were observed to form transluminal pillars andwalls, leading ultimately to vessel splitting and branch-ing (Figures 16–18). Concurrently, several small por-tions of the botryoidal tissue were cut off and separated

Figure 1. Drawing representing the cross-sectioned body of H. medicinalis. The body wall is made of an epithelium (E) wrapping a thick muscular

wall in which fiber layers are oriented circularly (C), obliquely and longitudinally (L). Longitudinal muscle fibers are organized in fields by

dorsoventral muscles (DV). Under the compact muscle wall, blood vessels (v) are located dorsally, ventrally and laterally. The botryoidal tissue

(B), surrounded by loose connective tissue, is located between the body wall and the gut (G).

Figure 2. Semithin cross section of the body of an unlesioned H. medicinalis used as control animal. The external epithelium (E), as well as

circular (C), oblique (O) and longitudinal (L) layers of muscles can be distinguished. Longitudinal muscle fields are tightly packed and separated

only by dorsoventral (DV) muscle fibers. The botryoidal tissue (B) is localized between the muscle layers and the gut (G). Scale bar, 250 lm.

Figure 3. Semithin section of H. medicinalis botryoidal tissue, composed by large clustered cells with granule-filled cytoplasm. Several botryoidal

cells are outlined. Scale bar, 10 lm.

Figures 4 and 5. Semithin and thin cross sections of the botryoidal tissue of leeches subjected to surgical explants. After 3–6 h from surgery, cells

loss their cord-like structure, visible in Figure 3, and lined a new cavity (c) via a dehiscence process. Scale bars, 10 lm (Figure 4), 2 lm (Figure 5).

Figure 6. Semithin cross section of the botryoidal tissue of a leech subjected to surgical stimulation. Stimulated cells have incorporated BrdU

(arrowheads), indicating DNA replication. c: cavity. Scale bar, 10 lm.

Figure 7. Semithin cross section of the botryoidal tissue of a leech subjected to surgical stimulation. The botryoidal tissue acquires a lumen and

the small cavity (c) is lined not only by granular botryoidal cells (B) but also by flattened endothelial-like cells (E). Scale bar, 0.3 lm.

Figures 8 and 9. TEM. Detail of a contact area between a botryoidal cell (B) and an endothelial-like cell (E) (Figure 8), and between two adjacent

endothelial-like cells (Figure 9). In both cases adjacent cells are linked by desmosome-like junctions (arrowheads). c: vessel cavity. Scale bars,

0.3 lm.

c


from the part of the same tissue involved in theangiogenic process by interposed, newly-synthesizedextracellular matrix (ECM) (Figures 19 and 20). Theseevents were previously shown to be accompanied bymassive secretion of iron granules (contained in the

cytoplasm of botryoidal cells) in the lumen of newly-forming vessels [6].

About 24 h after surgery, an extensive network of newvessels, spanning throughout the formerly avascularmuscle tissue – and ultimately throughout the entire


body wall – was evident in the explant area of all treatedleeches (Figures 21–23).

The angiogenic process is controlled by a finelyregulated cross-talk of specific adhesion molecules,which are not only involved in cell–cell interactionsbut also in cell–ECM interactions. During this process,ECM components have been shown to be frequentlyupregulated [15–19]. In particular, the involvement ofaVb3, an integrin which binds ligands like vitronectinand fibronectin, has been clearly documented [16, 20,21]. In our experiments, aVb3 was specifically over-expressed on the cell surface of endothelial cells inHirudo growing vessels (Figures 24a, b). This effect wasparalleled by increased expression of fibronectin inlesioned leeches (Figure 25), compared to untreatedcontrols (Figure 26).

Angiogenic effect of VEGF, bFGF and GM-CSF inHirudo

When unlesioned leeches were injected with recombi-nant human VEGF (group (2)a), extensive angiogenesisoccurred in treated areas (Figures 27 and 28). Theangiogenic process was seen to involve the same celltypes (i.e. granular botryoidal cells and endothelial-likecells) and the same sequence of events which wereobserved in surgically-lesioned leeches. The angiogenicprocess was characterized by botryoidal tissue dehi-scence (Figure 27) and by the modification of the cellshape, thus lining the lumen of new vessels. The processinvolved extensive proliferation and migration of endo-thelial-like cells, which ultimately shaped the boundariesof the entire new vessel (Figure 27). A similar effect wasobserved when leeches were treated with recombinanthuman bFGF (group (2)b) (Figure 29).

Immunocytochemistry revealed on the surface of allgrowth factor-stimulated endothelial-like cells the pres-ence of VEGF receptor-1 (VEGFR-1/flt-1), VEGFreceptor-2 (VEGFR-2/KDR) and bFGF receptor 1(data not shown), as described in surgically-lesionedleeches (Figures 13–15).

Angiogenesis induced in unlesioned leeches by injec-tion of both VEGF (Figure 28) and bFGF (Figure 29)

was more marked when compared to treatment withrecombinant GM-CSF (Figure 30) [(group (2)c)]. Inter-estingly, the endothelial cell marker integrin aVb3 wasoverexpressed during GM-CSF-induced neovasculariza-tion (Figure 31). Angiogenesis was not observed incontrol leeches injected with vehicle (PBS).

Effect of anti-angiogenic agents on neovascularizationinduced in H. medicinalis by a surgical stimulus

Leeches within group (4) were subjected to surgicalstimulation of angiogenesis (as in group (3)) andsubsequently treated with the anti-angiogenic com-pounds angiostatin, endostatin and mitomycin. Admin-istration of recombinant angiostatin markedly inhibitedthe formation of new vessels (Figure 32). Althoughpositive staining for BrdU suggested that botryoidaltissue cells were in S-phase (Figure 33), botryoidal andendothelial cells were essentially unable to form newvessels, and the dehiscence process was limited andincomplete (Figure 34). About 24 h after treatment newvessels were scarce (Figure 32), when compared to eithersurgically-lesioned leeches, or leeches treated withVEGF, bFGF and GM-CSF (Figures 21, 22, 28–30).Recombinant yeast endostatin was also effective ininhibiting angiogenesis. However, endostatin was lesseffective than angiostatin, as few smaller vessels couldstill be observed within the muscle tissue (Figure 35).Analogous results were obtained with mitomycin treat-ment. Interestingly, when compared to controls, mito-mycin treatment markedly affected the intracellularorganization of fibroblasts (Figures 36–38) and, as aconsequence, the production of collagen.

Figure 39 shows the quantitation of angiogenesisin leeches subjected to surgical stimulation and tosubsequent treatment with vehicle (panel a), angiosta-tin (panel b), endostatin (panel c) or mitomycin (panel d).

Discussion

In this paper we provide evidence that the leechH. medicinalis could be a novel in vivo model for studyingangiogenesis.

Figure 10. Cryosections of the botryoidal tissue of leeches subjected to surgical stimulation. The CD34 antigen, specifically expressed on vascular

endothelium, is visible on flattened endothelial-like cells (arrowheads) interposed among botryoidal cells (B) (panels a, b) (see also Figure 7 to

understand the position of botryoidal tissue cells). The positively fluorescent images have been obtained by overlying the fluorescence images onto

transmission images. Negative controls are shown in panels c and d, transmission images are separated from negative black controls showing

absence of fluorescence. Pictures were taken 6 h after surgery. c: vessel cavity.

Figure 11. Cryosections of the botryoidal tissue of leeches subjected to surgical stimulation. Immunofluorescence staining of VE-cadherin in

dehiscent botryoidal tissue. VE-cadherin, typical of endothelial cells, is expressed by endothelial-like cells lining the new vessel cavity (c) (panel a),

and by endothelial-like cells transversally dividing the lumen (panel b). Pictures were taken 6 h after surgery. Panel c, d: control.

Figure 12. Semithin section. After 6 h from explant, a finger-like pre-vascular structure (arrowheads) delimited by endothelial-like cells (E)

sprouting from dehiscent botryoidal cells (B), is visible. c: vessel cavity. Scale bar, 10 lm.

Figure 13. Cryosections. Botryoidal tissue from surgically stimulated leeches, immunostained with an antibody against the VEGFR-1/flt-1

(panel a). Endothelial-like cells show marked immunoreactivity. Panels b,c: controls. B: botryoidal cell; c: vessel cavity.

Figure 14. Cryosections. Botryoidal tissue from surgically stimulated leeches, immunostained with an antibody against the VEGFR-2/KDR

(panel a). Endothelial-like cells show marked immunoreactivity. Panel b: control. c: vessel cavity.

Figure 15. Cryosections. Botryoidal tissue from surgically stimulated leeches, immunostained with an antibody against the bFGF receptor 1

(panel a). Endothelial-like cells and cells in the vessel cavity (c) show marked immunoreactivity (panel a). Panel b: control.

c


First, we have observed a striking similarity betweennew vessel formation patterns in leeches and vertebrates.In addition, the different steps of angiogenesis in Hirudoare clear and easily detectable because they involve theentire – mostly muscular – body wall in which, in

normal conditions, blood vessels are virtually absent.This feature allows easy, unambiguous detection andmonitoring of stimulation or inhibition of angiogenesis.Moreover, due to the small size of animals, theaccomplishment of the angiogenic process in Hirudo is


Figures 16–18. Semithin and thin sections of surgically stimulated botryoidal tissue. After shaping the vascular lumen, endothelial-like cells (E)

split the vessel (v) by forming transluminal pillars (arrowheads). The presence of endothelial-like cells is evident in the TEM photograph

(Figure 18, arrowheads). B: botryoidal cells; F: fibroblast. Scale bars, 10 lm (Figures 16 and 17), 2 lm (Figure 18).

Figures 19 and 20. Semithin and thin sections of surgically stimulated botryoidal tissue. During neovascularization small portions of the

botryoidal tissue, not directly involved in the process, are cut off from forming vessels. This separation process is accomplished by ECM de novo

synthesis and deposition (arrowheads). B: botryoidal cells. E: endothelial-like cell. Scale bars, 10 lm (Figure 19), 1.5 lm (Figure 20).


Figures 21 and 22. About 24 h after surgery a large amount of new vessels (v, arrowheads) occupies the entire thickness of Hirudo body wall.

Vessels coming from the botryoidal tissue outgrow among muscle fields (M) toward the surface. Scale bars, 100 lm (Figure 21), 25 lm(Figure 22).

Figure 23. TEM. Detail of Figure 22 showing a new vessel (v) containing circulating cells (arrowheads). The vessel is surrounded by connective

tissue (ct) and muscle fibers (M). Scale bar, 1 lm.

Figure 24. Cryosections stained with an antibody anti-aVb3 integrin. Immunoreactive endothelial-like cells (arrowheads) are interposed among

botryoidal cells (B) in Hirudo growing vessels (v). Controls are shown in panel b.

Figures 25 and 26. Cryosections of stimulated (Figure 25) and unlesioned (Figure 26) leeches, probed with a fibronectin antibody. The expression

of fibronectin increases in lesioned leeches (Figure 25) when compared to quiescent (Figure 26) animals. M: muscle, v: vessel.


fast, since the growth of new vessels affects in 24 h theentire, about 2 mm-thick body wall, which in mamma-lians represents indeed a very small portion of anyaffected district. Furthermore, the response of Hirudoto a variety of treatments like explantation or growth

factor administration, suggests that both surgical andbiochemical stimuli may evoke in botryoidal tissue cellsangiogenic signalling pathways analogous to thoseobserved in vertebrates. Whereas in control, unstimu-lated animals, both granular-botryoidal and endothelial-


like cells formed stable cords and ropes of clusteredcells, in treated animals solid cellular cords were shownto acquire a lumen, since botryoidal tissue cells wereable to shape a luminal cavity through a dehiscenceprocess. This process occurred because botryoidal tissuecells underwent marked cell shape modifications: thin-ning, flattening and tapering of these cells allowedvessels to increase their diameter and length.

It is generally assumed that in adult vertebrates theoutgrowth of a network of capillaries is achieved by

endothelial cell sprouting, a process well described invertebrate wound healing [22–24]. In our study we haveevidenced that in adult leeches the growth of new vesselsis characterized by a combination of two distinct events,i.e. an initial vasculogenic step followed by extensiveangiogenesis. In fact, immediately after a surgical orbiochemical stimulus we have observed the transforma-tion of the botryoidal tissue into new immature vessels(vasculogenic step), from whom new branches andcapillaries can further outgrow, following a ‘classic’

Figure 27. Semithin section of the botryoidal tissue of unlesioned leeches injected with VEGF. The angiogenic process involves botryoidal (B)

and endothelial-like cells (E). Extensive proliferation and migration of endothelial cells results in the formation of large vessels (v). Pictures were

taken 24 h after surgery. Scale bar, 10 lm.

Figures 28–30. Semithin sections of unlesioned leeches treated with VEGF (Figure 28), bFGF (Figure 29) and GM-CSF (Figure 30), as

described in experimental procedures. In all cases extensive angiogenenesis (arrowheads) is induced by growth factor treatment. Pictures were

taken 24 h after surgery. Scale bars, 50 lm (Figure 28), 60 lm (Figures 29 and 30).

Figure 31. Cryosection. Immunofluorescence analysis of the botryoidal tissue of unlesioned leeches treated with GM-CSF. Endothelial-like cells

are stained with anti-aVb3 antibody. v: vessel; B: botryoidal cell.

Figures 32–34. Semithin and thin sections. Angiostatin treatment inhibits the growth of new vessels in leeches subjected to surgical stimulation of

angiogenesis. The body wall is mainly composed by layers of muscle fibers. In particular, fields of muscle fibers (M) are in close contact as in

control unlesioned leeches, and no vessels are visible between fibers. Even though botryoidal cells (B) (encircled area of Figure 32) are positively

stained by BrdU (detail shown in Figure 33) (a proliferation index, on botryoidal cells), they are unable to start the dehiscence process responsible

for new vessel lumen formation (Figure 34, arrowheads). Scale bars: 50 lm (Figure 32), 2 lm (Figure 34).

Figure 35. Semithin section of endostatin-treated leeches subjected to surgical stimulation of angiogenesis: scarce, small vessels are visible

(arrowheads). M: muscle fibers. Scale bar, 30 lm.

b

Figures 36–38. TEM. Fibroblast damage induced by mitomycin treatment (Figure 38) is evident when compared to controls (Figure 36). In

particular, the organization of collagen fibrils of stimulated leeches (Figure 37) is lost in treated specimens (Figure 38). Collagen: C. Scale bars,

1 lm (Figures 36 and 38), 0.25 lm (Figure 37).


angiogenic pattern (i.e. an extensive development of newvessels from pre-exhisting ones) (see Figure 40 summa-rizing the whole process).

In our model, surgically or biochemically stimulatedangiogenesis was always massive and very evidentwithin the muscular body wall.

As a tentative explanation of the occurrence ofangiogenesis in districts of Hirudo which are virtuallyavascular in normal conditions, we speculate that newvessels can act as a piping system to rapidly transfer inthe surgically lesioned area a large amount of cells

involved in immune defence and wound healing (Fig-ures 15 and 23).

An interesting outcome of our investigation showedthat in leeches the injection of human VEGF, bFGFor GM-CSF promoted vascular growth, migration anddifferentiation. Our findings indicate that leeches canrespond to the same growth factors involved in angio-genesis in vertebrates, thus suggesting that this could beanother example of evolutionary conservation so com-mon in the animal kingdom. VEGF and bFGF areconsidered as major growth factors promoting angio-

Figure 39. Quantitative evaluation of angiogenesis in leeches subjected to stimulation of angiogenesis, followed by treatment with vehicle or

inhibitors of angiogenesis. Vascular surface density is expressed as surface occupied by blood vessels in tissue sections taken at 0.4, 0.8, 1.2, 1.6,

2.0 and 15.0 mm (control) from the center of lesioned/treated areas. Leeches treated with vehicle (panel a), angiostatin (panel b), endostatin

(panel c), and mitomycin (panel d) after surgical stimulation.

Figure 40. Proposed model of the angiogenic process in Hirudo. Left to right: from solid cords of botryoidal tissue cells a vascular lumen is

shaped through a dehiscence process. During this process, the new vessel cavity enlarges, and endothelial cells proliferate and emerge from the

botryoidal tissue, ultimately lining the lumen. At later stages (right), new vessel branches outgrow from pre-exhisting vessels. B: botryoidal cells,

arrowheads: endothelial cells, c: new vessel cavity, v: new vessel.


genesis in a variety of experimental models [25–28]. Thepotent activity of these factors in H. medicinalis isexplained by the expression on the surface of flattened,endothelial-like cells of the botryoidal tissue of specificgrowth factor receptors. Interestingly, immunocyto-chemistry revealed that the VEGFRs 1 and 2 (VEG-FR-1, also known as flt-1 and VEGFR-2, also known asflk/KDR) were overexpressed in endothelial-like cellsundergoing active neovascularization within surgically/biochemically stimulated botryoidal tissue. In ananalogous fashion, the type-1 bFGF receptor wasoverexpressed in endothelial-like cells (Figure 15a).Interestingly, these receptors are located on the surfaceof endothelial cells, but are absent from the membranesof botryoidal cells. Despite the fact that botryoidal cellscooperate with endothelial cells in the formation of newblood vessels, this cell-specific expression pattern ofvascular growth factor receptors suggests that prolifer-ation of botryoidal cells might rely on different, inde-pendent signal transduction networks.

The cytokine GM-CSF is produced by macrophagesand by other immunocompetent cells, and has beenshown to regulate the migration and the proliferation ofendothelial cells [7]. Since macrophage-like cells arepresent in leeches, and since these cells usually accumu-late in lesioned districts of surgically treated animals[26], we speculate that an analogous factor could beinvolved in neovascularization of wounds in Hirudo.

Angiogenesis is promoted and regulated by a varietyof soluble growth factors and by interaction of specificendothelial cell adhesion molecules with a number ofexternal elements, located within the ECM [15, 18, 29–33]. In particular, the pivotal role of the aVb3 integrinhas been extensively documented in vertebrates [22, 34]:these cell adhesion molecules, expressed on the surfaceof endothelial cells in growing vessels, have beenimplicated in migration through the ECM. Interestingly,a parallel analogous process seems to occur during themigration of newly-forming vessels in Hirudo ECM. Instimulated leeches both integrins and ECM componentswere found to be involved in the angiogenic process. Infact, fibronectin, which plays a major function in celladhesion and migration, was found to be overexpressedin stimulated leeches when compared to control animals(Figures 25, 26). In an analogous fashion, the expressionof the integrin aVb3 (a receptor for several ligands,including vitronectin and fibronectin) was increased instimulated leeches.

In this study we have also shown that three differ-ent, potent inhibitors of angiogenesis were able tomarkedly reduce the formation of new vessels insurgically stimulated Hirudo. Administration of angio-statin and endostatin efficiently suppressed the out-growth of new vessels within the body wall of treatedanimals, probably by directly targeting the botryoidaltissue. It is tempting to speculate that angiostatin, likesomatostatin, can act on botryoidal cells (always BrdU+

in stimulated leeches). A recent report by Brevini-Gandolfi et al. [35] suggests the possibility that soma-

tostatin may interfere with the transcription levels ofTopoisomerase II a: impaired synthesis of this enzyme islikely to affect cell cycle dynamics, thus inhibiting cellproliferation.

Mitomycin, a drug used in routine ophtalmic surgeryto avoid the formation of collagenous opacifyingscares, was used in our research as an anti-angiogeniccompound. This drug is known to inhibit neovascu-larization by blocking fibroblast activity and collagensynthesis, a crucial step in vessel migration. Treatmentwith mitomycin suppressed the neovascularization ofH. medicinalis body wall, likely by acting on fibro-blasts (Figures 36 and 38) and, as a consequence,on the collagenous matrix. The organization of colla-gen fibrils results dramatically changed by mitomy-cin treatment, when compared to controls (Figures 37,38).

The response to mytomycin and the expression ofaVb3 suggests that the formation of new vessels inHirudo involves the ECM as a key-component of theangiogenic process.

In conclusion, the following evidences lead us tobelieve that H. medicinalis can be a reliable model forstudying the angiogenic and vasculogenic processes, andfor testing/screening activators and inhibitors of newvessel formation:

1. the modality of neovascularization in leeches issurprisingly similar to the one occurring in verte-brates. In addition, in Hirudo new vessel formation isa sequential process, involving an initial vasculogenicstep, followed by angiogenesis. This feature allowsthe study (a) of vasculogenesis and angiogenesisas distinct, separated processes; (b) of the transi-tion from vasculogenesis to angiogenesis; (c) ofthe thorough neovascularization of the leech bodywall;

2. both botryoidal and endothelial-like cells were shownto be in S-phase, and to be promptly responsive togrowth factor stimulation induced by VEGF, bFGFand GM-CSF;

3. different angiogenesis inhibitors (angiostatin, en-dostatin and mitomycin) have shown to be active inthe Hirudo model.

4. massive angiogenesis in stimulated leeches can beclearly, easily and unambiguously evaluated, becauseit involves within a short time (24 h) the entire bodywall thickness of Hirudo, a predominantly musculardistrict where vessels are almost completely absent innormal conditions.

In future investigations we will focus on the charac-terization of the molecular mechanisms leading botry-oidal tissue cells to change their shape and function,thus cooperating to new vessel formation. We are alsowilling to characterize the leech receptors which haveshown both to respond to human VEGF, and to reactwith human-specific anti-flt-1 and anti-flk/KDR anti-sera. An ongoing study is focusing on the feasibility of


Hirudo as a model for testing ‘naked DNA’ vectors forangiogenesis gene therapy.

Acknowledgement

We are grateful to Maria Luisa Guidali for her excellenttechnical assistance.

References

1. Mann KH. Leeches (Hirudinea): Their Structure, Physiology,

Ecology and Embryology. Oxford: Pergamon Press 1962.

2. Sawyer RT. Leech Biology and Behavior 1: Anatomy, Physiology

and Behaviour. Oxford: Oxford Scientific Publications 1986.

3. Sawyer RT, Fitzgerald SW. Leech Biology. Vol 1. Oxford: Oxford

University Press 1981.

4. Fischer E. The myelo-erythroid nature of the chloragogenous-like

tissues of the annelids. Comp Biochem Physiol 1993; 106: 449–53.

5. Bradbury S. The botryoidal and vaso-fibrous tissue of the leech

Hirudo medicinalis. Quart J Microsc Sci 1959; 100: 483–98.

6. de Eguileor M, Grimaldi A, Tettamanti G et al. Ultrastructural

and functional versatility of the hirudinean botryoidal tissue.

Tissue & Cell 2001; 33: 332–41.

7. Bussolino F, Wang JM, Defilippi P et al. Granulocyte- and

granulocyte–macrophage-colony stimulating factors induce human

endothelial cells tomigrate and proliferate.Nature 1989; 337: 471–3.

8. Santamaria JA, Marı̀-Beffa M, Santos-Ruiz L et al. Incorporation

of bromodeoxyuridine in regenerating fin tissue of the goldfish

Carassius auratus. J Exp Zool 1996; 275: 300–7.

9. Moore RD, Mumaw V, Shonberg MD. Optical microscopy of

ultrathin tissue sections. J Ultrastruct Res 1960; 4: 113–6.

10. Miller JB, Aidley DJ. Two rates of relaxation in the dorsal

longitudinal muscle of leech. J Exp Biol 1973; 58: 91–103.

11. Geiger B, Tokuyasu KT, Dutton A et al. Vinculin an intracellular

protein localized at specialized sites where microfilament bundles

terminate at cell membranes. Proc Natl Acad Sci USA 1980; 77:

4127–31.

12. Barth PJ, Weingartner K, Kohler HH et al. Assessment of the

vascularization in prostatic carcinoma: A morphometric investi-

gation. Human Pathol 1996; 27: 1306–10.

13. Weibel ER. Stereological Methods. I. Practical Methods for

Biological Morphometry. London: Academic Press 1979.

14. Cruz-Orive LM, Weibel ER. Recent stereological methods for cell

biology. A brief survey. Am J Physiol 1990; 258: L148–56.

15. Beck L, D’Amore PA. Vascular development: Cellular and

molecular regulation. FASEB J 1997; 11: 365–73.

16. Brooks PC, Montgomery AMP, Rosenfeld M et al. Integrin alpha

v beta 3 antagonists promote tumor regression by inducing

apoptosis of angiogenic blood vessels. Cell 1994; 79: 1157–64.

17. Carmeliet P. Mechanism of angiogenesis and arteriogenesis. Nat

Med 2000; 6: 389–95.

18. Ruoslahti E, Engvall E. Integrins and vascular extracellular matrix

assembly. J Clin Invest 1997; 99: 1149–52.

19. Wayner EA, Orlando RA, Cheresh DA. Integrins alpha v beta 3

and alpha v beta 5 contribute to cell attachment to vitronectin but

differentially distribute on the cell surface. J Cell Biol 1991; 113:

919–29.

20. Enenstein J, Kramer RH. Confocal microscopic analysis of

integrin expression on the microvasculature and its sprouts in the

neonatal foreskin. J Invest Dermatol 1994; 103: 381–6.

21. IngberDE. Fibronectin controls capillary endothelial cell growth by

modulating cell shape. Proc Natl Acad Sci USA 1990; 87: 3579–

83.

22. Clark RAF. Regulation of fibroplasia in cutaneous wound repair.

Am J Med Sci 1993; 306: 42–8.

23. Jiang WG, Hiscox S. Hepatocyte growth factor/scatter factor, a

cytokine playing multiple and converse roles. Histol Histopathol

1997; 12: 537–55.

24. Landeen LK, Ziegler FC, Halberstadt C. Characterization of a

human dermal replacement. Wounds 1992; 5: 167–75.

25. Kim JK, Bing I, Armanini M et al. Inhibition of vascular

endothelial growth factor-induced angiogenesis suppresses tumour

growth in vivo. Nature 1993; 362: 841–46.

26. Senger DR, Claffey KP, Benes JE et al. Angiogenesis promoted

by vascular endothelial growth factor: regulation through alpha 1

beta 1 and alpha 2 beta 2 integrins. Proc Natl Acad Sci USA 1997;

94: 13612–17.

27. Neufeld G, Gospodarowicz D. Basic and acidic fibroblast growth

factors interact with the same cell surface receptors. J Biol Chem

1986; 261: 5631–7.

28. Rahmoune H, Chen H, Gallagher JT et al. Interaction of heparan

sulfate from mammary cells with acidic fibroblast growth factor

(FGF) and basic FGF. J Biol Chem 1998; 273: 7303–10.

29. de Eguileor M, Tettamanti G, Grimaldi A et al. Histopathological

changes after induced injury in leeches. J Invert Pathol 1999; 74:

14–28.

30. Vlodavsky I, Fuks Z, Ishai-Michaeli R et al. Extracellular matrix-

resident basic fibroblast growth factor: implication for the control

of angiogenesis. J Cell Biochem 1991; 45: 167–76.

31. Moses MA, Klagsbrun M, Shing Y. The role of growth factors in

vascular cell development and differentiation. Int Rev Cytol 1995;

161: 1–48.

32. Carmeliet P, Collen D. Role of vascular endothelial growth factor

and vascular endothelial growth factor receptors in vascular

development. Curr Top Microbiol Immunol 1999; 237: 133–

158.

33. Shimo T, Nakanishi T, Nishida T et al. Connective tissue growth

factor induces the proliferation, migration, and tube formation of

vascular endothelial cells in vitro, and angiogenesis in vivo. J

Biochem 1999; 126: 137–45.

34. Friedlander M, Theesfeld CL, Sugita M et al. Involvement of

integrins alpha v beta 3 and alpha v beta 5 in ocular neovascular

diseases. Proc Natl Acad Sci USA 1996; 93: 9764–69.

35. Brevini-Gandolfi TAL, Cillo F, Favetta LA et al. Somatostatin

up-regulates Topoisomerase II Alpha expression and affects

LNCaP cell cycle. Mol Cell Endocrinol 2001; 176: 103–10.


Documents

Hirudo medicinalis: a new model for testing activators and inhibitors of angiogenesis