Hovering between death and life: Natural apoptosis and phagocytes in the blastogenetic cycle of the colonial ascidian Botryllus schlosseri

Developmental and Comparative Immunology 34 (2010) 272–285

Contents lists available at ScienceDirect

Developmental and Comparative Immunology journal homepage: www.elsevier.com/locate/dci

Hovering between death and life: Natural apoptosis and phagocytes in the blastogenetic cycle of the colonial ascidian Botryllus schlosseri Francesca Cima a, Lucia Manni a,*, Giuseppe Basso b, Elena Fortunato b, Benedetta Accordi b, Filippo Schiavon a, Loriano Ballarin a a b

Department of Biology, University of Padova, Padova, Italy Department of Pediatry, University of Padova, Padova, Italy

A R T I C L E I N F O

A B S T R A C T

Article history: Received 15 May 2009 Received in revised form 6 October 2009 Accepted 9 October 2009 Available online 28 October 2009

Colonies of the compound ascidian Botryllus schlosseri undergo recurrent generation changes during which massive, natural apoptosis occurs in zooid tissues: for this reason the species is emerging as an interesting model of invertebrate chordate, phylogenetically related to vertebrates, for studies of apoptosis during development. In the present work, we carried out a series of morphological, cytofluorimetrical and biochemical analyses, useful for a better characterization of Botryllus apoptosis. Results are consistent with the following viewpoints: (i) both intrinsic and extrinsic pathways, probably connected by the BH3-only protein Bid, are involved in cell death induction; (ii) phagocytes, once loaded with senescent cells, frequently undergo apoptosis, probably as a consequence of oxidative stress caused by prolonged respiratory burst, and (iii) senescent phagocytes are easily recognized and ingested by other phagocytes, responsible for their clearance. In addition, results suggest the conservation of apoptosis induction mechanisms throughout chordate evolution. ß 2009 Elsevier Ltd. All rights reserved.

Keywords: Apoptosis Asexual reproduction Botryllus Phagocytes Tunicates

1. Introduction Cell death by apoptosis is a fundamental biological process required for the correct sculpturing of developing organs and the controlled elimination of unwanted cells in morphogenesis, regeneration, tissue renewal and maturation of the immune system [1–5]. Apoptosis is characterized by a series of morphological changes, such as cytoplasm and nuclear condensation, leading to cell shrinking, blebbing, internucleosomal cleavage of chromatin and exposure of cell surface molecules enabling the recognition and the removal of dying cells by phagocytes [6]. It implies the triggering of a series of biochemical events culminating in the activation of initiator and effector cysteinyl aspartate proteases, known as caspases. The latter are normally present in healthy cells as inactive zymogens, with little or no protease activity, and are converted to active enzymes during apoptosis so that extensive proteolysis occurs during the process [7–9]. Proapoptotic stimuli favor the assembly of large complexes that allow the clustering of initiator zymogens and their cleavage to active enzymes which, in turn, convert effector pro-enzymes to active caspases [8]. Initiator procaspases are characterized by the

* Corresponding author at: Dipartimento di Biologia, Universita` di Padova, Via Ugo Bassi 58/B, 35100 Padova, Italy. Tel.: +39 049 8276252; fax: +39 049 8276199. E-mail address: [email protected] (L. Manni). 0145-305X/$ – see front matter ß 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.dci.2009.10.005

presence of the caspase recruitment domain (CARD) which allows their aggregation by adaptor proteins such as Fas-associated death domain protein (FADD) or apoptotic protease-activating factor-1 (APAF-1), and, lastly, their autoactivation [7–9]. In mammals, apoptosis can be either intrinsically induced by mitochondrial damage, leading to the release of cytochrome c in the cytosol and the setting up of apoptosomes which activate the initiator caspase-9, or extrinsically triggered by the interaction of death receptors, such as Fas (CD95), with their ligands, e.g., FasL (CD95L), that allows the formation of the death-inducing signaling complex (DISC) and the activation of the initiator caspase-8. In both cases, initiator caspase-9 and -8 activate effector caspases which are responsible of the majority of the events occurring in cells having entered an apoptotic pathway [7,8]. Tunicates are invertebrate chordates, mainly represented by ascidians, phylogenetically related to vertebrates. In the solitary ascidian Ciona intestinalis, a reference species for developmental and cell biology studies, the genome has been fully sequenced and homologous of vertebrate genes involved in apoptosis have been identified [10–14]. This suggests a high degree of conservation of the apoptotic machinery throughout the evolution of chordates. Botryllus schlosseri is a cosmopolitan colonial ascidian, easily found in shallow temperate waters, which is emerging as a valuable model organism for the study of variety of biological processes such as sexual and asexual reproduction, regeneration, allorecognition, immunobiology [15].

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

A colony is a clone as it derives from the settlement and metamorphosis of a tadpole-like larva into a founder zooid, and grows by asexual reproduction through the continuous production of new buds. Colonies are formed by many filter-feeding zooids, grouped in star-shaped systems, bearing palleal buds which, in turn, produce budlets, so that three blastogenetic generations are usually present. Zooids, buds and budlets are connected by the colonial circulatory system which assures well-synchronized development and tight coordination among colony members [15]. At regular intervals (weekly at the temperature of 20 8C), colonies undergo a generation change or take-over, lasting 24– 36 h, during which adult zooids are progressively replaced by growing buds; in the meantime budlets become buds and a new budlet generation appears. Therefore, a colonial blastogenetic cycle can be defined beginning with the opening of the siphons of a new zooid generation, which starts its filtering activity, and ending with the take-over, when adult zooids cease filtering, contract and are gradually resorbed [15]. During take-over, diffuse, natural apoptosis occurs in zooid tissues, which renders this species an interesting model also for the study of this process. In addition, a colony is virtually immortal so that many cyclical apoptotic events can be studied during its lifespan. The clearance of dying cells is assured by circulating phagocytes which are massively recruited and infiltrate zooid tissues and engulf senescent cells. Phagocytes are represented by spreading hyaline amoebocytes (HA), able to recognize and ingest foreign particles or cells, and round macrophage-like cells (MLC) deriving from HA which withdraw their cytoplasmic projections after the ingestion of non-self material [16]. At the generation change, a significant change in the distribution of circulating phagocytes with respect to mid-cycle, i.e., phases of the blastogenetic cycle that are more than one day from the preceding and following take-over [15,17], occurs: the frequency of circulating HA falls from 25–42% to 12–25%; at the same time, the percentage of MLC with ingested materials inside their vacuoles rises from 4–10% to 20–30% [16,18]. In addition, during the blastogenetic cycle, the amount of dying hemocytes with apoptotic features increases abruptly from 2–5% at mid-cycle to about 30% at take-over. They are characterized by DNA fragmentation, phosphatidylserine (PS) exposure on the outer layer of plasma membrane and activation of caspase-9 and -3 [18,19]; recognition of PS is required for the phagocytes to clear dying cells [18]. New young, undifferentiated cells, released from unidentified hemopoietic sites, enter the circulation at this phase [19]. However, despite the increasing number of studies on apoptosis in Botryllus, uncertainties and doubts still persist on the biochemical pathways involved, the modalities of recognition of effete cells and the fate of phagocytes. With the aim to clarify some of these aspects, we carried out new morphological, cytofluorimetrical and biochemical studies, useful for a better characterization of Botryllus apoptosis. Our results indicate that both intrinsic and extrinsic pathways are involved in cell death induction and are probably connected by Bid, a member of the ‘‘BH3-only protein’’ subgroup of the Bcl-2 family [20]. In addition, they suggest that oxidative stress represent the key event in triggering the apoptotic cascade, at least in phagocytes. 2. Materials and methods 2.1. Animals Colonies of B. schlosseri were collected in the Lagoon of Venice and allowed to adhere to glass slides (5 cm  5 cm). They were reared in aquaria filled with filtered seawater (FSW) at the temperature of 19 8C, fed with Liquifry Marine (Interpet, Dorking, England) and the water was changed every other day.

273

2.2. Hemocyte collection Blood was collected with glass micropipettes from the peripheral vessel of colonies (which had been previously rinsed in 12.9 mM Na-citrate in FSW, pH 7.5, to prevent hemocyte clumping), that were punctured with fine tungsten needles. Cells were collected with glass micropipettes and centrifuged at 780  g for 10 min. Pellets were resuspended in FSW, in 1.5-ml vials, to a final concentration of 106 cells/ml. Sixty microlitres of hemocyte suspension were placed in the centre of culture chambers prepared as described elsewhere [22,23] and left to adhere to coverslips for 30 min at room temperature. Cell mortality was evaluated with the Trypan blue exclusion assay [24] and was less than 5% after 2 h of incubation at room temperature. At least three colonies (10–20 systems in size) were used for each experiment; each colony was previously cut into subclones of 4–5 systems and each subclone was bled during the course of the blastogenetic cycle, in order to obtain blood samples from takeover and mid-cycle phases. 2.3. Transmission electron microscopy Colonies at take-over were fixed in 1.5% glutaraldehyde buffered with 0.2 M sodium cacodylate buffer (CB), pH 7.4, plus 0.29 M NaCl. After washing in CB and postfixation in 1% OsO4 in 0.2 M cacodylate buffer, specimens were dehydrated and embedded in Epon Araldite. Transversal and sagittal serial sections (1 mm) of zooids were counterstained with Toluidine blue; thin sections (90 nm) were stained with uranyl acetate and lead citrate. Micrographs were taken with a Hitachi H-600 electron microscope operating at 75 kV. Photographs were digitalized with an Epson Perfections Scanner 1200S and were collected and typeset in Corel Draw X3. 2.4. Immunocytochemistry on hemocytes Hemocyte monolayers were fixed in 4% paraformaldehyde plus 0.1% glutaraldehyde in 0.4 M cacodylate buffer containing 0.29 M NaCl and 29 mM sucrose, treated for 30 min with 1% H2O2 in phosphate-buffered saline (PBS: 0.13 M NaCl, 2.7 mM KCl, 10 mM Na2PO4, 1.7 mM KH2PO4; pH 7.4) to block endogenous peroxidase, incubated for 30 min in 5% powdered milk and 5% fetal calf serum in PBS to reduce nonspecific binding and finally incubated overnight with 10 mg/ml primary polyclonal antibodies. The following primary polyclonal antibodies, raised against human antigens, were assayed: anti-Bcl-2, anti-Bax, anti-caspase-3, anticaspase-7, anti-caspase-8, anti-Fas and anti-FasL. They were purchased from Santa Cruz Biotech (anti-Bcl-2 and anti-Bax), Oncogene (anti-Fas, anti-caspase-3), GeneTex (anti-FasL), Sigma (anti-caspase-7), Calbiochem (anti-caspase-8). In control preparations, antibodies were pre-incubated for 2 h with colony lysates or HEP-G2 cells as described below. Hemocytes were then washed in PBS, incubated in goat biotinylated anti-rabbit-IgG antibody (Santa Cruz Biotech, 10 mg/ ml) for 30 min, washed again and incubated for 30 min in avidin– biotin–peroxidase complex (ABC, Vector Laboratories). After thorough washing in PBS, they were finally incubated for 5 min in a solution of 0.63 mM 3,30 -diaminobenzidine (DAB), containing 4% hydrogen peroxide, and mounted with Acquovitrex (Carlo Erba). In controls, primary antibodies were substituted with absorbed primary antibodies prepared as described below; pre-immune serum was used in negative controls. Positive sites appeared brown. 2.5. Lectin cytochemistry Fixed hemocytes were treated for 30 min with 1% H2O2 in PBS followed by 30 min in 5% powdered milk and 5% fetal calf serum in

274

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

PBS and incubated for 60 min in 5 mg/ml of the following biotinlabeled lectins in PBS: wheat germ agglutinin (WGA, specific for Nacetyl-b-D-glucosamine), concanavalin A (ConA, recognizing internal and nonreducing terminal a-mannosyl groups), Ricinus communis agglutinin (RCA, able to bind b-D-galactosides), and Limax flavus lectin (LFA, specific for sialic acid). All these lectins were purchased from Vector Labs with the exception of LFA (EY Labs). Slides were then washed and incubated for 30 min in avidin– biotin complex (Vector) according to the manufacturer’s instructions, treated for 5 min in a solution of 0.63 mM DAB containing 4% hydrogen peroxide and mounted with Acquovitrex. In controls, lectins were omitted. Positive sites appeared brown.

Fas antibody or biotin-conjugated anti-FasL antibody followed by incubation with FITC-conjugated streptavidin. Excitation was set at 525 nm and the emission filters were at 570 nm. Surface exposure of PS was measured by flow cytometry with a Coulter Cytomics FC500 by incubating freshly collected living cells with Annexin-V-FITC in ISO (5  106 cells/ml), for 15 min, according to the manufacturer’s instructions (Annexin-V Fluos, Roche Diagnostic). Simultaneously, the cells were stained with propidium iodide, which enters cells that have altered membrane permeability and allows discrimination between necrotic and apoptotic cells. Excitation was set at 488 nm and the emission filters were at 525 nm. All the analyses were repeated three times and representative results were shown in figures.

2.6. Phagocytosis assay 2.9. Spectrophotometric and cytochemical assays for peroxides Hemocytes (6.5  106 cells/ml) collected from mid-cycle colonies were left to adhere to coverslips in culture chambers, as described above, and then incubated in yeast (Saccharomyces cerevisiae)-containing FSW (yeast:hemocyte ratio = 10:1), according to Ballarin et al. [22], for 30 min. Monolayers were then repeatedly washed in FSW, to remove uningested yeast cells. Expression of PS on hemocyte surface was revealed with the Annexin-V-FLUOS Staining kit (Roche). After incubation with yeast cells, hemocytes were incubated for 15 min in 10 ml fluoresceinconjugated Annexin-V and 10 ml propidium iodide in 1 ml FSW, according to the manufacturer’s instructions. Cells incubated in the absence of yeast cells were used as controls. In order to assess whether apoptotic phagocytes can be recognized and ingested by other phagocytes, hemocyte monolayers challenged with yeast were exposed for 10 min at 25 8C, in the dark, to 4 mM red fluorescent phagocyte-linker PKH26 lipophilic dye (PKH26-PCL kit, prepared for Sigma by Zynaxis, Inc.) diluted with Diluent B, provided with the kit and to which 0.29 M NaCl was added. Membrane labeling was stopped by rinsing twice with FSW and new hemocytes from the same donor colony were added to the culture. After 60 min of incubation, samples were observed under LM equipped for fluorescence with and excitation filter for Texas Red of 550 nm and a barrier filter of 590 nm. In another experimental series, we incubated hemocytes with yeast for a brief ‘‘pulse’’ period of 5 min and then cultures were repeatedly washed and incubated in FSW for additional 5, 10, 15, 30 and 60 min. Cells were then fixed as described above, washed in PBS, and stained for 5 min in 10% Giemsa solution; the percentages of HA and MLC in the culture were then evaluated. 2.7. TUNEL assay To reveal DNA fragmentation, fixed hemocytes were incubated in the permeabilization solution (0.1% Triton X-100 in 3.4 mM Nacitrate) for 2 min at 4 8C. Samples were then rinsed twice with PBS and incubated in the terminal dUTP nick-end labeling (TUNEL) reaction mixture (in situ cell death detection kit, Roche) for 60 min at 37 8C, according to the manufacturer’s instructions. Subsequently, hemocytes were incubated with a peroxidase-conjugated anti-fluorescein-isothiocyanate (FITC) antibody, stained with 0.63 mM DAB in PBS containing 4% hydrogen peroxide, mounted with Acquovitrex and observed under the light microscope. The presence of fragmented DNA was revealed by dark brown staining. 2.8. Cytofluorimetric analysis The presence of membrane proteins immunoreactive to antiFas and anti-FasL was evaluated by flow cytometry with a Coulter Epics XL-MCL. Living cells (5  106 cells/ml), resuspended in isotonic salt solution (ISO: 20 mM Tris, 0.5 M NaCl; [25]), were previously incubated with either phycoerythrin-conjugated anti-

Hemolymph, collected from subclones of the same colony at mid-cycle and take-over, was centrifuged at 780  g for 10 min. Phenol red (Sigma) and peroxidase (grade II, Boheringer Mannheim) were added to 20 ml of the supernatant, referred to as blood plasma (BP), to a final concentration of 0.25% and 250 U/ml, respectively, in the wells of a microtiter plate. After 10 min of incubation, NaOH was added to a final concentration of 0.45 N and the absorbance was read at 620 nm with a microplate reader (modified after Pick [26]). Standard curves were prepared for the expression of results as nmoles of peroxides by means of dilutions of H2O2 from 1 to 100 mM. The protein content of BP was determined according to Bradford [27]. The presence of intracellular peroxides was detected using the cell-permeable compound dichlorodihydrofluorescein diacetate (DCFH-DA) which is converted to cell-impermeable 20 ,70 -dichlorodihydrofluorescein (DCFH) by cytosol esterases. DCFH is easily oxidized by peroxides to 20 ,70 -dichlorofluorescein (DCF) a highly fluorescent compound. Once they had adhered to the glass slides, living hemocytes were incubated for 30 min in the dark with 5 mM DCFH-DA in FSW. Cells were then washed three times for 5 min in FSW and directly observed under a Leitz Dialux 22 fluorescence microscope equipped with a I2/3 (490 nm) filter block. 2.10. Caspase activity assay Whole blood from large colonies (10 systems) that had been previously blotted dry was collected and centrifuged at 780  g for 10 min. Supernatants were discarded and pellets were resuspended in 100 ml of lysis buffer from colorimetric activity assay kits for caspase-3 and -8 (Chemicon). After a 10-min incubation, samples were centrifuged at 10,000  g for 10 min. Supernatants, called hemocyte lysates, were collected and their protein content evaluated as described above. In the wells of a 96-well microtiter plate, 50 ml of hemocyte lysate were incubated for 1 h at 37 8C with 10 ml of the specific colorimetric substrate N-acetyl-Asp-Glu-ValAsp-p-nitroaniline or N-acetyl-Ile-Glu-Thr-Asp-p-nitroaniline, for caspase-3 and -8, respectively, according to the manufacturer’s instructions. The release of p-nitroaniline was measured at 405 nm in a microplate reader. A p-nitroaniline reference curve was obtained by serial dilution of a 10 mM standard solution in dimethylsulfoxide. One unit of caspase activity was defined as the amount of enzyme able to cleave 1 nmole substrate per hour at 37 8C. Results were expressed as specific activities (units/mg proteins) and normalized for values obtained at mid-cycle. 2.11. Immunoblot analysis Immunoblot analysis was carried out on subclones, 2–3 systems in size, of various colonies at mid-cycle and take-over. SDS polyacrylamide (15%) slab gel electrophoresis was performed

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

according to the method of Laemmli [28]. In order to have detectable protein bands, whole colony homogenates were used. Colonies were transferred in lysis buffer (50 mM Tris–HCl, 0.25 M sucrose, 1% SDS, 1 mg/ml pepstatin, 1 mg/ml leupeptin, 40 mg/ml PMSF, 2 mm Na-orthovanadate, 10 mM NaF, 0.1% NP-40, 5 mM EDTA, 5 mM N-ethylmaleimide), subjected to sonication at 0 8C in a Branson 1200 sonifier at 50% duty cycles for 1.5 min, and centrifuged at 10,000  g. The supernatants were then frozen in liquid nitrogen and stored at 80 8C until use. The protein content of the supernatants was determined as previously described; each well of the gel received a volume of supernatant equivalent to 10 mg of proteins. Gels were run at 10 mA/gel for approximately 3.5 h and were then stained with Coomassie blue. Proteins were transferred to 0.45 mm Electran nitrocellulose membrane (BDH) according to Towbin et al. [29], with 25 mM Tris, 160 mM glycine, 20% methanol, 0.7 mM SDS as transfer buffer. After blotting, membranes were thoroughly washed in Tris-buffered saline (TBS: 50 mM Tris–HCl, 150 mM NaCl, pH 7.4), incubated for 30 min in TBS containing 5% powdered milk (Sigma) and probed overnight with 1 mg/ml rabbit polyclonal primary antibodies. The following antibodies, raised against human antigens, were used in immunoblot analysis: anti-caspase-3, anti-caspase-7, anti-caspase-8, anti poly-ADP-ribose polymerase-1 (PARP), anti-Bid, anti-FADD and anti-Fas. They were purchased from Calbiochem (anti-caspase-3), Sigma (anti-caspase-7 and -8), Cell signaling (anti-FADD and antiPARP), GeneTex (anti-Bid), Oncogene (anti-Fas). After further extensive washing in TBS containing 0.05% Tween 20 (TTBS), membranes were overlaid for 1 h with goat anti-rabbit IgG conjugated with peroxidase (BioRad), diluted 1/10,000 in TTBS according to the manufacturer’s instructions. Immunogenic bands were revealed using the Super Signal West Pico Chemiluminescent Substrate kit (Pierce). In control series, diluted primary antibodies were pre-incubated overnight in colony lysates, prepared as described above, from colonies at take-over as a source of apoptotic cytoplasmatic proteins. The mixtures were then used for immunobot analysis. In the case anti-Fas, 10 ml of the diluted primary antibody were pre-incubated overnight with 107 HEP-G2 cells from human hepatocellular carcinoma, which constitutively express Fas on their surface. The supernatant was then collected and used in the immunoassay. For negative controls, primary antibodies were replaced with pre-immune serum, of the same origin and at the same dilution. The resulting signal was recorded on an X-ray film. The density of the bands, in pixels, was evaluated and compared using the ImageJ software (http://rsbweb.nih.gov/ij/). 2.12. Assay for cytochrome c Hemocytes, collected from large (200–300 zooids) colonies as described above, were snap-frozen by immersion in liquid nitrogen and then thawed by immersion of the vial in water kept at 37 8C. The process, which disrupts the plasma membrane, was repeated five times. The resulting suspensions were then centrifuged at

275

18,000  g for 20 min at 4 8C in order to sediment microsomes and mitochondria (modified after Enari et al. [30]). The supernatants, called cytosolic extracts, were collected and stored at 80 8C until their use in immunoblot analysis, as described above, with polyclonal anti-cytochrome c antibody (BD Pharmingen). In order to exclude that the presence of cytochrome c was related to mitochondrial damage, the presence of activity of citrate synthase – the initial enzyme of the tricarboxylic acid cycle and an exclusive marker of the mitochondrial matrix – was measured by following, at 412 nm, the color of thionitrobenzoic acid, which is generated from 5,50 -dithiobis-(2-nitrobenzoate) present in the reaction of citrate synthesis, and caused by the deacetylation of Acetyl-CoA [31]. 2.13. Statistical analysis Experiments were replicated at least three times. Data were expressed as means  SD. At least 300 cells in 10 optical fields at 1250 were counted in each experiment for the evaluation of the fraction of phagocytosing or immunopositive cells. Frequencies were subjected to angular transformation. Means were compared by Student’s t-test. 3. Results 3.1. The Botryllus blastogenetic cycle Three blastogenetic generations are usually present in Botryllus colonies: adult, filter-feeding zooids, primary buds on zooids and secondary buds (budlets) on buds [15]. A staging method of the blastogenetic development of a new individual, from budlet to adult zooid, was introduced by Berrill [32] and modified by Sabbadin [21]. Since the development of zooids, buds and budlets in a colony is highly coordinated, the colonial developmental phase can be univocally defined by the developmental stages of its zooids, buds and budlet and can be expressed, as indicated by Sabbadin [21] and recently discussed by Manni et al. [15], by a series of three numbers separated by slashes, each referring to the developmental stages of zooids, buds and budlets, respectively. The main developmental phases in the blastogenetic cycle are listed in Table 1: it starts with phase 9/6/0, a brief period during which only two generations are present in colonies, and continues with 9/7/1, 9/8/2–5, ending with phase 11/8/6, which represents take-over. Phases from 9/8/2 to 9/8/4, more than one day from the preceding and following take-over, are collectively called midcycle phases [15,17] (Fig. 1). 3.2. Immunoblot analysis reveals changes in the expression of apoptotic markers during the colonial blastogenetic cycle The antibodies used in immunoblot analysis recognized specific, single protein bands, with a molecular weight comparable

Table 1 Developmental phases in the colonial blastogenetic cycle of Botryllus schlosseri (modified according to Manni et al. [15]). Phase

Events

9/6/0 9/7/1 9/8/2 9/8/3 9/8/4 9/8/5 11/8/6

Onset of a new blastogenetic cycle. Brief interval after take-over, with colonies containing newly formed adults and their buds, without budlets. Stigmata primordia in primary buds. Appearance of secondary bud rudiments as thickenings of the atrial wall of primary buds. Beginning of heartbeat in primary buds. Budlet primordia in the form of a hemisphere that progressively skews towards the anterior end of the parent. Stigmata perforation in primary buds. Secondary buds form double-layered vesicles. Beginning of organogenesis in secondary buds. Pigment cells in the outer epithelium of primary buds. Branchial and peribranchial chambers recognizable in secondary buds. Onset of take-over: shrinkage of adult zooids and closure of their siphons. Massive apoptosis of the internal tissues and clearance of apoptotic cell and corpses by infiltrating phagocytes. End of heartbeat in zooids. Growth of primary buds to adult size. Endostyle differentiation and appearance of intestine rudiments in secondary buds.

Colonial developmental phases are expressed by a series of three numbers separated by slashes, each referring to the developmental stages of zooids, buds and budlets, respectively, as defined by Sabbadin [21].

276

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

Fig. 1. Schematic drawing of B. schlosseri colonial blastogenetic cycle; ventral view. Days from the opening of the siphons of a new zooid generation are in brackets. b: bud; bl: budlet; en: endostyle; z: zooid.

to that of their mammalian antigens, which decreased in intensity when absorbed antibodies were used. Some antibodies (anticaspase-7, anti-FADD) recognized additional, nonspecific bands which did not change their intensity when colony lysates from mid-cycle and take-over were compared or when absorbed

antibodies were used. In all the cases, specific bands disappeared or were greatly diminished in density when the diluted primary antibodies were incubated with lysates of colonies at take-over or, in the case of anti-Fas, with human cells constitutively expressing Fas. No bands were observed in negative controls (Fig. 2). Immunoblot analysis of colony lysates at mid-cycle with anticaspase-8 antibodies identified a protein band at 54 kDa, which was greatly reduced in intensity (83.2  5.4%) at take-over. The intensity of the bands at 21 and 38 kDa, recognized by antibodies against caspase-3 and caspase-7, respectively, increased at take-over (5.4  1.2- and 2.4  0.7-fold, respectively) with respect to mid-cycle (Fig. 2). A single band of 22 kDa, was recognized by the anti-Bid antibody: it underwent a 47.8  2.5% decrease at take-over when compared to mid-cycle (Fig. 2). A fourfold (3.8  0.9) increase in the density of the electrophoretic band recognized by the anti-Fas antibody and a twofold (2.1  0.5) increase in that of the band immunopositive to the anti-FADD antibody was observed at takeover when compared with mid-cycle (Fig. 2). The anti-PARP antibody recognized two bands of 116 and 89 kDa, respectively, in the electrophoretic pattern of whole colony homogenate: the upper band was more intense at mid-cycle and decreased to 29.7  3.2% of the original value at take-over; the lower band was very faint at midcycle and showed a 3.5  1.3-fold increase at take-over (Fig. 2). 3.3. Cytochrome c is differently expressed in mid-cycle and take-over In the cytosol extracts of hemocytes, the anti-cytochrome c antibody revealed a single band of 37 kDa, the density of which

Fig. 2. Immunoblot analysis of whole colony homogenates (cytosolic extract in the case of anti-cytochrome c) of the expression proteins recognized by antibodies raised against some apoptotic markers at mid-cycle (MC) and take-over (TO). NC: negative control. Results with absorbed primary antibodies (TO absorbed) are also shown.

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

277

Fig. 3. Representative flow cytometric comparison of blood cells distribution at mid-cycle and take-over. The analysis was carried out three times and the population of cells with high side and forward scatter is evidenced in (a), whereas the analysis of phosphatidylserine expression in the cell population is shown in (b). Annexin-V: Annexin-V fluorescence; FS: forward scattering; PI: propidium iodide fluorescence; SS: side scattering.

increased 5.1  0.8-fold at take-over with respect to mid-cycle (Fig. 2). No detectable activity of citrate synthase was observed in extracts from either mid-cycle or take-over (data not shown). 3.4. Giant, granular hemocytes increase their frequency at take-over According to flow cytometric analysis, the population of circulating cells with high forward and side scatter amounted to 3.7% of total hemocytes at mid-cycle; none of them were recognized by Annexin-V. Conversely, at take-over the frequency of these hemocytes exceeded 20% of circulating cells and almost 90% of them resulted labeled by Annexin-V (Fig. 3). 3.5. The production of soluble peroxides increases at the generation change The concentration of peroxides in BP at take-over (19.97  0.91 nmol/mg protein) was significantly (p < 0.001) higher with respect to mid-cycle (10.2  1.14 nmol/mg protein). The frequency of fluorescent cells after incubation of hemocytes in DCFH-DA was significantly (p < 0.05) increased at take-over (12.1  1.8%) with respect to mid-cycle (5.9  1.9%). Most of the fluorescent cells were represented by giant MLC and labeling was located in the cytoplasm (Fig. 4a). 3.6. Change of expression of apoptotic markers on hemocytes during the blastogenetic cycle The number of hemocytes showing cytoplasmic immunopositivity to anti-Bcl-2 antibodies amounted to about 10% in mid-cycle phases (Fig. 5) and most of the cells labeled by the anti-Bcl-2 antibody were represented by hyaline amoebocytes (Fig. 4b). The

fraction of positive hemocytes significantly (p < 0.05) decreased during the take-over. The opposite was true for Bax: cytoplasm immunoreactivity to anti-Bax antibody was observed in about 6% of hemocytes in mid-cycle and this fraction significantly (p < 0.001) increased to about 13% at take-over (Fig. 5) and the majority of the cells immunopositive to the anti-Bax antibody was represented by large MLC containing engulfed material in their vacuoles (Fig. 4d). The fraction of hemocytes recognized by antibodies raised against caspase-3 and -7 was significantly increased (p < 0.001) at take-over when compared with mid-cycle whereas the opposite was true for cells recognized by anticaspase-8 antibodies (Fig. 5). Most hemocytes recognized by the above antibodies were phagocytes (Fig. 4f, h and j). In controls with absorbed antibodies, hemocytes were not or only slightly stained (Fig. 4c, e, g, i, k, m, o and q). No labeling was observed in negative controls. Significant (p < 0.01) increases were also observed when the activity of both caspase-3 and -8 from hemocyte lysates at takeover were compared with mid-cycle (Table 2).

Table 2 Specific activity of caspase-3 and -8 from haemocyle lisates during mid-cycle and take-over. Asterisks mark significant differences. Caspase

Activity (U/mg protein) Mid-cycle

Take-over

Caspase-3 Caspase-8

39.3  2.1 5.1  2.0

120.3  10.8** 32.4  4.2**

One unit of caspase activity is defined as the quantity of enzyme able to cleave 1 nmole/h of the colorimetric substrate Ac-DEVD-pNA or Ac-IETD-pNA, for caspase3 and -8, respectively at 37 8C. ** p < 0.01.

278

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

Fig. 4. Histochemical and immunocytochemical analysis on Botryllus circulating hemocytes. (a) MLC (asterisk), with three vacuoles containing ingested cells, labeled by DCFH-DA; arrowhead: unlabeled morula cell. (b–q) Hemocytes labeled by antibodies raised against Bcl-2 (b), Bax (d), caspase-8 (f), caspase-7 (h), caspase-3 (j), Fas (l), FasL (n and p) and their respective controls (c, e, g, i, k, m, o and q). (r and s) Cytochemical analysis with LFA. The surfaces of hyaline amoebocytes and young, round hemocytes are recognized by the lectin (r), whereas labeling disappear from MLC (s). (t) MLC containing TUNEL-positive cells having ingested, in turn, other senescent cells (asterisks); arrowheads: nucleus. (u and v) Annexin-V-positive MLC having ingested yeast cells in vitro. (w) MLC with ingested PKH26-labeled cells. Scale bars: 10 mm.

3.7. Membrane proteins recognized by anti-Fas and -FasL increase their expression at take-over The percentage of hemocytes recognized by the anti-Fas antibody, as measured by flow cytometry, amounted to about

4% at mid-cycle and increased to 20% at take-over (Fig. 6). This matches immunohistochemical results indicating 5% of hemocytes recognized by the anti-Fas antibody at mid-cycle and 19% at the generation change (Fig. 5). In all considered phases, hemocytes expressing molecules recognized by the anti-Fas antibody were mainly represented by MLC (Fig. 4l), although some labeled cytotoxic morula cell could be found. As regards FasL, the fraction of immunopositive cells, as estimated by cytometric analysis, reached a maximum value of about 50% at take-over but only amounted to 30% at mid-cycle (Fig. 6). This was confirmed by immunocytochemical analysis indicating 33% of immunopositive cells at mid-cycle and 54% at the generation change (Fig. 5). The great majority of cells recognized by the antibody was represented by immunocytes, both phagocytes and morula cells (Fig. 4n and p). In both cases, hemocytes were not or only slightly stained by absorbed antibodies (Fig. 5m and o). No labeling was observed in negative controls. 3.8. MLC are not recognized by LFA at take-over

Fig. 5. Percentage of haemocytes recognized by antibodies against apoptotic markers at mid-cycle (light grey bars) and take-over (dark grey bars). Significant differences with respect to mid-cycle are marked by asterisks. *p < 0.05; ***p < 0.001.

WGA, RCA and ConA recognized all the hemocytes and no differences were observed in the labeling pattern when mid-cycle and take-over were compared. Conversely, LFA strongly labeled the surface of young round hemocytes and HA, (Fig. 4r), the prevailing phagocyte-type at mid-cycle, whereas it did not recognize MLC, abundant at take-over (Fig. 4s).

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

279

Fig. 6. Cytofluorimetric analysis of the expression of molecules recognized by anti-Fas and anti-FasL antibodies on hemocytes at mid-cycle and take-over.

3.9. Morphological changes of phagocytes at take-over Frequently, at take-over, MLC contained TUNEL-positive MLC with ingested cells inside their vacuoles (Fig. 4t). Often, MLC were in close contact through their plasma membranes and formed small aggregates (Fig. 7a). In some cases, MLC showed typical signs of apoptosis, since their nuclear chromatin was compacted into uniformly dense masses that abut on the nuclear envelope (Fig. 7b and c). Large magnifications of contact region between two MLC showed that plasma membranes were separated by narrow intercellular clefts crossed by fibrous material (Fig. 7d–f). Analysis of a large number of ultrathin sections revealed processes of internalization of cells heavily filled with phagosomes by other MLC. In some cases, dying phagocytes were partially surrounded with long, narrow projections or pseudopodia of phagocytising MLC (Fig. 7g); in other cases, senescent MLC were completely engulfed by active MLC. The same specialized junctional areas recognized in clustered MLC were found in phagocytes interacting with dying MLC (Fig. 7h and i). Different MLC, not necessarily adherent, occasionally contacted with their pseudopodia the same apoptotic cell. 3.10. Ingestion of foreign cells induces apoptosis in phagocytes and their clearance by other phagocytes To evaluate whether apoptosis can be induced in Botryllus phagocytes upon their ingestion of foreign cells, fluorescent Annexin-V was used to recognize cells expressing PS on their surface after the incubation of hemocytes with yeast cells, in vitro. A fraction of 4.3  0.4% of phagocytes (MLC) with ingested yeast resulted fluorescein-labeled on their surface after 60 min of incubation with target cells (Fig. 4u and v), whereas only 0.4  0.1% were positive in controls. When hemocytes were incubated with PKH26-labeled hemocytes from the same colony, previously incubated with yeast cells, MLC containing labeled, yeast-containing phagocytes were observed (Fig. 4w). In addition, when hemocytes were challenged with yeast for only 5 min and the percentage of phagocytes with

ingested material was evaluated in the following 60 min, the fraction of MLC cells increased from 1 to 12%, whereas that of HA decreased from 31 to 19% (Fig. 8). 4. Discussion Research on apoptosis and related phenomena, mainly based on in vitro systems, has greatly increased in the last few decades, leading to great enhancement of our knowledge on programmed cell death, its biochemical control and fate of effete cells. However, we still need simple model organisms useful for suitable in vivo approaches and to replace vertebrate species, the use of which is limited by ethical restrictions. From this point of view, the colonial ascidian B. schlosseri is a reliable model species for studying apoptosis as colonies are characterized by recurrent generation changes during which diffuse, natural cell death occurs in zooid tissues [17–19,21,33–35]. These events mark the end of the colonial blastogenetic cycles, the length of which is temperaturedependent and regulated by diffusible humoral factors [21,36,37]. In addition, cell death and blastogenesis are closely related, suggesting the existence of cross-talk between old zooids and developing buds [21,37–39]. Dead cells and corpses are quickly ingested by circulating phagocytes infiltrating senescent tissues [18,33,37,40], which allows the study of this interaction in an invertebrate species closely related to vertebrates, and alternative to Drosophila and Coenorhabditis. However, despite the abundance of studies on programmed cell death at take-over in the last two decades (see [35] for a review), Botryllus apoptosis and its molecular control are still little characterized and uncertainties persist on the biochemical pathways involved as well as the nature of the cyclical signal(s) inducing recurrent cell death. With the aim of contributing to fill this gap, we performed a series of morphological studies on colony sections, immunoblot analyses on whole colony or hemocyte lysates, and immunocytochemical and cytofluorimetric studies on hemocytes. As already stated [19], the latter represent a good reference tissue for the study of cellular events during the generation change as they are easy to obtain and their collection

280

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

Fig. 7. Transmission electron microscopy of MLC at take-over. (a) Aggregated MLC with phagosomes of different size (arrowheads); engulfed apoptotic nuclei are visible (asterisks). Scale bar: mm 30. (b and c) Apoptotic phagocyte with phagosomes (asterisks) and nucleus (n) with irregular profile and dense chromatin along the nuclear envelope; rer: rough endoplasmic reticulum. The squared area in (b) is enlarged in (c). Scale bar 20 mm in (b) and 10 mm in (c). (d–f) Two adhering MLC rich in phagosomes (asterisks); note the strictly apposed plasma membranes (arrowheads), progressively enlarged in (e and f). Scale bar: 50, 15 and 2 mm in (d–f), respectively. (g–i) Phagocyte in the process of internalization, by means of long pseudopods (p), of a dying macrophage-like cell (dMLC); the presence of large phagosomes (asterisks) characterizes both the cells. Scale bar: 20 mm. Boxed areas bordered by continuous line are enlarged in (h) (scale bar: 1.5 mm) and (i) (scale bar: 3 mm) to show points of close membrane apposition between the two MLC (arrowheads). The boxed area bordered by dotted line is enlarged in the inset to show the pseudopodium of the phagocytosing cell; n, nucleus of engulfing phagocyte. Scale bar: 10 mm.

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

Fig. 8. Percentage of HA (light grey bars) and MLC (dark grey bars) at different times after a 5-min challenge with yeast cells. Significant differences with respect to controls are marked by asterisks. *p < 0.05; **p < 0.01; ***p < 0.001.

does not involve serious damage to the colonies, which can be repeatedly bled during their lifespan. 4.1. The molecular machinery of apoptosis A variety of molecular markers, such as TUNEL and comet assays for DNA fragmentation, Trypan blue exclusion assay for plasma membrane permeability, Annexin-V labeling for phosphatidylserine on the plasma membrane external layer, were previously used to demonstrate the occurrence of massive apoptosis during the take-over in B. schlosseri; during this period, a marked activation of caspase-3 and -9 also occurs [18,19,35]. The different expression of both initiator and effector caspases during the generation change was suggested in the present work by immunoblot, and immunocytochemical analysis. Although we do not exactly know the nature of the antigens recognized by the antibodies, we hypothesize the presence of caspase genes, orthologous to the mammalian caspase genes, in the Botryllus genome which, unfortunately, is still largely unknown. This idea is supported by the presence of orthologue genes for caspases and other genes of the apoptotic machinery in lower invertebrates, which suggests a high degree of conservation of the complex apoptotic pathways in metazoan evolution [41–44]. In particular, multiple genes for different caspases, showing homology with mammalian caspase genes, have been described in the solitary ascidians C. intestinalis and Ciona savignyi, and other tunicates [10,11,13]. In colonies homogenates, the antibody against initiator caspase-8 recognized a band of 54 kDa which almost disappeared at take-over. We hypothesize that this protein corresponds to the inactive zymogen, as suggested by the molecular size similar to mammalian procaspase-8 [45], which disappears from the cell cytoplasm when it is converted to active enzyme. In addition, our protein lies within the size range of 500 and 700 amino acids reported for the putative procaspase-8 homologues of C. intestinalis and C. savignyi [13] and the band disappeared when the antibody was previously incubated with homogenates from colonies at takeover which contain high number of apoptotic cells. Moreover, immunocytochemical analysis indicates that the majority of immunopositive cells are represented by hyaline amoebocytes and their frequency significantly decreases at take-over with respect to mid-cycle, in accordance with previous data indicating a reduction of this cell-type during the generation change [19]. In agreement with the above assumption, an increase in the specific activity of caspase-8, at take-over with respect to mid-cycle was measured in Botryllus hemocyte lysates. Specific caspase-8 activity has also been detected during the development of the solitary ascidian Boltenia villosa [46] and putative genes for initiator caspase-8 have been described in the amphioxus genome [47].

281

The activation of initiator caspases (the change in expression of a protein, ascribable to active caspase-9, during the blastogenetic cycle, was already demonstrated in a previous paper [19]) leads to consequent activation of effector caspases: this is in agreement with the increased density of the bands recognized by antibodies raised against caspase-3 and -7, two of the major effector caspases [9]. In our opinion, the 21-kDa band, immunoreactive to anticaspase-3 antibody, represents active caspase-3 as it corresponds to the size of the mammalian enzyme and an increase in caspase-3 activity has been observed during take-over in haemocytes lysates. The 38-kDa band recognized by anti-caspase-7 antibody corresponds in size to the inactive mammalian proenzyme and results overexpressed during the generation change as recently reported for caspase-3 proform [19]. In both cases, the use of absorbed antibodies leads to a marked decrease in the intensity of both the recognized bands and staining in immunocytochemical analysis. As caspase-3 and -7 show high sequence similarity, we cannot say if the two bands responsive to anti-caspase-3 and anti-caspase-7 antibodies represent two different gene products or the precursor and the active form of the same gene product: caspase genes with similarity to both mammalian caspase-3 and -7 are considered ancestral founder genes from which the variety of vertebrate effector caspases derived and have been reported in Ciona, larvaceans and amphioxus [11,13,48]. The presence of active caspases at the generation change is also demonstrated by the cleavage of the protein recognized by the anti-PARP antibody, probably a PARP orthologue as suggested by the molecular size of both the active and the cleaved form, resembling the corresponding forms of mammalian PARP. In this case, the use of the absorbed antibody results in the complete disappearance of the recognized bands. PARP, known as an early marker of apoptosis [49,50], is a key enzyme in DNA repair and its inhibition promotes DNA fragmentation and cell death via apoptosis [51,52]. Putative PARP orthologues have been annotated in the genome of Ciona savignyi (ENSCSAVG00000004832 and ENSCSAVG00000004652, respectively) and of the cephalochordate Branchiostoma floridae (BRAFLDRAFT_125672). In addition, the cleavage of a PARP orthologue in ascididemnin-induced apoptosis, resulting in electrophoretic bands of 116 and 89 kDa, has been reported in the ascidian Cystodytes dellechajei [53]. As regards caspase activation, the fivefold increase in the density of the 37-kDa band in cytosol extracts, recognized by anticytochrome c antibody at take-over, indicates the involvement of mitochondria in the process. We can exclude the possibility that the release of cytochrome c is due to mitochondrial damage as the activity of citrate synthase, a specific marker of the mitochondrial matrix, was never detected in cytosol extracts. The observed significant increase in the fraction of hemocytes expressing molecules immunoreactive to anti-Bax antibody and the parallel decrease in cells immunostained by anti-Bcl-2 indicate that, at take-over, cells are more prone to undergo apoptosis than in midcycle phases. This matches the idea that cell survival is controlled by the balance between pro- and anti-apoptotic factors, controlling the release of cytochrome c, as in Vertebrates [54]. A similar balance between undefined apoptotic and anti-apoptotic signals has been invoked to explain the regulation of the colonial blastogenetic cycle in Botryllus [36] and finds confirmation in the recent observation of an increase in the expression of proteins recognized by anti-Bax and a parallel decrease in anti-Bcl-2 immunoreactivity in cells of the digestive system during the generation change [55]. In addition, putative orthologues of Bax and Bcl-2 are expressed during C. intestinalis development [12,14]. At take-over, we also observed an increase in the density of the electrophoretic bands recognized by anti-Fas and anti-FADD antibodies, the molecular weights of which resembles those of mammalian specific antigens, suggesting the involvement of the

282

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

extrinsic, receptor-mediated pathway in the induction of apoptosis. In both cases, the specificity of the antibody recognition is suggested by the observed decrease of the intensity of the band recognized by anti-Fas and anti-FADD when antibodies absorbed with HEP-G2 cells or homogenate from colony at take-over, respectively, were used. Fas (CD95) is a type I membrane protein (45 kDa in mammals) and is one of the best-known death receptors. It recognizes and binds its ligand, FasL, a type II membrane protein (40 kDa in mammals), and a member of the tumor necrosis factor (TNF) family. The interaction with the ligand induces clustering of the receptor cytoplasmic death domains and recruitment of adaptor proteins (FADD) and procaspase-8 to constitute the death-inducing signal complex (DISC), where active caspase-8 forms by autoproteolysis of procaspase-8 [56–58]. Homologues of death receptors and FasL have been detected in the C. intestinalis genome [11,59]. The occurrence of receptormediated apoptosis, at least in hemocytes, is supported by cytofluorimetric and immunocytochemical analysis, indicating an increase in the fraction of circulating cells recognized by antiFas and anti-FasL at take-over. In mammals, the intrinsic and extrinsic caspase-activating pathways can cross-talk through caspase-8-mediated proteolysis of the BH3-only protein Bid. Truncated Bid allows oligomerization of the pro-apoptotic proteins Bax or Bak in the outer mitochondrial membrane and the consequent leakage of cytochrome c [4,7,20,60,61]. Our observations indicating decreased density of the band recognized by anti-Bid antibody at take-over suggest the occurrence of such an event also in Botryllus apoptosis. Although a homologous Bid was not found in the Ciona genome [46,59], this may be due to a lack of significant motifs allowing reliable predictions based on sequences [11]. 4.2. Efferocytosis: a commitment to death? Ascidian blood consists of colorless plasma and circulating hemocytes, represented by immunocytes, either phagocytes (HA and MLC) or cytotoxic morula cells (MC), storage cells (pigment cells and nephrocytes) and undifferentiated hemoblasts [16]. From previous studies, we know that a fraction of hemocytes, up to 30% of circulating cells, undergoes apoptosis at take-over and is replaced by new, young cells entering the circulation at this phase [19]. Cytofluorimetric analysis showed that, at take-over, the population of circulating cells with high forward and side scatter, amounting to less than 4% at mid-cycle, rises to a value around 20% at take-over. From these characteristics and the fact that these cells are almost absent at mid-cycle, we infer they are represented by MLC containing vacuoles with ingested materials in their cytoplasm. This assumption matches the results of our immunocytochemical analyses on hemocytes with anti-caspase-3, anticaspase-7, anti-Bax and anti-Fas, showing that the majority of cells recognized by these antibodies are represented by MLC and their number, at take-over, ranges from 12 to 26%, according to type of experiment. In addition, cells of this population do not express PS on the outer leaflet of the plasma membrane at mid-cycle; conversely, almost 90% of the hemocytes in this cell population are labeled by Annexin-V at take-over. This gives interesting indications on the nature of circulating senescent cells and suggests that, as phagocytosis of dying cells (efferocytosis) occurs, phagocytes themselves are directed towards apoptosis. Electron microscopy clearly demonstrated the occurrence, at take-over, of circulating MLC with condensed chromatin inside their nuclei, a typical sign of apoptosis, and the death of phagocytes after the ingestion of dying cells has already been reported during the metamorphosis of B. schlosseri [62]. This may be the consequence of an excessive oxidative stress due to the production of high levels of reactive oxygen species during the sustained respiratory burst related to

phagocytosis. Oxidative stress is a good apoptosis inducer [63], as it leads to over expression of p53 which, in turn, activates senescence genes and pro-apoptotic Bcl-2 family proteins which make motochondria permeable [64,65] with consequent release of cytochrome c. Proteins of the p53 family, are expressed in C. intestinalis [64,66]. However, stress can also lead to increased expression of Fas on the surface of injured cells [67]. In line with the above hypothesis, our data indicates that, in addition to their increase in number at take-over, most MLC contain detectable cytosol peroxides even in the presence of reduced glutathione and glutathione peroxidase, two of the most important intracellular antioxidant systems [68] that have been previously demonstrated in Botryllus phagocytes [69,70], and, at take-over, the concentration of these oxygen derivatives in BP increases twofold with respect to mid-cycle. The abundance of oxidation products in the residual zooids, which persist in the form of dark brown spheres in the centre of colonial systems [55], is in agreement with the occurrence of oxidative stress during take-over. The induction of apoptosis by excessive oxidative burst consequent to phagocytosis has been reported in mammalian neutrophils [68,71,72] and is suggested by our results with in vitro yeast phagocytosis indicating that, after ingesting yeast cells, MLC can express PS. Our observation that MLC represent the great majority of hemocytes recognized by the anti-Fas antibody, the number of which, from mid-cycle to take-over, increases fivefold to reach 20% of total hemocytes, agrees with the above results and suggest that, after engulfing senescent cells, phagocytes undergo metabolic modifications which direct them towards apoptosis, and the appearance of membrane Fas is one of these changes. The lower fraction of cells recognized by the anti-Fas antibody at mid-cycle is probably due to the clearance of hemocytes committed to death once they interact with cells expressing membrane ligands for the death receptor. The number of hemocytes expressing surface molecules recognized by anti-FasL antibody peaks at take-over, so we may argue that these molecules are involved in the induction of cell death. Cells expressing such surface molecules are represented by immunocytes, both phagocytes and cytotoxic morula cells which, therefore, play a key role also in the coordination of cell death by apoptosis during the take-over. The 20% increase of immunopositive cells is probably due to the new cell population entering the circulation at this phase [19] most of which replace dying phagocytes. 4.3. Fate of senescent MLC The presence, at take-over, of circulating phagocytes containing TUNEL-positive MLC inside their vacuoles indicate that, once directed towards apoptosis, cells are quickly removed from the circulation through phagocytosis thus contributing to form the bulk of nutrients required to support bud development to adulthood [17,37]. In addition, in vitro experiments with PKH26, a phagocyte-specific dye, indicate that senescent phagocytes can be ingested by other phagocytes, in agreement with what reported for aged human leukocytes [73]. Experiments with short yeast challenges support the above view as they indicate that the frequency of MLC continues to increase even in the absence of yeast: this can be explained assuming that, after the ingestion of foreign material or senescent cells, MLC undergo apoptosis and are ingested by other phagocytes, presumably HA that acquire a MLC morphology, as previously indicated [16,22]. Electron microscopy observations indicate that, during takeover, MLC can closely interact with each other and active MLC extends their pseudopods over apoptotic phagocytes, leading to the formation of intracellular phagosomes, as indicated in Fig. 9. During efferocytosis, healthy and effete phagocytes closely interact, with the formation of a series of junctional complexes

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

Fig. 9. Explicatory scheme of phagocyte efferocytosis in B. schlosseri, as deduced by light and electron microscopy analysis, showing its occurrence through a ‘‘zipperlike’’ mechanism involving thin pseudopodia of active MLC and the formation of specialized junctional areas between the two plasma membranes.

283

erythrocytes associated with a loss of sialic acid-bearing glycoproteins was reported in beta-thalassaemia [83]. Since neither adult zooid nor growing buds open their siphons during take-over, in this period a colony does not feed. The appearance of a new generation of budlets and the maturation to adulthood of buds occur only after the beginning of take-over, suggesting that a colony relies on the recycling of nutrients deriving from efferocytosis for the progression of blastogenetic development. The observed delay in the maturation of a new generation when adults prolong their lifespan [21,36,37] and the precocious bud maturation when take-over is anticipated under stress conditions [17,36,37] confirm the above hypothesis. In addition, the requirement of appropriate phagocytosis for the onset of a new blastogenetic cycle is indicated by the severe impairment of bud development when phagocyte activity is inhibited [39]. The importance of the availability of adequate quantity of nutrients for bud development is suggested by both the higher budding activity, leading to colonies with more than three blastogenetic generations, when only one bud (instead of two or three) per zooid is experimentally left in a colony [37] and the increased size of remaining buds when most of them were experimentally removed from a colony [17,37]. As colonies enter a new blastogenetic cycle, circulating MLC decrease their frequency and empty their vacuoles [84]. Therefore, we can argue that products of the digestion of dying MLC contribute part of nutrients which enable the colony to sustain bud development from stage 8 to stage 9, characterized by an abrupt increase in size [17,21]. Collectively, our results confirm that Botryllus is a useful model species for in vivo studies of apoptosis and its role in the organism as a whole. They are consistent with the following viewpoints: during the take-over massive apoptosis occurs in tissues of adult zooid and hemocytes, involving both Fas-like death receptors and cytochrome c release from mitochondria. The intrinsic and extrinsic caspase activation pathways are linked by a Bid-like protein. At least in the case of hemocytes, senescent cells express various ‘‘eat-me’’ signals for their clearance by circulating phagocytes. As far as the latter are concerned, the ingestion of effete cells render them prone to apoptosis, probably because of oxidative stress consequent to sustained respiratory burst. Future efforts and a better knowledge of Botryllus genome/ transcriptome will certainly confirm the importance of this invertebrate chordate in cell death research. Acknowledgements

in which plasma membranes are separated by a narrow cleft filled with dense, fibrous material. This fits the known mechanism of dying cell removal through efferocytosis, which implies the formation of a series of focal contacts along the contact surface of the active phagocyte, mediated by integrins, which assure a zipper-like progression of pseudopods [74]. The involvement of integrins in Botryllus phagocytosis has been suggested by previous experiments [75] and a homologue of a6 integrin has been reported in the compound ascidian Polyandrocarpa misakiensis [76]. MLC lose the ability to bind LFA, specific for sialic acids [77,78] – which intensely labels the surface of young, undifferentiated hemocytes, likely hemoblasts, and HA – indicating a change in surface glycosylation, probably consequent to the removal of terminal sialic acid by neuraminidase. Analogously to what happens in Vertebrates [79,80], this can represent a further signal for the recognition and clearance of effete cells, in addition to PS and thrombospondin receptor, the involvement of which was previously demonstrated [18]. A similar relationship between apoptosis induction, reduction in sialic acid expression on cell surface and increase in sialidase activity has been observed in mammalian cells [81,82] and a faster clearance of circulating

Authors wish to thank Mr. M. Del Favero for technical help and the two anonymous referees who helped to improve the quality of the manuscript with their observations and suggestions. This research was supported by a grant from the University of Padova (CPDA084343) to LB and GB. References [1] Jacobson MD, Weil M, Raff MC. Programmed cell death in animal development. Cell 1997;88:347–54. [2] Horvitz HR. Genetic control of programmed cell death in the nematode C. elegans. Cancer Res 1999;59:1701–6. [3] Baehrecke EH. How death shapes life during development. Nat Rev Mol Cell Biol 2002;3:779–87. [4] Opferman JT, Korsmeyer SJ. Apoptosis in the development and maintenance of the immune system. Nat Immunol 2003;4:410–5. [5] Penaloza C, Lin L, Lockshin RA, Zakeri Z. Cell death in development: shaping the embryo. Histochem Cell Biol 2006;126:149–58. [6] Kerr JF, Wyllie AH, Currie AR. Apoptosis: a basic biological phenomenon with wide-ranging implications in tissue kinetics. Br J Cancer 1972;26:239–57. [7] Hengartner M. The biochemistry of apoptosis. Nature 2000;407:770–6. [8] Shi Y. Mechanisms of caspase activation and inhibition during apoptosis. Mol Cell 2002;9:459–70. [9] Taylor RC, Cullen SP, Martin SJ. Apoptosis: controlled demolition at the cellular level. Nat Rev Mol Cell Biol 2008;9:231–41.

284

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

[10] Chambon J-P, Suole J, Pomies P, Fort P, Sahuquet A, Alexandre D, et al. Tail regression in Ciona intestinalis (Protochordate) involves a caspase-dependent apoptosis event associated with ERK activation. Development 2002;129: 3105–14. [11] Terajima D, Shida K, Takada N, Kasuya A, Rokhsar D, Satoh N, et al. Identification of candidate genes encoding the core components of the cell death machinery in the Ciona intestinalis genome. Cell Death Differ 2003;10:749–53. [12] Takada N, Yamaguchi H, Shida K, Terajima D, Satou Y, Kasuya A, et al. The cell death machinery controlled by Bax and Bcl-XL is evolutionary conserved in Ciona intestinalis. Apoptosis 2005;10:1211–20. [13] Weill M, Philips A, Chourrout D, Fort P. The caspase family in urochordates: distinct evolutionary fates in ascidians and larvaceans. Biol Cell 2005;97:857–66. [14] Baghdiguian S, Martinand-Mari C, Mangeat P. Using Ciona to study developmental programmed cell death. Semin Cancer Biol 2007;17:147–53. [15] Manni L, Zaniolo G, Cima F, Burighel P, Ballarin L. Botryllus schlosseri: a model ascidian for the study of asexual reproduction. Dev Dyn 2007;236:335–52. [16] Ballarin L, Cima F. Cytochemical properties of Botryllus schlosseri haemocytes: indications for morpho-functional characterisation. Eur J Histochem 2005;49:255–64. [17] Lauzon RJ, Ishizuka KJ, Weissman IL. A cyclical, developmentally-regulated death phenomenon in a colonial urochordate. Dev Dyn 1992;194:71–83. [18] Cima F, Basso G, Ballarin L. Apoptosis and phosphatidylserine-mediated recognition during the take-over phase of the colonial life-cycle in the ascidian Botryllus schlosseri. Cell Tissue Res 2003;312:369–76. [19] Ballarin L, Menin A, Tallandini L, Matozzo V, Burighel P, Basso G, et al. Haemocytes and blastogenetic cycle in the colonial ascidian Botryllus schlosseri: a matter of life and death. Cell Tissue Res 2008;331:555–64. [20] Strasser A. The role of BH3-only proteins in the immune system. Nature Rev Immunol 2005;5:189–200. [21] Sabbadin A. Osservazioni sullo sviluppo, l’accrescimento e la riproduzione di Botryllus schlosseri (Pallas), in condizioni di laboratorio. Boll Zool 1955;22:243–65. [22] Ballarin L, Cima F, Sabbadin A. Phagocytosis in the colonial ascidian Botryllus schlosseri. Dev Comp Immunol 1994;18:467–81. [23] Ballarin L, Cammarata M, Cima F, Grimaldi A, Lorenzon S, Malagoli D, et al. Immune-neuroendocrine biology of invertebrates: a collection of methods. Invertebr Survival J 2008;5:192–215. [24] Gorman A, McCarthy J, Finucane D, Reville W, Cotter T. Morphological assessment of apoptosis. In: Cotter TG, Martin SJ, editors. Techniques in apoptosis. A user’s guide. London: Portland; 1997. p. 1–20. [25] Edds KT. Morphological and cytoskeletal transformations in sea urchin coelomocytes. In: Cohen WD, editor. Blood cells of marine invertebrates experimental systems in cells, vol. 4. New York: AR Liss; 1985. p. 53–74. [26] Pick E. Microassays for superoxide and hydrogen peroxide production and Nitroblue Tetrazolium reduction using an enzyme immunoassay microplate reader. Methods Enzymol 1986;132:407–21. [27] Bradford MM. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein dye binding. Anal Biochem 1976;72:248–54. [28] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–5. [29] Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 1979;76:4350–4. [30] Enari M, Hase A, Nagata S. Apoptosis by a cytosolic extract from Fas-activated cells. EMBO J 1995;14:5201–8. [31] Srere PA. Citrate synthase. Methods Enzymol 1969;13:3–11. [32] Berrill NJ. The development of the bud in Botryllus. Biol Bull 1941;80:169–84. [33] Burighel P, Schiavinato A. Degenerative regression of the digestive tract in the colonial ascidian Botryllus schlosseri (Pallas). Cell Tissue Res 1984;235:309–18. [34] Tiozzo S, Ballarin L, Burighel P, Zaniolo G. Programmed cell death in vegetative development: apoptosis during the colonial life-cycle of the ascidian Botryllus schlosseri. Tissue Cell 2006;38:193–201. [35] Ballarin L, Burighel P, Cima F. A tale of death and life: natural apoptosis in the colonial ascidian Botryllus schlosseri (Urochordata, Ascidiacea). Curr Pharm Des 2008;14:138–47. [36] Watanabe H. Studies on the regulation in fused colonies in Botryllus primigenus (Ascidiae Compositae). Sci Rep Tokyo Bunrika Daigaku B 1953;7:83–198. [37] Lauzon RJ, Kidder SJ, Long P. Suppression of programmed cells death regulated the cyclical degeneration of organs in a colonial urochordate. Dev Biol 2007;301:92–105. [38] Sabbadin A. Analisi sperimentale dello sviluppo delle colonie di Botryllus schlosseri (Pallas) [Ascidiacea]. Arch It Anat Embriol 1958;63:178–221. [39] Lauzon RJ, Ishizuka KJ, Weissman IL. Cyclical generation and degeneration of organs in a colonial urochordate involves crosstalk between old and new: a model for development and regeneration. Dev Biol 2002;301:92–105. [40] Voskoboynik A, Rinkevich B, Weiss A, Moiseeva E, Reznick AZ. Macrophage involvement for successful degeneration of apoptotic organs in the colonial urochordate Botryllus schlosseri. J Exp Biol 2004;207:2409–16. [41] Bergmann A, Agapite J, Steller H. Mechanisms and control of programmed cell death in invertebrates. Oncogene 1998;17:3215–23. [42] Wiens M. Grundlegende Mechanismen der Apoptose in den einfaschsten Invertebraten, den Porifera. Z Gerontol Geriat 2004;37:190–9. [43] Dunn SR, Phillips WS, Spatafora JW, Green DR, Weis VM. Highly conserved caspase and Bcl-2 homologues from the sea anemone Aiptasia pallida: lower metazoans as models for the study of apoptosis evolution. J Mol Evol 2006;63:95–107.

[44] Seipp S, Wittig K, Stiening B, Bo¨ttger A, Leitz T. Metamorphosis of Hydractinia echinata (Cnidaria) is caspase-dependent. Int J Dev Biol 2006;50:63–70. [45] Qin ZH, Wang Y, Kikly KK, Sapp E, Kegel KB, Aronin N, et al. Pro-caspase-8 is predominantly localized in mitochondria and released into cytoplasm upon apoptotic stimulation. J Biol Chem 2001;276:8079–86. [46] Bates WR. Cellular features of an apoptotic form of programmed cell death during the development of the ascidian, Boltenia villosa. Zool Sci 2004;21:553– 63. [47] Huang S, Yuan S, Guo L, Yu Y, Li J, Wu T, et al. Genomic analysis of the immune gene repertoire of amphioxus reveals extraordinary innate complexity and diversity. Genome Res 2008;18:1112–26. [48] Bayascas JR, Yuste VJ, Benito E, Garcia-Ferna`ndez J, Comella JX. Isolation of AmphiCASP-3/7, an ancestral caspase from amphioxus (Branchiostoma floridae). Evolutionary considerations for vertebrate caspases. Cell Death Differ 2002;9:1078–89. [49] Kaufmann SH, Desnoyers S, Ottaviano Y, Davidson NE, Poirier GG. Specific proteolytic cleavage of poly(ADP-ribose) polymerase: an early marker of chemotherapy-induced apoptosis. Cancer Res 1993;53:3976–85. [50] Wang Z-Q, Stingl L, Morrison C, Jantsch M, Los M, Schulze-Osthohh K, et al. PARP is important for genomic stability but dispensable in apoptosis. Genes Dev 1997;11:2347–58. [51] Oliver FJ, Menissier-de Murcia J, de Murcia GJ. Poly(ADP-ribose) polymerase in the cellular response to DNA damage, apoptosis, and disease. Am J Hum Genet 1999;64:1282–8. [52] Schreiber V, Dantzer F, Ame´ J-C, de Murcia G. Poly(ADP-ribose): novel functions for an old molecule. Nature Rev Mol Cell Biol 2006;7:517–28. [53] Dassonneville L, Wattez N, Baldeyrou B, Mahieu C, Lansiaux A, Banaigs B, et al. Inhibition of topoisomerase II by the marine alkaloid ascididemin and induction of apoptosis in leukemia cells. Biochem Pharmacol 2000;60:527–37. [54] Krajewska M, Mai JK, Zapata JM, Ashwell KWS, Schendel SL, Reed JC, et al. Dynamics of expression of apoptosis-regulatory proteins Bid, Bcl-2, Bcl-X, Bax and Bak during development of murine nervous system. Cell Death Differ 2002;9:145–57. [55] Cima F, Ballarin L. Apoptosis and pattern of Bcl-2 and Bax expression in the alimentary tract during the colonial blastogenetic cycle of Botryllus schlosseri (Urochordata, Ascidiacea). Ital J Zool 2009;76:28–42. [56] Nagata S, Golstein P. The Fas death factor. Science 1995;267:1449–56. [57] Krammer PH. CD95’s deadly mission in the immune system. Nature 2000;407:789–95. [58] Askenasy N, Yolcu ES, Yaniv I, Shirwan H. Induction of tolerance using Fas ligand: a double-edged immunomodulator. Blood 2005;105:1396–404. [59] Dehal P, Satou Y, Campbell RK, Chapman J, Degnan B, De Tomaso A, et al. The draft genome of Ciona intestinalis: insight into chordate and vertebrate origins. Science 2002;298:2157–67. [60] Borner C. The Bcl-2 protein family: sensors and checkpoints for life-or-death decisions. Mol Immunol 2003;39:615–47. [61] Oh KJ, Barbuto S, Pitter K, Morash J, Walensky LD, Korsmeyer SJ. A membranetargeted Bid Bcl-2 homology 3 peptide is sufficient for high potency activation of Bax in vitro. J Biol Chem 2006;281:36999–7008. [62] Schiaffino S, Burighel P, Nunzi MG. Involution of the caudal musculature during metamorphosis in the ascidian, Botryllus schlosseri. Cell Tissue Res 1974;153:293–305. [63] Kuwabara M, Asanuma T, Niwa K, Inanami O. Regulation of cell survival and death signals induced by oxidative stress. J Clin Biochem Nutr 2008;43:51–7. [64] Vousden KH. Lane DP p53 in health and disease. Nat Rev Mol Cell Biol 2007;8:275–83. [65] Wang P, Yu J, Zhang L. The nuclear function of p53 is required for PUMAmediated apoptosis induced by DNA damage. Proc Natl Acad Sci USA 2007;104:4054–9. [66] Nedelcu AM, Tan C. Early diversification and complex evolutionary history of the p53 tumor suppresor gene family. Dev Genes Evol 2007;217:801–6. [67] Suhara T, Fukuo K, Sugimoto T, Morimoto S, Nakahashi T, Hata S, et al. Hydrogen peroxide induces up-regulation of Fas in human endothelial cells. J Immunol 1998;160:4042–7. [68] Splettstoesser WD, Schuff-Werner P. Oxidative stress in phagocytes: ‘‘The enemy within’’. Microsc Res Technol 2002;57:441–55. [69] Cima F, Dominici D, Ballarin L, Burighel P. Influence of TBT on activity of detoxifying enzymes from haemocytes of a colonial ascidian. Fresen Environ Bull 2002;11:573–7. [70] Cima F, Ballarin L. TBT-sulfhydryl interaction as a cause of immunotoxicity in Tunicates. Ecotoxicol Environ Saf 2004;58:386–95. [71] Watson RW, Redmond HP, Wang JH, Condron C, Bouchier-Haydes D. Neutrophils undergo apoptosis following ingestion of Escherichia coli. J Immunol 1996;156:3986–92. [72] Yan SR, Bortolussi R, Issekutz TB, Issekutz AC. Increased chemoattractant induced neutrophil oxidative burst, accelerated apoptosis, and dysregulated tyrosine phosphorylation associated with lifelong bacterial infections. Clin Immunol 2005;117:36–47. [73] Savill JS, Henson PM, Haslett C. Phagocytosis of aged human neutrophils by macrophages is mediated by a novel ‘‘charge-sensitive’’ recognition mechanism. J Clin Invest 1989;84:1518–27. [74] Krysko DV, Brouckaert G, Vandenabeele P, D’Herde K. Mechanisms of internalization of apoptotic and necrotic L929 cells by a macrophage cell line studied by electron microscopy. J Morphol 2003;258:336–45. [75] Ballarin L, Scanferla M, Cima F, Sabbadin A. Phagocyte spreading and phagocytosis in the compound ascidian Botryllus schlosseri: evidence for an integrin-

F. Cima et al. / Developmental and Comparative Immunology 34 (2010) 272–285

[76] [77]

[78] [79]

like, RGD-dependent recognition mechanism. Dev Comp Immunol 2002;26: 345–54. Kawamura K, Sugino YM. Cell adhesion in the process of asexual reproduction of Tunicates. Microsc Res Technol 1999;44:269–78. Miller RL, Collawn Jr JF, Fish WW. Purification and macromolecular properties of a sialic acid-specific lectin from the slug Limax flavus. J Biol Chem 1982;257:7574–80. Lehmann F, Tiralongo E, Tiralongo J. Sialic acid-specific lectins: occurrence, specificity and function. Cell Mol Life Sci 2006;63:1331–54. Witting A, Mu¨ller P, Herrmann A, Kettenmann H, Nolte C. Phagocytic clearance of apoptotic neurons by microglia/brain macrophages in vitro: involvement of lectin-, integrin-, and phosphatidylserine-mediated recognition. J Neurochem 2000;75:1060–70.

285

[80] Dini L, Pagliara P, Carla` EC. Phagocytosis of apoptotic cells by liver: a morphological study. Microsc Res Technol 2002;57:530–40. [81] Azuma Y, Taniguchi A, Matsumoto K. Decrease in cell surface sialic acid in etoposide-treated Jurkat cells and the role of cell surface sialidase. Glycoconjugate J 2000;17:301–6. [82] Balcan E, S¸ahin M. Apoptosis-sialic acid relationship at different developmental stages of murine thymus. Pak J Biol Sci 2006;9:1852–60. [83] Basu S, Banerjee D, Chandra S, Chakrabarti A. Loss of phospholipid membrane asymmetry and sialylated glycoconjugates from erythrocyte surface in haemoglobin E beta-thalassaemia. Br J Haematol 2008;141: 92–9. [84] Cima F, Ballarin L, Sabbadin A. New data on phagocytes and phagocytosis in the compiund ascidian Botryllus schlosseri. Ital J Zool 1996;63:357–64.