Oxidative stress and apoptotic events during thermal stress in the symbiotic sea anemone, Anemonia viridis Sophie Richier 1 ,Ce ´ cile Sabourault 1 , Juliette Courtiade 1 , Nathalie Zucchini 3 , Denis Allemand 1,2 and Paola Furla 1 1 UMR 1112 UNSA-INRA ROSE, Nice-Sophia Antipolis University, Nice, France 2 Centre Scientifique de Monaco, Monaco 3 UMR 1112 UNSA-INRA ROSE, Sophia-Antipolis, France Over the past several decades, symbiotic invertebrates such as cnidarians, sponges and mollusks have been regularly affected by a phenomenon known as ‘bleach- ing’. This event has been observed all around the world and involves principally the mass expulsion of unicel- lular photosynthetic symbionts from animal tissue. Bleaching phenomenon has been widely reported in the ecologically and economically important tropical corals reef, but several other invertebrates such as giant clams, gorgonians and sea anemones have also been affected. Previous studies have established a causal link between environmental stresses, such as elevated tem- perature, ultraviolet light, pathogen infection or pol- lution, and symbiosis disruption (reviewed in [1]). Mass symbiont expulsion has, however, been most fre- quently associated with elevated seawater temperature, generally considered to be the primary stress causing worldwide bleaching [2]. General mechanisms have been proposed to explain the thermal sensitivity of symbiotic cnidarians including symbiont photoinhibi- tion [3–6], cell degradation [7] and cell death [8,9]. However, despite the importance of the phenomenon, the underlying molecular mechanisms associated with symbiosis breakdown remained undetermined. Oxidative stress is one molecular pathway that has been suggested to cause bleaching. A pro-oxidant per- iod is experienced daily by invertebrates harboring pho- tosynthetic symbionts due to the high concentration of oxygen produced throughout photosynthesis [10–13]. The light-dependent hyperoxic state induces high fluxes of reactive oxygen species (ROS) such as O À 2 and OH • [14] produced largely from mitochondria and chloro- plasts. This increase in ROS is counterbalanced by an efficient antioxidant capacity in the host and symbiont cells [12,13,15,16]. The first hypothesis of oxidative stress involvement in the bleaching event was proposed, Keywords apoptosis; bleaching; caspase; cnidarian; oxidative stress Correspondence P. Furla, UMR 1112 ROSE, Nice-Sophia Antipolis University, Parc Valrose, BP 71, F-06108 Nice Cedex 2, France Fax: +33 4 92 07 65 63 Tel: +33 4 92 07 68 30 E-mail: furla@unice.fr (Received 6 April 2006, revised 30 June 2006, accepted 11 July 2006) doi:10.1111/j.1742-4658.2006.05414.x Symbiosis between cnidarian and photosynthetic protists is widely distri- buted over temperate and tropical seas. These symbioses can periodically breakdown, a phenomenon known as cnidarian bleaching. This event can be irreversible for some associations subjected to acute and ⁄ or prolonged envi- ronmental disturbances, and leads to the death of the animal host. During bleaching, oxidative stress has been described previously as acting at mole- cular level and apoptosis is suggested to be one of the mechanisms involved. We focused our study on the role of apoptosis in bleaching via oxidative stress in the association between the sea anemone Anemonia viridis and the dinoflagellates Symbiodinium species. Characterization of caspase-like enzymes were conducted at the biochemical and molecular level to confirm the presence of a caspase-dependent apoptotic phenomenon in the cnidarian host. We provide evidence of oxidative stress followed by induction of caspase-like activity in animal host cells after an elevated temperature stress, suggesting the concomitant action of these components in bleaching. Abbreviations AFC, 7-amino-4-trifluromethylcoumarin; CARD, caspase recruitment domain; CHO, adelhyde; DEVD, Asp-Glu-Val-Asp; IETD, Ile-Glu-Thr-Asp; ROS, reactive oxygen species; TUNEL, dUTP nick end labeling. 4186 FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS and later supported, by Lesser and coworkers [17–20]. These studies demonstrated the role of ROS production in the temperature-induced bleaching. During thermal stress, although the enzymatic antioxidant defenses are induced [18,21–23], the additional amount of ROS pro- duction causes a large increase in cellular damage such as protein carbonylation [21,22,24], lipid peroxidation [24] and DNA degradation [23]. ROS could be involved in cell death by two path- ways: (a) they could cause oxidative stress that leads to massive cellular damage [25] and they could be involved in necrosis or in so-called postmitochondrial phase of apoptosis [26]; or (b) they could be involved in the initiation phase of apoptosis contributing to cell death signaling [27]. Programmed cell death is known to model tissue dur- ing embryogenesis, to remove damaged cells, protect against pathogen infection, and regulate cell numbers and tissue homeostasis. Program cell death is character- istic of all multicellular animals and can be extended now to the most basal metazoan phyla as porifera and cnidaria with occurrence of apoptosis and homologues of caspases and Bcl2 proteins [28]. Moreover, apoptosis has been remarkably well conserved throughout meta- zoan phyla both in terms of morphological features and of the genes controlling the process. Recently, mor- phological indicators of programmed cell death or apoptosis have been observed in a symbiotic sea anem- one, Aiptasia pallida, subjected to thermal stress [8,9], leading to the proposal of a new molecular pathway for bleaching induction. Furthermore, DNA cleavage analysis [8,23] and increased expression of p53, a pro- apoptotic protein expression [23], confirmed tempera- ture-induced DNA damage in symbiotic cnidarians, which in turn could activate the apoptotic cascade. Upstream to specific morphological modifications, apoptosis is also characterized by activation of highly selective cysteine aspartate-specific proteases, known as ‘caspases’, which are constitutively expressed as pro- enzymes with little catalytic activity and are activated following apoptotic stimulation. Evidence of caspase 3- like expression in cnidarians was first obtained in Hydra vulgaris by Cikala et al. [29] with caspase activ- ity measurements and gene characterization. Recently, evidence of caspase-like involvement in Hydractinia echinata metamorphosis [30] and a caspase gene in the sea anemone Aiptasia pallida [31] has been shown. To date however, no relation between heat stress and ca- spase activity has ever been established. In this study, we examined the biological effects of heat stress on the sea anemone Anemonia viridis, living in symbiosis with the unicellular dinoflagellate, Symb- iodinium sp. commonly known as zooxanthellae. The first aim of this study was to characterize caspase-like activity and clone a putative caspase cDNA in sea anemone tissues. In the second part of the work, we tested the effect of hyperthermal stress on antioxidant induction and on apoptotic markers (caspase-like acti- vation and ⁄ or DNA degradation) in order to demon- strate the concomitant involvement of oxidative stress and apoptosis in a thermally induced bleaching event. Results Detection of caspase-like activity in tissue extracts of A. viridis In order to test for the presence of caspase-like activity in the symbiotic sea anemone A. viridis, Asp-Glu-Val- Asp (DEVD)-dependent (Fig. 1A) and Ile-Glu-Thr-Asp (IETD)-dependent (Fig. 1B) protease activities were tested in animal host cells (ectoderm and gastroderm) and in freshly isolated zooxanthellae. For both sub- strates, high protease activities were measured in the animal host while only low activities were measured in the freshly isolated zooxanthellae extracts. Moreover, in the host extracts, IETD substrate presented a two- fold higher rate of 7-amino-4-trifluoromethylcoumarin (AFC) cleavage than DEVD-AFC substrate. Addition of the inhibitors DEVD-adelhyde (CHO) and IETD-CHO specific for the caspases 3 and 8, respectively, completely abolished both protease activit- ies in the ectodermal and gastrodermal tissue extracts (Fig. 2). Table 1 summarizes the IC 50 for both inhibi- tors, obtained by incubating extracts in each specific substrate. In both tissue extracts, DEVD- and IETD- dependent protease activities showed the same sensitivity for the DEVD-CHO competitive substrate. Surprisingly, the DEVD-CHO inhibitor had a higher effect on IETD-dependent protease activity (3 nm) than on the DEVD-dependent protease activity (15–20 nm). For both tissue extracts and protease substrates, the IETD-CHO inhibitor showed higher IC 50 values with a predictably higher sensitivity of IETD-dependent activity to this inhibitor. Although inhibition of IETD-depend- ent protease activity by IETD-CHO was similar in ectodermal and gastrodermal cells, inhibition of DEVD- dependent protease activity by the same inhibitor was lower in ectodermal cells than in gastrodermal cells. Identification of a caspase-like cDNA from A. viridis To confirm the presence of caspases in A. viridis tissues, cDNA encoding a caspase 3-like protein was isolated from ectodermal cells. Using a PCR S. Richier et al. Apoptosis and oxidative stress in bleaching FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS 4187 approach with degenerate primers based on two highly conserved caspase 3 domains, we obtained a 1627 bp sequence named AvCasp3 (accession number DQ097195) containing an open reading frame of 1239 bp (Fig. 3). The predicted amino acid sequence of 413 amino acids (Fig. 3) is highly conserved with vertebrate caspase 3 sequences (Fig. 4) and was there- fore named caspase 3-like. By homology with known vertebrate caspases, we determined that the long form of this sequence contains a prodomain, a large (p20) and a small (p10) subunit. We identified two potential cleavage sites at aspartate residues 164 and 172 for cleaving the prodomain, and a potential cleavage site at Asp306 for the cleavage between the large and small subunits. The prodomain presents a caspase recruitment domain (CARD) consisting of six alpha helices (Fig. 3) [32]. The large subunit contains highly conserved LS ⁄ THG and QACXG sequences sur- rounding histine (His255) and cysteine (Cys294) resi- dues of putative active site. The substrate binding site is highly conserved and composed of Arg337, Ser343, Gln296 and Arg198 (Fig. 3). Large and small subunit sequences from various vertebrate and invertebrate caspase 3 or caspase 7 sequences were aligned with the A. viridis caspase 3- like sequence. The phylogenetic comparison (Fig. 4) shows that AvCasp3 and other cnidarian sequences delineate a specific branch more closely related to exe- cutioner vertebrate caspases 3 or 7 than to caspases from other invertebrates models (Drosophila melano- gaster and Caenorhabditis elegans). Effect of heat stress on apoptosis-like induction in animal tissue of A. viridis In order to study the effect of a heat stress (+8 °C above ambient) on caspase-like activities, DEVD- and IETD-dependent protease activities were measured in the animal extracts of A. viridis throughout the stress (7 days at 25 °C). In the ectodermal tissue, the DEVD- dependent protease activity decreased while IETD- dependent activity did not vary significantly (Fig. 5A). In the gastrodermal tissue (Fig. 5B), both activities increased. The DEVD-dependent activity was twofold higher than controls (17 °C) after 48 h of stress, while the IETD-dependent activity was 1.5-fold higher. After 7 days at 25 °C, activities in both tissues were restored to control levels. To confirm the induction of a specific caspase-like activity and not a generic protease activity, all meas- urements were also performed in the presence of 1 lm DEVD-CHO and 10 lm IETD-CHO (data not shown). In these conditions, caspase-like activity was totally abolished. The induction of an apoptosis-like phenomenon in A. viridis subjected to a heat stress was confirmed by the analysis of DNA fragmentation in anemone tissues, using a dUTP nick end labeling (TUNEL) assay. Figure 6 shows the increase of end- labeling DNA after 48 h of thermal stress (25 °C). DNA fragmentation occurred largely in the gastroder- mal tissue harboring the zooxanthellae (Fig. 6B). Effect of heat stress on antioxidant defenses and bleaching in animal tissue of A. viridis The occurrence of oxidative stress in the animal tissue (ectoderm and gastroderm) and freshly isolated zoox- anthellae of A. viridis was monitored during heat stress (+8 °C above the control temperature) by the A 0 2 4 6 8 10 12 Ectoderm Gastroderm Zooxanthellae DEVD-AFC cleavage (pmol . min -1 ) B 0 2 4 6 8 10 12 Ectoderm Gastroderm Zooxanthellae IETD-AFC cleavage (pmol . min -1 ) Fig. 1. Caspase-like activities in host epithelial tissues (ectoderm and gastroderm) and zooxanthella extracts of A. viridis maintained in control condition (+17 °C). Caspase 3-like (A) and caspase 8-like (B) activities were assayed by fluorometric method using, respect- ively, Ac-DEVD-AFC and Ac-IETD-AFC as substrates. One hundred and twenty-five micrograms of protein have been tested for ecto- derm, gastroderm and zooxanthella extracts. Results are expressed as means ± SE of at least six independent tissue extractions from distinct sea anemones. Apoptosis and oxidative stress in bleaching S. Richier et al. 4188 FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS measurement of oxygen radical-scavenging capacities (Fig. 7). The ectodermal antioxidant capacity did not change significantly over the stress period while the gastrodermal antioxidant capacity increased starting at 6 h, peaking at 24 h at 2.5-fold higher than control values and decreasing after 48 h. In the zooxanthellae, the oxygen radical-scavenging capacity decreased signi- ficantly after 24 h. Because the stressed organisms present an evident loss of pigmentation during the stress, concomitant analyses have been conducted on whole tentacles of A. viridis to highlight the bleaching event. Figure 8 shows a rapid decrease of chlorophyll (a+c 2 ) content in the first days of the stress period that became significant with a two times decrease at the end of the kinetic. Discussion In this study, we have investigated a pathway for sym- biosis breakdown (bleaching) in the symbiotic associ- ation A. viridis during an elevated temperature stress. We have also demonstrated connections between oxi- dative stress and host programmed cell death during the bleaching event. Characterization of caspase-like activities in A. viridis In order to test for the presence of programmed cell death or apoptosis in A. viridis subjected to heat stress, we measured protease activities using mammalian caspase substrates. In control conditions, we measured high specific protease activities in the animal compo- nent (ectoderm and gastroderm) of A. viridis while, in freshly isolated zooxanthellae an activity was almost undetectable. The presence of high caspase-like activ- ities, in animal tissues of control animals, could be 0 20 40 60 80 100 120 0.1 1 1 0 100 1000 10000 100000 Inhibitor concentration (n M ) Ac-IETD-AFC cleavage activity (%) B 0 20 40 60 80 100 120 140 160 0.1 1 1 0 100 1000 10000 100000 Inhibitor concentration (n M ) Ac-DEVD-AFC cleavage activity (%) A Ectodermal extracts 0 20 40 60 80 100 120 0.1 1 10 100 1000 10000 100000 Inhibitor concentration (n M ) Ac-IETD-AFC cleavage activity (%) D Ac-DEVD-AFC cleavage activity (%) 0 20 40 60 80 100 120 140 160 0.1 1 1 0 100 1000 10000 100000 Inhibitor concentration (n M ) C Gastrodermal extracts Fig. 2. Inhibition of DEVD- and IETD-dependent protease activities in animal tissue of A. viridis by commercial synthetic peptide inhibitors. One hundred and twenty-five micrograms of ectodermal (A,B) and gastrodermal (C,D) extracts were incubated with the fluorochromic caspase substrate Ac-DEVD-AFC (A,C) or Ac-IETD-AFC (B,D) and with the competitive inhibitors Ac-DEVD-CHO (d) or Ac-IETD-CHO (s). Results are expressed as means ± SE of at least six independent tissue extractions from distinct sea anemones. Table 1. IC 50 of the competitive substrate inhibitors Ac-DEVD-CHO and Ac-IETD-CHO on the DEVD-dependent and IETD-dependent protease activities in the two animal tissue extracts of A. viridis. Tissue extracts Protease substrates Ac-DEVD-CHO IC 50 (nM) Ac-IETD-CHO IC 50 (nM) Ectodermal Ac-DEVD-AFC 20 45190 Ac-IETD-AFC 3 172 Gastrodermal Ac-DEVD-AFC 15 495 Ac-IETD-AFC 3 105 S. Richier et al. Apoptosis and oxidative stress in bleaching FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS 4189 related to the high regeneration ability of cnidarians. In fact, the role of apoptosis in development and regeneration has been determined not only in verte- brates (i.e., bone regeneration; reviewed in [33]) but also in invertebrates such as cnidarians and flatworms [34–36]. Apoptosis is considered a necessary character- istic of all self-renewing tissues and its presence has been detected not only in stressed organisms but also in healthy ones. Mire and Venable [34] reported that up to 10% of cells from the sea anemone Haliplanella lineata contained TUNEL-labeled nuclei even under control conditions. Moreover, protease activities related to animal extracts display several properties characteristic of caspases, the critical central molecules of apoptotic pathways. First, they were activated by two polypep- tides, DEVD and IETD, which are used to distinguish some of the caspase classes in mammalian cells. DEVD-AFC is generally cleaved by caspase 3, which belongs to executioner caspases [37] while, IETD-AFC Fig. 3. Nucleotide and deduced amino acid sequence of caspase 3-like cDNA of A. viridis (AvCasp3). Putative prodomain sequence appears in italic characters, the small subunit in regular type. The large (p20) subunit in bold and the small subunit fit between the two domains. Residues boxed are the component of substrate binding site. Asterisks indicate the His and Cys residues of the putative active site located in the large subunit. The six a -helix components of the prodomain are underlined. Apoptosis and oxidative stress in bleaching S. Richier et al. 4190 FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS is cleaved by caspase 8, an initiator caspase [38]. In control specimens of A. viridis, IETD-dependent prote- ase activity was two times higher than the DEVD- dependent one, suggesting a predominantly caspase 8-like activity in the animal tissue. Secondly, the spe- cificity of detected caspases has been tested using their respective competitive substrates: DEVD-CHO and IETD-CHO [39]. The inhibition of the two protease activities by competitive substrates strengthens the involvement of a caspase-like activity, in animal tissue of A. viridis, avoiding interference by generic proteas- es. In the two animal cell layers, the effect of the two inhibitors was similar. Although we measured a higher efficiency of DEVD-CHO for IETD-dependent activity in both tissues of A. viridis (Table 1), in the literature this inhibitor was found to be highly effective on both caspase 3 and caspase 8 activities but still more speci- fic to caspase 3 [39]. DEVD peptide, which was devel- oped as a caspase 3 inhibitor, is also a fairly potent inhibitor of caspases 1, 4 and 7, and is not conse- quently selective for a particular caspase [40]. This seems to indicate that tetrapeptide-based inhibitors are unlikely to achieve the specificity required to allow selective inhibition of caspases. However, compared to results related to inhibitor specificity performed on mammalian cells, we can conclude that there are at least two original caspase-like activities in the animal tissue of A. viridis. Previous work has already highlighted the presence of a caspase-like activity in the hydrozoan Hydra vul- garis using DEVD substrate and DEVD-CHO inhib- itor. The presence of such an enzyme in cnidarians was confirmed by gene sequencing first, in H. vulgaris with 3A and 3B Casp, sharing a high degree of identity with, respectively, C. elegans CED3 and human Casp 3, respectively [29]. More recently a caspase 3- like cDNA has been sequenced in both the anemone A. pallida sharing high identity with 3B casp of H. vul- garis [31] and in Hydractinia vulgaris [30]. We also confirmed the presence of a caspase 3-like cDNA in A. viridis host tissues. We obtained a full-length cDNA sequence from ectodermal tissue with a deduced amino acid sequence that is closely related to vertebrate exe- cutioner caspases 3 ⁄ 7. All of the conserved residues involved in the catalytic mechanism of caspases are present in AvCasp3, as well as cleavage sites identified by homology with vertebrate caspases. However, this sequence also possess a long prodomain homologous to the CARD domain [41], mostly similar to the initi- ator caspases 2, 8, 9 and 10 [42]. The characteristics have been described in Acasp from the sea anemone A. pallida [31]. Acasp (large and small subunits) shares an 81% identity and 91% similarity with AvCasp3 but only a 40% identity and 57% similarity with the caspase 3B from H. vulgaris. As cnidarian caspase 3-like sequences shared both characteristics of execu- Rattus norvegicus Casp3 Mus musculus Casp3 Homo sapiens Casp3 Gallus gallus Casp3 Xenopus laevis Casp3 Danio rerio Casp3 Salmo salar Casp3B Xenopus laevis Casp7 Rattus norvegicus Casp7 Homo sapiens Casp7 Hydra vulgaris Casp3B Aiptasia pallida Casp3-like Anemonia viridis Casp3-like D. melanogaster Casp3 Homo sapiens Casp8 Mus musculus Casp8 C. elegans Casp3 100 100 95 44 97 81 91 99 85 74 29 33 45 35 0. 5 Fig. 4. Phylogenetic comparison of A. viridis caspase with caspase sequences from ver- tebrates, invertebrates caspase 3 and from vertebrates caspase 7. Vertebrate caspase 8 sequences have been used as an outgroup. The tree was derived from alignments of p10 and p20 domains excluding the pro- domain. S. Richier et al. Apoptosis and oxidative stress in bleaching FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS 4191 tioner (caspases 3 and 7) and initiator caspases (CARD domain, caspases 2, 8, 9, 10), this suggests that cnidarian caspase 3-related enzymes may be considered as potential ancestors of other metazoan and vertebrate executioner caspases [31]. This has also been suggested for H. vulgaris caspase 3-like [29] and by Wiens et al. [43] for sponge caspase 3-like enzymes. Furthermore, the caspase 3-like gene we described in A. viridis appears more related to vertebrates than to other invertebrate biological models such as nematodes and flies [44]. This could be explained by the basal position of cnidarians on the metazoan tree and by the extensive gene loss in protostomes. Caspase-like activity and thermal stress After 2 days of heat stress, the increase in at least two caspase-like activities detected in the animal tissue and the DNA fragmentation induction in the gastrodermal cells suggest the involvement of apoptotic events dur- ing the first hours of high temperature treatment. It also confirms previous work, which has already dem- onstrated the induction of apoptosis in cnidarians sub- jected to heat stress [8,9]. Because the present data constitute the first evidence of caspase activation under heat stress in cnidarian, further experiments are required to exclude the hypothesis of involvement of this latter enzyme in mechanisms other than the apop- totic cascade. Nonapoptotic functions of caspase 3 have been described recently in human nervous tissue [45,46]. Dunn et al. [8,9] have recently reported an increase in morphological apoptotic indicators in the sea anemone Aiptasia sp. incubated at high tempera- ture. These authors reported a high frequency of cells with apoptosis-like morphology predominantly in the gastrodermal host cells and from the first hour of exposure. A similar time-dependent pattern has been Ectoderm Gastroderm control condition (17°C) stress condition (25°C) 20 µm Mesoglea A B Fig. 6. Temperature-induced DNA fragmen- tation in tentacle tissue of A. viridis. DNA fragmentation in specimen maintained at 17 °C (A, control condition) or incubated at 25 °C for 48 h (B, stress condition) was revealed by TUNEL staining with DAB ⁄ H 2 O 2 substrate. Arrows indicate the different cell layers. 0 50 100 150 200 250 300 0 6 24 48 168 Time (hours) Gastrodermal rate of AFC cleavage (% of control) B ** 0 50 100 150 200 250 300 0 6 24 48 168 Time (hours) A *** DEVD-dependent protease activity IETD-dependent protease activity Ectodermal rate of AFC cleavage (% of control) Fig. 5. Temperature-induced protease activity in host epithelial tis- sue of A. viridis. One hundred and twenty-five micrograms of ecto- dermal (A) and gastrodermal (B) tissue extracts were incubated with the Ac-DEVD-AFC (Caspase 3; black bars) and Ac-IETD-AFC (Caspase 8, white bars) substrates. The assays were performed during the increase temperature treatment (25 °C) and are expressed as percentage of control (temperature incubation 17 °C) ± SE of at least five independent tissue extractions from dis- tinct sea anemones. Asterisks indicate significant differences between control and stress conditions (P<0.05; ANOVA). Apoptosis and oxidative stress in bleaching S. Richier et al. 4192 FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS observed in our study, however, the apoptotic events appear later in the heat stress (with a high activity reported at 48 h). The difference between the two stud- ies could be related to a species-specific sensitivity or to the temperature range. Moreover, the later induc- tion of necrotic events observed by Dunn et al. [9] could be correlated with the caspase-like activity decrease measured in A. viridis gastrodermal tissue after 7 days of treatment. The decrease in ectodermal caspase-like activity observed after 6 h was, however, not related to necrotic cell death because it concerned only DEVD-dependent protease activity. It could be the result of caspase inhibition by a still undetermined mechanism. Implication of different molecules is suggested, such as inhibitor of apoptosis protein and heat shock proteins [47,48]. Caspase-like activity and oxidative stress After the finding of caspase-like activities and a response by these enzymes to heat stress, parallel ana- lyses were conducted to follow the occurrence of oxi- dative stress in the stressed organisms. Variations of antioxidant defenses and caspase activities were then compared. An increase in caspase 8-like activity appears after 24 h of gastrodermal antioxidant defense induction. High protein damage has been shown in previous stud- ies and supports the occurrence of an oxidative stress period in heat stressed A. viridis [24]. We suggest that apoptosis induction could be the consequence of the previous oxidative stress event, a phenomenon that is well established in vertebrates [27]. Increases in ROS are the consequence of electron transport chain impair- ments, principally in mitochondria and chloroplasts and can directly and ⁄ or indirectly cause caspase activa- tion (reviewed in [25,27,49]). The delay observed in our study between the antioxidant induction and the caspase-like activation largely supports this hypothesis. Bleaching event and oxidative stress Several studies have underlined the effect of thermal stress on symbiotic cnidarians [1,2] and several hypo- theses have been suggested to explain the mechanisms of symbiosis breakdown, focused on the respective implication of both partners in the phenomenon. In the zooxanthella, heat stress has been demonstrated to reduce the photosynthetic rate by decreasing the effi- ciency of the photosystem II [4–6] and ⁄ or by causing damage to the Calvin cycle [50,51]. Moreover, in situ degeneration of zooxanthellae has been reported in corals [52–54] and sea anemones [8,9,55]. Heat stress was also documented to induce host cell degeneration [8,9,53,56] and ⁄ or gastrodermal detachment [7,52,57]. Nevertheless, independently of the resulting effect of the thermal stress, all mechanisms point to the involve- ment of oxidative stress in the early stages of symbiosis breakdown. In fact, several authors have suggested the involvement of ROS production in the zooxanthella photoinhibition (reviewed in [1]), gastrodermal cell detachment [7], and host and symbiont degeneration [9,23]. However, molecular mechanisms linked to ther- mal stress induction, production of ROS and its phy- siological consequences (e.g., photoinhibition and cell degeneration) are still unclear. In previous work, Richier et al. [24] showed the occurrence of oxidative 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 6 24 48 168 Time (hours) Relative antioxidant activity * * * * Ectodermal extracts Zooxanthellae extracts Gastrodermal extracts Fig. 7. Temperature-induced antioxidant activity in tentacle tissue of A. viridis. Relative antioxidant activities were measured in 1 lg of ectodermal (black bars), gastrodermal (grey bars) and zooxanthel- la (white bars) extracts by fluorometric assay during the increase temperature treatment (25 °C) and are expressed relative to control condition (temperature incubation 17 °C) ± SE of at least five inde- pendent tissue extractions from distinct sea anemones. Asterisks indicate significant differences between control and stress condi- tions (P<0.05; ANOVA). 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0 50 100 150 200 Time (hours) Chlorophyll (a+c 2 ) content (µg · mg -1 ) * Fig. 8. Chlorophyll a and c 2 content in A. viridis total tentacle extract during thermal stress (25 °C). Results are expressed as means ± SE of at least three independent chlorophyll extractions from distinct sea anemone tentacles. Asterisks indicate significant differences between control and stress conditions (P<0.05; ANOVA). S. Richier et al. Apoptosis and oxidative stress in bleaching FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS 4193 stress in the animal host and its time course of appear- ance. While symbiotic sea anemones seem to be more resistant to thermal stress than nonsymbiotic species, there was nonetheless an increase in oxidative attack on proteins, as evidenced by the carbonylation of pro- tein in A. viridis after a thermal increase of 8 °Cin gastrodermal cells where zooxanthellae are housed [24]. The present study supports these previous results and shows the induction of antioxidant defenses exclu- sively localized within the animal compartment and more precisely in gastrodermal cells. These results sug- gest the induction of antioxidant defenses in the gastro- dermal compartment in order to counteract the increase in cellular damage. By contrast, a decrease in global antioxidant defenses in the zooxanthellae was observed during the experiment. Previous results obtained by Lesser and coworkers [18,19] and Richier et al. [24] showed a slight increase in the activity of the antioxidant enzymes superoxide dismutase and ascor- bate peroxidase in zooxanthellae. Nevertheless, a glo- bal decrease in zooxanthellae antioxidant defenses during thermal stress would suggest two hypotheses: (a) a dysfunction of zooxanthella metabolism induced by necrosis or programmed cell death as suggested by Dunn et al. [8,9]; or (b) a decrease in antioxidant def- enses following a chlorophyll decrease as demonstrated by Shick et al. [58]. In this thermal stress experiment (+8 °C), we have observed a significant decrease of chlorophyll content after 7 days incubation. Moreover, visible bleaching of the experimental animals occurred as the heat stress incubation progressed. This result supports the induction of a bleaching event as a conse- quence of an oxidative stress period. In conclusion, our results contribute to the under- standing of the mechanisms involved in coral bleaching events in host cells of symbiotic cnidarians. Gastroder- mal cells appear to be the predominant location of thermal stress impact. In the first hour of stress, the gastrodermal cells undergo oxidative stress, which is rapidly followed by apoptotic events and completed by occurrence of bleaching. The gastrodermal cell death is then hypothesized to be responsible of zooxanthella expulsion and ⁄ or gastrodermal cell detachment. Fur- ther investigation will, however, be necessary to link gastrodermal cell death with zooxanthella photoinhibi- tion and expulsion. Finally, this work contributes to the investigation of the evolutionary conservation and the role of apoptosis in basal metazoans. Experimental procedures Unless otherwise specified, all chemicals were obtained from Sigma-Aldrich (St Louis, MO). Biological materials Specimens of the Mediterranean sea anemone, Anemonia viridis (Forska ˚ l), were collected in Villefranche-sur-mer (France) and maintained in a closed-circuit seawater aquar- ium at 17 ± 1 °C. Half of the aquarium seawater was changed every week. A metal halide lamp (HQI-TS, 400 W; Philips, Eindhoven, the Netherlands) provided light, at a constant saturating irradiance of 250 lmolÆquanta m )2 Æs )1 on a 12 h light ⁄ 12 h dark regime. Experimental designs Three tanks, each containing one anemone, were heated from 17 ± 1 °C (control condition) to 25 ± 1 °C (stress condition) over 2 h and maintained at 25 °C for 7 days. During the course of the experiment, all tanks were main- tained under identical illumination conditions (250 lmolÆ quanta m )2 Æs )1 ,12h⁄ 12 h L ⁄ D) and the sea anemones were not fed. Five to ten tentacles from each specimen were sam- pled from each aquarium after 6, 24, 48 and 168 h after the initial temperature increase. The experiment was repeated twice with distinct sea anenome specimens. Tissue extractions The three cellular compartments of the symbiotic associ- ation (host ectoderm, host gastroderm and zooxanthellae) were extracted from tentacles according to Richier et al. [13], avoiding any contamination between zooxanthellae protein and the host gastrodermal cell. Each extract was prepared at 4 °C in the appropriate medium for the subse- quent analyses. Oxygen radical-scavenging assay The oxygen radical-scavenging activities of different tissue extracts were determined using a fluorometric assay accord- ing to Naguib [59]. Tissue extractions were performed in a medium containing 50 mm phosphate buffer (pH 7.0), 2mm phenylmethylsulfonyl fluoride and 10 lgÆml )1 prote- ase inhibitor cocktail (P8340, Sigma). The volume corres- ponding to 1 lg of protein extract was then incubated with 75 mm phosphate buffer (pH 7.0), 15 nm fluorescein as fluorescent probe, 3 mm 2,2¢-azobis (2-amidino-propane) dihydrochloride as the peroxyl radical generator, and 1 lm Trolox as antioxidant standard. The decay of fluorescence signal was recorded by a spectrofluorometer (Safas, Monaco) at an excitation ⁄ emission wavelength of 520 ⁄ 495 nm every minute for a total duration of 45 min. Caspase assay Caspase 3-like and caspase 8-like activities were assayed fluorometrically using the specific substrates Ac-DEVD- Apoptosis and oxidative stress in bleaching S. Richier et al. 4194 FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS AFC (N-acetyl-Ile-Glu-Val-Asp-7-amino-4-trifluoromethyl- coumarin) and Ac-IETD-AFC (N-acetyl-Ile-Glu-Thr-Asp- 7-amino-4-trifluoromethylcoumarin), respectively (Biosource International, Cliniscience, Montrouge, France). Tissue extractions were performed in a medium containing 25 m m Hepes (pH 7.5), 5 mm MgCl 2 ,5mm 1,4-dithiothreitol, 2mm phenylmethylsulfonyl fluoride and 10 lgÆml )1 prote- ase inhibitor cocktail (P8340, Sigma). A quantity of 125 lg of ectodermal, gastrodermal and zooxanthella extracts was incubated in a reaction buffer containing 50 lm specific probe, 100 mm Hepes (pH 7.5), 10% (v ⁄ v) sucrose, 0.1% (v ⁄ v) CHAPS, 10 mm dithiothreitol and dimethylsulfoxide. The AFC fluorescence was measured in a spectrofluorome- ter (Safas, Monaco) at an excitation ⁄ emission wavelength of 400 ⁄ 505 nm every 3 min for 21 min for animal compart- ments and extended to 60 min for zooxanthellae extracts. Caspase-like activities were expressed in picomol of AFC cleavage per minute, according to a standard curve obtained from AFC (A8401, Sigma). For inhibition experi- ments, the competitive peptide (inhibitor) Ac-DEVD-CHO or Ac-IETD-CHO were added to the reaction buffer prior the addition of the fluorometric substrate. RNA extraction RNA from the ectodermal compartment was extracted from six tentacles of A. viridis according to a modified pro- tocol of Trizol extraction (Invitrogen, Carlsbad, CA). The tentacles were isolated and immediately dried with blotting paper. The gastrodermal cells were scraped at 4 °C with forceps in order to effectively separate the ectodermal cell layer, which was dissolved in 2 mL Trizol using a glass ho- mogenizer. Homogenate was then processed in accordance with Invitrogen recommendations, followed by final addi- tional treatment using chloroform. After the precipitation step using 70% ethanol, the RNA pellet was finally air dried and dissolved in RNAse-free water. The RNA was quantified by measuring the absorbance at 260 nm (Safas UVmc2 spectrophotometer). RT-PCR Total RNA from the ectodermal fraction of A. viridis was reverse transcribed using oligodT primer and Superscript II reverse transcriptase (Invitrogen). Degenerate primers were designed from two highly conserved regions of caspase 3 amino acid sequences from phylogenetically different organisms (Hydra vulgaris AAF98012, Homo sapiens AAH15799, Aiptasia pallida DQ218058): CniCaspF, 5¢-CAYGGNGARGARGGRAT-3¢ and CniCaspR 5¢-AT NGANGGDATYTGYTTYTT-3¢). A volume of 0.5 lLof ectodermal cDNA was PCR amplified using CniCaspF (300 nm), CniCaspR (300 nm) and Platinum TAQ poly- merase (Invitrogen). PCR products were analyzed by 2% (w ⁄ v) agarose gel electrophoresis, then subcloned into pGEM-T Easy vector (Promega, Madison, WI) and se- quenced (Macrogen Inc, Seoul, South Korea). Rapid amplification of cDNA ends To further obtain the full-length cDNA sequence of A. vir- idis caspase, 5¢⁄3¢ RACE-PCR kit and Expand Long Template DNA Polymerase (both from Roche, Mannheim, Germany) were used. For 3¢-RACE, the gene specific prim- ers were: AvCasp1F (5¢-CTTGGCGAAACTCAGTCAAT GG-3¢) and AvCasp2F (5¢-CTGCTGACAATGATGACG AGAG-3¢). For 5¢-RACE, the gene specific primers were: AvCasp1R (5¢-GTCAGCAGATCTGTGGTTTTG-3¢), AvCasp2R (5¢-CCATTG ACTGA GTTTCG CCAA G-3¢). PCR products were cloned into pGEM-T Easy vector and sequenced. Sequence analysis The blast sequence analysis program [60] was used for ini- tial comparisons. Multiple alignment of large (p20) and small (p10) caspase 3 domains from vertebrates (Homo sapiens P42574, Mus musculus P70677, Rattus norvegicus NP037054, Gallus gallus AF083029, Danio rerio AAH78310, Salmo salar AAY28972, Xenopus laevis P55866) and invertebrates (Hydra vulgaris AAF98012, Aiptasia pallida DQ218058, Drosophila melanogaster AAD54071, Caenorhabditis elegans P42573), of vertebrate caspase 7 sequences (Homo sapiens AAH15799, Rattus norvegicus NP071596, Xenopus laevis AAH78049) and of vertebrate caspase 8 sequences as an outgroup (Homo sap- iens Q14790, Mus musculus AF067834) was performed with clustalx program [61]. A phylogenetic tree was derived from alignments using the Neighbour Joining method and the mega3 software [62]. Secondary structure (alpha heli- ces) of the corresponding peptide was predicted using the PSIPRED server [63]. TUNEL assay DNA fragmentation was identified in situ by terminal de- oxynucleotidyl transferase mediated dUTP nick end labe- ling (TUNEL) labeling [64]. Tentacle bags of A. viridis [65] were fixed with 4% (w ⁄ v) paraformaldehyde in a fixation buffer (450 mm NaCl, 10 mm KCl, 58 mm MgCl2, 100 mm Hepes pH 7.8) overnight at 4 °C. Tissues were then dehy- drated in ethanol series, cleared with toluene and embedded in paraplast. Sections of 8 lm-thick were attached to Silane-Prep slides, deparaffinized with xylene, and rehydrat- ed in ethanol series. Sections were successively incubated in proteinase K (1 ngÆml )1 in TE buffer) for 15 min at room temperature and in TdT buffer (140 mm cacodylate, 1 mm cobalt chloride, 30 mm Tris pH 7.4) for 5 min. End-labeling was carried out in 45 lL TUNEL-labeled dUTP (Roche S. Richier et al. Apoptosis and oxidative stress in bleaching FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS 4195 [...]...Apoptosis and oxidative stress in bleaching S Richier et al Diagnostic) and 5 lL terminal deoxynucleotidyl transferase (TdT, Roche Diagnostic) solution for 30 min at 37 °C Sections were then washed in the stop solution (300 mm NaCl, 30 mm sodium citrate) and incubated with biotin horseradish peroxidase complex (TUNEL POD, Roche Diagnostic) for 15 min After two 15 min washes in NaCl ⁄ Pi, the detection... mechanisms underlying temperature-induced bleaching in the tropical sea anemone Aiptasia pulchella J Exp Biol 204, 3443–3456 Dunn SR, Bythell JC, Le Tissier MDA, Burnett WJ & Thomason JC (2002) Programmed cell death and cell necrosis activity during hyperthermic stress- induced bleaching of the symbiotic sea anemone Aiptasia sp J Exp Mar Biol Ecol 272, 29–53 Dunn SR, Thomason JC, Le Tissier MDA & Bythell JC... (1989) The effect of sudden changes in temperature, light and salinity on the population density and export of zooxanthellae from the reef corals Stylophora pistillata Esper and Seriatopora hystrix Dana J Exp Mar Biol Ecol 129, 279–303 4 Iglesias-Prieto R, Matta JL, Robins WA & Trench R (1992) Photosynthetic response to elevated temperature in the symbiotic dinoflagellate Symbiodinium microadriaticum in. .. Heat stress induces different forms of cell death in sea anemones and their endosymbiotic algae depending on temperature and duration Cell Death Differ 11, 1213–1222 D’Aoust BG, White R, Wells JM & Olsen DA (1976) Coral–algal associations: capacity for producing and sustaining elevated oxygen tensions in situ Undersea Biomed Res 3, 35–40 Crossland CJ & Barnes DJ (1977) Gas-exchange studies with the. .. Woodley CM (2002) Oxidative stress and seasonal coral bleaching Free Rad Biol Med 33, 533–543 Brown BE, Downs CA, Dunne RP & Gibb SW (2002) Exploring the basis of thermotolerance in the reef coral Goniastrea aspera Mar Ecol Prog Ser 242, 119–129 Lesser MP & Farrell JH (2004) Exposure to solar radiation increases damage to both host tissues and algae symbionts of corals during thermal stress Coral Reefs... protein family (IAPs): an emerging therapeutic target in cancer Semin Cancer Biol 14, 231–243 48 Beere HM (2004) The stress of dying’: the role of heat shock proteins in the regulation of apoptosis J Cell Sci 117, 2641–2651 49 Skulachev VP (1998) Cytochrome c in the apoptotic and antioxidant cascades FEBS Lett 423, 275–280 50 Jones RJ, Hoegh-Guldberg O, Larkum AWD & Schreiber U (1998) Temperature-induced... Stanislaus D (2001) Osteoblast apoptosis and bone turnover J Bone Miner Res 16, 75–84 Mire P & Venable S (1999) Programmed cell death during longitudinal fission in a sea anemone Invertebr Biol 118, 319–331 Apoptosis and oxidative stress in bleaching 35 Miller DJ & Ball EE (2000) The coral Acropora: what it can contribute to our knowledge of metazoan evolution and the evolution of developmental processes... presented as means ± standard errors and normalized by the protein content of each fraction Protein was determined by the Biorad (Hercules, CA) protein assay using BSA as standard protein The results were validated by one-way anova with Fisher post hoc test and were considered statistically significant when P < 0.05 10 11 Acknowledgements We are grateful to the Daniel Jouvance Company and the Provence Alpes... Temperature-induced bleaching of corals begins with impairment of the CO2 fixation mechanism in zooxanthellae Plant Cell Environ 21, 1219–1230 FEBS Journal 273 (2006) 4186–4198 ª 2006 The Authors Journal compilation ª 2006 FEBS 4197 Apoptosis and oxidative stress in bleaching S Richier et al 51 Leggat W, Whitney S & Yellowlees D (2004) Is coral bleaching due to the instability of the zooxanthellae dark reactions?... 25 26 27 28 29 30 31 32 33 34 activities of protective enzymes against active oxygen Coral Reefs 8, 225–232 Lesser MP (1996) Elevated temperatures and ultraviolet radiation cause oxidative stress and inhibit photosynthesis in symbiotic dinoflagellates Limnol Oceanogr 41, 271–283 Lesser MP (1997) Oxidative stress causes coral bleaching during exposure to elevated temperature Coral Reefs 16, 187–192 Downs . Oxidative stress and apoptotic events during thermal stress in the symbiotic sea anemone, Anemonia viridis Sophie Richier 1 ,Ce ´ cile Sabourault 1 , Juliette Courtiade 1 , Nathalie Zucchini 3 ,. be one of the mechanisms involved. We focused our study on the role of apoptosis in bleaching via oxidative stress in the association between the sea anemone Anemonia viridis and the dinoflagellates. [7,52,57]. Nevertheless, independently of the resulting effect of the thermal stress, all mechanisms point to the involve- ment of oxidative stress in the early stages of symbiosis breakdown. In fact,